Matthews' Plant Virology
'The farther b a c k w a r d s y o u can look, the farther forward y o u are likely to see.'...
498 downloads
2291 Views
100MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Matthews' Plant Virology
'The farther b a c k w a r d s y o u can look, the farther forward y o u are likely to see.' Winston Churchill
Matthews' Plant Virology Fourth Edition
Roger Hull Emeritus Research Fellow John Innes Center Norwich Research Park Colney, Norwich
ELSEVIER ACADEMIC PRESS
AMSTERDAM
9 BOSTON
PARIS e SAN DIEGO
9 HEIDELBERG
9 SAN FRANCISCO
9 LONDON 9 SINGAPORE
9 NEW YORK 9 SYDNEY
9 OXFORD 9 TOKYO
This book is printed on acid-free paper Copyright 92002, 1991, 1981, 1970, Elsevier (USA). All rights reserved Reprinted 2004 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 written permission of the publisher Permissions may be sought directly from Elsevier's Science & Technology Rights Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: permissions@ elsevier.co.uk. You may also complete your request on-line via the Elsevier homepage (http://www.elsevier.com), by selecting 'Customer Support' and then 'Obtaining Permissions' Elsevier Academic Press 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.elsevier.com Elsevier Academic Press 84 Theobald's Road, London W C l X 8RR, UK http ://www.elsevier.com Library of Congress Catalog Number: 2001089791 British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0-12-361160-1
Typeset by J & L Composition Filey, North Yorkshire Printed and bound by Krips, The Netherlands 02 03 04 05 06 07 08 9 8 7 6 5 4 3 2
About the Author
Roger Hull graduated in Botany from Cambridge University in 1960, and subsequently studied plant virus epidemiology at London University's Wye College, gaining a PhD in 1964. He lectured on agricultural botany there between 1960 and 1965. He was seconded to Makerere University in Kampala, Uganda in 1964 where he taught, and learnt, tropical agricultural botany and studied the epidemiology of groundnut rosette disease. By watching aphids land on groundnut plants he gained an understanding of the edge effect of spread of virus into the field. In 1965 Roger Hull joined Roy Markham at the ARC Virus Research Unit in Cambridge where he worked on biophysical and biochemical characterization of a range of viruses, especially alfalfa mosaic virus. This work continued when he moved to the John Innes Institute, Norwich with Roy Markham in 1968. There Dr Hull became a project leader and
deputy head of the Virus Research Department. In 1974 he spent a sabbatical year with Bob Shepherd in the University of California, Davis where he worked on the characterization of cauliflower mosaic virus. There he was introduced to the early stages of molecular biology which changed the direction of his research. On returning to the John Innes Institute he applied a molecular biological approach to the study of cauliflower mosaic virus elucidating that it replicated by reverse transcription, the first plant virus being shown to do so. Involvement with the Rockefeller Rice Biotechnology Program reawakened his interest in tropical agricultural problems and he led a large group studying the viruses of the rice tungro disease complex. He also promoted the use of transgenic technology to the control of virus diseases and was in the forefront in discussing biosafety issues associated with this approach. Moving from rice to bananas (plantains) his group was among those who discovered that the genome of banana streak badnavirus was integrated into the host genome and in certain cultivars was activated to give episomal infection- another first for plant viruses. He retired at the statutory age in 1997. Dr Hull is an Honorary Professor at East Anglia University and Peking University, a Doctoris Honoris Causa at the University of Perpignan, France, and a Fellow of the American Phytopathological Society. Today Dr Hull is an Emeritus Research Fellow at the John Innes Centre and still continues research on banana streak virus. He is involved in promoting the uptake of transgenic technology by developing countries as one approach to alleviating food insecurity. His other interests are gardening, bird watching, travelling and his children and grandchildren.
This Page Intentionally Left Blank
Contents
Chapter 1
Chapter 2
About the Author Preface Introduction I. Historical Background II. Definition of a Virus III. About this Edition Nomenclature and Classification of Plant Viruses I. Nomenclature A. Historical aspects B. Systems for classification C. Families, genera, species and groups D. Plant virus families, genera and orders E. Use of virus names II. Criteria Used for Classifying Viruses A. Structure of the virus particle B. Physicochemical properties of virus particles C. Properties of viral nucleic acids D. Viral proteins E. Serological relationships F. Activities in the plant G. Methods of transmission III. Families and Genera of Plant Viruses A. Family Caulimoviridae B. Family Geminiviridae C. Family Circoviridae D. Family Reoviridae E. Family Partitiviridae F. No family G. Family Rhabdoviridae H. Family Bunyaviridae I. No family J. Family Bromoviridae K. Family Comoviridae L. Family Potyviridae M. Family Tombusviridae vii
V
xix 1 1 9 11 13 13 13 14 15 19 19 21 21 21 21 22 24 25 26 27 27 28 28 29 29 30 30 30 31 31 32 33 33
viii
CONTENVS
Chapter 3
Chapter 4
Chapter 5
N. Family Sequiviridae O. Family Closteroviridae P. Family Luteoviridae Q. Floating genera IV. Retroelements A. Family Pseudoviridae B. Family Metaviridae VI. Viruses of Lower Plants A. Viruses of algae B. Viruses of fungi C. Viruses of ferns D. Viruses of gymnosperms E. Summary VI. Discussion Disease Symptoms and Host Range I. Economic Losses due to Plant Viruses II. Macroscopic Symptoms A. Local symptoms B. Systemic symptoms C. Agents inducing virus-like symptoms D. The cryptoviruses III. Histological Changes A. Necrosis B. Hypoplasia C. Hyperplasia IV. Cytological Effects A. Methods B. Effects on cell structures C. Virus-induced structures in the cytoplasm D. Cytological structures resembling those induced by viruses E. Discussion V. The Host Range of Viruses A. Limitations in host range studies B. Patterns of host range C. The determinants of host range VI. Discussion and S u m m a r y Purification and Composition of Plant Viruses I. Introduction II. Isolation A. Choice of plant material III. Components A. Nucleic acids B. Proteins C. Other components in viruses D. Discussion and s u m m a r y Architecture and Assembly of Virus Particles I. Introduction II. Methods A. Chemical and biochemical studies B. Methods for studying size of viruses
35 35 36 37 40 41 41 42 42 43 44 44 44 44 47 47 48 48 49 53 56 56 56 56 57 58 58 59 62 66 67 67 68 69 69 73 75 75 75 76 86 87 100 104 106 109 109 109 109 110
CONTENTS
Chapter 6
C. Fine structure determination: electron microscopy D. X-ray crystallographic analysis E. Neutron small-angle scattering F. Mass spectrometry G. Serological methods H. Methods for studying stabilizing bonds III. Architecture of Rod-Shaped Viruses A. Introduction B. Tobamovirus genus C. Tobravirus genus D. Other helical viruses IV. Assembly of Rod-Shaped Viruses A. TMV B. Other rod-shaped viruses V. Architecture of Isometric Viruses A. Introduction B. Quasi-equivalence C. Possible icosahedra D. Clustering of subunits E. 'True' and 'quasi' symmetries F. Bacilliform particles VI. Small Icosahedral Viruses A. Subunit structure B. Virion structure C. The arrangement of nucleic acid within icosahedral viruses VII. More Complex Isometric Viruses VIII. Enveloped Viruses A. Rhabdoviridae B. Tospoviruses IX. Assembly of Icosahedral Viruses A. Bromoviruses B. Alfalfa mosaic virus C. Other viruses D. RNA selection during assembly of plant reoviruses X. Discussion and Summary Genome Organization I. Introduction II. General Properties of Plant Viral Genomes A. Information content B. Economy in the use of genomic nucleic acids C. The functions of viral gene products D. Non-coding regions III. Plant Viral Genome Organization IV. Double-Stranded DNA Viruses A. Family Caulimoviridae V. Single-Stranded DNA Viruses A. Family Geminiviridae B. Family Circoviridae VI. Double-Stranded RNA Viruses A. Family Reoviridae
ix
111 113 114 114 114 116 117 117 118 123 124 126 126 134 134 134 135 136 137 138 138 138 138 139 157 159 160 160 162 163 163 165 165 167 168 171 171 171 171 172 172 174 174 174 174 180 180 183 183 183
X
CONTENTS
Chapter 7
Chapter 8
B. Family Partitiviridae C. Genus Varicosavirus VII. Negative-Sense Single-Stranded RNA Genomes A. Family Rhabdoviridae B. Family Bunyaviridae VIII. Positive-Sense Single-Stranded RNA Genomes A. Family Bromoviridae B. Family Comoviridae C. Family Potyviridae D. Family Tombusviridae E. Family Sequiviridae F. Family Closteroviridae G. Family Luteoviridae H. Floating genera IX. Summary and Discussion Expression of Viral Genomes I. Introduction II. Virus Entry and Uncoating A. Virus entry B. Uncoating of TMV C. Uncoating of bromoviruses D. Uncoating of SBMV E. Uncoating of TYMV F. Discussion III. Viral Genome Expression A. Structure of the genome B. Defining functional ORFs C. Recognizing activities of viral genes D. Matching gene activities with functional ORFs IV. Synthesis of mRNAs A. Negative-sense single-stranded RNA viruses B. Double-stranded RNAviruses C. DNA viruses V. Plant Viral Genome Strategies A. The eukaryotic protein-synthesizing system B. Virus strategies to overcome eukaryotic translation constraints C. Control of translation D. Discussion E. Positive-sense ssRNA viruses that have more than one strategy F. Negative-sense single-stranded RNA viruses G. Double-stranded RNAviruses H. DNA viruses VI. Discussion Virus Replication I. Introduction II. Host Functions Used by Plant Viruses A. Components for virus synthesis B. Energy C. Protein synthesis D. Nucleic acid synthesis
187 187 187 187 188 189 189 194 196 198 202 203 205 207 221 225 225 226 226 226 229 230 230 231 232 233 235 238 240 244 244 245 246 253 253 254 272 276 276 289 289 289 290 293 293 293 293 293 293 294
CONTENTS
Chapter 9
E. Structural components of the cell III. Methods for Studying Viral Replication A. In vivo systems B. In vitro systems IV. Replication of Positive-Sense Single-Stranded RNA Viruses A. Viral templates B. Replicase C. Sites of replication D. Mechanism of replication E. Replication of brome mosaic virus F. Replication of cucumber mosaic virus G. Replication of alfalfa mosaic virus H. Replication of tobacco mosaic virus I. Replication of potyviruses J. Replication of Comoviridae K. Replication of turnip yellow mosaic virus L. Replication of other (+)-strand RNA viruses M. Discussion V. Replication of Negative-Sense Single-Stranded RNA Viruses A. Plant Rhabdoviridae B. Tospoviruses VI. Replication of Double-Stranded RNA Viruses A. Plant Reoviridae VII. Replication of Reverse Transcribing Viruses A. Reverse transcriptase B. Replication of 'caulimoviruses' C. Replication of 'badnaviruses' VIII. Replication of Single-Stranded DNA Viruses A. Methods for studying geminivirus replication B. In vivo observations on geminiviruses C. Rolling-circle replication D. Geminivirus replication E. Nanovirus replication IX. Mutation and Recombination A. Mutation B. Recombination C. Defective and defective interfering nucleic acids and particles X. Mixed Virus Assembly XI. Discussion Induction of Disease 1" Virus Movement through the Plant and Effects on Plant Metabolism I. Introduction II. Movement and Final Distribution A. Routes by which viruses move through plants B. Methods for studying virus movement C. Transport across nuclear membranes D. Cell-to-cell movement E. Time of movement from first infected cells F. Rate of cell-to-cell movement G. Long-distance movement
xi
294 294 294 302 304 305 306 310 310 310 315 316 319 322 324 326 330 333 333 333 335 336 336 339 339 340 344 345 345 345 345 346 351 352 352 353 363 368 371 373 373 373 374 374 376 377 396 396 397
xii
CONTENTS
Chapter 10
H. Rate of systemic movement I. Movement in the xylem J. Final distribution in the plant K. Host factors L. Discussion III. Effects on Plant Metabolism A. Experimental variables B. Nucleic acids and proteins C. Lipids D. Carbohydrates E. Cell wall compounds F. Respiration G. Photosynthesis H. Transpiration I. Activities of specific enzymes J. Hormones K. Low-molecular-weight compounds L. Summary IV. Processes Involved in Symptom Induction A. Sequestration of raw materials B. Effects on growth C. Effects on chloroplasts D. Mosaic symptoms E. The role of membranes V. Discussion Induction of Disease 2: Virus-Plant Interactions I. Introduction II. Definitions and Terminology of Host Responses to Inoculation A. R genes III. Steps in the Induction of Disease A. Ability of virus to replicate in initial cell B. Ability of virus to move out of first cell C. Hypersensitive local response D. HR induced by TMV in N-gene tobacco E. Other viral-host hypersensitive responses F. Host protein changes in the hypersensitive response G. Other biochemical changes during the hypersensitive response H. Systemic necrosis I. Programed cell death and plant viruses J. Local acquired resistance K. Systemic acquired resistance L. Wound healing responses M. Antiviral factors N. Ability of virus to spread through various barriers O. Systemic host response P. Development of mosaic disease Q. Symptom severity R. Recovery IV. Inherent Host Response A. Gene silencing
401 403 403 408 410 411 411 413 415 415 417 418 418 423 423 424 424 426 426 426 428 431 432 434 435 437 437 437 438 439 440 442 442 443 445 448 449 450 450 450 451 454 455 455 455 460 461 462 463 463
CONTENTS
Chapter 11
B. Transcriptional and post-transcriptional gene silencing C. Genes involved in post-transcriptional gene silencing D. Mechanism of post-transcriptional gene silencing E. PTGS systemic signaling F. Induction and maintenance G. PTGS in virus-infected plants H. Suppression of gene silencing I. Other mechanisms of avoiding PTGS J. Discussion V. Influence of Other Agents A. Viroids and satellite RNAs B. Defective interfering nucleic acids C. Other associated nucleic acids D. Cross-protection E. ~oncurrent protection F. Interactions between unrelated viruses G. Interactions between viruses and fungi VI. Discussion and Summary Transmission 1: By Invertebrates, Nematodes and Fungi I. Introduction II. Transmission by Invertebrates A. Arthropoda B. Nematoda C. Relationships between plant viruses and invertebrates III. Aphids (Aphididae) A. Aphid life cycle and feeding habits B. The vector groups of aphids C. Aphid transmission by cell injury D. Types of aphid-virus relationship E. Non-persistent transmission F. Semi-persistent transmission G. Bimodal transmission H. Persistent transmission IV. Leafhoppers and Planthoppers (Auchenorrhyncha) A. Structure and life cycle B. Kinds of virus-vector relationship C. Semi-persistent transmission D. Persistent transmission V. Whiteflies (Aleyrodidae) A. Whiteflies B. Begomoviruses C. Closteroviruses and criniviruses VI. Thrips (Thysanoptera) A. Thrip anatomy B. Tospovirus transmission C. Virus-vector relationship D. Route through the thrips VIII Other Sucking and Piercing Vector Groups A. Mealybugs (Coccoideaand Pseudococcoidea) B. Bugs (Miridae and Piesmatidae)
xiii
464 466 467 469 469 470 471 474 475 475 475 475 476 477 478 478 48O 481 485 485 485 485 486 486 487 487 491 491 491 493 499 500 501 506 506 507 508 508 513 513 514 514 515 515 516 516 517 518 518 518
xiv
CONTENTS
Chapter 12
Chapter 13
VIII. Insects with Biting Mouthparts A. Vector groups and feeding habits B. Viruses transmitted by beetles C. Beetle-virus relationships IX. Mites (Arachnida) A. Eriophyidae B. Tetranychidae X. Pollinating Insects XI. Nematodes (Nematoda) A. Criteria for demonstrating nematode transmission B. Nematode feeding C. Virus-nematode relationships D. Virus-vector molecular interactions XII. Fungi A. In vitro fungal transmission B. In vivo fungal transmission XIII. Discussion and Summary Transmission 2: Mechanical, Seed, Pollen and Epidemiology I. Mechanical Transmission A. Source and preparation of inoculum B. Applying the inoculum II. Factors Influencing the Course of Infection and Disease A. The plant being inoculated B. Development of disease C. Viral nucleic acid as inoculum D. Nature and number of infectible sites E. Number of particles required to give an infection F. Mechanical transmission in the field G. Abiotic transmission in soil H. Summary and discussion III. Direct Passage in Living Higher Plant Material A. Through the seed B. By vegetative propagation C. By grafting D. By dodder E. Summary and discussion IV. Ecology and Epidemiology A. Biological factors B. Physical factors C. Survival through the seasonal cycle D. Disease forecasting E. Conclusions New Understanding of the Functions of Plant Viruses I. Introduction II. Early Events III. Mid-stage Events A. Host and virus translation B. Host and virus replication C. Spatial factors in virus expression and replication D. Plant viruses and cytoskeletal elements
518 518 519 519 520 520 522 522 522 523 523 524 525 526 526 527 527 533 533 533 534 535 536 538 541 542 544 545 546 546 546 546 554 5.54 5'55
555 555 556 572 576 577 578 583 583 584 585 585 585 586 588
CONTENTS
Chapter 14
Chapter 15
IV. Late Events V. Systemic Interactions with Plants VI. Discussion Viroids, Satellite Viruses and Satellite RNAs I. Viroids A. Classification of viroids B. Pathology of viroids C. Structure of viroids D. Replication of viroids E. Molecular basis for biological activity F. Diagnostic procedures for viroids II. Satellite Viruses and Satellite RNAs A. Satellite plant viruses B. Satellite RNAs (satRNAs) C. Satellite DNAs D. Complex-dependent viruses E. Discussion Methods for Assay, Detection and Diagnosis I. Introduction II. Methods Involving Biological Activities of the Virus A. Infectivity assays B. Indicator hosts for diagnosis C. Host range in diagnosis D. Symptom-related methods E. Methods of transmission in diagnosis F. Cytological effects for diagnosis G. Mixed infections H. Preservation of virus inoculum III. Methods Depending on Physical Properties of the Virus Particle A. Stability and physicochemical properties B. Ultracentrifugation C. Electron microscopy D. Chemical assays for purified viruses E. Assay using radioisotopes IV. Methods Depending on Properties of Viral Proteins A. Serological procedures B. Methods for detecting antibody-virus combination C. Collection, preparation and storage of samples D. Monoclonal antibodies E. Phage-displayed single-chain antibodies F. Serologically specific electron microscopy G. Fluorescent antibody H. Neutralization of infectivity I. Electrophoretic procedures V. Methods Involving Properties of the Viral Nucleic Acid A. Type and size of nucleic acid B. Cleavage patterns of DNA C. Hybridization procedures D. Polymerase chain reaction VI. Discussion and Summary
XV
590 590 591 593 593 593 593 596 598 606 607 608 609 614 625 626 626 627 627 628 628 632 633 634 634 634 635 635 636 636 637 638 640 640 641 641 647 655 656 657 657 659 660 660 661 661 662 663 671 673
XVi
CONTENTS
Chapter 16
Chapter 17
Control and Uses of Plant Viruses I. Introduction II. Removal or Avoidance of Sources of Infection A. Removal of sources of infection in or near the crop B. Virus-free seed C. Virus-free vegetative stocks D. Propagation and maintenance of virus-free stocks E. Modified planting and harvesting procedures III. Control or Avoidance of Vectors A. Air-borne vectors B. Soil-borne vectors IV. Protecting the Plant from Systemic Disease A. Mild strain protection (cross-protection) B. Satellite-mediated protection C. Antiviral chemicals V. Conventional Resistance to Plant Viruses A. Kinds of host response B. Genetics of resistance to viruses C. Tolerance D. Use of conventional resistance for control VI. Transgenic Protection Against Plant Viruses A. Introduction B. Natural resistance genes VII. Pathogen-Derived Resistance A. Protein-based protection B. Nucleic acid-based protection C. Other forms of transgenic protection D. Field releases of transgenic plants E. Potential risks associated with field release of virus transgenic plants VIII. Discussion and Conclusions IX. Possible Uses of Viruses for Gene Technology A. Viruses as gene vectors B. Viruses as sources of control elements for transgenic plants C. Viruses for presenting heterologous peptides D. Viruses in functional genomics of plants E. Summary and discussion Variation, Evolution and Origins of Plant Viruses I. Strains of Viruses A. Quasi-species B. Virus strains II. Criteria for the Recognition of Strains A. Structural criteria B. Serological criteria C. Biological criteria D. Discussion III. Isolation of Strains A. Strains occurring naturally in particular hosts B. Isolation from systemically infected plants C. Selection by particular hosts or conditions of growth D. Isolation by means of vectors
675 675 676 676 679 680 685 686 690 691 696 698 699 700 701 702 702 704 707 707 712 712 712 713 714 718 724 727 728 730 731 731 735 736 739 740 743 743 743 744 744 744 750 757 761 762 762 762 762 763
CONTENTS
IV.
V.
VI.
VII.
VIII. IX. X.
XI.
XII.
XIII.
XIV.
E. Isolation of artificially induced mutants F. Isolation of strains by molecular cloning The Molecular Basis of Variation A. Mutation (nucleotide changes) B. Recombination C. Deletions and additions D. Nucleotide sequence re-arrangement E. Re-assortment of multi-particle genomes F. The origin of strains in nature Constraints on Variation A. Muller's ratchet B. Does Muller's ratchet operate with plant viruses? Virus Strains in the Plant A. Cross-protection B. Selective survival in specific hosts C. Loss of infectivity for one host following passage through another D. Double infections in vivo E. Selective multiplication under different environmental conditions Correlations Between Criteria for Characterizing Viruses and Virus Strains A. Criteria for identity B. Strains and viruses C. Correlations for various criteria Discussion and S u m m a r y Speculations on Origins and Evolution Types of Evolution A. Microevolution and macroevolution B. Sequence divergence or convergence C. Modular evolution D. Evidence for virus evolution Sources of Viral Genes A. Replicases B. Proteinases C. Coat proteins D. Cell-to-cell m o v e m e n t proteins E. Suppressors of gene silencing Origins of Viruses, Viroids and Satellites A. Origins of viruses B. Origin of viroids C. Origin of satellite viruses and nucleic acids Selection Pressures for Evolution A. Maximizing the variation B. Controlling the variation C. Adaptation to niches D. Rates of evolution Co-evolution of Viruses with their Hosts and Vectors A. Co-evolution of viruses, host plants and invertebrate vectors B. Evolution of angiosperms and insects C. Horizontal transmission through plants of viruses infecting only insects
xvii
763 764 764 764 765 766 766 766 767 767 767 768 768 768 768 769 770 770 770 770 771 771 774 775 776 776 776 777 777 791 792 793 793 793 794 794 794 796 798 799 799 799 802 802 804 804 804 804
xviii
CONTENTS D. Affinities of viruses that replicate in both insects and plants E. Adaptation of plant viruses to their present invertebrate vectors XV. Discussion and Summary Appendix 1A Appendix 1B Appendix 2A Appendix 2B Appendix 3 References Index Plate section appears between pages 74 and 75.
805 806 807 813 838 850 852 854 857 983
Preface
Plant Virology is synonymous with the name of R.E.F. Matthews, who wrote the first three editions of this standard text. It was a great loss, not only to the plant virology community but also to the scientific community as a whole, that Dick Matthews died in 1995. Obituaries to Dick Matthews published at that time by Bellamy (Virology 1995; 209, 287; Virology 1995; 211, 598; Arch. Virol. 1995; 140, 1885) and Harrison
(Biographical Memoires of Fellows of the Royal Society 1999; 45, 297-313) describe his contribution to plant virology. This edition is dedicated to his memory. New editions of Plant Virology have been published at'10-year intervals, and the fourth edition follows this timing. As was noted in the prefaces for previous editions, each has chronicled everincreasing major advances in the subject. The last decade has been no exception--if anything, the rate of progress has increased almost exponentially. This is illustrated in the graph of annual numbers of publications shown in Fig. 1.3. The advances have been due to several technologies, including the ability to clone and manipulate plant viral genomes, be they RNA or DNA, the ability to express viral (and other) sequences integrated (transformed) into the plant genome, and non-destructive techniques for observing the behavior of the virus within the plant cell. Over the last 1:0 years, the classification of plant virus'es has'been rationalized with the general acceptance of taxa such as genera and species. With the increase of taxonomic information, this has led to, and is continuing to raise, difficulties of definition, which the International Committee on the Taxonomy of Viruses will
have to resolve. For instance, it is becoming increasingly apparent that there is considerable nucleotide variation in isolates of certain viruses that cause similar symptoms in a specific (crop) plant. Should these be considered as one or several species and, if the latter, are there common criteria for viruses from different genera? The wealth of data used in virus taxonomy has now allowed 977 species in 70 genera to be recognized, compared with the 590 species (viruses) and 35 genera (groups) of 10 years ago. The genomes of representatives of all but one of the genera have now been fully sequenced. From the sequence data has come a greater understanding of the genes that viruses encode and how these genes are expressed in a controlled manner within the plant. The sequence data have also given a clearer understanding into virus evolution and especially the role played by recombination. The expression of viral sequences integrated in plant genomes by transformation techniques has broadened the understanding of viral gene function and opened up the field on using viral sequences to confer protection against target viruses. This, in turn, has revealed a previously unknown generic resistance system in plants and other organisms, and is opening up the way to capitalize on this system in areas as different as disease resistance and genomics. These rapid developments over the past decade have necessitated a substantial rewriting and reorganization of this edition. However, I have considered it important to retain material from the third edition giving description of phenomena studied over the years as these can, and
PREFACE
do, form the basis of understanding newly recognized mechanisms. I hope to have expressed the d y n a m i s m of the subject and I have tried in various places to point to future directions that may prove to be scientifically profitable. The chapters are now arranged to lead the reader through the subject logically, building on the information in previous chapters. I have started with an introduction to the subject and a description of each group of viruses, the principles of the architecture of their particles and their genome organizations. This lays the ground for the molecular information given in subsequent chapters, such as the mechanisms by which viral genomes are expressed and replicated, and how the genomes interact with host genomes. The descriptions of how viruses move from host to host is followed by a chapter that brings together the various interactions involved in the full functioning of a virus. After a chapter on virus-like agents such as viroids and satellites, virus detection, control and evolution are discussed.
XX
Over 60% of the more than 360 illustrations are new, including several in color, as in the first edition. The reference list has been expanded from about 3000 in the previous edition to about 4500 in this. In the spirit of remembering important contributions to the subject, many of the important 'older' references are retained. In other places, references to reviews are used to limit the overall number. References are also given to some non-virological subjects that are important in understanding the interactions of viruses with their hosts and vectors. I am greatly indebted to a large number of colleagues for their helpful discussion on various topics and for access to pre-publication material. My eternal gratitude goes to my wife, Jennifer, who has tolerated the 'piles of papers' all over the house and who has given me continuous encouragement. Roger Hull May 2001
This Page Intentionally Left Blank
C H A P T E R
1
Introduction
I. H I S T O R I C A L B A C K G R O U N D
TABLE 1.1 Tulipomania: the goods exchanged for one bulb of Viceroy tulip
The scientific investigation of plant diseases now k n o w n to be caused by viruses did not begin until the late nineteenth century. However, there are much earlier written and pictorial records of such diseases. The earliest k n o w n written record describing what was almost certainly a virus disease is a poem in Japanese written by the Empress Koken in 752 AD and translated by T. Inouye as follows:
4 tons of wheat 8 tons of rye 4 fat oxen 8 fat pigs 12 fat sheep 2 hogsheads of wine
In this village It looks as if frosting continuously For, the plant I saw In the field of summer The color of the leaves were yellowing The plant, identified as Eupatorium tindleyanum, has been found to be susceptible to TLCV*, which causes a yellowing disease (Osaki et al.,
~985). In Western Europe in the period from about 1600 to 1660, m a n y paintings and drawings were made of tulips that demonstrate flower symptoms of virus disease. These are recorded in the Herbals of the time (e.g. Parkinson, 1656) and some of the earliest in the still-life paintings of artists such as Ambrosius Bosschaert. During this period, blooms featuring such striped patterns were prized as special varieties leading to the p h e n o m e n o n of 'tulipomania' (see Blunt, 1950; Pavord, 1999). The trade in infected tulip bulbs resulted in hyperinflation with bulbs exchanging hands for large amounts of money or goods (Table 1.1).
* A c r o n y m s of virus n a m e s are s h o w n in A p p e n d i x 1.
4 barrels of beer 2 barrels of butter 1000 lb cheese 1 bed with accessories 1 full dress suit 1 silver goblet
One of the earliest written accounts of an u n w i t t i n g experimental t r a n s m i s s i o n of a virus is that of Lawrence (1714). He described in detail the transmission of a virus disease of jasmine by grafting. This description was incidental to the main purpose of his experiment, which was to prove that sap must flow within plants. The following quotation from Blair (1719) describes the p r o c e d u r e and demonstrates, rather sadly, that even at this protoscientific stage e x p e r i m e n t e r s were already indulging in arguments about priorities of discovery.
The inoculating of a strip'd Bud into a plain stock and the consequence that the Stripe or Variegation shall be seen in a few years after, all over the shrub above and below the graft, is a full demonstration of this Circulation of the Sap. This was first observed by Mr. Wats at Kensington, about 18 years ago: Mr. Fairchild performed it 9 years ago; Mr. B radly says he observ'd it several years since; though Mr. Lawrence would insinuate as if he had first discovered it. (Lawrence, 1714) In the latter part of the nineteenth century, the idea that infectious disease was caused by microbes was well established, and filters were
2
1 INTRODUCTION
available that would not allow the known bacterial pathogens to pass. Mayer (1886) (Fig. 1.1) described a disease of tobacco that he called Mosaikkrankheit. He showed that the disease could be transmitted to healthy plants by inoculation with extracts from diseased plants. Iwanowski (1892) demonstrated that sap from tobacco plants displaying the disease described by Mayer was still infective after it had been passed through a bacteria-proof filter candle. This work did not attract much attention until itwas repeated by Beijerinck (1898) (Fig. 1.1) who described the infectious agent as 'contagium vivum fluidum' (Latin for contagious living fluid) to distinguish it from contagious corpuscular agents (Fig. 1.2). The centenary of Bejerinck's discovery, which was considered to be the birth of virology, was marked by several publications and celebratory meetings (see Zaitlin, 1998; Bos, 1999a, 2000a; Harrison and Wilson, 1999; Scholthof et al., 1999a; van Kammen, 1999). Baur (1904) showed that the infectious variegation of Abutilon could be transmitted by grafting, but not by mechanical inoculation. Beijerinck and Baur used the term 'virus' in describing the causative agents of these diseases, to contrast them with bacteria. The term 'virus' had been used as more or less synonymous with bacteria by earlier workers. As more diseases of this sort were discovered, the
unknown causative agents came to be called 'filterable viruses'. The papers by Mayer, Iwanowski, Beijerinck and Baur have been translated into English by Johnson (1942). Between 1900 and 1935, many plant diseases thought to be due to filterable viruses were described, but considerable confusion arose because adequate methods for distinguishing one virus from another had not yet been developed. The original criterion of a virus was an infectious entity that could pass through a filter with a pore size small enough to hold back all known cellular agents of disease. However, diseases were soon found that had virus-like symptoms not associated with any pathogen visible in the light microscope, but that could not be transmitted by mechanical inoculation. With such diseases, the criterion of filterability could not be applied. Their infectious nature was established by graft transmission and sometimes by insect vectors. Thus, it came about that certain diseases of the yellows and witches' broom type, such as aster yellows, came to be attributed to viruses on quite inadequate grounds. Many such diseases are now known to be caused by mycoplasmas (phytoplasma and spiroplasmas), and a few, such as ratoon stunting of sugarcane, by bacteria. An important practical step forward was the recognition that some viruses could be transmitted from plant to plant by insects. Fukushi
Fig. 1.1 Left: Martinus Willem Beijerinck (1851-1931). Right: Adolf Eduard Mayer (1843-1942). Photographs courtesy of the historical collection, Agricultural University, Wageningen, Netherlands.
I.
HISTORICAL
BACKGROUND
3
Fig. 1.2 Page from Lab Journal of W. M. Beijerinck from 1898 relating to TMV. ( 9 Kluyver Institute). Courtesy of Lesley Robertson, Curator, Kluyver Laboratory Collection, Delft University of Technology. (1969) records the fact that in 1883 a Japanese rice grower transmitted what is now k n o w n to be RDV by the leafhopper Recelia dorsalis. However, this work was not published in any available form and so had little influence. Kunkel (1922) first reported the transmission of a virus by a planthopper; within a decade, many insects were reported to be virus vectors. During most of the period between 1900 and 1935, attention was focused on the description of diseases, both macroscopic symptoms and cytological abnormalities as revealed by light microscopy, and on the host ranges and methods of transmission of the disease agents. Rather ineffective attempts were made to refine filtration methods in order to define the size of viruses more closely. These were almost the only aspects of virus disease that could be studied with the techniques that were available. The influence of various physical and chemical agents on virus infectivity was investigated,
but methods for the assay of infective material were primitive. Holmes (1929) showed that the local lesions produced in some hosts following mechanical inoculation could be used for the rapid quantitative assay of infective virus. This technique enabled properties of viruses to be studied much more readily and paved the way for the isolation and purification of viruses a few years later. Until about 1930, there was serious confusion by most workers regarding the diseases produced by viruses and the viruses themselves. This was not surprising, since virtually nothing was k n o w n about the viruses except that they were very small. Smith (1931) made an important contribution that helped to clarify this situation. Working with virus diseases in potato, he realized the necessity of using plant indicat o r s - p l a n t species other than potato, which w o u l d react differently to different viruses present in potatoes. Using several different and
4
1 INTRODUCTION
novel biological methods to separate the viruses, he was able to show that many potato virus diseases were caused by a combination of two viruses with different properties, which he named X and Y. Virus X was not transmitted by the aphid Myzus persicae, whereas virus Y was. In this way, he obtained virus Y free of virus X. Both viruses could be transmitted by needle inoculation, but Smith found that certain solanaceous plants were resistant to virus Y. For example, by needle inoculation of the mixture to Datura stramonium, he was able to obtain virus X free of virus Y. Furthermore, Smith observed that virus X from different sources fluctuated markedly in the severity of symptoms it produced in various hosts. To quote from Smith (1931):
There are two factors, therefore, which have given rise to the confusion which exists at the present time with regards to potato mosaic diseases. The first is the dual nature, hitherto unsuspected, of so many of the potato virus diseases of the mosaic group, and the second is the fluctuation in virulence exhibited by one constituent, i.e., X, of these diseases. Another discovery that was to become important was Beale's (1928) recognition that plants infected with tobacco mosaic contained a specific antigen. Gratia (1933) showed that plants infected with different viruses contained different specific antigens. Chester (1935, 1936) demonstrated that different strains of TMV and PVX could be distinguished serologically. He also showed that serological methods could be used to obtain a rough estimate of virus concentration. Since Fukushi (1940) first showed that RDV could be passed through the egg of a leafhopper vector for many generations, there has been great interest in the possibility that some viruses may be able to replicate in both plants and insects. It is now well established that plant viruses in the families Rhabdoviridae and Reoviridae and the Tenuivirus, Tospovirus and Marafivirus genera multiply in their insect vectors as well as in their plant hosts.
The high concentration at which certain viruses occur in infected plants and their relative stability turned out to be of crucial importance in the first isolation and chemical characterization of viruses, because methods for extracting and purifying proteins were not highly developed. In 1926, the first enzyme, urease, was isolated, crystallized, and identified as a protein (Sumner, 1926). The isolation of others soon followed. In the early 1930s, workers in various countries began attempting to isolate and purify plant viruses using methods similar to those that had been used for enzymes. Following detailed chemical studies suggesting that the infectious agent of TMV might be a protein, Stanley (1935) announced the isolation of this virus in an apparently crystalline state. At first Stanley (1935, 1936) considered that the virus was a globulin containing no phosphorus. Bawden et al. (1936) described the isolation from TMV-infected plants of a liquid crystalline nucleoprotein containing nucleic acid of the pentose type. They showed that the particles were rod-shaped, thus confirming the earlier suggestion of Takahashi and Rawlins (1932) based on the observation that solutions containing TMV showed anisotropy of flow. Best (1936) noted that a globulin-like protein having virus activity was precipitated from infected leaf extracts when they were acidified, and in 1937 he independently confirmed the nucleoprotein nature of TMV (Best, 1937b). Electron microscopy and X-ray crystallography were the major techniques used in early work to explore virus structure, and the importance of these methods has continued to the present day. Bernal and Fankuchen (1937) applied X-ray analysis to purified preparations of TMV. They obtained accurate estimates of the width of the rods and showed that the needleshaped bodies produced by precipitating the virus with salt were regularly arrayed in only two dimensions and, therefore, were better described as paracrystals than as true crystals. The isolation of other rod-shaped viruses, and spherical viruses that formed true crystals, soon followed. All were shown to consist of protein and pentose nucleic acid.
I. HISTORICAL BACKGROUND
Early electron micrographs (Kausche et al., 1939) confirmed that TMV was rod-shaped and provided approximate dimensions, but they were not particularly revealing because of the lack of contrast between the virus particles and the supporting membrane. The development of shadow-casting with heavy metals (Mtiller, 1942; Williams and Wycoff, 1944) greatly increased the usefulness of the method for determining the overall size and shape of virus particles. However, the coating of metal more or less obscured structural detail. With the development of high-resolution microscopes and of negative staining in the 1950s, electron microscopy became an important tool for studying virus substructure. From a comparative study of the physicochemical properties of the virus nucleoprotein and the emptyviral protein shell found in TYMV preparations, Markham (1951) concluded that the RNA of the virus must be held inside a shell of protein, a view that has since been amply confirmed for this and other viruses by X-ray crystallography. Crick and Watson (1956) suggested that the protein coats of small viruses are made up of numerous identical subunits arrayed either as helical rods or as a spherical shell with cubic symmetry. Subsequent X-ray crystallographic and chemical work has confirmed this view. Caspar and Klug (1962) formulated a general theory that delimited the possible numbers and arrangements of the protein subunits forming the shells of the smaller isodiametric viruses. Our recent knowledge of the larger viruses with more complex symmetries and structures has come from electron microscopy using negativestaining and ultrathin-sectioning methods. Until about 1948, most attention was focused on the protein part of the viruses. Quantitatively, the protein made up the larger part of virus preparations. Enzymes that carried out important functions in cells were known to be proteins, and knowledge of pentose nucleic acids was rudimentary. No function was known for them in cells, and they generally were thought to be small molecules. This was because it was not recognized that RNA is very susceptible to hydrolysis by acid, by alkali, and by enzymes that commonly contaminate virus preparations. Markham and Smith (1949) isolated TYMV and
5
showed that purified preparations contained two classes of particle, one an infectious nucleoprotein with about 35% of RNA, and the other an apparently identical protein particle that contained no RNA and that was not infectious. This result clearly indicated that the RNA of the virus was important for biological activity. Analytical studies (e.g. Markham and Smith, 1951) showed that the RNAs of different viruses have characteristically different base compositions while those of related viruses are similar. About this time, it came to be realized that viral RNAs might be considerably larger than had been thought. The experiments of Hershey and Chase (1952), which showed that when Escherichia coIi was infected by a bacterial virus, the viral DNA entered the host cell while most of the protein remained outside, emphasized the importance of the nucleic acids in viral replication. Harris and Knight (1952) showed that 7% of the threonine could be removed enzymatically from TMV without altering the biological activity of the virus, and that inoculation with such dethreonized virus gave rise to normal virus with a full complement of threonine. A synthetic analog of the normal base guanine, 8azaguanine, when supplied to infected plants was incorporated into the RNA of TMV and TYMV, replacing some of the guanine. The fact that virus preparations containing the analog were less infectious than normal virus (Matthews, 1953c) gave further experimental support to the idea that viral RNAs were important for infectivity. However, it was the classic experiments of Gierer and Schramm (1956), Fraenkel-Conrat and Williams (1955) and Fraenkel-Conrat (1956) that demonstrated the infectivity of naked TMV RNA and the protective role of the protein coat. These discoveries ushered in the era of modern plant virology. The remainder of this section summarizes the major developments of the past 45 years. Subsequent chapters in this book deal with many of the developments in more detail. The first amino acid sequence of a protein (insulin) was established in 1953. Not long after this event, the full sequence of 158 amino acids in the coat protein of TMV became known (Anderer et al., 1960; Tsugita et al., 1960; Wittmann and Wittmann-Liebold, 1966). The
6
1 INTRODUCTION
sequence of many naturally occurring strains and artificially induced mutants was also determined at about the same time. This work made an important contribution to establishing the universal nature of the genetic code and to our understanding of the chemical basis of mutation. Brakke (1951, 1953) developed density gradient centrifugation as a method for purifying viruses. This has been an influential technical development in virology and molecular biology. Together with a better understanding of the chemical factors affecting the stability of viruses in extracts, this procedure has allowed the isolation and characterization of many viruses. The use of sucrose density gradient fractionation enabled Lister (1966, 1968) to discover the bipartite nature of the TRV genome. Since that time, density gradient and polyacrylamide gel fractionation techniques have allowed many viruses with multipartite genomes to be characterized. Their discovery, in turn, opened up the possibility of carrying out genetic reassortment experiments with plant viruses (Lister, 1968; van Vloten-Doting et al., 1968). Density gradient fractionation of purified preparations of some other viruses has revealed non-infectious nucleoprotein particles containing subgenomic RNAs. Other viruses have been found to have associated with them satellite viruses or satellite RNAs that depend on the 'helper' virus for some function required during replication. With all of these various possibilities, it is in fact rather uncommon to find a purified virus preparation that contains only one class of particle. The 1960s can be regarded as the decade in which electron microscopy was a dominant technique in advancing our knowledge about virus structure and replication. Improvements in methods for preparing thin sections for electron microscopy allowed completed virus particles to be visualized directly within cells. The development and location of virus-induced structures within infected cells could also be studied. It became apparent that many of the different groups and families of viruses induce characteristic structures, or viroplasms, in which the replication of virus components and
the assembly of virus particles take place. Improved techniques for extracting structural information from electron microscope images of negatively stained virus particles revealed some unexpected and interesting variations on the original icosahedral theme for the structure of 'spherical' viruses. There were further developments in the 1970s. Improved techniques related to X-ray crystallographic analysis and a growing knowledge of the amino acid sequences of the coat proteins allowed the three-dimensional structure of the protein shells of several plant viruses to be determined in molecular detail. For some decades, the study of plant virus replication had lagged far behind that of bacterial and vertebrate viruses. This was mainly because there was no plant system in which all the cells could be infected simultaneously to provide the basis for synchronous 'one-step growth' experiments. However, following the initial experiments of Cocking (1966), Takebe and colleagues developed protoplast systems for the study of plant virus replication (reviewed by Takebe, 1977). Although these systems had significant limitations, they greatly increased our understanding of the processes involved in plant virus replication. Another important technical development has been the use of in vitro protein-synthesizing systems such as that from wheatgerm, in which many plant viral RNAs act as efficient messengers. Their use allowed the mapping of plant viral genomes by biochemical means to begin. During the 1980s, major advances were made on improved methods of diagnosis for virus diseases, centering on serological procedures and on methods based on nucleic acid hybridization. Since the work of Clark and Adams (1977), the ELISA technique has been developed with many variants for the sensitive assay and detection of plant viruses. Monoclonal antibodies against TMV were described by Dietzen and Sander (1982) and Briand et al. (1982). Since that time, there has been a very rapid growth in the use of monoclonal antibodies for many kinds of plant virus research and for diagnostic purposes. The late 1970s and the 1980s also saw the
I.
start of application of the powerful portfolio of molecular biological techniques to developing other approaches to virus diagnosis, to a great increase in our understanding of the organization and strategy of viral genomes, and to the development of techniques that promise novel methods for the control of some viral diseases. The use of nucleic acid hybridization procedures for sensitive assays of large numbers of samples has been greatly facilitated by two techniques: (1) the ability to prepare doublestranded cDNA from a viral genomic RNA and to replicate this in a plasmid grown in its bacterial host, with the batches of cDNA labeled radioactively or with non-radioactive reporter molecules to provide a sensitive probe; and (2) the dot blot procedure, in which a small sample of a crude plant extract containing virus is hybridized with labeled probe as a spot on a sheet of nitrocellulose or other material. The polymerase chain reaction, also dependent on detailed knowledge of genome sequences, is being increasingly used in virus diagnosis. Our understanding on the genome organization and functioning of viruses has come from the development of procedures whereby the complete nucleotide sequence of viruses with RNA genomes can be determined. Of special importance has been the ability to prepare in vitro infectious transcripts of RNA viruses derived from cloned viral cDNA (Ahlquist et al., 1984b). This has allowed techniques such as sitedirected mutagenesis to be applied to the study of genome function. Nucleotide sequence information has had, and continues to have, a profound effect on our understanding of many aspects of plant virology, including the following: (1) the location, number, and size of the genes in a viral genome; (2) the amino acid sequence of the known or putative gene products; (3) the molecular mechanisms whereby the gene products are transcribed; (4) the putative functions of a gene product, which can frequently be inferred from amino acid sequence similarities to products of known function coded for by other viruses; (5) the control and recognition sequences in the genome that modulate expression of viral genes and genome replication; (6) the understanding of the structure and replication of viroids and of the satellite RNAs found
HISTORICAL
BACKGROUND
7
associated with some viruses; (7) the molecular basis for variability and evolution in viruses, including the recognition that recombination is a widespread phenomenon among RNA viruses and that viruses can acquire host nucleotide sequences as well as genes from other viruses; and (8) the beginning of a taxonomy for viruses that is based on evolutionary relationships. In the early 1980s, it seemed possible that some plant viruses, when suitably modified by the techniques of gene manipulation, might make useful vectors for the introduction of foreign genes into plants. Although this has been achieved for several genes in model experiments, the concept has only been demonstrated to any practical significance in a few cases. However, some plant viruses have been found to contain regulatory sequences that can be very useful in other gene vector systems. In the early decades of this century, attempts to control virus diseases in the field were often ineffective. They were mainly limited to attempts at general crop hygiene, roguing of obviously infected plants, and searches for genetically resistant lines. Developments since this period have improved the possibilities for control of some virus diseases. The discovery of two kinds of soil-borne virus vectors (nematodes, Hewitt et al., 1958; fungi, Grogan et al., 1958) opened the way to possible control of a series of important diseases. Increasing success has been achieved with a range of crop plants in finding effective resistance or tolerance to viruses. Heat treatments and meristem tip culture methods have been applied to an increasing range of vegetatively propagated plants to provide a nucleus of virus-free material that then can be multiplied under conditions that minimize reinfection. Such developments frequently have involved the introduction of certification schemes. Systemic insecticides, sometimes applied in pelleted form at the time of planting, provide significant protection against some viruses transmitted in a persistent manner by aphid vectors. Diseases transmitted in a non-persistent manner in the foregut or on the stylets of aphids have proved more difficult to control. It has become increasingly apparent that effective control of virus disease in a particular crop in a given area usually requires an
8
1 INTRODUCTION
integrated and continuing program involving more than one kind of control measure. However, such integrated programes are not yet in widespread use. Cross-protection (or mild-strain protection) is a phenomenon in which infection of a plant with a mild strain of a virus prevents or delays infection with a severe strain. The phenomenon has been used with varying success for the control of certain virus diseases, but the method has various difficulties and dangers. PowellAbel et al. (1986) considered that some of these problems might be overcome if plants could be given protection by expression of a single viral gene. Using recombinant DNA technology, they showed that transgenic tobacco plants expressing the TMV coat protein gene either escaped infection following inoculation or showed a substantial delay in the development of systemic disease. These transgenic plants expressed TMV coat protein mRNA as a nuclear event. Seedlings from self-fertilized transformed plants that expressed the coat protein showed delayed symptom development when inoculated with TMV. Thus, a new approach to the control of virus diseases emerged. Since these experiments, the phenomenon has been shown to be widespread and two basic types of protection have been recognizedmone based on the expression of the gene product and the other being RNA-based. Both of these are leading to economically useful protection against specific viruses in several crops but are raising various non-scientific and ethical questions about the acceptability of this approach. The late 1980s and 1990s was a period where molecular biological techniques were applied to a wide range of aspects of plant virology. As well as those areas described above, reverse genetics is being used to elucidate the functions of viral genes and control sequences. This approach, together with others such as yeast systems for identifying interacting molecules and to transform plants to express viral genes, and coupled with the ability to label viral genomes in such a manner that their sites of function within the cell are known, is revealing the complexities of the interactions between viruses and their hosts. The advances in plant
genome sequencing are identifying plant genes that confer resistance to viruses. A major advance in the late 1990s arising from the work on transformation of plants with viral sequences was the recognition that plants have a generic defense system against 'foreign' nucleic acids. Coupled with this is the identification of viral genes that suppress this defense system. In recent years, considerable progress has been made in the development of a stable and internationally accepted system for the classification and nomenclature of viruses. Nine hundred and seventy seven plant viruses have been placed in 14 families and 70 genera. The 14 virus families and most (but not all) of the genera are very distinctive entities. They possess clusters of physical and biological properties that often make it quite easy to allocate a newly isolated virus to a particular family or genus. The rapidly expanding information on nucleotide sequences of viruses infecting plants, invertebrates, vertebrates and microorganisms is emphasizing, even more strongly than in the past, the essential unity of virology. The time is therefore ripe for virologists to consider more grouping into higher taxa. These advances in our understanding of plant viruses, how they function and how this knowledge can be applied to their control has resulted in a burgeoning of papers on the subject. This is illustrated in Fig. 1.3 which shows the numbers of papers that have virus + mosaic and virus + mottle in their titles, abstracts and key words. Obviously this survey does not include all the plant virus literature, as papers devoted to viruses such as those with 'streak' or 'stripe' in their names would not be included. In spite of our greatly increased understanding of the structure, function and replication of viral genomes, there is still a major deficiency. We have little molecular understanding of how an infecting virus induces disease in the host plant. The processes almost certainly involve highly specific interactions between viral macromolecules (proteins and nucleic acids) and host macromolecular structures. At present, we appear to lack the appropriate techniques to make further progress. Perhaps an understanding of disease processes will be the
II. DEFINITIONOF A VIRUS exciting area of virology in the first decade of the twenty-first century. More details of the historical development of plant virology are given by Zaitlin and Palukaitis (2000); and a collection on seminal papers on TMV, which have led many of the conceptual advances, has been published by Scholthof et al. (1999a). Hull et al. (1989) provide a useful directory and dictionary of viruses and terms relating to virology.
II. D E F I N I T I O N OF A V I R U S In defining a virus, we have to consider those of all organisms. The size of viral nucleic acids ranges from a monocistronic mRNA in the satellite virus of tobacco necrosis virus (STNV) to a genome larger than that of the smallest cells (Fig. 1.4). Before attempting to define what viruses are, we must consider briefly how they differ from cellular parasites on the one hand and transposable genetic elements on the other. The three simplest kinds of parasitic cells are the Mycoplasmas, the Rickettsiae and the Chlamydiae. Mycoplasmas and related organisms are not visible by light microscopy. Cells are 150300 nm in diameter with a bilayer membrane, but no cell wall. They contain ribosomes and DNA. They replicate by binary fission, and
9
some that infect vertebrates can be grown in vitro. Their growth is inhibited by certain antibiotics. The Rickettsiae, for example the agent of typhus fever, are small, non-motile bacteria, usually about 300 nm in diameter. They have a cell wall, plasma membrane, and cytoplasm with ribosomes and DNA strands. They are obligate parasites and were once thought to be related to viruses, but they are definitely cells because (1) they multiply by binary fission, and (2) they contain enzymes for ATP production. The Chlamydiae, for example the agent causing psittacosis, include the simplest known type of cell. They are obligate parasites and lack an energy-generating system. They have two phases in their life cycle. Outside the host cell they exist as infectious elementary bodies about 300 nm in diameter. These bodies have dense contents, no cell wall, and are specialized for extracellular survival. The elementary body enters the host cell by phagocytosis. Within 8 hours it is converted into a much larger noninfectious reticulate body. This is bounded by a bilayer membrane derived from the host. The reticulate body divides by binary fission within this membrane, giving thousands of progeny within 40-60 hours. The reticulate bodies are converted to elementary bodies, which are released when the host cell lyses.
No. of publications
I000t 800, 600 400 200
1980
1985
1990 Year
1995
2000
Fig. 1.3 Publications on 'mottle + virus' and 'mosaic + virus' from 1980 to 2000.
10
1 INTRODUCTION
A
Viruses
1 I0'
..acterio
1 10 5
i 10 6
| I0 z
Escherichla coil
r animals some plants
Fungi~~.
03
B
Many plants, some a n i m a . ~ ~
1 IO s
Nucleotides or nucleotide pairs
Vsccinla Virus 19 I0
4 I0 m
I I0 ~
DNA of 1"2
Bacteriophage T ; ~
l
t I
Tobacco Mosaic Virus
Bacteriophage M13
Fig. 1.4 Size comparisons of different organisms. (a) Organisms classified according to genome size. The vertical axis gives an approximate indication of relative numbers of species (or viruses) within the size range of each group. Modified from Hinegardner (1976). (b) Size comparisons between a bacterium, several viruses and the viroid. From Diener (1999), with kind permission of the copyright holder, 9 Springer-Verlag.
There are several criteria that do not distinguish all viruses from all cells: 1. Some pox viruses are bigger than the elem e n t a r y bodies of Chlamydiae. 2. The presence of D N A and RNA is not a distinguishing feature. M a n y viruses have double-stranded (ds) D N A like that of cells, and in some the D N A is bigger than in the
Chlamydiae. 3. A rigid cell envelope is absent in viruses a n d mycoplasmas. 4. G r o w t h outside a living host cell does not occur w i t h viruses or w i t h m a n y groups of obligate cellular parasites, for e x a m p l e ,
Chlamydiae.
Adenovirus
9
Polyoms virus
9
Pollovlrus Bacteriophsge 12
DNA of polyoma virus
s RNA of bectodophage t2 9. . . . . . . . . .
r
I
-
5. An e n e r g y - y i e l d i n g s y s t e m is absent in viruses and Chlamydiae. 6. Complete d e p e n d e n c e on the host cell for amino acids, etc., is found with viruses and some bacteria. There are three related criteria that do appear to distinguish all viruses from all cells: 1. There is no continuous m e m b r a n e separating viral parasite and host during intracellular replication. Cellular parasites that replicate inside a host cell a p p e a r always to be separated from host-cell cytoplasm by a c o n t i n u o u s b i l a y e r m e m b r a n e (see Fig. 11.24).
Ill. ABOUT THIS EDITION
2. There is no protein-synthesizing system in viruses. 3. Replication of viruses is by synthesis of a pool of components, followed by assembly of many virus particles from the pool. Even the simplest cells replicate by binary fission. Plasmids are autonomous extrachromosomal genetic elements found in many kinds of bacteria. They consist of closed circular DNA. Some can become integrated into the host chromosome and replicate with it. Some viruses infecting prokaryotes have properties like those of plasmids and, in particular, an ability to integrate into the host cell chromosome. However, viruses differ from plasmids in the following ways: 1. Normal viruses have a particle with a structure designed to protect the genetic material in the extracellular environment and to facilitate entry into a new host cell. 2. Virus genomes are highly organized for specific virus functions of no k n o w n value to the host cell, whereas plasmids consist of genetic material often useful for survival of the cell. 3. Viruses can cause death of cells or disease in the host organism but plasmids do not. We can now define a virus as follows: A virus is a set of one or more nucleic acid template molecules, normally encased in a protective coat or coats of protein or lipoprotein, that is able to organize its own replication only within suitable host cells. It can usually be horizontally transmitted between hosts. Within such cells, virus replication is (1) dependent on the host's protein-synthesizing machinery, (2) organized from pools of the required materials rather than by binary fission, (3) located at sites that are not separated from the host cell contents by a lipoprotein bilayer membrane, and (4) continually giving rise to variants through various kinds of change in the viral nucleic acid. To be identified positively as a virus, an agent must normally be shown to be transmissible and to cause disease in at least one host. However, the Cryptovirus group of plant viruses is an exception. Viruses in this group rarely cause detectable disease and are not transmissible by any mechanism except through the seed or pollen. The structure and replication of viruses have the following features:
11
1. The nucleic acid may be DNA or RNA and single- or double-stranded. If the nucleic acid is single-stranded it may be of positive or negative sense. (Positive sense has the sequence that would be used in an mRNA for translation to give a viral-coded protein.) 2. The mature virus particle may contain polynucleotides other than the genomic nucleic acid. 3. Where the genetic material consists of more than one nucleic acid molecule, each may be housed in a separate particle or all may be located in one particle. 4. The genomes of viruses vary widely in size, encoding between 1 and about 250 proteins. Plant viral genomes are at the small end of this range, encoding between 1 and 12 proteins. The viral-coded proteins may have functions in virus replication, in virus movement from cell to cell, in virus structure, in transmission by invertebrates or fungi, and in suppression of host defense systems. 5. Viruses undergo genetic change. Point mutations occur with high frequency as a result of nucleotide changes brought about by errors in the copying process during genome replication. Other kinds of genetic change may be due to recombination, reassortment of genome pieces, loss of genetic material, or acquisition of nucleotide sequences from unrelated viruses or the host genome. 6. Enzymes specified by the viral genome may be present in the virus particle. Most of these enzymes are concerned with nucleic acid synthesis. 7. Replication of many viruses takes place in distinctive virus-induced regions of the cell, k n o w n as viroplasms. 8. Some viruses share with certain non-viral nucleic acid molecules the property of integration into host-cell genomes and translocation from one integration site to another. 9. A few viruses require the presence of another virus for their replication.
III. A B O U T THIS E D I T I O N This edition follows many of the features of previous editions but has been reorganized to
12
1 INTRODUCTION
take account of the greater understanding of h o w viruses function. The first chapters describe the basic features of viruses, their classification, the symptoms they cause, how they are purified, what they are made of and the structure of their particles. The next two chapters recount h o w viruses express their genetic information and replicate themselves. This is followed by several chapters discussing the interactions between viruses and their hosts and vectors in disease manifestation and transmission, culminating in a chapter on how these functions are integrated and how some of them have been put to other uses. The last set of chapters deal with other virus-like sequences, the detection, control and evolution of plant
viruses. It is hoped that this will form a logical sequence and will reveal the breadth and dynamism of the subject. In such a dynamic subject, there has been an avalanche of publications over the last 10 years since the previous edition (see Fig. 1.3). In many cases, I have referred to review papers on specific topics where the original papers on that topic can be found. I have retained many of the older references from the previous edition as these describe phenomena and results that can assist in the interpretation of the new phenomena that are being unveiled. The older references also put a perspective on the subject which can be s w a m p e d by the new 'in vogue' topics.
C
H
A
P
T
E
R
2
Nomenclature and Classification of Plant Viruses 1. Viruses can exist as different strains, which m a y cause very different s y m p t o m s in the same host plant. 2. Different viruses m a y cause very similar s y m p t o m s on the same host plant. 3. Some diseases m a y be caused by a mixture of two unrelated viruses.
I. N O M E N C L A T U R E A. Historical aspects In all studies of natural objects, h u m a n s have an innate desire to n a m e and to classify. Virologists are no exception. Virus classification, as with all other classifications, arranges objects showing similar properties into groups and, even t h o u g h this m a y be a totally artificial and h u m a n driven activity without any natural base, it does have certain properties:
J. Johnson in 1927 and in subsequent w o r k stressed the need for using some criteria other than disease s y m p t o m s and host plants for identifying viruses. He suggested that a virus should be n a m e d by a d d i n g the w o r d virus and a n u m ber to the c o m m o n n a m e for the host in which it was first found; for example, tobacco virus I for TMV. Johnson and H o g g a n (1935) compiled a descriptive key based on five characters: modes of transmission, natural or differential hosts, longevity in vitro, thermal death point, and distinctive or specific symptoms. About 50 viruses were identified and placed in groups. K. M. Smith (1937) outlined a scheme in which the k n o w n viruses or virus diseases were divided into 51 groups. Viruses were n a m e d and grouped according to the generic n a m e of the host in w h i c h t h e y w e r e first found. Successive m e m b e r s in a group were given a number. For example, TMV was Nicotiana virus 1, and there were 15 viruses in the Nicotiana virus group. Viruses that were quite unrelated in their basic properties were p u t in the same group. Although Smith's list served for a time as a useful catalog of the k n o w n viruses, it could not be regarded as a classification. H o l m e s (1939) p u b l i s h e d a classification based primarily in host reactions and m e t h o d s of transmission. He used a Latin b i n o m i a l trinomial system of naming. For example, TMV
9 It gives a structured a r r a n g e m e n t of the organisms so that the h u m a n m i n d can comp r e h e n d them more easily. 9 It helps w i t h c o m m u n i c a t i o n b e t w e e n virologists. 9 It enables properties of new viruses to be predicted. 9 It could reveal possible e v o l u t i o n a r y relationships. In theory, it is possible to consider the problems of n a m i n g and classifying viruses as separate issues. In practice, however, n a m i n g soon comes to involve classification. Early workers generally gave a virus a n a m e derived from the host plant in which it was found together with the most conspicuous disease s y m p t o m , for example, tobacco mosaic virus. Viruses were at first thought of as stable entities, and each disease condition in a particular host species was considered to be due to a different virus. However, by the early 1930s three i m p o r t a n t facts began to be recognized: Virus acronyms are given in Appendix 1.
13
14
2 N O M E N C L A T U R E AND C L A S S I F I C A T I O N OF PLANT VIRUSES
became Marmor tabaci, Holmes (Marmor meaning marble in Latin). His classification was based on diseases rather than the viruses, and thus 53 of the 89 plant viruses considered by Holmes fell in the genus Marmor, which contained viruses k n o w n even at that time to differ widely in their properties. Between 1940 and 1966, various schemes were proposed either for plant viruses only or for all viruses. None of these schemes was adopted by any significant number of virologists. It became increasingly apparent that a generally acceptable system of nomenclature and classification could be developed only through international co-operation and agreement, with the opinions of a majority of working virologists being taken into account. At the International Congress for Microbiology held in Moscow in 1966, the first meeting of the International Committee for the Nomenclature of Viruses was held, consisting of 43 people representing microbiological societies of many countries. An organization was set up for developing an internationally agreed taxonomy and nomenclature for all viruses. Rules for the nomenclature of viruses were laid down. The subsequent development of the organization, now k n o w n as the International Committee for Taxonomy of Viruses (ICTV), has been s u m m a r i z e d (Matthews, 1983a; 1985a,b,); the ICTV has presented seven reports (Wildy, 1971; Fenner, 1976; Matthews, 1979, 1982; Francki et aI., 1991; Murphy et al., 1995; van Regenmortel et al., 2000) and published intermediate reports in the Archives of Virology. The main features of the agreed nomenclature and taxonomy as they apply to plant viruses are considered in the following sections.
B. Systems for classification Organisms may be classified in two general ways. One is the classic monothetic hierarchical system applied by Linnaeus to plants and animals. This is a logical system in which decisions are made as to the relative importance of different properties, that are then used to place a taxon in a particular phylum, order, family, genus, and so on. Maurin et al. (1984) proposed a classification system of this
sort that embraces viruses infecting all kinds of hosts. While such systems are convenient to set up and use, there is, as yet, no sound basis for them as far as viruses are concerned. The major problem with any such system is that we have no scientific basis for rating the relative importance of all the different characters involved. For example, is the kind of nucleic acid (DNA or RNA) more important than the presence or absence of a b o u n d i n g lipoprotein membrane? Is the particle symmetry of a small RNA virus (helical rod or icosahedral shell) more important than some aspect of genome strategy during virus replication? An alternative to the hierarchical system was proposed by Adanson (1763). He considered that taxa were best derived by considering all available characters. He made a series of separate classifications, each based on a single character, and then examined how many of these characters divided the species in the same way. This gave divisions based on the largest number of correlated characters. The method is laborious and has not been much used until recently because at least about 60 equally weighted independent qualitative characters are needed to give satisfactory division (Sneath, 1962; Harrison et al., 1971) (Table 2.1). Although the availability of computers has renewed interest in this kind of classification, so far they have not really been used for this to any great extent. In practice, some weighting of characters is inevitably involved even if it is limited to decisions as to which characters to leave out of consideration and which to include. Its main advantages at present may be to confirm groupings arrived at in other ways and to suggest possibly unsuspected relationships that can then be checked by further experimental work. The rapid accumulation of complete nucleotide sequence information for m a n y viruses in a range of different families and groups is having a profound influence on virus taxonomy in at least three ways: 1. Most of the virus families delineated by the ICTV, mainly on morphological grounds, can now be seen to represent clusters of viruses with a relatively close evolutionary origin.
I. NOMENCLATURE
TABLE 2.1 Descriptors used in virus taxonomy I. Virion properties A. Morphology properties of virions 1. Size 2. Shape 3. Presence or absence of an envelope or peplomers 4. Capsomeric symmetry and structure
B. Physical properties of virions 1. Molecular mass 2. Buoyant density 3. Sedimentation coefficient 4. pH stability 5. Thermal stability 6. Cation (Mg2+, Mn 2§ Ca 2§ stability 7. Solvent stability 8. Detergent stability 9. Radiation stability C. Properties of the genortze 1. Type of nucleic acid, DNA or RNA 2. Strandedness: single-stranded or double-stranded 3. Linear or circular 4. Sense: positive, negative or ambisense 5. Number of segments 6. Size of genome or genome segments 7. Presence or absence and type of 5' terminal cap 8. Presence or absence of 5' terminal covalently-linked polypeptide 9 Presence or absence of 3' terminal poly(A) tract (or other specific tract) 10. Nucleotide sequence comparisons D. Properties of proteins 1. Number 2. Size 3. Functional activities (especially virion transcriptase, virion reverse transcriptase, virion hemagglutinin, virion neuraminidase, virion fusion protein) 4. Amino acid sequence comparisons E. Lipids 1. Presence or absence 2. Nature F. Carbohydrates 1. Presence or absence 2. Nature II. Genome organization and replication 1. Genome organization 2. Strategy of replication of nucleic acid 3. Characteristics of transcription 4. Characteristics of translation and post-translational processing 5. Sites of accumulation of virion proteins, site of assembly, site of maturation and release 6. Cytopathology, inclusion body formation III. Antigenic properties 1. Serological relationships 2. Mapping epitopes
15
IV. Biological properties 1. Host range, natural and experimental 2. Pathogenicity, association with disease 3. Tissue tropisms, pathology, histopathology 4. Mode of transmission in nature 5. Vector relationships 6. Geographic distribution From Fauquet (1999), with permission. 2. With the d i s c o v e r y of p r e v i o u s l y u n s u s p e c t e d genetic similarities b e t w e e n viruses infecting different host groups, the unity of virology is n o w quite a p p a r e n t . The stalling a p p r o a c h of s o m e p l a n t virologists to the application of families, genera a n d species to plant viruses is no longer tenable. 3. G e n o t y p i c i n f o r m a t i o n is n o w m o r e i m p o r tant for m a n y aspects of virus t a x o n o m y t h a n p h e n o t y p i c characters. H o w e v e r , there are s o m e limitations to the use of g e n o t y p e (sequence) d a t a alone. A n i m p o r t a n t one is that w i t h the p r e s e n t state of k n o w l e d g e it is v e r y difficult, or impossible, to predict p h e n o t y p i c p r o p e r t i e s of a virus b a s e d on sequence data alone. For e x a m p l e , if w e h a v e t w o viral n u c l e o t i d e sequences differing in a nucleotide at a single site, w e could not, in m o s t cases, d e d u c e f r o m this i n f o r m a t i o n alone that one led to m o s a i c disease a n d the other to lethal necrosis in the s a m e host. H o w e v e r , w i t h increasing u n d e r s t a n d i n g of gene function, w e c o u l d n o w d e c i d e f r o m the n u c l e o t i d e s e q u e n c e s alone w h i c h of t w o r h a b d o v i r u s e s replicated in plants a n d insects a n d w h i c h in v e r t e b r a t e s a n d insects. A n y classification of viruses s h o u l d be b a s e d not only on e v o l u t i o n a r y history, as far as this can be d e t e r m i n e d f r o m the g e n o t y p e , b u t s h o u l d also be useful in a practical sense. Most of the p h e n o t y p i c characters u s e d t o d a y in virus classification will r e m a i n i m p o r t a n t e v e n w h e n the n u c l e o t i d e s e q u e n c e s of m o s t viral genomes have been determined.
C. Families, genera, species and groups At a m e e t i n g in Mexico City in 1970, the ICTV (then called the ICNV) a p p r o v e d the first taxa for viruses ( M a t t h e w s , 1983a). There w e r e t w o families a n d type g e n e r a for these, plus 22 genera not placed in families for viruses infecting
16
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF P L A N T VIRUSES
vertebrates, invertebrates, or bacteria. Sixteen taxa designated as groups were approved for viruses infecting plants. While other virologists subsequently moved quite rapidly to develop a system of families and genera for the viruses, plant virologists clung to the notion of groups. Some plant virologists had especial difficulty with the species concept, believing that it should not be applied to viruses. Their main reason is that because viruses reproduce asexually the criterion of reproductive isolation cannot be used as a basis for defining virus species (e.g. Milne, 1988). The main building block of a biological classification is the species. The pragmatic view of Davis and H e y w o o d (1963) in discussing angiosperm taxonomy was that: There is no universally correct definition (of a species) and progress in understanding the species problem wilt only be reached if we concentrate on the problem of what we shall treat as a species for any particular purpose.
The species concept applied to plants is also discussed by Wagner (1984). In day-to-day practice, virologists use the concept of a 'virus' as being a group of fairly closely related strains, variants or pathovars. A virus defined in this way is essentially a species in the sense suggested by Davis and Heywood, and defined by the ICTV. In 1991, the ICTV accepted the concept that viruses exist as species, adopting the following definition (van Regenmortel, 1990): A viral species is a potythetic class of viruses that constitutes a replicating lineage and occupies a particular ecological niche.
The species has formed the basis of modern virus classification being firmed up in subsequent ICTV reports, especially the seventh in which a 'list of species-demarcating criteria' is provided for each genus. This enables viruses to be differentiated as species and tentative species. Guidelines to the demarcation of a virus species are given in van Regenmortel et at. (1997).
With the species forming the basis of the classification system, they can be grouped into other taxa on various criteria. To date, the taxonomic levels of Order, Family and Genus have been defined by the ICTV (Table 2.2) and it is likely that there will be pressure for further higher and intermediate taxa. For example, Hull (1999b) argued that the development of an overall classification for viruses and elements that replicate by reverse transcription necessitates creating the taxa of Class and Sub-order. A detailed discussion of virus classification, the currently accepted taxa and how the ICTV operates is given in Fauquet (1999) and by van Regenmortel et al. (2000). 1. Delineation of viruses (species) The delineation of the kinds of virus that exist in nature, that is to say virus species, is a practical necessity, especially for diagnostic purposes relating to the control of virus diseases. TABLE 2.2 Criteria demarcating different virus taxa I. Order Common properties between several families including: Biochemical composition Virus replication strategy Particle structure (to some extent) General genome organization II. Family Common properties between several genera including: Biochemical composition Virus replication strategy Nature of particle structure Genome organization HI. Genus Common properties with a genus including: Virus replication strategy Genome size, organization and/or number of segments Sequence homologies (hybridization properties) Vector transmission IV. Species Common properties within a species including: Genome rearrangement Sequence homologies (hybridization properties) Serological relationships Vector transmission Host range Pathogenicity Tissue tropism Geographical distribution
From Fauquet (1999), with permission.
I. NOMENCLATURE
For some 20 years, B. D. Harrison and A. F. Murant acted as editors for the Association of Applied Biologists (AAB) to produce a series of some 354 descriptions of plant viruses and plant virus groups and families. Each description was written by a recognized expert for the virus or group (genus) in question. The two editors used common-sense guidelines devised by themselves to decide whether a virus described in the literature is a new virus, or merely a strain of a virus that has already been described. When they published a new virus description, they were, in effect, delineating a new species of virus. The AAB descriptions of plant viruses are widely used by plant virologists and accepted as a practical and effective taxonomic contribution. The descriptions now cover an increasing proportion of the k n o w n plant viruses and are being updated and extended as a CD-ROM produced by the AAB (see Appendix 3). Viruses are now recognized as 'species' or 'tentative species' which are viruses that have not yet been sufficiently characterized to ensure that they are distinct and not strains of an existing virus or do not have the full characteristics of the genus to which they have been assigned. Of the 977 plant viruses listed in the ICTV seventh report, 701 are true species and 276 tentative species. 2. Virus strains and isolates A c o m m o n problem is to determine whether a new virus is truly a new species or a strain of an existing species. Conversely, what was considered to be a strain may, on further investigation, turn out to be a distinct species. This is because of the population structure of viruses which, because of continuous production of errors in replication, can be considered a collection of quasispecies. The concept of quasispecies is discussed in more detail in Chapter 17 (Section I.A). Various characters are considered in designating a strain. These are described in more detail in relation to virus variation and evolution in Chapter 17. 3. Naming of viruses (species) Questions of virus nomenclature have generated more heat over the years than the much
17
more practically important problems of how to delineate distinct virus species. At an early stage in the development of m o d e r n viral taxonomy, a proposal was made to use cryptograms to add precision to vernacular names of viruses (Gibbs et al., 1966). The proposal was not widely adopted except, for a time, among some plant virologists. When a family or genus is approved by the ICTV, a type species or type virus is designated. Some virologists favor using the English vernacular name as the official species name. Using part of a widely k n o w n vernacular name as the official species name may frequently be a very suitable solution, but it could not always apply (e.g. with newly discovered viruses). Other virologists favor serial numbering for viruses (species). The experience of other groups of microbiologists is that, while n u m bering or lettering systems are easy to set up in the first instance, they lead to chaos as time passes and changes have to be made in taxonomic groupings. The idea of Latinized binomial names for viruses was supported by the ICTV for m a n y years but never implemented for any viruses. The proposal for Latinization has now been withdrawn. In successive editions of the reports of the ICTV (e.g. Matthews, 1982), virus names in the index have been listed by the vernacular n a m e (usually English) followed by the family or genus name; for example, tobacco mosaic Tobamovirus, Fiji disease Fijivirus, and lettuce necrotic yellows Rhabdovirus. This method for naming a plant virus is becoming increasingly used in the literature. 4. Acronyms or abbreviations Abbreviations of virus names have been used for m a n y years to make the literature more easy to read and more succinct to present. The abbreviation is usually in the form of an acronym using the initial letters of each word in the virus name. As the designation of the acronym was by the author of the paper, it was leading to m u c h overlap and confusion. For instance, AMV was used to designate alfalfa mosaic virus and arabis mosaic virus and could also justifiably be used for abutilon mosaic virus, agropyron mosaic virus, alpina mosaic
18
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF P L A N T VIRUSES
virus, Alstromeria mosaic virus, Alternantha mosaic virus, A n e i l e m a mosaic virus or A n t h o x a n t h u m mosaic virus. Therefore, in 1991 the Plant Virus Section of the ICTV initiated a rationalization of plant virus acronyms (Hull et al., 1991) and has subsequently u p d a t e d the list regularly (Fauquet and Martelli, 1995; Fauquet and Mayo, 1999). The designation of the abbreviations is based on the following principles: 9 A b b r e v i a t i o n s s h o u l d be as simple as possible. 9 An abbreviation m u s t not duplicate any other previously coined and still in current usage. 9 The w o r d 'virus' in a name is abbreviated as *VI. 9 The w o r d 'viroid' in a name is abbreviated as 'Vd'. A set of guidelines is laid out in Fauquet and Mayo (1999). A l t h o u g h these, and the acronyms derived from them, are not officially sanctioned by the ICTV, the acronyms are used in the seventh report (van Regenmortel et al., 2000) and it is h o p e d the plant virology comm u n i t y will adopt them. The guidelines are: 1. W h e n similar virus names contain the terms 'mosaic' and 'mottle', 'M' is chosen for 'mosaic' and 'Mo' for 'mottle' except where Guidelines 7 a n d / o r 8 apply. For example, Cowpea mosaic virus and Cowpea mottle virus are a b b r e v i a t e d as CPMV and CPMoV respectively. However, there are still a few abbreviations in which 'Mo' is not used for 'mottle', as there was no need to do so. These will r e m a i n u n c h a n g e d unless change becomes necessary. 2. The w o r d 'ringspot' is abbreviated as 'RS' in many, but not all, instances even if 'R' could have sufficed. 3. The w o r d ' s y m p t o m l e s s ' is abbreviated as 'SU in many, but not all, instances even if 'S' could have sufficed. 4. The second or third letter, or sometimes the second consonant or last letter, of the host plant name, in lower case, can be used to differentiate certain conflicting abbreviations; e.g. CdMV for Cardamom mosaic virus. 5. Abbreviations that use the same letters, but
differ only by the case used (upper or lower), should be avoided. 6. Abbreviations for single w o r d s should not normally exceed two letters. 7. Secondary letters in abbreviations are omitted w h e n their use w o u l d make the abbreviation excessively long, generally more than five letters. For example, CGMMV is preferred to the longer CGMoMV, even though that complies with Guideline 1. 8. Abbreviations in current and widespread use are retained except Where their use could cause confusion. However, abbreviations can and have been changed in c o m m o n usage. The purpose of the advisory list is to suggest targets for harmonization of abbreviations. It is accepted that some abbreviations are unlikely ever to be changed. 9. When a particular combination of letters has been adopted for a particular plant species, new abbreviations for virus names containing the same host will normally use the same combination. 10. When several viruses have the same name and are differentiated by a number, the abbreviations will have a h y p h e n between the letters and numbers, e.g. Plantain virus 6 and Plantain virus 8 are abbreviated as 'P1V-6' and 'P1V-8'. 11. When viruses are distinguished by a letter, this letter is a d d e d at the end of the abbreviation without a hyphen, e.g. Plantain virus X is abbreviated P1VX, in agreement with most of the cases in c o m m o n usage in plant virology (PVY, PVX, PVM, PVS, etc.). 12. When viruses are distinguished by their geographic origin or any other combination of letters, a m i n i m u m n u m b e r of letters (two or three) is a d d e d to the virus abbreviation and a h y p h e n is used between the two sets of letters w h e n the are a d d e d after the word 'virus'; e.g. Tomato yellow leafcurl virus from Thailand is abbreviated 'TYLCV-Th'. (Abbreviations for country names are given in A p p e n d i x 1.) 13. When a virus name comprises a disease n a m e and the w o r d s 'associated virus', these are abbreviated as 'aV'. For example, Grapevine leafroll associated virus 2 is abbreviated to 'GLRVaV-2'.
I. NOMENCLATURE
14. In some instances, where a plant name is abbreviated as two capital (upper case) letters, this usage is retained as an exception; e.g. 'CPRMV' for Cowpea rugose mosaic virus. The current preferred list of abbreviations or acronyms is given in Appendix 1.
D. Plant virus families, genera and orders The current classification of plant viruses is given in Appendix 2 and shown diagrammatically in Fig. 2.1. There is no formal definition for a genus, but it is usually considered as 'a population of virus species that share common characteristics and are different from other populations of species'. The criteria for demarcating a genus are listed in Table 2.2. Currently there are 70 genera of plant viruses recognized. In some cases (e.g. the Rhabdoviridae), there are n u m e r o u s viruses which obviously belong to that family but for which there is insufficient information to place them either in existing genera or for creating new genera; these are listed as 'unassigned'. Genera are n a m e d either after the type species (e.g. Caulimovirus after cauliflower mosaic virus) or are given a descriptive name, often from a Greek or Latin word, for a major feature of the genus (e.g. Closterovirus from the Greek Kkcoa~rlp (kloster) which is spindle or thread, descriptive of the virus particle shape; and Geminivirus from the Latin geminus meaning twins to describe the particles). Similarly, genera are grouped together into families on common characteristics (Table 2.2). There are 14 families recognized for plant viruses (Appendix 2); some, such as Reoviridae and Rhabdoviridae, are in c o m m o n with animal virus families. Twenty two of the genera have not yet been assigned to families and are termed 'floating genera'. The acquisition of further data on these 'floating genera', together with changing attitudes on virus classification, will no doubt lead to the designation of further plant virus families. The family is either n a m e d after the type m e m b e r genus (e.g. Caulimoviridae n a m e d after the genus Caulimovirus), or given a descriptive name, as with the genus, for a major feature of the family
19
(e.g. Geminiviridae, descriptive of the virus particles). Only three orders have been accepted thus far by the ICTV. The Mononegavirales contains, amongst other families, the Rhabdoviridae in which there are two plant virus families. E. U s e of virus names The ICTV sets rules, which are regularly revised, on virus nomenclature and the orthography of taxonomic names (see Pringle, 1998; van Regenmortel ef al., 2000). The last word of a species is 'virus' and suffix (ending) for a genus is ' . . . v i r u s ' , for a subfamily is ' . . . virinae', for a family is ' . . . viridae', and for an order is ' . . . virales'. In formal taxonomic usage the virus order, family, subfamily, genus and species names are printed in italics (or underlined) with the first letter being capitalized; other words in species names are not capitalized unless they are proper nouns or parts of proper nouns. Also, in formal use, the name of the taxon should precede the name being used, e.g. 'the family Caulimoviridae', 'the genus Mastrevirus', and 'the species Potato virus Y'. In informal use, the family, subfamily, genus and species names are written in lower case Roman script, the taxon does not include the formal suffix and the taxonomic unit follows the name being used; e.g. 'the caulimovirus family', 'the mastrevirus genus', and the 'potato virus Y species'. In even less formal circumstances, but widely used, the taxonomic unit is omitted and the taxon for higher taxa can be in the plural; e.g. 'caulimoviruses', 'mastreviruses', 'potato virus Y'. Informal usage arises from practicalities and can lead to the adoption of more formal use. For instance, the genus Badnavirus was not adopted in 1991 but was used widely in the literature and was adopted in the 1995 ICTV report. However, the year 2000 report limited its use to certain D N A viruses with bacilliform particles excluding RTBV. As will be apparent in this book, it is necessary to distinguish the reverse transcribing DNA viruses that have isometric particles from those that have bacilliform particles; the informal usage with be ' c a u l i m o v i r u s e s ' for the former a n d 'badnaviruses' for the latter.
20
2 NOMENCLATURE AND CLASSIFICATION OF PLANT VIRUSES
Fig. 2.1 The families and genera of viruses infecting plants. Outline diagrams are d r a w n approximately to scale. From van Regenmortel et al. (2000), with permission.
II. CRITERIA USED FOR CLASSIFYING VIRUSES
II. CRITERIA USED FOR CLASSIFYING VIRUSES There are two interconnected problems involved in attempting to classify viruses. First, related viruses must be placed in genera and higher taxa. For this purpose, the more stable properties of the v i r u s m s u c h as amount and kind of nucleic acid, particle morphology, and genome s t r a t e g y m a r e most useful. The second problem is to be able to distinguish between related viruses and give some assessment of degrees of relationship within a group. Properties of the virus for which there are many variants are more useful for this purpose. These include symptoms, host range, and amino acid composition of the coat protein. Certain properties, such as serological specificity and amino acid sequences, may be useful both for defining groups and for distinguishing viruses within groups. Two general problems must be borne in mind w h e n considering the criteria to be used for virus classification. One is the problem of w e i g h t i n g m t h e relative importance of one character as c o m p a r e d to another. Some weighting of characters is inevitable, whatever system is used to place viruses in taxa. The second problem concerns the extent to which a difference in one character depends on a difference in another. For example, a single-base change in the nucleic acid may result in an amino acid replacement in the coat protein, which in turn alters serological specificity and electrophoretic mobility of the virus. The same base change in another position in a codon may not affect the gene product at all. The preceding discussion relates to phenotypic characters. As the complete nucleotide sequences of more and more viruses are being determined, virus genotype is becoming increasingly important. What weight is to be given to sequence data in virus classification? A geneticist may say that the evolutionary history of a virus and therefore its relationships with other viruses are completely defined, as far as we will be able to k n o w it, by its nucleotide sequence. This is certainly true in a theoretical sense, but there are also practical aspects to be considered as discussed in Section I.B.
21
The overall properties used in classifying a virus are listed in Table 2.1. Properties that are useful for characterizing strains of a virus are discussed in Chapter 17 (Section II).
A. Structure of the virus particle The importance of virus structure in the classification of viruses is summarized in Fig. 2.1. With isometric viruses, particle morphology, as revealed by electron microscopy, has not proved as generally useful as with the rod-shaped viruses. This is mainly because m a n y isometric particles lie in the same size range (about 25-30 n m in diameter) and are of similar appearance unless preparations and photographs of high quality are obtained (Hatta and Francki, 1984). Where detailed knowledge of s y m m e t r y and arrangement of subunits has been obtained by X-ray analysis or high-resolution electron micro- scopy, these properties give an important basis for grouping isometric viruses. For large viruses with a complex morphology, the structure of the particle as revealed by electron microscopy gives valuable information indicating possible relationships.
B. Physicochemical properties of virus particles Physicochemical properties and stability of the virus particle have sometimes played a part in classification. These properties are discussed in Chapter 15 (Section III). Non-infectious empty viral protein shells and minor noninfectious nucleoproteins are characteristic of certain groups (e.g. tymoviruses). The existence of these components reflects the stability of the protein shell in the absence of the full-length viral RNA.
C. Properties of viral nucleic acids The organization and strategy of viral genomes, as discussed in Chapter 6, is n o w of prime importance for the placing of viruses into families, and genera, or for the establishment of a new family or genus. The luteo-viruses provide a good example (Mayo and D'Arcy, 1999). They were first formally recognized as a group comprising BWYV, SbDV and the RPV, RMV
22
z NOMENCLATURE AND CLASSIFICATION OF PLANT VIRUSES
and SGV strains of BYDV in the second ICTV report (Fenner, 1976). Over the next few years groupings of these viruses increased in a variable manner, depending upon the criteria used. By the sixth report (Murphy et al., 1995) the previous 'luteovirus group' was named as the genus Luteovirus and the species divided into two subgroups typified by BYDV-PAV and PLRV. By this time, many of the luteoviral genomes had been sequenced and it was becoming very apparent that there needed to be a rethink on the classification of this group (D'Arcy and Mayo, 1997). Not only were there differences in genome organizations but there were different polymerase enzymes in different species, some having a 'carmovirus-type' polymerase and others a 'sobemovirus-type' polymerase (see Chapter 8, Section IV, for polymerases). Based mainly on these sequence data the group was divided into three genera (see Section III.P) that were contained in the family Luteoviridae (Mayo and D'Arcy, 1999; van Regenmortel et al., 2000). Another example is the delimitation of the genera within the family Caulimoviridae. Previously, four of the genera (Cautimovirus, 'Soybean chlorotic virus-like', 'Cassava vein mosaic virus-like' and 'Petunia vein clearing virus-like') were classed in the genus Caulimovirus, and two (Badnavirus and 'Rice tungro bacilliform virus-like') in the genus Badnavirus (Murphy et al., 1995). However, because of differences in genome organization these have now been separated into different genera (see Appendix 2) in the family Caulimoviridae in recognition of some basic similarities in genome organization and their replication involving reverse transcription (van Regenmortel et al., 2000).
D. Viral proteins The properties of viral proteins, and in particular the amino acid sequences, are of great importance in virus classification at all levels: for delineating strains, as discussed in Chapter 17 (Section II.A.2); for viruses, as discussed in this section; and for indicating evolutionary relationships between families and groups of viruses.
DISTINCT MEMBERS
//, " l
14~ >-
12-
I
Z 10,,0 w
6,
u.
4,
10
20
30
40
~
50 60 70 % HOMOLOGY
80
90
100
Fig. 2.2 Demarcation between the extent of amino acid sequence homologies in coat proteins amongst distinct i n d i v i d u a l p o t y v i r u s e s (left-hand distribution) and between strains of the same virus (right-hand peak). The 136 possible pairings between 17 strains of eight distinct viruses were analyzed. The homologies between distinct viruses had a mean value of 54.1% and a standard deviation of 7.29%, while the homologies between strains of individual viruses showed a mean of 94.5% and standard deviation of 2.56%. The dashed curves show that all values for distinct viruses and strains fall within +__3 standard deviations from their respective mean values. From Shukla and Ward (1988), with permission.
Coat protein amino acid sequence homologies have been used to distinguish between distinct potyviruses and strains of these viruses (Fig. 2.2) and to estimate degrees of relationship within the group (Fig. 2.3). Dendrograms indicating relationships within a virus group have also been based on amino acid composition of the coat proteins, for example, with the tobamoviruses (Gibbs, 1986). A dendrogram based on peptide patterns obtained from in vitro translated 126-kDa proteins of eight tobamoviruses was very similar to that obtained with coat proteins (Fraile and GarciaArenal, 1990). Once a set of viruses such as the tobamoviruses has been delineated based on coat protein amino acid composition, new isolates may sometimes be readily placed as a strain of an existing virus (e.g. Creaser et al., 1987). Using a statistical procedure known as principal component analysis, Fauquet et al. (1986) showed that groupings of 134 plant viruses and strains obtained using the amino acid composition of their coat proteins (Fig. 2.4) correlated well with the groups established by the ICTV (Matthews, 1982). However, phylogenetic relationships based on one protein may not be similar to that on
II.
BaYMV WSMV TVMV SbMV-V TEV-NAT TEV-HAT SCMV-SC MDMV-B MDMV-A SorgMV JGMV MDMV-KS1 ZYMV-NAT ZYMV-C ZYMV-F WMV-2FC WMV-2 SbMV-N PStV BCMV-NVRS BCMV-3 BCMV-5 PWV-TB PWV-S PWV-K PWV-M PWV-K PRSV-P PRSV-W PepMoV PVY-N PVY-IO PVY-D PVY-43 PVY-18 ~ e __L_
l VIRUSES
-
-
PVY-I LMV PSbMV PPV-R PPV-AT PPV-NAT PPV-D BYMV-G BYMV-S BYMV-CS ClYVV
STRAINS
Fig. 2.3 Relationship between some potyviruses and their strains. The dendrogram is based on the amino acid sequence of the core of the viral coat proteins, omitting the more variable terminal sequences. The core regions were aligned, and the percentages of different amino acids in each sequence (that is, their distance) were calculated. Gaps were counted as a 21st amino acid. The distances were then converted into the dendrogram using the neighbor-joining method of Saitou and Nei (1987). The divisions in the horizontal scale represent 10% differences. The broken vertical line indicates a possible boundary between virus species and strains. From Matthews (1991), with permission.
another protein. This is suggestive of recombination b e t w e e n viruses (see Chapter 8, Section IX.B).
C R I T E R I A USED FOR C L A S S I F Y I N G VIRUSES
23
In recent years, it has p r o v e d easier to obtain the sequence of nucleotides in genomic nucleic acids than that of the amino acids in the gene products. Thus, it is possible to identify amino acid sequence homologies b e t w e e n potential gene products, as has been done for some geminiviruses ( M u l l i n e a u x et al., 1985) a n d potyviruses (Domier et al., 1987). Such studies have s h o w n that there m a y be amino acid sequence similarities b e t w e e n different taxonomic groups of viruses in their non-structural proteins. Coat proteins, on the other hand, have an amino acid composition that is characteristic of a virus group (Fig. 2.4), and there are usually no significant a m i n o acid s e q u e n c e similarities b e t w e e n groups. Thus, in spite of the fact that coat proteins represent only a small fraction of the information in most viral genomes, they appear to be the most useful gene products for delineating m a n y distinct viruses and virus groups. However, functional equivalence of viral coat proteins m a y not always be associated with a m i n o acid sequence similarities. Thus, the coat proteins of AMV and TSV are required to activate the genomic RNAs to initiate infection (see Chapter 8, Section IV.G). The coat proteins of these two viruses are able to activate each other's genome, recognizing the same sequence of nucleotides in the RNAs, but there is no o b v i o u s a m i n o acid s e q u e n c e s i m i l a r i t y b e t w e e n t h e m (Cornelissen and Bol, 1984). An i m p o r t a n t goal for using properties of viral proteins in classification is to k n o w the three-dimensional structure of the protein of at least one m e m b e r of a group. This then allows a m i n o acid substitutions in different viruses in a group to be correlated w i t h biological function. This w a s a c h i e v e d for TMV a n d six viruses related to it (Altschuh et aI., 1987). The a m i n o acid sequence homologies of these seven t o b a m o v i r u s e s r a n g e d from about 28% to 82%. Twenty-five residues are conserved in all seven sequences (Fig. 2.5). Twenty of these conserved residues are concentrated in two locations in the molecule: at low radius in the TMV rod near the RNA binding site (residues 36-41, 88-94 a n d 113-120) and at high radius forming a h y d r o p h o b i c core (residues 61-63 and 144-145). Where viruses w i t h i n this set of seven differ in sequence, the
24
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF PLANT VIRUSES
146 222
HOR
POTY
),~
>3 CLOSTERO ILAR
NEPO
DIANTHO
TOMBUS
9
TYMO
,AMV
Fig. 2.4 Three-dimensional diagram illustrating factors 2, 3 and 4 of a principal-components analysis of 122 data sets of plant virus coat proteins compared by their amino acid composition. The three axes contain 62% of the information. The positions of the viruses on axis 2 are indicated by the sizes of the circles. The numbers within the circles are codes for individual viruses. From Fauquet et al. (1986), with kind permission of the copyright holder, 9 Karger, Basel. differences are often complementary. For example, a m o n g buried residues, a change to a large side-chain in one position may be compensated by a second change to a smaller one in a n e i g h b o r i n g a m i n o acid. This s t u d y with tobamoviruses, while clearly delineating functionally critical regions of the molecule, did not lead to any clear evidence of particular evolutionary relationships within the group. Other viral gene products have been used in taxonomy. As described in C h a p t e r 8 (Section IV.B), there are major groups of RNAd e p e n d e n t RNA polymerases. However, a re-
evaluation of these suggested that this criterion was insufficient to support evolutionary groupings of RNA viruses (Zanotti et al., 1996).
E. Serological relationships Serological methods for determining relationships and their limitations are discussed in Chapter 17 (Section II.B). In the past, these methods have been the most important single criterion for placing viruses in related groups. Members of some groups m a y be all serologically related (e.g. tymoviruses), whereas, in
II,
C R I T E R I A USED FOR C L A S S I F Y I N G V I R U S E S
61
94 I I I 1 I
144
88 38..-, [,
128
%
108
36
25
others, only a few may show any serological relationship (e.g. nepoviruses). In the groups of rod-shaped viruses defined primarily on particle dimensions, m a n y viruses within groups are serologically related, and little or no serological relationship has been established b e t w e e n groups. It seems probable that serological tests will remain an important criterion upon which virus groups are based. Serological tests may also be used to estimate degrees of relationship between viruses in a group (e.g. Fig. 2.6). For both tombusviruses and tymoviruses there is no correlation between SDI values as illustrated in Fig. 2.6 and estimates of genome homology. For tobamoviruses, however, there are clear correlations b e t w e e n SDI values, amino acid sequences, and estimates of genome homology (Koenig and Gibbs, 1986). Serological methods can be used to designate a set of virus strains as constituting a new virus within an established group; for example, PMMoV virus in the genus T o b a m o v i r u s (Pares, 1988). When it comes to defining degrees of relationship within a group, borderline situations will sometimes be found. In that circumstance, additional criteria will be needed. Using a broadly cross-reactive antiserum against the viral coat protein core, Shukla et al. (1989d) showed that potyviruses transmitted by mites or whiteflies were serologically related to a definitive potyvirus with aphid vectors. Distinct potyviruses have often been difficult to
25
Fig. 2.5 Conserved amino acid residues in the coat proteins of seven tobamoviruses. The (xcarbon chain tracing of one subunit is shown viewed down the disk axis. The positions are marked for hydrophobic (m) and hydrophilic (O) residues, which are invariant in all seven viruses From Altschuh et al. (1987), with permission.
define because serological relationships have been found to be complex and often inconsistent. Shukla et al. (1989c) applied a systematic immunochemical analysis to some members of this group. This method involved the use of overlapping peptide fragments to define the parts of the protein combining with antibodies in particular antisera and to make it possible to develop both virus-specific and group-specific antisera. F. A c t i v i t i e s in t h e p l a n t The use of biological properties such as host range and s y m p t o m s in particular host plants as major criteria led to considerable confusion in the identification and classification of viruses. As knowledge of physical and chemical properties of viruses increased, the biological properties were rated as m u c h less important. However, biological properties must still be given some importance in classifying viruses. Macroscopic s y m p t o m differences will often reveal the existence of a different strain of a virus where no other criterion, except a full nucleotide sequence comparison, will do so. The detailed study of the cytology of infected cells made possible by electron microscopy has shown that m a n y groups of viruses cause characteristic inclusion bodies or other cytological abnormalities in the cells they infect (see
26
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF P L A N T VIRUSES
5.0 -7.5
3.0
AMCV
_ _
|
,
_
TBSV-BS
J~.
.~,bRSV Fig. 2.6 The use of serology to estimate degrees of relationship between viruses in a group. The diagram illustrates a classification of 10 tombusviruses with distances representing the mean serological differentiation index of reciprocal tests (RT-SDI). RT-SDI values have been rounded to the nearest 0.25 and, when in order to represent the relationships in two dimensions, the 'observed' and 'diagrammatic' (in parentheses) RT-SDI values differ, the two values are shown. CIRV, TVN and CybRSV are only distantly related to one another and to the other tombusviruses, which form a central cluster. In a multidimensional system, these three viruses would have to be arranged in planes above and below that of the central cluster. The arrows indicate the average distance of these viruses from the central cluster as a whole, but not from individual viruses. TVN has been renamed NRV. From Koenig and Gibbs (1986), with permission.
C h a p t e r 3, Section IV). U l t r a s t r u c t u r a l c h a n g e s in cells u s u a l l y a p p e a r to be m u c h m o r e stable c h a r a c t e r i s t i c s of virus g r o u p s t h a n m a c r o scopic s y m p t o m s . Certain tombusviruses can be d i s t i n g u i s h e d f r o m o t h e r m e m b e r s of the g r o u p by their c y t o p a t h o l o g i c a l effects, e s p e c i a l l y in Chenopodium quinoa (Russo et al., 1987). C r o s s - p r o t e c t i o n in the p l a n t has b e e n u s e d as a u s e f u l i n d i c a t o r of r e l a t i o n s h i p b e t w e e n v i r u s e s (see C h a p t e r 17, Section II.C.4).
G. Methods of transmission As d i s c u s s e d in C h a p t e r 11, m o s t v i r u s e s h a v e only o n e t y p e of vector, a n d u s u a l l y all the v i r u s e s w i t h i n a g r o u p h a v e the s a m e t y p e of vector. O v e r the last f e w years, this c h a r a c t e r has b e e n c o n s i d e r e d i m p o r t a n t in d e f i n i n g n e w g e n e r a w i t h i n , for e x a m p l e , t h e o r i g i n a l Potyvirus a n d Geminivirus g r o u p s . In general, the t y p e of v e c t o r a p p e a r s to be a stable character t h a t is u s e f u l in d e l i n e a t i n g m a j o r g r o u p s of
III.
viruses. However, within a group of viruses or virus strains some may be transmitted efficiently by a vector species, some inefficiently, and some not at all. Details of the way in which a virus is transmitted by a vector (e.g. non-persistent or persistent aphid-transmitted viruses, or on the surface or within the spores of a fungus) may provide further criteria for the grouping of viruses. However, under certain conditions a virus culture may lose the ability to be transmitted by a vector; for example, WTV (see Chapter 11, Section IV.D.2).
III. FAMILIES A N D GENERA OF P L A N T VIRUSES Below are given'thumb-nail' sketches of salient features of members of plant virus families and genera. More details of various features are given in subsequent chapters, in reviews and books referred to here and in the seventh ICTV report (van Regenmortel et al., 2000). Properties of these virus taxa are s u m m a r i z e d in Appendix 2.
A. Family Caulimoviridae (reviewed by Hohn, 1999; Schoelz and Bourque, 1999) This family contains all the plant viruses that replicate by reverse transcription. The viruses all have circular double-stranded DNA genomes with gaps or discontinuities at specific sites, one in one strand and between one and three in the other strand; these discontinuities represent priming sites for DNA synthesis during replication. Transcription is asymmetric to give a more-than-genome length RNA which is both the template for reverse transcription and the mRNA for at least some of the gene products. Other mRNAs may also be transcribed. See Chapter 8 (Section VII) for details of replication and Chapter 7 (Section V.H.1) for details of genome expression. There are six genera within this family, falling into two groups: the caulimoviruses that have isometric particles and the badnaviruses that have bacilliform particles. The genera are distinguished on their genome organization.
FAMILIES A N D G E N E R A OF P L A N T V I R U S E S
27
1. Genus Caulimovirus (type species: Cauliflower mosaic virus) Caulimoviruses have isometric particles of about 50 nm diameter with a T = 7 multilayered structure found within cytoplasmic inclusion bodies. The genome contains six open reading frames (ORFs), the 5' ones being expressed from the more-than-genome length RNA and the 3' ORF 6 from a separate mRNA. There are two other possible ORFs but no products have yet been found from them. Most species are transmitted by aphids in the semipersistent manner, transmission requiring a virus-coded helper protein. See Chapter 5 (Section VI.B.7) for details of particle structure, Chapter 6 (Section IV.A.1) for details of genome organization, Chapter 7 (Section V.H.1) for details of genome expression, and Chapter 11 (Section III.F) for details of transmission. 2. Genus 'Soybean chlorotic mottle virus-like' (type species: Soybean chlorotic mottle virus) (reviewed by Reddy and Richins, 1999) This resembles the genus Caulimovirus but with differences in genome organization (see Chapter 6, Section IV.A.2). 3. Genus 'Cassava vein mosaic virus-like' (type species: Cassava vein mosaic virus) (reviewed by de Kocho, 1999) This resembles the genus Caulimovirus but with differences in genome organization (see Chapter 6, Section IV.A.3). 4. Genus 'Petunia vein clearing virus-like' (type species: Petunia vein clearing virus) This resembles the genus Caulimovirus but with differences in genome organization (see Chapter 6, Section IV.A.4). Recent sequencing indicates that there is only one large ORF comprising ORFs 1 and 2 illustrated in Fig. 6.1 (K.R. Richert-POggeler and T. Hohn, personal communication). 5. Genus Badnavirus (type species: Commelina yellow mottle virus) (reviewed by Lockhart and Olszewski, 1999) The genus contains several viruses of importance to tropical agriculture. Badnaviruses have bacilliform particles usually about 130 • 30 n m
28
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF P L A N T VIRUSES
though there are some instances of longer particles (up to 900 nm long) not associated with cytoplasmic inclusion bodies. The DNA genome encodes three ORFs thought to be translated from the more-than-genome length RNA transcript. ORF III expresses a polyprotein processed to give, amongst other products, the virus coat protein, an aspartate protease, reverse transcriptase and RNaseH. Many members are transmitted in a semi-persistent manner by mealybugs. See Chapter 6 (Section IV.A.5) for details of genome organization.
manner (see Chapter 11, Section IV.D.1). They have genomes comprising a single component of ssDNA (see Chapter 6, Section V.A.2).
6. Genus 'Rice tungro bacilliform virus-like' (type species: Rice tungro bacilliform virus) (reviewed by Hull, 1996, 1999a) Together with RTSV, this genus causes rice tungro disease; it depends on RTSV for its leafhopper transmission. Particles are very similar to those of the genus Badnavirus, but the genome has four ORFs, the 3' one being translated from an mRNA spliced from the more-than-genome length RNA. See Chapter 5 (Section VI.B.5) for details of virion structure, Chapter 6 (Section IV.A.6) for details of genome organization, and Chapter 7 (Section V.H.1) for details of genome expression.
3. Genus Begomovirus (type species: Bean golden mosaic virus) This is the largest geminivirus family, members of which have narrow host ranges among dicotyledonous species. They are transmitted by whiteflies (see Chapter 11, Section V.B). Most members of this genus have their genomes divided between two DNA molecules (see Chapter 6, Section V.A.4), though the genomes of some members are monopartite.
B. Family Geminiviridae (reviewed by Buck, 1999a) This is one of the two families of plant viruses that have circular ssDNA genomes. The main distinguishing feature is that these genomes are contained in geminate virus particles consisting of two incomplete icosahedra (see Chapter 5, Section VI.B.3). The family comprises four genera that are distinguished from one another by their genome organization and their vectors. 1. Genus Mastrevirus (type species: Maize streak virus) (reviewed by Palmer and Rybicki, 1998) This genus contains the economically important MSV. All members have narrow host ranges and, with the exception of TYDV and BeYDV which infect dicotyledenous species, their host ranges are limited to species in the Poaceae. Mastreviruses are transmitted by leafhoppers in a circulative, non-propagative
2. Genus Curtovirus (type species: Beet curly top virus) The type member of this genus used to be an economically important virus of sugarbeet and has a wide host range. Curtoviruses are transmitted by leafhoppers in the circulative nonpropagative manner (see Chapter 11, Section IV.D.1). They have a monopartite ssDNA genome (see Chapter 6, Section V.A.3).
4. Genus Topocuvirus (type species: Tomato pseudocurly top virus) This is a very recently designated genus (Pringle, 1999a) that has been split off from the Curtovirus genus. It has a similar genome organization to the curtoviruses but it is transmitted by the treehopper, Micrutalis malleifera.
C. Family Circoviridae This family contains two genera, one of viruses that infect animals and the other of viruses that infect plants. They have small virus particles, 17-22 nm in diameter, with icosahedral symmetry that contain small circular ssDNA genomes. 1. Genus Nanovirus (type species: Subterranean clover stunt virus) This genus contains the economically important virus BBTV. The viruses have a genome of at least six circular ssDNA molecules of approximately I kb in size encapsidated in an icosahedral virion about 2 0 n m in diameter. See Chapter 6 (Section V.B.1) for genome organiza-
III.
tion and Chapter 8 (Section VIII.E) for the replication mechanism.
D. Family Reoviridae The Reoviridae comprises genera of viruses that infect vertebrates, invertebrates and plants. Some members are restricted to vertebrates but others infect both vertebrates and invertebrates. Those that infect plants also infect their invertebrate vectors. The virus particles are complex, made up of one, two or three distinct shells each with icosahedral symmetry. The particles of most of the genera have surface spikes. The genomes comprise 10, 11 or 12 segments of linear dsRNA depending upon the genus. There are three genera that infect plants which are distinguished on their structure and the n u m b e r of segments into which their genome is divided. 1. Genus Fijivirus (type species: Fiji disease virus) Fijiviruses have 10 dsRNA segments (see Chapter 6, Section VI.A.1, for genome organization) encapsidated in a double-shelled particle 65-70 n m in diameter (see Chapter V, Section VII.2, for structure). These viruses cause hypertrophy of the phloem leading to vein swelling and sometimes enations. They are transmitted in delphacid planthoppers in a circulative propagative manner and infect monocotyledonous plants of the Graminae and Liliaceae families. 2. Genus Oryzavirus (type species: Rice ragged stunt virus) Members of this genus have genomes comprising 10 dsRNA segments (see Chapter 6, Section VI.A.2, for genome organization) encapsidated in a double-shelled particle 75-80 nm in diameter (see Chapter 5, Section VII.3). They infect monocotyledonous plants of the family Graminae and are transmitted by delphacid planthoppers in the circulative propagative manner. Oryzaviruses resemble fijiviruses in having 10 dsRNA segments and being transmitted by delphacid planthoppers. The two genera are distinguished by particle size and by nucleic acid and serological similarities within, but not between, the genera.
F A M I L I E S A N D G E N E R A OF P L A N T V I R U S E S
29
3. Genus Phytoreovirus (type species: Wound tumor virus) Phytoreoviruses have 12 dsRNA genome segments (see Chapter 6, Section VI.A.3, for genome organization) encapsidated in a doubleshelled particle about 70 nm in diameter (see Chapter 5, Section VII.l). They infect both dicotyledonous and monocotyledonous species and are transmitted by leafhoppers in the circulative propagative manner.
E. Family Partitiviridae (reviewed by Milne and Marzachi, 1999) Members of the Partitiviridae have small isometric non-enveloped virions 30-40 n m in diameter with a dsRNA genome divided into two segments, the smaller encoding the coat protein and the larger the RNA-dependent RNA polymerase. Two of the genera of this family infect fungi and two infect plants. The plant-infecting viruses are known as 'cryptic viruses' as they cause no, or very few, symptoms. They are not graft-transmitted and they have no biological vector, but are highly seed-transmitted. Presumably, this leads to full systemic infection as they move through the plant as cells divide.
1. Genus Alphacryptovirus (type species: White clover cryptic virus 1) The 30-nm diameter isometric particles of members of this genus appear smooth in outline, lacking fine structural detail. The genome segments are 1.7 and 2 kbp in size and the capsid is made up of molecules of a single protein species of 55 kDa.
2. Genus Betacryptovirus (type species: White clover cryptic virus 2) The isometric particles are 38 nm in diameter, show prominent subunits, and contain genome segments of about 2.1 and 2.25 kbp. The protein subunit weight has not yet been determined.
30
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF P L A N T VIRUSES
E No family 1. Genus Varicosavirus (type species: Lettuce big-vein
virus) This genus has rod-shaped virions of 320-360nm length and 18 nm diameter that contain two dsRNA molecules encapsidated in multiple copies of a single protein species of about 48 kDa. They are transmitted by chytrid fungi (Olpidium).
G. Family Rhabdoviridae (reviewed
by
Jackson et al., 1999) Rhabdoviruses have characteristic bulletshaped or bacilliform membrane-enveloped particles. The outer surface has glycoprotein spikes that pass through the membrane. On the inner surface of the membrane is a layer of protein (the matrix protein) that contains the helically arranged nucleocapsid. The nucleocapsid is made up of (-)-strand RNA and nucleoprorein together with small amounts of other virus-coded proteins. Their structure is discussed in Chapter 5 (Section VIII.A). Rhabdoviruses infect vertebrates, invertebrates and plants. There are two genera of plant-infecting rhabdoviruses that resemble the animalinfecting viruses except in having an extra gene (in the few viruses that have been sequenced), which is thought to encode a protein for cellto-cell movement (for genome arrange see Chapter 6, Section VII.A). The plant-infecting rhabdoviruses are transmitted by insects in a circulative, propagative manner. The most common vectors are leafhoppers, planthoppers and aphids, though mite- and lacebugtransmitted viruses (one each) have been identified. 1. Genus Cytorhabdovirus (type species: Lettuce necrotic yellows virus) Members of this genus are thought to replicate in the perinuclear space or in the endoplasmic reticulum and to mature by budding into the cytoplasm. Many of the members have relatively narrow (about 60 nm) diameter particles.
2. Genus Nucleorhabdovirus (type species: Potato yellow dwarf virus) Members of this genus are thought to replicate in the nucleus and to mature by budding through the inner nuclear membrane into the perinuclear space. Many of the members have relatively broad (about 90nm) diameter particles. 3. Unassigned Many plant rhabdoviruses have not been assigned to a genus, as they have not been sufficiently characterized. Some of these appear to lack the outer membrane and essentially be just nucleoprotein 'cores' (Francki et al., 1985a).
H. Family Bunyaviridae The Bunyaviridae is a large family of viruses, most of which infect both vertebrates and invertebrates. There is one genus that infects plants and also its invertebrate vectors. The virions are spherical or pleomorphic with surface glycoprotein spikes embedded in a lipid bilayer envelope. Within this envelope the genome, comprising three RNA species, is associated with a nucleoprotein forming a nucleocapsid. The 5' and 3' terminal nucleotides of each segment are complementary, leading to the formation of 'panhandle' structures. The RNAs are either (-)-sense or have their coding regions in an ambisense arrangement (see Chapter 7, Section V.B.12). 1. Genus Tospovirus (type species: Tomato spotted wilt virus) (reviewed by Moyer, 1999) Tospoviruses have three ssRNA species, two of which have an ambisense strategy (see Chapter 6, Section VII.B.1), the largest being of (-)-sense. These are contained within membrane-bound particles (see Chapter 5, Section VIII.B, for structure). One of the gene products is involved in cell-to-cell spread of the virus (see Chapter 9, Section II.D.I.j) which distinguishes members of this genus from other Bunyaviruses. Tospoviruses are transmitted by thrips in a circulative, propagative manner (see Chapter 11, Section VI.B). TSWV has a host range of more than 925 species belonging to 70 botanical families.
Ili.
I. No family There are two genera that have affinities to the Bunyaviridae but for which there is not sufficient information to formally place them in that family 1. Genus Tenuivirus (type species: Rice stripe virus) (reviewed by Falk and Tsai, 1998; Falk, 1999) The virions of tenuiviruses so far characterized comprise very long filamentous nucleoproteins, 3-10 n m in diameter, which often form a circular outline. Although these nucleoproteins have various features of the genome organization that indicate a relationship to bunyaviruses, particularly the Phlebovirus genus, no m e m b r a n e - b o u n d particles have yet been detected. The genome comprises four or more segments, the largest having negative polarity and three or more species with an ambisense a r r a n g e m e n t (Chapter 6, Section VII.C). Tenuiviruses infect m e m b e r s of the Graminae and are transmitted by planthoppers in a circulative, propagative manner. 2. Genus Ophiovirus (type species: Citrus psorosis virus) Ophioviruses have genomes similar to those of tenuiviruses, being filamentous nucleocapsids about 3 n m in diameter and often taking on a circular form. The genome is divided into three segments forming nucleoproteins with a protein species of 43-50 kDa. No natural vectors have been identified for ophioviruses. Members of this genus infect both dicotyledonous and m o n o c o t y l e d o n o u s plant species.
FAMILIES A N D G E N E R A OF P L A N T VIRUSES
also encapsidate satellite or defective interfering RN As. 1. Genus Bromovirus (type species: Brome mosaic virus) (reviewed by Ahlquist, 1999) The virions of b r o m o v i r u s e s are isometric, about 27 n m in diameter and are stabilized by a p H - d e p e n d e n t protein:protein interaction and also a protein:RNA interaction. Above p H 7 the particles swell and become salt labile. The capsids are m a d e up of 180 copies of a single protein species of 20 kDa. The natural host range is narrow. All species are thought to be transmitted by beetles, and some species are possibly naturally mechanically transmitted (see Chapter 12, Section II.F).
2. Genus Cucumovirus (type species: Cucumber mosaic virus) (reviewed by Roossinck, 1999a) C u c u m o v i r u s e s have isometric particles, about 30 n m in diameter, that are stabilized by protein:RNA interactions and thus are salt labile. The capsid comprises 180 copies of a single protein species of about 24 kDa. CMV has an extensive host range infecting over 1000 species in more than 85 plant families. Other cucumoviruses have a n a r r o w e r host range. All are transmitted by aphids in a nonpersistent manner. CMV is divided into two subgroups (I and II) based on serology and nucleotide sequence identity (Table 2.3). PSV probably also has several subgroups. TABLE 2.3 Nucleotide sequence similarities between CMV subgroups and Cucumovirus species for each RNA. From Roossinck (1999a), with permission. Virus
J. Family Bromoviridae Members of this family have isometric particles with T = 3 icosahedral symmetry, 26-35 n m in diameter or bacilliform particles whose symmetry is based u p o n the icosahedron. The genomes of linear positive-sense ssRNA are divided between three molecules with genome organizations as s h o w n in Chapter 6 (Section VIII.A). The subgenomic RNA for coat protein is often also encapsidated; several m e m b e r s
31
RNA1 CMV-I CMV-II PSV RNA2 CMV-I CMV-II PSV RNA3 CMV-I CMV-II PSV
CMV-I
CMV-II
PSV
TAV
100
78 100
68 70 100
70 69 68
100
74 100
65 66 100
64 66 64
100
78 100
64 65 100
65 66 72
32
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF P L A N T VIRUSES
3. Genus Alfamovirus (type species: Alfalfa mosaic virus) (reviewed by Bol, 1999a) The virions of this monospecific genus are bacilliform, 18 nm in diameter and 30-57 nm long depending on the RNA species, encapsidated in a protein of 24 kDa (see Chapter 5, Section VI.B.2, for details of structure). They are stabilized primarily by protein:RNA interactions and thus are salt labile. The presence of coat protein or the subgenomic RNA encoding it is required for virus replication (see Chapter 8, Section IV.G). Because of sequence similarities, it has been suggested that AMV is included in the genus Ilarvirus (Scott et al., 1998). AMV has a wide host range and is transmitted by aphi~ls in a non-persistent manner. 4. Oenus Ilarvirus (type species: Tobacco streak virus) (reviewed by Bol, 1999a) Ilarvirus particles have quasi-isometric shapes varying from being roughly spherical to being bacilliform. They are stabilized by protein:RNA bonds and thus are salt labile. As with AMV, the coat protein (24-26 kDa) or its subgenomic mRNA is required for replication; AMV coat protein can be substituted for ilarvirus coat protein. Viruses in this genus infect mainly woody plants and are transmitted by pollen and seeds. Ilarviruses are divided into seven or eight subgroups or clusters based on serological relationships (Bol, 1999a). 5. Genus Oleavirus (type species: Olive latent virus 2) Virions of this monospecific genus have different shapes and sizes ranging from quasi-spherical with a diameter of 26 nm to bacilliform of diameter 18 nm and lengths of 55, 48, 43 and 37nm. They encapsidate the three genomic RNAs and a subgenomic RNA that is not the coat protein mRNA. The coat protein is 20 kDa. No natural vector is known for OLV-2. K. Family C o m o v i r i d a e Virions are isometric with T = 1 (pseudo T = 3) icosahedral symmetry (see Chapter 5, Section VI.B.6.a, for structure). The capsids are made up of one or two coat protein species and
encapsidate the genome comprising two positive-sense ssRNA species. The RNAs are expressed as polyproteins (for genome organization, see Chapter 6, Section VIII.B) that are processed to give the functional proteins (see Chapter 7, Section V.E.8).
1. Genus Comovirus (type species: Cowpea mosaic virus) (reviewed by Lomonossoff and Shanks, 1999) The capsids are constructed from two polypeptide species ( M r 40-45 kDa and 21-27 kDa) expressed from the smaller RNA species (see Chapter 6, Section VIII.B.1, for genome organization). Comoviruses are transmitted by Chrysomelid beetles. 2. Genus Fabavirus (type species: Broad bean wilt virus l) (reviewed by Cooper, 1999) As with comoviruses, the capsids of fabaviruses are constructed from two protein species of similar sizes to those of comoviruses and also expressed from the smaller RNA species (see Chapter 6, Section VIII.B.2, for genome organization). Fabaviruses are transmitted by aphids in a non-persistent manner. 3. Genus Nepovirus (type species: Tobacco ringspot virus) (reviewed by Murant et al., 1996; Mayo and Jones 1999a) Capsids of TRSV are composed of a single polypeptide species (52-60 kDa) while those of other nepoviruses may be made up of two or three smaller proteins. The genome organization (see Chapter 6, Section VIII.B.3) is similar to that of comoviruses. Nepoviruses often cause ringspot symptoms. Many are transmitted by longidorid nematodes (see Chapter 11, Section XI.E), some by pollen, one (BRAV) by mites and others have no known vector. The species in this genus are clustered into three subgroups 'a', 'b' and 'c', based on the length and packaging of RNA2 and on serological relationships (Murant, 1981).
llI.
FAMILIES AND GENERA OF PLANT VIRUSES
33
L. Family Potyviridae (reviewed by Shukla
4. Genus Rymovirus (type species: Ryegrass mosaic
et al., 1994; L6pez-Moya and Garcia, 1999)
virus)
Members of the Potyviridae have flexuous particles 650-900 nm long (genus Bymovirus with two particle lengths) and about 11-15 nm in diameter. The genome is positive-sense ssRNA with a VPg at the 5' end and a 3' poly(A) tract; the bymovirus genome is divided between two segments. The genome is expressed as a polyprotein (for genome organization, see Chapter 6, Section VIII.C) that is cleaved to functional proteins (see Chapter 7, Section V.B.I.b). All members of the Potyviridae form cylindrical inclusion bodies in infected cells (see Chapter 3, Section IV.C). They are transmitted by a variety of v e c t o r s ~ o n e of the characters that define the genus.
Members of this genus have particles of 690-720 nm length and 11-15 nm diameter that contain an RNA genome of 9-10 kb with the capsid comprising many copies of a single protein species of about 29 kDa. They are transmitted by eriophyid mites (see Chapter 11, Section IX.A).
1. Genus Potyvirus (type species: Potato virus Y) Virus particles are 6 8 0 - 9 0 0 n m long and 11-13 nm in diameter encapsidating a genome of about 9.7 kb with multiple copies of a single protein species of 30-47 kDa. Potyviruses are transmitted by aphids in the non-persistent manner using a helper component (see Chapter 11, Section Ill.E). Some of the members of this genus are seed transmitted (see Chapter 12, Section Ill.A). This is the largest of the genera of plant viruses (91 species and 88 tentative species) and contains some economically important virus such as PVY, BYMV, PPV and PRSV. 2. Genus Ipomovirus (type species: Sweet potato mild mottle virus) Virus particles are 800-950 nm long containing a genome of 10.8 kb encapsidated in multiple copies of a single protein species of about 38 kDa. The natural vector is the whitefly Bemisia tabaci with which the virus has a non-persistent relationship. 3. Genus Macluravirus (type species: Maclura mosaic
virus) Macluravirus particles are 650-675 nm long and contain an RNA of about 8 kb encapsidated in multiple copies of a single protein species of 33-34 kDa. They are transmitted by aphids in a non-persistent manner.
5. Genus Tritimovirus (type species: Wheat streak mosaic virus) Tritimovirus particles are 690-700 n m long and contain an RNA of about 8.5-9.6 kb in size, the capsid comprising multiple copies of a protein of about 32 kDa. T h e y are restricted to the Graminae and are transmitted by eriophyid mites possibly in a persistent manner (see Chapter 11, Section IX.A). 6. Genus Bymovirus (type species: Barley yellow mosaic virus) The virions are flexuous rods of 13 nm width and two modal lengths, 250-300 and 500-600 nm. The genome is divided between two RNA species, the longer particle containing RNA of 7.5-8.0 kb in size and the shorter particle RNA of 3.5-4.0 kb in size. The capsid is made up of multiple copies of a single protein species of 28-33 kDa. Members of this genus are restricted to the Graminae and are transmitted by the plasmodiophora fungus Polymyxa graminis.
M. Family Tombusviridae Viruses in this family have isometric icosahedral (T = 3) particles, 32-35 n m in diameter, with well-defined structure (see Chapter 5, Section VI.B.4.d). The particles contain a single species of positive-sense ssRNA with a size ranging from 3.7 to 4.7 kb. The genome organization differs between genera in this family (see Chapter 6, Section VIII.D). The unifying feature of the family is that each member has a highly conserved RNA-dependent RNA polymerase that is interrupted by an in-frame termination codon that is periodically suppressed (see Chapter 7, Section V.B.9).
34
2 N O M E N C L A T U R E AND C L A S S I F I C A T I O N OF PLANT VIRUSES
Individual members have relatively narrow host ranges and can infect either mono- or di-cotyledonous species, but not both. These viruses are relatively stable and many are found in natural environments such as surface waters and soils from which in many cases they can be acquired without the need for a biological vector.
of 32-35 nm diameter. The particles have a regular surface structure under the electron microscope, giving a granular appearance but not showing subunit arrangement of its T - - 3 icosahedral symmetry comprising 180 copies of a single protein species of 36-41 kDa.
1. Genus Tombusvirus (type species: Tomato bushy stunt virus) (reviewed by Rochon, 1999) Members of this genus have genomes of about 4.7 kb encapsidated in particles of 32-35 nm, the detailed structure of which has been well characterized (see Chapter 6, Section VIII.D.1). They are composed of 180 subunits of a single protein species of 41 kDa. The transmission of most viruses in this genus is soil-borne with no obvious biological vector. CNV is transmitted by the fungus Olpidium bornovanus.
5. Genus Machlomovirus (type species: Maize chlorotic mottle virus) (reviewed by Lommel, 1999b) This is a monospecific genus. MCMV particles are approximately 30 nm in diameter and comprise a genomic RNA of 4.4 kb and capsid protein subunits of 25 kDa. The genome organization is discussed in Chapter 6 (Section VIII.D.5). The host range of MCMV is restricted to members of the family Graminae. The virus is seed-transmitted and has been reported to be transmitted by chrysomelid beetles and thrips.
2. Genus Aureusvirus (type species: Pothos latent virus) The single species in this genus, PoLV, has a genome of 4.4 kb contained in particles of 30 nm diameter that have a rounded outline and knobby surface made up of 180 copies of a single 40-kDa protein species. PoLV has only been reported from southern Italy. Its natural transmission is through the soil or circulating solution in hydroponics. 3. Genus Avenavirus (type species: Oat chlorotic stunt virus) The single species in this genus, OCSV, has a coat protein significantly larger (48.2kDa) than those of most other members of the family Tombusviridae with a protruding domain. The isometric particles of 35 nm diameter contain a genome of 4114 nt with a genome organization described in Chapter 6 (Section VIII.D.3). OCSV has only been described in oats (Avena sativa). Its natural transmission is soil-borne possibly by zoosporic fungi. 4. Genus Carmovirus (type species: Carnation mottle virus) (reviewed by Qu and Morris, 1999) The 3.88-4.45 kb RNA genome of members of this genus is contained in isometric particles
6. Genus Necrovirus (type species: Tobacco necrosis virus A) (reviewed by Meulewaeter, 1999) Necrovirus virions are about 28 nm in diameter and are made up of 180 subunits of a 30-kDa capsid protein. The genome is linear ssRNA of 3.7 kb, the organization of which is shown in Chapter 6 (Section VIII.D.6). Necroviruses have wide host ranges including both monocotyledonous and dicotyledonous species. In nature the infection is usually restricted to the roots. These viruses are transmitted by the chytrid fungus, Otpidium brassicae (see Chapter 11, Section XII). 7. Genus Panicovirus (type species: Panicum mosaic virus) The 4.3 kb genomic RNA is contained in isometric particles of approximately 30 nm diameter comprising 180 copies of a single protein species of 26 kDa. As with the machlomoviruses, the panicovirus RNA-dependene RNA polymerase has an N-terminal extension fused to the rest of the polymerase that is conserved in the family Tombusviridae (see Chapter 6, Section VIII.D.7, for genome organization). The host range is restricted to the Graminae. Natural transmission is most likely mechanical.
III.
8. Genus Dianthovirus (type species: Carnation ringspot virus) (reviewed by Lommel, 1999a) Dianthoviruses differ from the rest of the Tombusviridae in that their genomes are divided between two species of ssRNA; the genome organization is shown in Chapter 6 (Section VIII.D.8). These two RNAs are contained in virions of 32-35 n m diameter that have icosahedral symmetry being made up of 180 subunits of a 37-38-kDa coat protein. The m o d e r a t e l y broad host range of dianthoviruses is restricted to dicotyledonous plant species. The virus appears to be readily transmitted through the soil but no biological vector is known.
FAMILIES A N D GENERA OF PLANT VIRUSES
O. Family Closteroviridae
35
(reviewed by
German-Retana et al., 1999) Members of the Closteroviridae have very flexuous filamentous particles about 12 n m in diameter, the length being characteristic of the genus. The genomic nucleic acid is linear, positive-sense ssRNA that may be mono- or bipartite, dependent on genus. The host ranges of individual species are usually narrow. These viruses are usually p h l o e m limited and cause yellowing-type symptoms or pitting or grooving of w o o d y stems. Transmission is a characteristic of the genus but they m a y be reclassified into aphid-, whitefly- and mealybug-transmitted genera.
N. Family Sequiviridae Members of this family have isometric particles, 30 nm in diameter, that contain positivesense ssRNA 9-12 kb in size. The virion is made up of three protein species of approximately 32, 26 and 23 kDa; these proteins are in equimolar amounts. The RNA encodes a polyprotein that is cleaved to give the functional proteins. 1. Genus Sequivirus (type species: Parsnip yellow fleck virus) (reviewed by Mayo and Murant, 1999) The RNA is about 10 kb and is not polyadenylated (see Chapter 6, Section VIII.E.1, for genome organization). The viruses are transmitted by aphids in a semi-persistent noncirculative manner; transmission of PYFV depends on a helper waikavirus, AYV. 2. Genus Waikavirus (type species: Rice tungro spherical virus) (reviewed by Gordon, 1999) The RNA is more than 11kb and has a 3' poly(A) tail. One or more small ORFs near the 3' end of the RNA have been reported but there is uncertainty as to whether these are expressed (see Chapter 6, Section VIII.E.2, for genome organization). Waikaviruses are transmitted either by leafhoppers or aphids in a semi-persistent noncirculative manner and transmission is thought to involve one or more virus-encoded helper proteins.
1. Genus Closterovirus (type species: Beet yellows virus) (reviewed by Bar-Joseph et al., 1997; GermanRetana et al., 1999) The flexuous virions are 1250-2000 n m long, with one end coated with an anomalous coat protein giving a 'rattlesnake' structure (see Fig. 5.12). For most species, there is a single characteristic particle length, but CTV also has shorter particles containing subgenomic or defective RNAs. Full-length particles contain a single RNA species of 15.5-19.3 kb in size, the genome organization of which is described in Chapter 6 (Section VIII.F.1). Natural transmission is by aphids, mealybugs or whiteflies (Trialeuroides), dependent on species. Both CTV and BYV cause important diseases.
2. Genus Crinivirus (type species: Lettuce infectious yellows virus) Crinivirus particles have two modal lengths, 700-900nm and 6 5 0 - 8 5 0 n m and, as with Closteroviruses, has an anomalous protein at one end giving a 'rattlesnake' structure; the major coat protein is 28-33 kDa. The genome is in two segments that are separately encapsidated. The natural vectors of criniviruses are whiteflies (Bemisia and Trialeuroides) which transmit in a semi-persistent manner.
36
2 NOMENCLATURE
AND
CLASSIFICATION
OF PLANT
P. Family Luteoviridae (reviewed
by Smith and Barker, 1999; Mayo and Miller, 1999; Miller, 1999) The viruses in this family have isometric icosahedral (T = 3) particles 25-30 n m in diameter that are hexagonal in outline and comprise 180 subunits of a single protein species of 21-23 kDa. The particles contain a single molecule of positive-sense ssRNA 5.7-5.9 kb in size. Genera are distinguished on genome organization (see Chapter 6, Section VIII.G). Many m e m b e r s of the Luteoviridae are phloem-limited, causing 'yellows'-type diseasesmhence the name luteovirus from the Latin 'luteus' meaning yellow. Transmission is by aphids in a specific, circulative, nonpropagative manner (see Chapter 11, Section III.H.l.a). These viruses often assist the transmission of other viruses (see Chapter 11, Section III.H.l.a). The sequencing of genomes of members of this family has led to considerable reorganization of the classification.
VIRUSES
does not spread in plants unless PEMV-2 is present. Unlike other m e m b e r s of the Luteoviridae, pea enation mosaic is transmissible mechanically, as well as by aphids, and spreads to mesophyll tissues; these two properties are conferred by PEMV-2. Virions of some strains of PEMV-1 + PEMV-2 also contain a satellite RNA. Aphid transmission of PEMV-1 + PEMV-2 is conferred by PEMV-1. Aphid transmissibility can be lost after multiple passages of mechanical transmission. Differences have been found in the electrophoretic profiles of the particles of aphid-transmissible and non-transmissible isolates (Hull, 1977b) (Fig. 2.7) but the cause of these differences has not been fully resolved. 4. Unassigned Eleven viruses that have properties suggestive of the family Luteoviridae have not been assigned to genera because of lack of full characterization.
(A)
1. Genus Luteovirus (type species: Barley yellow dwarf virus-PAV) Members of this genus are restricted to Graminae where they cause serious disease. 2. Genus Polerovirus (type species: Potato leafroll virus) This genus has members that infect dicotyledonous species and others that are restricted to monocotyledonous species. They also cause serious diseases.
L.._ _
Y
(B) 131
3. Genus Enamovirus (type species: Pea enation mosaic virus 1 ) Pea enation mosaic disease is caused by a complex of two viruses, PEMV-1 which is a member of the Luteoviridae, and PEMV-2 which is an umbravirus (Demler et al., 1993; 1994a). Two sizes of isometric particles are found, that of PEMV-1 (28 nm diameter) having icosahedral symmetry and that of PEMV-2 (25 n m diameter) having quasi-icosahedral symmetry (see Chapter 5, Section VI.B.4.f). Both types of particle are composed of coat protein encoded by PEMV-1. PEMV-1 can infect protoplasts but
132
33 B5 M
B6
~'
"x.B?
J
Fig. 2.7 Densitometer traces of the nucleoproteins of (a) a PEMV isolate not aphid transmissible, and (b) an aphid-transmissible PEMV isolate, electrophoresed in 3.4% polyacrylamide gels at 16 V/cm for 16 h using the buffer system described by Hull and Lane (1973). From Hull (1977b), with permission.
III.
Q. Floating genera There are currently 20 genera of (+)-strand ssRNA plant viruses that have not been placed in families. I will describe these in an order that brings together viruses with similar properties firstly rod-shaped viruses and then viruses with isometric particles. The first two genera have rigid rod-shaped particles and several similarities in genome organization. Genera 3-9 have flexuous rod-shaped particles with a single RNA genomic species; several of these have been reclassified relatively recently from larger groupings. Genera 10-14 have rod-shaped particles encapsidating divided genomes; several of these also have recently been reclassified. Genera 15-18 have isometric particles, genus 19 has bacilliform particles, and genus 20 does not self-encapsidate. It is likely that some of the arbitrary juxtapositions will be grouped into families as and when more information becomes available. 1. Genus Tobamovirus (type species: Tobacco mosaic virus) (reviewed by Lewandowski and Dawson, 1999; Knapp and Lewandowski, 2001) As described in Chapter 1, TMV was the first virus to be recognized as a pathogenic entity that differed from bacteria, and this virus has been in the forefront of many of the advances in plant virology. The centenary of its recognition was celebrated in 1999 with various meetings and publications (see Scholthof et al., 1999a; a special issue of Proc. Roy. Soc. Lond. B). In fact, this has led to the problem that TMV has the image of the archetypical plant virus. Tobamoviruses have rod-shaped particles 300-310 nm long and 18 nm in diameter; the structure of these particles has been studied in detail (see Chapter 5, Section III.B). The particles contain a single positive-strand ssRNA of 6.3-6.6 kb in size, the genome organization of which is described in Chapter 6 (Section VIII.H.1). The capsid comprises multiple copies of a single polypeptide species of 17-18 kDa. Tobamoviruses are very stable with purified preparations and leaf material retaining infectivity for more than 50 years. They are naturally transmitted mechanically (see Chapter 12, Section II.F).
FAMILIES A N D G E N E R A OF P L A N T VIRUSES
37
2. Genus Tobravirus (type species: Tobacco rattle virus) (reviewed by Visser et al., 1999) Tobraviruses have rod-shaped particles of 20-22 nm diameter and with two predominant lengths, 180-215nm and 46-115nm. The genome is positive-sense ssRNA and is divided between two species, the size of that in the long particles being about 6.8 kb and in the short particles ranging from 1.8 kb to about 4.5 kb (see Chapter 6, Section VIII.H.2, for genome organization). The capsid is made up of many copies of a single protein species of 22-24 kDa. Tobraviruses have wide host ranges including both mono- and dicotyledonous species. The natural vectors are Trichodorid nematodes (see Chapter 11, Section XI.E). PEBV has been reviewed by Boulton (1996). 3. Genus Potexvirus (type species: Potato virus X) (reviewed by AbouHaidar and Gellatly, 1999) Potexviruses have flexuous, filamentous particles 470-580 nm long and 13 nm in diameter. These virions contain a single species of linear, positive-sense ssRNA of about 5.9-7.0 kb encapsidated in multiple copies of a single species of coat protein of 18-27 kDa. The RNA is capped at the 5' end and is 3' polyadenylated. The genome organization is described in Chapter 6 (Section VIII.H.3). Viruses in this genus are transmitted by mechanical contact. 4. Genus Carlavirus (type species: Carnation latent virus) (reviewed by Zavriev, 1999) Members of this genus have slightly flexuous filamentous particles, 610-700nm long and 12-15 nm in diameter. These virions contain a single molecule of linear, positive-sense ssRNA of 7.4-8.5 nm in size that has a 3' poly(A) tract; some species encapsidate subgenomic RNAs in shorter particles. The capsid is composed of a single polypeptide species of 31-36 kDa. The genome organization is described in Chapter 6 (Section VIII.H.4). Most carlavirus species are transmitted by aphids in the non-persistent manner. However, CPMMV is transmitted by the whitefly Bemisia tabaci. CPMMV is also seed-transmitted, as are PeSV and RCVMV. Some species (e.g. CLV and PVS) are naturally mechanically transmitted.
38
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF P L A N T VIRUSES
5. Genus Allexivirus (type species: Shallot virus X) (see Kanyuka et al., 1992) The virions of allexiviruses are highly flexuous and filamentous, about 8 0 0 n m long and 12 nm in diameter. They contain a single species of linear, positive-sense ssRNA about 9.0 kb in size which is encapsidated in a single species of coat protein of 28-36 kDa. The genome organization is described in Chapter 6 (Section VIII.H.5). Allexiviruses have very narrow host ranges and are naturally transmitted by eriophyid mites. 6. Genus Capillovirus (type species: Apple stem grooving virus) (reviewed by Salazar, 1999) Capilloviruses have flexuous, filamentous rodshaped particles 640-700 nm long and 12 nm in diameter containing a linear, positive-sense ssRNA genome of 6.5-7.4 kb in size. The RNA is polyadenylated at the 3' end and is encapsidated in a single protein species of 24-27 kDa. The genome organization is described in Chapter 6 (Section VIII.H.6). No natural vectors are known for viruses in this genus; ASGV is seed-transmitted. 7. Genus Foveavirus (type species: Apple stem pitting virus) (see Jelkmann, 1994; Martelli and Jelkmann, 1998) Members of this genus have flexuous filamentous virions about 800 nm long and 12 nm in diameter. The virions comprise a single molecule of linear, positive-sense ssRNA, 8.4-9.3 kb in size and a single species of coat protein (28-44 kDa). The genomic RNA is 3' polyadenylated and its organization is described in Chapter 6 (Section VIII.H.7). The natural host range of species in this genus is restricted to one or a few plant species. There is no known natural vector for any of the species. 8. Genus Trichovirus (type species: Apple chlorotic leaf spot virus) (reviewed by German-Retana and Candresse, 1999) Members of this genus have very flexuous rodshaped particles ranging from 6 4 0 n m to 760 nm in length and having a diameter of
12 nm. The virions contain a single species of linear positive-strand ssRNA of about 7.5 kb in size with a polyadenylated 3' terminus which is encapsidated in many copies of a single polypeptide species of 22-27 kDa. The genome organization is described in Chapter 6 (Section VIII.H.8). The host ranges of trichoviruses are relatively narrow. No natural vector has yet been identified, although there is evidence for field spread of GINV. PVT is seed-transmitted in several hosts. 9. Genus Vitivirus (type species: Grapevine virus A) (reviewed by German-Retana and Candresse, 1999) Viruses in this genus have flexuous filamentous particles 725-825 nm long and 12 nm in diameter. The particles contain a single species of linear, positive-stranded ssRNA of about 7.6 kb in size, capped at the 5' terminus and 3' polyadenylated. The capsid comprises multiple copies of a single protein species of 18-22 kDa. The genome organization is described in Chapter 6 (Section VIII.H.9). The natural host range of individual virus species is restricted to a single plant species, though the experimental host range may be large. GVA and GVB are transmitted in the semi-persistent manner by mealybugs (genera Pseudococcus and Planococcus) and GVA is also transmitted by the scale insect Neopulvinaria innu~Jlerabilis. HLV is transmitted semipersistently by aphids in association with a helper virus. 10. Genus Furovirus (type species: Soil-borne wheat m~saic virus) (reviewed by Shirako and Wilson, 1999) The virions of furoviruses are rigid rods of about 20 nm diameter and two predominant lengths of 260-300 nm and 140-160 nm. The genome is bipartite, linear positive-sense ssRNA that in the longer particles is about 6-7 kb in size, and in the shorter particles 3.5-3.6 kb; shorter particles of SBWMV may contain populations of deletion mutants. The capsid comprises multiple copies of a single polypeptide of 19-21 kDa. The genome organization is shown in Chapter 6 (Section VIII.H.10).
III.
As the name of the type species and the sigla for the generic name (furo- = fungus-borne, rod-shaped) suggest, viruses in this genus are transmitted by fungi. SBWMV is transmitted by the plasmodiophora fungus Potymyxa graminis. 11. Genus Pecluvirus (type species: Peanut clump virus) (reviewed by Reddy et al., 1999; Shirako and Wilson, 1999) Viruses in this genus have rod-shaped virions of about 21 nm diameter and two predominant lengths of 245 and 190 nm. The particles contain linear positive-sense ssRNA of two sizes, that from the long particles being about 5.9 kb and that from the shore particles about 4.5 kb encapsidated in multiple copies of a single protein species of 23 kDa. The genome organization is described in Chapter 6 (Section VIII.H.11). Pecluroviruses are transmitted by the plasmodiophorid fungus Polymyxa graminis and by seed (in groundnuts). 12. Genus Pomovirus (type species: Potato mop-top virus) (reviewed by Torrance, 1999; Shirako and Wilson, 1999) Pomoviruses have rod-shaped particles 18-20 nm in diameter and of three predominant lengths, 290-310nm, 150-160nm and 65-80 nm. These contain a genome of linear, positive-sense ssRNA divided into three molecules of about 6, 3-3.5 and 2.5-3.0 kb in size encapsidated in multiple copies of a major 20-kDa coat protein. The genome organization is described in Chapter 6 (Section VIII.H.12). Viruses in this genus have narrow host ranges, so far limited to dicotyledonous species. They are transmitted in the soil by fungi; Spongospora subterranea and Polymyxa betae have been identified as vectors of PMTV and BSBV respectively. 13. Genus Benyvirus (type species: Beet necrotic yellow vein virus) (reviewed by Tamada, 1999) Benyviruses have rod-shaped particles of 20 nm diameter and four predominant lengths, about 390, 265, 100 and 85 nm. These contain the genome of linear, positive-sense ssRNA which is divided into four molecules of about
FAMILIES A N D G E N E R A OF PLANT VIRUSES
39
6.7, 4.6, 1.8 and 1.4 kb in size; some sources of BNYVV have a fifth RNA molecule of 1.3 kb in size. The largest two make up the infectious genome, the other RNAs being ancillary and influencing transmission and symptomatology (see Chapter 14, Section II.B.3.c). These RNAs are 5' capped and, unlike the RNAs genomes of other rod-shaped viruses, are 3' polyadenylated. The capsid is made up of a major protein species of 21-23 kDa. The genome organization is described in Chapter 6 (Section VIII.H.13). The type species of this genus, BNYVV, is economically very important in many sugarbeet growing areas. It and the other species in this genus, BSBMV, are transmitted by the fungus Polymyxa betae. 14. Genus Hordeivirus (type species: Barley stripe mosaic virus) (reviewed by Lawrence and Jackson, 1999) Hordeiviruses have rigid rod-shaped particles of about 20 nm diameter and ranging in length from 110 to 150 nm. The genome is divided between three molecules of linear, positivesense ssRNA designated ~, [~ and 7; ~ ranges in size between species from 3.7 to 3.9 kb, [~ from 3.1 to 3.6 kb and 7 from 2.6 to 3.2 kb; between strains of a given species the sizes of {, and [3 are relatively similar but that of }, can vary (see Chapter 6, Section VIII.H.14, for genome organizations). The capsid is composed of multiple copies of a single coat protein species of 17-18 kDa. There are no natural vectors known for any member of this genus with field spread thought to be by leaf contact. BSMV and LRSV are efficiently seed-transmitted, this providing primary foci from which secondary spread occurs. 15. Genus Sobemovirus (reviewed by Sehgal, 1999; Tamm and Truve, 2000) Members of this genus have isometric virions (T = 3) of about 30 nm made up of 180 subunits of a single capsid protein species of about 26-30 kDa. Each particle contains a single molecule of positive sense ssRNA of about 4.1-4.5 kb which has a VPg at the 5' end. The genome organization is described in Chapter 6 (Section VIII.H.15).
40
2 N O M E N C L A T U R E AN[) C L A S S I F I C A T I O N OF P L A N T VIRUSES
Several sobemoviruses are seed-transmitted, most are beetle-transmitted and one is transmitted by myrids (see Chapter 11, Section VII.B). 16. Genus Marafivirus (type species: Maize rayado
fino virus) Purified preparations of marafiviruses sediment as two components, B component that contains the viral genome and T component that contains no RNA. The icosahedral particles have a diameter of 28-32 nm and most contain a major protein of 21-22 kDa and a minor protein of about 24-28 kDa. The 6.5 kb genome in the B component is capped at the 5' end, polyadenylated at the 3' end and has a high content of cytidine. The genome organization is shown in Chapter 6 (Section VIII.H.16). Marafiviruses have narrow host ranges restricted to the family Gramineae. They are transmitted by leafhoppers and possibly replicate in their vectors. 17. Genus Tymovirus (type species: Turnip yellow mosaic virus) (reviewed by Gibbs, 1999a) The virions of tymoviruses are isometric, with diameter about 30 nm, and have icosahedral T = 3 symmetry with 180 subunits of a protein of 20 kDa. Both full (B component) and empty (T component) particles can be distinguished by electron microscopy and sedimentation. The full particles contain a single species of linear, positive-sense ssRNA of about 6.3 kb in size which has a high cytidine content, resembling marafiviruses. The genome organization of this RNA is described in Chapter 6 (Section VIII.H.17). The replication of the viral RNA is associated with vesicles in the periphery of chloroplasts (see Chapter 8, Section IV.K.2). Tymoviruses appear to be restricted to dicotyledonous hosts. They are transmitted by beetles of the families Chrysomelidae and
species of linear, positive-sense ssRNA of about 5.5, 2.2 and I kb in size. The larger two make up the viral genome (see Chapter 6, Section VIII.H.18, for genome organization); the smallest RNA is a subgenomic mRNA for the 30-kDa coat protein. The natural host range of RBDV is restricted to Rubus species. The virus is transmitted both vertically through seed and horizontally through pollen. 19. Genus Ourmiavirus (type species: Ourmia melon virus) (see Accotto et al., 1997) Viruses in this genus have bacilliform particles of 18 nm diameter and 62, 46, 37 and 30 nm length (see Fig. 5.26); the ends of the particles are conical (hemi-icosahedral). The genome is linear, positive-sense ssRNA divided into three segments of approximately 2.9, 1.1 and 1.0 kb (estimated from M r measured by gel electrophoresis). The capsid comprises copies of a single protein species of 25 kDa. No natural vectors have been identified. 20. Genus Umbravirus (type species: Carrot mottle virus) (reviewed by Robinson and Murant,
Curculionidae.
1999) Umbraviruses do not encode conventional coat proteins and their naturally transmitted particles are formed by association with a helper virus, usually from the family Luteoviridae, which contributes its coat protein. The unencapsidated umbravirus RNA is surprisingly stable, being mechanically transmissible, though sensitivity to organic solvents suggests that it might be associated with lipid membranes. Umbravirus genomes are linear positivestranded ssRNA of about 4 k b in size; the genome organization is shown in Chapter 6 (Section VIII.H.20). Members of this genus have very limited host ranges and, as noted above, their natural transmission is by aphids in association with a helper virus.
18. Genus Idaeovirus (type species: Raspberry bushy dwarf virus) (reviewed by Mayo and Jones,
IV. RETROELEMENTS
1999b) This monospecific genus has isometric virions about 33 nm in diameter that contain three
Retroelements have been grouped into viral retroelements, eukaryotic chromosomal non-
IV. R E T R O E L E M E N T S
viral retroelements, and bacterial chromosomal retroelements (see Hull and Covey, 1996). Plant viral retroelements, classified in the family Caulimoviridae, are pararetroviruses which do not include an integration phase in their replication cycle. Plant genomes (and those of organisms from other kingdoms) contain a variety of retroelements which resemble integrating animal retroviruses in genome organization and replication cycle but are not known to cause infections. These elements are known as retrotransposons. Structurally, they have long terminal repeat (LTR) sequences, which are involved in the integration into the host genome. As well as the basic RNA -+ DNA -* RNA replication, these elements have several other features in common. The enzyme complex of active retroelements comprises reverse transcriptase (RT) and ribonuclease H (RNaseH) and an open reading frame encoding a nucleic acid-binding protein termed gag for retroviruses and coat protein for pararetroviruses (see Mason et al., 1987). These elements also encode an aspartate proteinase and an integrase (int) (except pararetroviruses). Some of these retrotransposons form virus-like particles for which the ICTV has recently proposed two new families, Pseudoviridae and Metaviridae (van Regenmortel et al., 2000). A. F a m i l y Pseudoviridae This family comprises retrotransposons that form isometric virus-like particles containing linear positive-strand ssRNA capable of undergoing a retrovirus-like replication cycle. They form an intrinsic, and often significant, part of the genomes of many eukaryotic species, especially plants. In many cases the plant genome contains numerous integrated copies of apparently defective forms of these elements that mutated and were unable to replicate. The virion is thought to play an essential role in the replication cycle but is unlikely to be infectious and capable of horizontal spread. 1. Genus Pseudovirus (type species: Saccharomyces cerevisiae Ty 1 virus) The particles of members of this genus are 40-60 nm in diameter and are heterodisperse
41
structurally. Some particles with icosahedral symmetry have been recognized. The particles contain an RNA of 5.6 kb in size that has two ORFs. The 5' ORF encodes the coat protein (gag) and the 3' ORF (pol) has aspartate protease, integrase and reverse transcriptase activities. Both ORFs are expressed from the genomic RNA, the 3' ORF by frameshift from the C terminus of the 5' ORF (see Chapter 7, Section V.B.10, for frameshift). There are numerous pseudoviruses found in higher plants and some in lower plants such as yeasts. 2. Genus Hemivirus (type species: Drosophila melanogaster copia virus) Members of this genus have isometric particles but little is known of their structure. The 5.1 kb genome has one ORF that encodes the gag and pol functions. The expression is controlled by splicing which produces an mRNA for gag. Although hemiviruses resemble pseudoviruses and members of the Caulimoviridae in using initiator methionine tRNA as a primer for synthesis of (-)-strand DNA from the genomic RNA, they differ from these other viruses/elements in that they use only half the tRNA generated by cleavage in the anticodon loop. There are no species of this genus known yet in higher plants and only one in yeast.
B. Family Metaviridae The particles formed by members of this family are poorly characterized and often heterogeneous; they are frequently referred to as virus-like particles (VLPs). 1. Genus Metavirus (type species: Saccharomyces cerevisiae Ty3 virus) The heterogeneous VLPs of metaviruses are generally spherical, about 50 nm in diameter. They contain a polyadenylated RNA of about 5 kb. The genome of Saccharomyces cerevisiae Ty3 virus resembles that of pseudoviruses in having two ORFs but differ in the order of functional domains (protease, reverse transcriptase, integrase) in the pol ORF.
42
2 N O M E N C L A T U R E A N D C L A S S I F I C A T I O N OF PLANT VIRUSES
This genus contains species from higher plants (Lilium henryi dell virus) and from yeasts and other fungi. 2. Genus Errantivirus (type species: Drosophila melanogaster gypsy virus) Members of this genus have enveloped irregular particles of about 100 nm diameter containing an RNA of about 7.5 kb in size. The genome encodes three ORFs, a 5' ORF encoding the gag gene, a pol ORF with the functional domains (protease, reverse transcriptase, integrase) in a different order from those of Pseudoviridae, and an env (envelope) ORF. The pol ORF is expressed by frameshift from the gag ORF and the env ORF is expressed from a spliced mRNA. There are no errantiviruses known yet in higher or lower plants.
V. VIRUSES OF LOWER PLANTS A. Viruses of algae Two groups of viruses have been found infecting algae, large viruses placed in the family Phycodnaviridae and a virus that morphologically resembles TMV. There are also one or more uncharacterized viruses. 1. Large algal viruses (reviewed by van Etten, 1999; van Etten and Meints, 1999) Virus-like particles (VLPs) have been observed in thin sections of many eukaryotic algal species belonging to the Chtorophyceae, Rhodophyceae and the Phaeophyceae. The particles are polygonal in outline and vary in diameter from about 22 to 390 nm. Some have tails reminiscent of bacteriophage. The most studied viruses are those infecting Chlorella-like green algae. Members of this group are very diverse biochemically (van Etten et al., 1988). An important technical advance was made when van Etten et al. (1983) developed a plaque assay for a virus called Paramecium bursaria chlorella virus I (PBCV-1) infecting a culturable Chlorellalike alga. As a consequence, most is now known about the properties of this virus. The particles of PBCV-1 are large icosahedra with multilaminate shells surrounding an
electron-dense core. The outer capsid comprises 1692 capsomeres arranged in a T = 169 skew icosahedral lattice. Hair-like fibres extend from some of the particle vertices. The genome of PBCV-1 is dsDNA of 330 kb with covalently closed hairpin termini. It contains 701 potential coding ORFs that are arranged in 376, mostly overlapping, ORFs that are believed to encode proteins, and 325 minor ORFs that may or may not encode proteins. The encoded proteins include replication enzymes (e.g. DNA polymerase, DNA ligase and endonuclease), nucleotide metabolism enzymes (e.g. ATPase, thioredoxin and ribonuclease reductase), transcription factors (e.g. RNA transcription factors TFIIB and TFIIS and RNase III), enzymes involved in protein synthesis, modification and degradation (e.g. ubiquitin C-terminal hydrolase, translation elongation factor 3 and 26S protease subunit), phosphorylating enzymes, cell-wall degrading enzymes, DNA restriction and modification enzymes, sugar and lipid manipulation enzymes, and various structural proteins. Similar large DNA viruses have been found in marine algae (see van Etten and Meints, 1999). 2. Small algal viruses Skotnicki et al. (1976) described a virus infecting the eukaryotic alga Chara australis. The virus (CAV) has rod-shaped particles and some other properties like those of tobamoviruses. However, its genome is much larger (11 kb, rather than 6.4 kb for TMV), and about 7 kb of the genome has been sequenced, revealing other relationships (Matthews, 1991): (1) the coat protein of CAV has a composition closer to BNYVV, and to TRV than to TMV; (2) the GDD-polymerase motif of CAV is closest to that of BNYVV; and (3) the two GKT nucleotide-binding motifs found in CAV are arranged in a manner similar to that found in potexviruses. Thus, CAV appears to share features of genome organization and sequences found in several groups of rod-shaped viruses infecting angiosperms. It appears to have strongest affinity with the Furovirus genus and has no known angiosperm host. For these reasons it is
V. V I R U S E S OF L O W E R P L A N T S
most unlikely that CAV originated in a recent transfer of some rod-shaped virus from an angiosperm host to Chara. In its morphology, Chara is one of the most complex types of Charophyceae, which is a welldefined group with a very long geographical history (Round, 1984). Based on cytological and chemical similarities, land plants (embryophytes) are considered to have evolved from a charophycean green alga. Coleochaete, another of the more complex types among the Charophyceae, has been shown to contain lignin, a substance thought to be absent from green algae (Delwiche et al., 1989). Molecular genetic evidence supports a charophycean origin for land plants. Group II introns have been found in the tRNA a~ and tRNA i~e genes of all land plant chloroplast DNAs examined. All the algae and eubacteria examined have uninterrupted genes. Manhart and Palmer (1990) have shown that introns are present in three members of the Charophyceae in the same arrangement as in Marchantia, giving strong support to the view that they are related to the lineage that gave rise to land plants. Tree construction suggests that the Charophyceae may have acquired the introns 400 to 500 million years ago. Thus, the virus described in Chara is probably the oldest recorded virus infecting a plant on or near the lineage that ultimately gave rise to the angiosperms. Straight (c. 280 nm) and flexuous (c. 700900 nm) virus particles have been found associated with dieback symptoms in the brown alga Ektonia radiata in New Zealand (Easton et al., 1997).
B. Viruses of fungi Representatives of five viral families infect fungi, if one excludes yeast. Isometric particles of 60 nm diameter and containing a 16.8 kbp dsDNA are found in the aquatic fungus Rhizidomyces (Dawe and Kuhn, 1983). This virus belongs to the genus Rhizidovirus and appears to be transmitted in a latent form in the fungal zoospores being activated under stress conditions. As shown in Table 4.1, the majority of fungal viruses have dsRNA genomes. These viruses
43
can influence the biology of plant pathogenic fungi (McCabe and van Alfen, 1999). In many cases, they are cryptic but some induce the fungus (e.g. Ustilago maydis) to produce killer toxins and others reduce the virulence of a pathogenic fungus. DsRNA viruses of fungi are found in three families. The members of the Hypoviridae have pleomorphic vesicle-like particles 50-80 nm in diameter that contain a single molecule of dsRNA of about 9-13 kbp (reviewed by Nuss, 1999). The coding region may be divided into two ORFs or expressed as a single ORF depending upon species. Hypoviruses infect the chestnut blight fungus Cryphonectria parasitica, reducing its virulence on chestnut trees. The family Partitiviridae has two genera that infect fungi (reviewed by Ghabrial and Hillman, 1999). As with the plant-infecting partitiviruses (Section III.E), the isometric particles, 30--40 nm in diameter, contain two segments of dsRNA, the larger encoding the viral polymerase and the smaller the virus coat protein. These viruses are associated with latent infections of fungi. Members of the genus Totivirus, family Totiviridae, encapsidate a single molecule of dsRNA (4.6-6.7 kbp) in isometric particles, 4 0 n m in diameter (reviewed by Ghabrial and Patterson, 1999). These viruses are associated with latent infections of their fungal hosts. The one family of (+)-strand ssRNA genome viruses that infects fungi, the Barnaviridae, has bacilliform particles 1 8 - 2 0 n m wide and 48-53 nm long (reviewed by Romaine, 1999). The virion contain a single RNA molecule of 4 kb that has a VPg at the 5' end (Revill et al., 1998) and lacks a poly(A) tail at the 3' end. The genome contains four major ORFs (Revill et al., 1994). The genome arrangement is similar to that of poleroviruses (Section III.P.2). The function of ORF 1 is unknown. ORFs 2 and 3 make up the viral polymerase containing a protease, VPg and RdRp domain. ORF 4 encodes the 22kDa coat protein that is expressed from a subgenomic RNA (Revill et al., 1999). The one species in this genus, Mushroom bacilliform virus, infect Agaricus bisporus and A. campestris. It is possibly associated with La France disease of cultivated mushrooms.
44
e N O M E N C L A T U R E AN[-) CLASSIFI(.~ATION OF P L A N T V I R U S E S
C. Viruses of ferns No viruses have been reported from bryophytes. A virus with particles like those of a Tobravirus was found in hart's tongue fern (Phyllitis scolopendrium) by Hull (1968).
D. Viruses of gymnosperms A disease of Cycas revoluta has been shown to be due to a virus with a bipartite genome and other properties that place it in the Nepovirus group (Hanada et al., 1986). The virus was readily transmitted by mechanical inoculation to various Chenopodium species. It was also transmitted through the seed of these species. There have been a few reports of pines being infected experimentally with viruses from angiosperms (see Fulton, 1969; Jacobi et al., 1992). There are a few reports of naturally occurring virus-like diseases in other gymnosperms but the viral nature of the diseases has not been demonstrated (e.g. Schmelzer et al., 1966). However, because of the presence of substances such as tannins, there are technical difficulties in attempting to isolate viruses from gymnosperms.
E. Summary The existence of a (+)-sense ssRNA virus infecting the genus Chara suggests an ancient origin for this type of virus. Other than this example, the meager information about viruses infecting photosynthetic eukaryotes below the angiosperms can tell us very little about the age and course of evolution a m o n g the plant viruses. The cycads are regarded as living fossils, being in the record since early Mesozoic times. However, the Nepovirus found in Cycas revoluta is quite likely to have originated in a modern angiosperm, since it readily infects Chenopodium spp. The Phycodnavirus PBCV-1 infecting a Chlorella-like alga is much more likely to be of ancient origin. However, based on structure they do not appear to be primitive viruses. They are much larger and more complex than any k n o w n viruses infecting angiosperms, with a genome of about 300 kbp
and at least 50 structural proteins (Meints et at., 1986).
VI. DISCUSSION As noted in the introduction to this chapter, the classification system for viruses, though fully justified, is without a natural base. This is primarily because there is no time-related information on the evolution of, and relationships between, virus species and genera. Thus, one cannot distinguish with certainty between convergent and divergent evolution. There is further discussion on virus evolution in Chapter 17. Notwithstanding these limitations, an effective system for classifying plant viruses has been developed over the last 20 years or so. The development of this system has given rise to controversies and, no doubt, there will be others in the future (see Bos, 1999b, 2000b; Pringle, 1999b; van Regenmortel, 1999, 2000). This system is fulfilling most of the four criteria that I gave in Section I.A but, as pointed out above, there are some reservations on evolutionary relationships. The system is dynamic and is being modified and refined to take account of new research findings. New virus genera are being created and genera grouped together on common features into families. For instance, the grouping of luteoviruses into three genera in one family (described in Section III.P) has helped significantly with the understanding of these viruses. Other virus groups are being, or are likely to be, merged. For instance, it is becoming increasingly apparent that AMV shares many properties with the ilarviruses (Section III.J.3) and there is beginning to be a strong case for merging the Ilarvirus and Alfamovirus genera. However, if this occurs recognition will have to be given to the differences in particle structure between the two groups. One large grouping that needs further consideration is that of the 20 floating genera of (+)-strand ssRNA viruses. In Section Q, I have brought together genera that appear to have c o m m o n properties to highlight the possibilities for creation of new taxa.
vl. t~iScussloN With the increasing masses of information, there is beginning to be pressure for the creation of higher taxa. The example of viruses (and other elements) that involve reverse transcription in their replication is noted in Section I.C. This would bring together viruses from different kingdoms, especially eukaryotic. There are, of course, virus groups such as the Reoviridae and Rhabdoviridae that span the plant and animal kingdoms. The real need now is to
45
develop systems that reflect commonalities in viruses with (+)-strand ssRNA genomes. As is discussed further in Chapters 7 and 8, viruses are faced with various problems to overcome in infecting their host, and common solutions have been found for plant- and animalinfecting viruses. This should be reflected in the classification. The current state of plant virus taxonomy is reviewed by Mayo and Brunt (2001).
This Page Intentionally Left Blank
C
H
A
P
T
E
R
Disease Symptoms and Hos t Range
Viruses are economically important only when they cause some significant deviation from normal in the growth of a plant. For experimental studies, we are usually dependent on the production of disease in some form to demonstrate biological activity. S y m p t o m a t o l o g y was particularly important in the early days of virus research, before any of the viruses themselves had been isolated and characterized. Dependence on disease symptoms for identification and classification led to much confusion, because it was not generally recognized that many factors can have a marked effect on the disease produced by a given virus (see Chapter 10). Most virus names in common use include terms that describe an important s y m p t o m in a major host or the host from which the virus was first described. There is a vast literature describing diseases produced by viruses. This has been summarized by Smith (1937, 1972), Bos (1978), Holmes (1964), and in the C M I / A A B Descriptions of Plant Viruses issued from 1970 onwards and edited by B. D. Harrison and A. F. M u r a n t m n o w available on CD-ROM from the Association of Applied Biologists (see Appendix 3). Some viruses under appropriate conditions may infect a plant without producing any obvious signs of disease. Others may lead to rapid death of the whole plant. Between these extremes, a wide variety of diseases can be produced. Virus infection does not necessarily cause disease at all times in all parts of an infected plant. We can distinguish six situations in which obvious disease may be absent: (1) infection with a very mild strain of the virus; (2) a
tolerant host; (3) nonsterile 'recovery' from disease symptoms in newly formed leaves; (4) leaves that escape infection because of their age and position on the plant; (5) dark green areas in a mosaic pattern (discussed in Chapter 10, Section III.O.4); and (6) plants infected with cryptic viruses (see Chapter 2, Section III.E).
I. E C O N O M I C LOSSES D U E TO P L A N T VIRUSES (reviewed by Waterworth and Hadidi, 1998) One of the main driving forces for the detailed studies on plant viruses described in this book is the impact that the diseases they cause have on crop productivity worldwide; yet it is difficult to obtain firm data on the actual losses themselves. The losses due to fungal and bacterial pathogens are well documented and it is c o m m o n to see lists of loss estimates attributable to specific named fungi or bacteria. In these compendia, the losses due to viruses are lumped together in categories such as 'virus diseases' and 'all other' or 'miscellaneous' diseases. However, viruses are responsible for far greater economic losses than is generally recognized. This lack of recognition is due to several factors, especially their insidious nature. Virus diseases are frequently less conspicuous than those caused by other plant pathogens and last for much longer. This is especially true for perennial crops and those that are vegetatively propagated. One further problem with attempting to assess losses due to virus diseases on a global basis is that most of the data are from small comparative trials rather than widescale comprehensive surveys. Even the small trials do not
Virus acronyms are given in A p p e n d i x 1.
47
48
3 DISEASE SYMPTOMS AND HOST RANGE
necessarily give data that can be used for more global estimates of losses. This is for several reasons, including: (1) variation in losses by a particular virus in a particular crop from year to year; (2) variation from region to region and climatic zone to climatic zone; (3) differences in loss assessment methodologies; (4) identification of the viral etiology of the disease; (6) variation in the definition of the term 'losses'; and (6) complications witl'i other loss factors. Several publications give g u i d a n c e on approaches to overcome these problems. These include one published by FAO on loss assessm e n t m e t h o d s (Chiarappa, 1971) and w i t h chapters on the rationale and concepts of crop loss assessments and modeling of crop growth and yield for loss assessment (Teng, 1987). Others have discussed the nature of crop losses (Main, 1983), identification and assessment of losses (McKenzie, 1983), crop destruction and classification of crop losses (Main, 1977), assessments of plant diseases and losses (James, 1974), m e t h o d s for determining losses (Bos, 1982), and terms and concepts of crop losses (Nutter et at., 1993). In addition to the obvious detrimental effects such as reduced yields and visual product quality, virus infections often do not induce noticeable disease but influence their effects on plants in a variety of more subtle ways. Table 3.1 identifies some of the ways in which viruses can d a m a g e crop plants. From this it can be seen that the effects of virus infection extend into areas far beyond the actual reduction in yield and quality. Loss estimates do not take account of these indirect factors. In spite of all these limitations, there have been various collections of loss data (e.g. Hull and Davies, 1992; Waterworth and Hadidi, 1998). Some examples are given in Table 3.2.
II. MACROSCOPIC SYMPTOMS A. Local symptoms Localized lesions that develop near the site of entry on leaves are not usually of any economic significance but are important for biological assay (see Chapter 12). Infected cells m a y lose
TABLE 3.1 Some types of direct and indirect damage associated with plant virus infections Reduction in growth Yield reduction (including symptomless infection) Crop failure Reduction in vigour Increased sensitivity to frost and drought Increased predisposition to attack by other pathogens and pests Reduction in quality or market value Defects of visual attraction: size shape, color Reduced keeping quality Reduced consumer appeal: grading, taste, texture, composition Reduced fitness for propagation Cost of attempting to maintain crop health Cultural hygiene on farm including vector control Production of virus-free propagation materials Checking propagules and commodities on export/import (quarantine programs) Eradication programs Breeding for resistance Research, extension and education Data from Waterworth and Hadidi (1998), with permission. chlorophyll and other pigments, giving rise to chlorotic local lesions (Plate. 3.1). The lesion m a y be almost white or merely a slightly paler shade of green than the rest of the leaf. In a few diseases, for example in older leaves of tomato inoculated with TBSV, the lesions retain more chlorophyll than the s u r r o u n d i n g tissue. For m a n y host-virus combinations, the infected cells die, giving rise to necrotic lesions. These vary from small pinpoint areas to large irregular spreading necrotic patches (Plate 3.2). In a third type, ring spot lesions appear. Typically, these consist of a central group of dead cells. Beyond this, there develop one or more superficial concentric rings of dead cells with normal green tissue between them (Plate 3.3). Some ring spot local lesions consist of chlorotic rings rather than necrotic ones. Some viruses in certain hosts show no visible local lesions in the intact leaf, but w h e n the leaf is cleared in ethanol and stained with iodine, 'starch lesions' m a y become apparent. Viruses that produce local lesions w h e n inoculated mechanically onto leaves m a y not do so w h e n introduced by other means. For example,
If.
MAC)ROSCTC3PI(~ S Y M P T O M S
49
TABLE 3.2 Some examples of crop losses due to viruses Crop
Virus
Countrie s
Loss/ y ear
Rice
Tungro Ragged stunt Hoja blanca Barley yellow dwarf Barley yellow dwarf Potato leafroll Potato virus Y Potato virus X Beet yellows Beet mild yellows Citrus tristeza African cassava mosaic Tomato spotted wilt" Cocoa swollen shoot
SE Asia SE Asia S. and C. America UK UK UK
$1.5 • 109 $1.4 • 10s $9.0 • 106 s • 106 s • 10~' s • 107
UK
s
Worldwide Africa Worldwide Ghana
s $2 • $1 • 1.9 •
Barley Wheat Potato Sugarbeet Citrus Cassava Many crops Cocao
• 10{' • 10~' 109 ] 09 10s trees ~,
" Data from Prins and Goldbach (1998), with permission; references to other data given in Hull and Davies (1992), with permission. bNumber of trees eradicated over about 40 years.
BYV
produces
necrotic
local
lesions
on
Chenopodium capitatum, but does not do so w h e n the virus is introduced by the aphid
Myzus persicae feeding on p a r e n c h y m a cells (Bennett, 1960). However, AMV does produce local lesions following aphid transmission.
B. Systemic symptoms The following sections s u m m a r i z e the major kinds of effects p r o d u c e d by systemic virus invasion. It should be borne in m i n d that these various s y m p t o m s often a p p e a r in combination in particular diseases, and that the pattern of disease d e v e l o p m e n t for a particular h o s t - v i r u s combination often involves a sequential develo p m e n t of different kinds of s y m p t o m s . 1. Effects on plant size Reduction in plant size is the most general s y m p t o m i n d u c e d by virus infection (Plate 3.4). There is probably some slight general stunting of growth even with ' m a s k e d ' or 'latent' infections where the systemically infected plant shows no obvious sign of disease. For example, mild strains of PVX infecting potatoes in the field m a y cause no obvious s y m p t o m s , and carefully designed experiments were necessary to show that such infection reduced tuber yield by about 7-15% (Matthews, 1949d). The degree of stunting is generally correlated w i t h the
severity of other s y m p t o m s , particularly w h e r e loss of chlorophyll from the leaves is concerned. Stunting is usually almost entirely due to reduction in leaf size and internode length. Leaf n u m b e r m a y be little affected. In p e r e n n i a l d e c i d u o u s plants such as grapes, there m a y be a delayed initiation of g r o w t h in the spring (e.g. Gilmer et al., 1970). Root initiation in cuttings from virus-infected plants m a y be reduced, as in c h r y s a n t h e m u m s (Horst et al., 1977). In vegetatively p r o p a g a t e d plants, stunting is often a progressive process. For example, virusinfected s t r a w b e r r y plants and tulip bulbs m a y become smaller in each successive year. Stunting m a y affect all parts of the plant more or less equally, involving a reduction in size of leaves, flowers, fruits, and roots and shortening of petioles a n d internodes. Alternatively, some parts m a y be considerably more stunted than others. For example, in little cherry disease, fruits remain small o w i n g to reduced cell division, in spite of a p p a r e n t l y a m p l e leaf growth. A reduction in total yield of fruit is a c o m m o n feature and an i m p o r t a n t economic aspect of virus disease. The lower yield m a y s o m e t i m e s be due to a reduction in both size and n u m b e r of fruits (e.g. H a m p t o n , 1975). In a few diseases, for example p r u n e dwarf, fruits m a y be greatly reduced in n u m b e r but of larger size than normal. Healthy cherry trees
50
3 DISEASE SYMPTOMS AND HOST RANGE
pollinated with pollen from trees infected with this virus (Way and Gilmer, 1963) or necrotic ring spot virus (V6rtesy and Ny6ki, 1974) had a reduced fruit set. Seed from infected plants may be smaller than normal, germination may be impaired, and the proportion of aborted seed may be increased (Walkey et al., 1985). 2. Mosaic patterns and related symptoms One of the most common obvious effects of virus infection is the development of a pattern of light and dark green areas giving a mosaic effect in infected leaves. I will describe the phenotypes in this section, and in Chapter 10 (Section Ill.P) I discuss the current molecular understanding of mosaic pattern formation. The detailed nature of the pattern varies widely for different host-virus combinations. In dicotyledons, the areas m a k i n g up the mosaic are generally irregular in outline. There may be only two shades of color i n v o l v e d m dark green and a pale or yellow-green, for example. This is often so with TMV in tobacco (Plate 3.5), or there may be many different shades of green and yellow, as with TYMV in Chinese cabbage (Plate 3.6). The junctions between areas of different color may be sharp and such diseases resemble quite closely the mosaics produced by inherited genetic defects in the chloroplasts. AbMV is a good example of this type (Plate 3.7). TYMV in Chinese cabbage may approach genetic variegation in the sharpness of the mosaic pattern it produces. The borders between darker and lighter areas may be diffuse (Plate 3.8). If the lighter areas differ only slightly from the darker green, the mottling may be difficult to observe as with some of the milder strains of PVX in potato. In mosaic diseases infecting herbaceous plants, there is usually a fairly well-defined sequence in the development of systemic symptoms. The virus moves up from the inoculated leaf into the growing shoot and into partly expanded leaves. In these leaves, the first symptoms are a 'clearing' or yellowing of the veins; as described in Chapter 10 (Section III.O.2), it is suggested that this s y m p t o m is an optical illusion. However, chlorotic vein-banding is a true s y m p t o m and may be very faint or may give striking emphasis to the pattern of veins (Plate
3.9). Vein-banding may persist as a major feature of the disease. In leaves that are past the cell division stage of leaf expansion when they become infected (about 4-6 cm long for leaves such as tobacco and Chinese cabbage) no mosaic pattern develops. The leaves become uniformly paler than normal. In the oldest leaves to show mosaic, a large number of small islands of dark green tissue usually appear against a background of paler color. The mosaic areas may be confined to the youngest part of the leaf blade, that is, the basal and central region. Although there may be considerable variation in different plants, successively younger systemically-infected leaves show, on the average, mosaics consisting of fewer and larger areas. The mosaic pattern is laid down at a very early stage of leaf development and may remain unchanged, except for general enlargement, for most of the life of the leaf. In some mosaic diseases the dark green areas are associated mainly with the veins to give a dark green vein-banding pattern. In monocotyledons, a common result of virus infection is the production of stripes or streaks of tissue lighter in color than the rest of the leaf. The shades of color vary from pale green to yellow or white, and the more or less angular streaks or stripes run parallel to the length of the leaf (Plate 3.10). The development of the stripe diseases found in monocotyledons follows a similar general pattern to that found for mosaic diseases in dicotyledons. One or a few leaves above the inoculated leaf that were expanded at time of inoculation show no stripe pattern. In the first leaf to show striping, the pattern is relatively fine and may occur only in the basal (younger) portion of the leaf blade. In younger leaves, stripes tend to be larger and occur throughout the leaf. The patterns of striping are laid d o w n at an early stage and tend to remain unchanged for most of the life of the leaf. Yellowed areas may become necrotic as the leaf ages. A variegation or 'breaking' in the color of petals commonly accompanies mosaic or streak symptoms in leaves. The breaking usually consists of flecks, streaks or sectors of tissue with a color different from normal (Plates 3.11 and
11. M A C R O S ( ~ ' O P I C S Y M P T O M S
3.12). The breaking of petal color is frequently due to loss of anthocyanin pigments, which reveals any underlying coloration due to plastid pigments. In a few instances, for example in tulip color-adding virus, infection results in increased pigmentation in some areas of the petals. Nectar guides in petals are often invisible to h u m a n s (but visible to bees) because they involve pigments that absorb strongly only in the ultraviolet region (Thompson et al., 1972). The effects of virus infection on these nectar guides and on the behavior of honeybees do not appear to have been studied. Infected flowers are frequently smaller than normal and may drop prematurely. Flower breaking may sometimes be confused with genetic variegation, but it is usually a good diagnostic feature for infection by viruses producing mosaic symptoms. In a few plants, virus-induced variegation has been valued commercially. Thus, as described in Chapter 1 (Section I), at one time virus-infected tulips were prized as distinct varieties. As with the d e v e l o p m e n t of mosaic patterns in leaves, color-breaking in the petals may develop only in flowers that are smaller than a certain size when infected. Thus, in tulips inoculated with TBV less than 11 days before blooming no break symptoms developed even though virus
51
was present in the petals (Yamaguchi and Hirai, 1967). Virus infection may reduce pollen production and decrease seed set, seed size and germination (e.g. Hemmati and McLean, 1977) (Fig. 3.1). Fruits formed on plants showing mosaic disease in the leaves may show a mottling, for example, zucchinis infected with CMV (Plate 3.13). In other diseases, severe stunting and distortion of fruit may occur. Seed coats of infected seed may be mottled. 3. Yellow diseases Viruses that cause a general yellowing of the leaves are not as numerous as those causing mosaic diseases, but some, such as the viruses causing yellows in sugarbeet, are of considerable economic importance. The first sign of infection is usually a clearing or yellowing of the veins (Plate 3.14) in the younger leaves followed by a general yellowing of the leaves (Plate 3.15). This yellowing may be slight or severe. No mosaic is produced, but in some leaves there may be sectors of yellowed and normal tissue. In strawberry yellow edge disease, yellowing is largely confined to the margins of the leaf. When severe, a yellows disease m a y lead to a total loss of the crop (e.g. Weidemann et al., 1975).
Fig. 3.1 Effect of TRSV infection on germination and germ tube growth of soybean pollen. The pollen grains were germinated overnight in 30% sucrose. (A) Infected; (B) healthy. From Yang and Hamilton (1974), with permission.
52
3 DISEASE SYMPTOMS AN[) HOST RANGE
4. Leaf rolling Virus infection can result in leaf rolling, which is usually u p w a r d s (Plate 3.16) or occasionally downwards. Pronounced epinasty of leaf petioles m a y sometimes be a p r o m i n e n t feature. 5. Ring spot diseases A marked s y m p t o m in many virus diseases is a pattern of concentric rings and irregular lines on the leaves (Plate 3.17) and sometimes also on the fruit (Plates 3.18 and 3.19). The lines may consist of yellowed tissue or may be due to death of superficial layers of cells, giving an etched appearance. In severe diseases, complete necrosis through the full thickness of the leaf lamina may occur. With the ring spot viruses, such as TRSV, there is a strong tendency for plants to recover from the disease after an initial shock period. Leaves that have developed symptoms do not lose these, but younger growth may show no obvious symptoms in spite of the fact that they contain virus. Ring spot patterns may also occur on other organs, for example bulbs (Asjes et al., 1973), and the tuber necrotic ring spot strain of PVY causes rings on potato tubers, often around the eyes, that become sunken and necrotic (see Beczner et al., 1984; Weidemann and Maiss, 1996). 6. Necrotic diseases Death of tissues, organs or the whole plant is the main feature of some diseases. Necrotic patterns may follow the veins as the virus moves into the leaf (Plate 3.20). In some diseases, the whole leaf is killed. Necrosis may extend fairly rapidly throughout the plant. For example, with PVX and PVY in some varieties of potatoes, necrotic streaks appear in the stem. Necrosis spreads rapidly to the growing point, which is killed, and subsequently all leaves may collapse and die. Wilting of the parts that are about to become necrotic often precedes such systemic necrotic disease. 7. Developmental abnormalities Besides being generally smaller than normal, virus-infected plants may show a wide range of developmental abnormalities. Such changes
may be the major feature of the disease or may a c c o m p a n y other s y m p t o m s . For example, uneven growth of the leaf lamina is often found in mosaic diseases. Dark green areas may be raised to give a blistering effect, and the margin of the leaf may be irregular and twisted (Plate 3.21). In some diseases, the leaf blade may be more or less completely suppressed, such as in tomatoes infected with CMV a n d / o r TMV (Plate 3.22) (Francki et al., 1980a). Some viruses cause swellings in the stem, which may be substantial in woody plants, such as in cocoa swollen shoot disease. Another group of growth abnormalities is known as enations. These are outgrowths from the upper or lower surface of the leaf usually associated with veins (Plate 3.23). They may be small ridges of tissue, or larger, irregularly shaped leaflike structures, or long filiform outgrowths. Conversely, normal outgrowths may be suppressed. For example, a potyvirus infection in Datura metel causes the production of fruits lacking the normal spines (Rao and Yaraguntiah, 1976). Viruses may cause a variety of tumor-like growths. The tumor tissue is less organized than with enations. Some consist of wart-like o u t g r o w t h s on stems or fruits. The most studied tumors are those produced by WTV which are characteristic of this disease (Plate 3.24). In a systemically infected plant, external tumors appear on leaves or stems where w o u n d s are made. In infected roots they appear spontaneously, beginning development close to cells in the pericycle that are w o u n d e d when developing side-roots break through the cortex. The virus may also cause many small internal tumors in the phloem of the leaf, stem and root (Lee and Black, 1955). Stem deformation such as stem splitting and scar-formation is caused by some viruses in some w o o d y plants. Virus infection of either the rootstock or scion can cause necrosis a n d / o r failure of the graft union (Plate 3.25). One of the unusual symptoms of BSV in some M u s a cultivars is that the fruit bunch emerges from the side of the p s e u d o s t e m instead of from the top of it (Plate 3.26). This is due to necrosis of the cigar leaf.
II. MACROSC;OPIC SYMPTOMS
8. Wilting Wilting of the aerial parts frequently followed by death of the whole plant may be an important feature (e.g. in virus diseases of chickpea; Kaiser and Danesh, 1971). 9. Recovery from disease Not uncommonly, a plant shows disease symptoms for a period and then new growth appears in which s y m p t o m s are milder or absent, although virus may be still present. This commonly occurs with Nepovirus infections. Many factors influence this recovery phenomenon. The stage of development at which a plant is infected may have a marked effect on the extent to which symptoms are produced. For example, tobacco plants inoculated with BCTV develop disease symptoms. This stage is frequently followed by a recovery period. If very young seedling plants are inoculated, a proportion of these might never show clear symptoms, even though they can be shown to contain virus (Benda and Bennett, 1964). The environment can also affect recovery from disease as can host species or variety and virus strain. The molecular aspects of disease recovery are discussed in Chapter 10 (Section Ill.R). 10. Reduced nodulation Various workers have described a reduction in the number, size and fresh weight of nitrogenfixing Rhizobium nodules induced by virus infection in legumes (e.g. AMV in alfalfa: Ohki et al., 1986; CMV in pea: Rao et al., 1987). In general, overall nitrogen fixation is reduced by virus infection. With AMV infection in Medicago, nitrogen fixation per unit of nodule fresh weight is the same as in healthy plants, but infected plants produce less nodule tissue. Hence, nitrogen fixation per plant is reduced (Dall et al., 1989). 11. Genetic effects Infection with BSMV induces an increase in mutation rate in Zea mays and also a genetic abnormality k n o w n as an aberrant ratio (AR) (Sprague et al., 1963; Sprague and McKinney, 1966, 1971). For example, w h e n a virus-infected pollen parent with the genetic constitution A1A 1, PrPr, SuSu was crossed with a homozygous
53
recessive line (ala 1, prpr, susu) resistant to the virus, a low frequency of the progeny lines gave significant distortion from the expected ratios for one or more of the genetic markers. (Al-a 1 = presence or absence of aleurone color; Pr-pr = purple or red aleurone colour; Su-su = starchy or sugary seed.) This AR effect was observed only when the original pollen parent was infected and showing virus symptoms on the upper leaves. The AR effect is inherited in a stable manner, with a low frequency of reversion to normal ratios. It is inherited in plants where virus can no longer be detected. WSMV also induces the AR effect (Brakke, 1984). The genetics of the AR effect are complex, and probably more than one phenomenon is involved. There is no evidence that a cDNA copy of part or all of the viral genome is incorporated into the host genome. The AR effect has been reviewed by Brakke (1984). Further work may show that similar phenomena occur in other virus-infected plants. C. A g e n t s i n d u c i n g v i r u s . l i k e s y m p t o m s Disease symptoms, similar to those produced by viruses, can be caused by a range of physical, chemical and biological agents. Such diseases may have interesting factors in c o m m o n with virus-induced disease. The activities of such agents have sometimes led to the erroneous conclusion that a virus was the cause of a particular disease in the field. Further confusion may arise w h e n a disease is caused by the combined effects of a virus and some other agent. 1. Small cellular parasites A group of diseases characterized by symptoms such as general yellowing of the leaves, stunting, witches-broom growth of axillary shoots, and a change from floral to leaf-type structures in the flowers (phyllody) was for m a n y years thought to be caused by viruses. Diseases of this type are not transmissible by mechanical means. They were considered to be caused by viruses because (1) no bacteria, rickettsiae, fungi or protozoa could be implicated; (2) they were graft transmissible; (3) they were transmitted by leafhoppers; and (4) some, at least, could be transmitted by dodder.
54
3 DISEASE SYMPTOMS AND HOST RANGE
Mycoplasmas and spiroplasmas belong to the Mycoplasmatales. They are characterized by a bounding unit membrane, pleomorphic form, the absence of a cell wall, and complete resistance to penicillin. Rickettsia-like organisms are distinguished from the Mycoplasmatales by the possession of a cell wall and by their in vivo susceptibility to penicillin. They have been found in the phloem of various species. From the pioneering experiments of Doi et al. (1967) and Ishiie et al. (1967) a new branch of plant pathology was opened up by the demonstration that mycoplasmas (now termed phytoplasmas), spiroplasmas and rickettsia-like organisms can cause disease in plants (e.g. Windsor and Black, 1973). These agents are generally confined to the phloem or xylem of diseased plants. They are too small to be readily identified by light microscopy. The widespread availability of electron microscopy provided the key technique that has allowed their importance in plant pathology to be recognized. Phytoplasma diseases of plants are described in Maramorosch and Raychaudhuri (1988) and recent advances in their study by Davis and Sinclair (1998) and Lee et al., (1998a). They are classified on molecular properties such as 16S ribosomal RNA (see Marcone et al., 1999; Webb et al., 1999). A simultaneous virus and phytoplasma infection may give rise to a more severe disease than either agent may alone. For example, OBDV and the phytoplasm agent of aster yellows are both confined to the phloem, and both can be transmitted by the leafhopper Macrosteles fascifrons (Stal.). In a mixed infection, they cause more severe stunting than either agent alone (Banttari and Zeyen, 1972).
systemically through the plant and produce virus-like symptoms. Calligypona pellucida (F.) (Homoptera) produces salivary toxins that cause general retardation of growth and inhibition of tillering. Only females produce the toxins (Nuorteva, 1962). Eryophyid mites feeding on clover may induce a mosaic-like mottle in the younger leaves. One mite is sufficient to induce such symptoms. These mites are small and may be overlooked since they burrow within the leaf. A virus-like mosaic disease of wax myrtle (Myrica cerifera) was found to be caused by a new species of eryophyid mite (Elliot et al., 1987). Maize plants showing wallaby ear disease are stunted and develop galls on the underside of the leaves, which are sometimes darker green than normal. This virus-like condition has been shown to be due to a toxin produced by the leafhopper Cicadulina bimaculata (Ofori and Francki, 1983).
2. Bacteria Some bacteria can cause virus-like symptoms. For instance, mechanical inoculation of cowpea leaves can give rise to necrotic local lesions caused by unknown, adventitious bacteria.
4. Genetic abnormalities Numerous cultivated varieties of ornamental plants have been selected by horticulturists because they possess heritable leaf variegations or mosaics. These are often due to maternally inherited plastid defects. The variegated patterns produced sometimes resemble virusinduced mosaics quite closely. However, in mosaics due to plastid mutation, the demarcation between blocks of tissue of different colors tends to be sharper than with many virus-induced mosaics. Other virus symptoms may be mimicked by genetic abnormalities. For example, Edwardson and Corbett (1962) described a 'wiry' mutant in Marglobe tomatoes that gave an appearance similar to that of plants infected with strains of CMV and TMV. All the leaves of the mutant resembled the symptoms of virus infection in that upper leaves lacked a lamina as shown in Plate 3.21. However, the disease could not be transmitted by grafting, and genetic experiments suggested that a pair of recessive genes control the mutant phenotype.
3. Toxins produced by arthropods Insects and other arthropods feeding on plants may secrete very potent toxins, which move
5. Transposons Transposons are mobile elements that can move about the plant (or animal) genome.
II.
There are two types of transposons, what may be termed true transposons such as the Ac/Ds and MuDR/Mu maize transposons (reviewed by Walbot, 2000) and retrotransposons, which are described in more detail in Chapter 2 (Section IV). If transposons move into a gene or genes that control leaf or flower color during the development of that organ, they can cause flecks or streaks of different colors that resemble virus symptoms. This can be especially noticeable in flowers where one can find apparent flower color break most likely caused by the movement of a transposon. 6. Nutritional deficiencies Plants may suffer from a wide range of nutritional deficiencies that cause abnormal coloration, discoloration or death of leaf tissue. Some of these conditions can be fairly easily confused with symptoms due to virus infection. For example, magnesium and iron deficiency in soybeans leads to a green banding of the veins with chlorotic interveinal areas. The yellowing, however, is more diffuse than is usual in virus infection and is usually in the older leaves. In sugarbeet, yellowing and necrosis due to m a g n e s i u m deficiency may be similar to the disease produced by BYV. Potassium and m a g n e s i u m deficiency in potatoes produces
MACROSCOPIC
SYMPTOMS
55
marginal and interveinal necrosis similar to that found in certain virus diseases. 7. High temperatures Growing plants at substantially higher temperatures than n o r m a l m a y induce virus-like symptoms. When N. glutinosa plants were held at 37.8 ~ for 4-8 days and then returned to 22 ~, new leaves displayed a pattern of mosaic, veinclearing, chlorosis, and other abnormalities with a resemblance to virus infection (see Fig. 3.2). These symptoms gradually disappeared in newer leaves, but could be induced in the same plants again by a second treatment at high t e m p e r a t u r e (John and Weintraub, 1966). Mechanical inoculation and grafting tests to various hosts and electron microscopy failed to reveal the presence of a virus in heated plants. 8. Hormone damage Commercially used hormone weedkillers may produce virus-like s y m p t o m s in some plants. Tomatoes and grapes are particularly susceptible to 2,4-dichlorophenoxyacetic acid (2,4-D). Growth abnormalities in the leaves caused by 2,4-D bear some resemblance to certain virus infections in these and other hosts. C o m p o u n d s related to 2,4-D can cause almost complete suppression of mesophyll development, giving
Fig. 3.2 Virus-like symptoms induced in Nicotiana glauca following a period of growth at high temperature (37.8~ From John and Weintraub (1966), with permission.
56
3 DISEASE SYMPTOMS A N D H O S T R A N G E
a plant with a 'shoestring' appearance. Alternatively, vein growth may be retarded more than the mesophyll. Mesophyll may then bulge out between the veins to give an appearance not unlike leaf curl diseases. 9. Insecticides Certain insecticides have been reported to produce leaf symptoms that mimic virus infection (e.g. Woodford and Gordon, 1978). 10. Air pollutants Many air pollutants inhibit plant growth and give rise to symptoms that could be confused with a virus disease. For example, chimney gases from a cement factory caused Zea mays seedlings to become stunted and yellowed, with necrotic areas and curled leaf margins (Cireli, 1976).
D. The cryptoviruses Cryptoviruses escaped detection for many years because most of them cause no visible symptoms or, in a few situations, very mild symptoms. They are not transmissible mechanically or by vectors, but are transmitted efficiently in pollen and seed. They occur in very low concentrations in infected plants (reviewed by Boccardo et al., 1987). Nevertheless, they have molecular characteristics that might be expected of diseaseproducing viruses. As described in Chapter 6 (Section VI.B), the genome consists of two dsRNA segments and these viruses share some properties with the reoviruses. There is no indication, other than the low concentration at which they occur, as to why they cause symptomless infection.
III. HISTOLOGICAL CHANGES The macroscopic symptoms induced by viruses frequently reflect histological changes within the plant. These changes are of three main typesmnecrosis, hypoplasia and h y p e r p l a s i a m t h a t may occur singly or together in any particular disease. For
example, all three are closely linked in the citrus exocortis disease (Fudl-Allah et al., 1971).
A. Necrosis Necrosis as the major feature of disease was discussed in Section II.B.6. In other diseases, necrosis may be confined to particular organs and tissues and may be very localized. It commonly occurs in combination with other histological changes. It may be the first visible effect or may occur as the last stage in a sequence. For example, necrosis of epidermal cells or of midrib parenchyma may be caused by lettuce mosaic virus in lettuce (Coakley et al., 1973). Necrosis caused by TNV is usually confined to localized areas of the roots (e.g. Lange, 1975). Late infection of virus-free tomato plants with TMV may give rise to internal necrosis in the immature fruits (e.g. Taylor et al., 1969). In the potato leaf roll disease, the phloem develops normally but is killed by the infection. Necrosis may spread in phloem throughout the plant, but is limited to this tissue (Shepardson et al., 1980). In Pelargonium infected with TRSV, histological effects seen by light microscopy were confined to reproductive tissues (Murdock et at., 1976). Pollen grain abortion and abnormal and aborted ovules were common. The symptoms could be confused with genetic male sterility.
B. Hypoplasia Leaves with mosaic symptoms frequently show hypoplasia in the yellow areas. The lamina is thinner than in the dark green areas, and the mesophyll cells are less differentiated with fewer chloroplasts and fewer or no intercellular spaces (Fig. 3.3). In stem-pitting disease of apples, pitting is shown on the surface of the wood when the bark is lifted. The pitting is due to the failure of some cambial initials to differentiate cells normally, and a wedge of phloem tissue is formed that becomes e m b e d d e d in newly formed xylem tissue (Hilborn et al., 1965). The affected phloem becomes necrotic. The major anatomical effect of ASGV in apple stems is the disappearance of the
III.
HISTOLOGICAL
CHANGES
57
Reduced size of pollen grain and reduced g r o w t h of pollen tubes from virus-infected pollen m a y be r e g a r d e d as hypoplastic effects (Fig. 3.1). A variety of other effects on pollen grains has been described (e.g. Haight and Gibbs, 1983).
C. Hyperplasia 1. Cells are larger than normal Vein-clearing s y m p t o m s are due, with some viruses at least, to e n l a r g e m e n t of cells near the veins (Esau, 1956). The intercellular spaces are obliterated, and since there is little chlorophyll present the tissue m a y become a b n o r m a l l y translucent. 2. Cell division in differentiated cells Some viruses such as PVX m a y produce islands of necrotic cells in potato tubers. The tuber m a y respond with a typical w o u n d reaction in a zone of cells a r o u n d the necrotic area. Starch grains disappear and an active cambial layer develops (Fig. 3.4). Similarly, in a white halo zone s u r r o u n d i n g necrotic local lesions induced by TMV in N. glutinosa leaves, cell division o c c u r r e d in m a t u r e palisade cells (Wu, 1973).
Fig. 3.3 Histological and cytological effects of TMV in tobacco. (A) Section through palisade cells of a dark green area of a leaf showing mosaic. Cells are essentially normal. (B) Section through a nearby yellow-green area. Cells are large and undifferentiated in shape. Nuclei are not centrally located as in the dark green cells. Bar = 20 ~m. (Courtesy of P. H. Atkinson.) c a m b i u m in the region of the groove. N o r m a l p h l o e m and xylem elements are replaced by a largely undifferentiated p a r e n c h y m a (Plebe et al., 1975).
Fig. 3.4 Section through a potato tuber infected with PVX, showing a cork cambial layer being developed near a group of necrotic cells (bottom left).
5S
3 DISEASE SYMPTOMS AND HOST RANGE
3. Abnormal division of cambial cells The vascular tissues appear to be particularly prone to virus-induced hyperplasia. In the diseased shoots found in swollen shoot disease of cocoa, abnormal amounts of xylem tissue are produced but the cells appear structurally normal (Posnette, 1947). In plants infected by BCTV, a large number of abnormal sieve elements develop, sometimes associated with companion cells. The arrangement of the cells is disorderly and they subsequently die (Esau, 1956; Esau and Hoefert, 1978). OBDV causes abnormalities in the development of phloem in oats, involving hyperplasia and limited hypertrophy of the phloem procambium (Zeyen and Banttari, 1972). In crimson clover infected by WTV, there is abnormal development of phloem cambium cells. Phloem parenchyma forms meristematic tumor cells in the phloem of leaf, stem and root (Lee and Black, 1955). Galls on sugarcane leaves arise from Fijivirusinduced cell proliferation. This gives rise in the mature leaf to a region in the vein where the vascular bundle is grossly enlarged (Fig. 3.5). Two main types of abnormal cell are present: lignified gall xylem cells and non-lignified gall phloem (Hatta and Francki, 1976). Hyperplastic growth of phloem was marked in plum infected with PPV (Buchter et al., 1987).
IV. CYTOLOGICAL EFFECTS The cytological effects of viruses have been a subject of interest ever since the early searches with light microscopes for causative organisms in diseased tissues. About the beginning of the twentieth century, these studies led to the discovery of two types of virus inclusion: amorphous bodies or 'X bodies' and crystalline inclusions. The X bodies resemble certain microorganisms. Some workers erroneously considered that they were in fact the parasite or a stage in the life cycle of the parasite causing the disease. These early conclusions were not entirely wrong since many of the X bodies are in fact virus-induced structures in the cell where the components of viruses are synthesized and assembled.
A. Methods Light microscopy is still important in the study of cytological abnormalities for several reasons: 1. Much greater areas of tissue can be scanned, thus ensuring that samples taken for electron microscopy are representative. 2. It may allow electron microscopic observations to be correlated with earlier detailed work on the same material using light microscopy. 3. Observations on living material can be made using both phase and bright-field illumination. Improvements in procedures for the fixing, staining and sectioning of plant tissues over the past 40 years and the widespread availability of high-resolution electron microscopes has led to a substantial growth in our knowledge about the cytological effects of viruses in cells. As in other fields of biology, electron microscopy is providing a link between macroscopic and light microscope observations on the one hand and molecular biological and biochemical studies on the other. Examination of stained thin sections remains the standard procedure, but freeze-fracturing can give useful information on virus-induced membrane changes (Hatta et al., 1973). Scanning electron microscopy is of less value in the study of virus diseases but is sometimes useful (Hatta and Francki, 1976). It must be remembered that small differences in conditions under which plants are grown before sampling and in the procedure used to prepare tissue for electron microscopy can have a marked effect on the appearance and stability of organelles and virus-induced structures (e.g. Langenberg, 1982). Chilling of tissue before fixation may improve the preservation of very fragile virusinduced structures (Langenberg, 1979). To relate any observed cytological effects to virus replication, it is very useful to be able to follow the time course of events in infected cells. In principle, protoplasts infected in vitro should provide excellent material for studying the time course of events. However, various limitations are becoming apparent:
IV. CYTOLOGICAL EFFECTS
59
Fig. 3.5 Structure of leaf galls on sugarcane infected with FDV. (A) Transverse section of vascular tissue in a leaf vein from a healthy sugarcane plant, showing the xylem (x) and phloem (p) tissues. (B) A transverse section of vascular tissues of a vein on a galled leaf of an FDV-infected sugarcane plant, showing the gall phloem (gp) and gall xylem (gx), in addition to normal phloem (p) and xylem (x) tissues. (Bars = 0.5 pm). (C) Part of a sugarcane leaf infected with FDV, showing small and large gall (arrows). (D) A diagram of the tissue distribution in the vein of an FDV-infected sugarcane leaf showing normal and gall tissues. From Egan et al. (1989), with permission. 1. Some ultrastructural features seen in TMVinfected tobacco leaf cells were not o b s e r v e d d u r i n g TMV r e p l i c a t i o n in p r o t o p l a s t s (Otsuki et al., 1972a). 2. Crystalline inclusion bodies in TMV-infected leaf cells w e r e d e g r a d e d in protoplasts m a d e f r o m such leaves (F/3glein et aI., 1976). 3. D e v e l o p m e n t of cytological changes in protoplasts infected in vitro m a y be m u c h less sync h r o n o u s t h a n i n d i c a t e d by g r o w t h curves.
The uses of fluorescent dyes a n d confocal m i c r o s c o p y are described in C h a p t e r 9 (Section
II.B). B. Effects on cell structures 1. Nuclei C y t o p a t h o l o g i c a l effects of virus infection h a v e been well illustrated by Francki et al. (1985a,b). M a n y viruses h a v e no detectable cytological
60
~ DISEASE SYMPTOMS AND HOST RANGE
effects on nuclei. Others give rise to intranuclear inclusions of various sorts and m a y affect the nucleolus or the size and shape of the nucleus, even t h o u g h they appear not to replicate in this organelle. Shikata and M a r a m o r o s c h (1966) found that in pea leaves and pods infected with PEMV, particles a c c u m u l a t e first in the nucleus. During the course of the disease, the nucleolus disintegrates. Masses of virus particles accumulate in the nucleus and also in the cytoplasm. PEMV also causes vesiculation in the perinuclear space (De Zoeten et al., 1972). Virus particles of several small isometric viruses accumulate in the nucleus (as well as the cytoplasm). They may exist as scattered particles or in crystalline arrays (e.g. SBMV: Weintraub and Ragetli, 1970; TBSV: Russo and Martelli, 1972). Masses of viral protein or empty viral protein shells have been observed in nuclei of cell infected with several tymoviruses (Hatta and Matthews, 1976) (Fig. 3.6). Crystalline plate-like inclusions were seen by light microscopy in cells infected with severe etch in tobacco (Kassanis, 1939). The plates were birefringent w h e n viewed sideways and were a very regular feature of severe etch infection. Intranuclear inclusions have been described for some other potyviruses. The v i r u s - c o d e d proteins i n v o l v e d in potyvirus nuclear and cytoplasmic inclusions are discussed in Chapter 6 (Section VIII.C.I.b).
Electron-lucent lacunae a p p e a r e d in the nucleolus of Nicotiana cells infected with PVA ( E d w a r d s o n and Christie, 1983). For some r h a b d o v i r u s e s , viral cores a p p e a r in the nucleus and accumulate in the perinuclear space (see Fig. 8.19). Geminiviruses cause m a r k e d h y p e r t r o p h y of the nucleolus, which m a y come to occupy three-quarters of the nuclear volume. Fibrillar rings of deoxyribonucleoprotein appear, and masses of virus particles accumulate in the nucleus (e.g. Rushing et al., 1987). An isolate of CaMV has been described in which nuclei become filled with virus particles and greatly enlarged (Gracia and Shepherd, 1985). The virus particles were not e m b e d d e d in the matrix protein found in cytoplasmic viroplasms. 2. Mitochondria The long rods of TRV may be associated with the mitochondria in infected cells (Harrison and Roberts, 1968) (see Fig. 8.17) as are the isometric particles of BBWV-1 (see Fig. 3.9C) (Hull and Plaskitt, 1974). The mitochondria in cells of a range of host species, and various tissues, infected with CGMMV develop small vesicles b o u n d e d by a m e m b r a n e and lying within the peri-mitochondrial space and in the cristae (Hatta et al., 1971). Aggregated mitochondria have been o b s e r v e d in Datura cells infected by the potyvirus HMV (Kitajima and Lovisolo, 1972), Fig. 3.6 The intracellular sites of TYMV and TYMV protein accumulation. (A) Small crystalline array of TYMV particles associated with electron-lucent area in the cytoplasm of a stage C cell subjected to plasmolysis. Bar = 500 nm; r, ribosomes; v, TYMV particles. From Hatta and Matthews (1974), with permission. (B) Accumulation of TYMV coat protein (electron-lucent material, L) in the nucleus of a stage D cell not subjected to plasmolysis. Bar = 500 nm. Modified from Hatta and Matthews (1976).
IV.
but there was no indication that these aggregates were involved in virus synthesis. The development of abnormal membrane systems within mitochondria has been described for several virus infections (Francki, 1987). They have no established relation to virus replication and are probably degenerative effects. For example, in some tombusvirus infections multivesiculate bodies appear in the cytoplasm. These have been shown to develop from greatly modified mitochondria (Di Franco et al., 1984; Di Franco and Martelli, 1987). In infections with other tombusviruses, multivesiculate bodies originate from modified peroxisomes (Martelli et al., 1984). In Sonchus infected with BYSV, virus particles are found in phloem cells. The flexuous rod-shaped particles are frequently inserted into the cristae of the mitochondria (Esau, 1979). 3. Chloroplasts The small peripheral vesicles and other changes in and near the chloroplasts closely related to TYMV replication are discussed in Chapter 8 (Section IV.K.2). TYMV infection can cause many other cytological changes in the chloroplasts, most of which appear to constitute a structural and biochemical degeneration of the organelles. The exact course of events in any mesophyll cell depend's on: (1) the developmental stage at which it was infected, (2) the strain of virus infecting, (3) the time after infection, and (4) the environmental conditions (Matthews, 1973; Hatta and Matthews, 1974). In inoculated leaves, the chloroplasts become rounded and clumped together in the cell. There is little effect on grana or stroma lamellae. The chloroplasts become cup-shaped, with the opening of the cup generally facing the cell wall. Starch grains accumulate. 'White' strains of the virus cause degeneration of the grana in inoculated leaves. In expanded leaves above the inoculated leaf that become fully infected without the appearance of mosaic symptoms, the effects of infection on chloroplasts are similar to those seen in inoculated leax}es. In leaves that are small at time of infection and that develop the typical mosaic, a variety of different pathological states in the chloroplasts can readily be distinguished by light
CYTOLCK}ICAL EFFECTS
61
microscopy in fresh leaf sections. Islands of tissue in the mosaic showing various shades of green, yellow and white contain different strains of the virus, which affect the chloroplasts in recognizably distinct ways (see Fig. 9.18). In dark green islands of tissue, which contain very little virus, chloroplasts appear normal. The most important changes in the chloroplasts seen in tissue types other than dark green are: (1) color, ranging from almost normal green to colorless; (2) clumping to a variable extent; (3) presence of large vesicles; (4) fragmentation of chloroplasts; (5) reduction in granal stack height; (6) presence of osmiophilic globules; and (7) arrays of phytoferritin molecules. Some of these abnormalities are illustrated in Fig. 9.18. Different strains of TYMV produce particular combinations of abnormalities in the chloroplasts. In blocks of tissue of one type, almost all cells show the same abnormalities and these persist at least for a time as the predominating tissue type when inoculations are made to fresh plants. In contrast to the small peripheral vesicles, which appear to be induced by all tymoviruses in the chloroplasts of infected cells, none of the changes noted is an essential consequence of tymovirus infection; nor can they be regarded as diagnostic for the group. For example, no clumping of the chloroplasts occurs in cucumbers infected with OkMV. On the other hand, clumping of chloroplasts is induced by TuMV in Chenopodium (Kitajima and Costa, 1973). Several viruses outside the Tymovirus genus induce small vesicles near the periphery of the chloroplasts. These vesicles differ from the Tymovirus type in that they do not appear to have necks connecting them to the cytoplasm. In Datura leaves infected with TBSV, the thylakoid membranes undergo varied and marked rearrangements (Bassi et al. 1985). For most of these viruses, the vesicles or other changes appear to be degenerative consequences of infection. However, for BSMV, the vesicles appear to be associated with virus replication (Lin and Langenberg, 1984a). In many infections, the size and number of starch grains seen in leaf cells are abnormal. In
62
3 DISEASE SYMPTOMS A N D H O S T R A N G E
mosaic diseases, there is generally speaking less starch than normal, but in some diseases (e.g. sugarbeet curly top and potato leaf roll) excessive amounts of starch may accumulate. Similarly in local lesions induced by TMV in cucumber cotyledons, chloroplasts become greatly enlarged and filled with starch grains (Cohen and Loebenstein, 1975.) (See also Section IV.D.) 4. Cell walls The plant cell wall tends to be regarded mainly as a physical supporting and barrier structure. In fact, it is a distinct biochemical and physiological compartment containing a substantial proportion of the total activity of certain enzymes in the leaf (Yung and Northcote, 1975). Three kinds of abnormality have been observed in or near the walls of virus-diseased cells: 1. Abnormal thickening, owing to the deposition of callose, may occur in cells near the edge of virus-induced lesions (e.g. Hiruki and Tu, 1972). Chemical change in the walls may be complex and difficult to study (Faulkner and Kimmins, 1975). 2. Cell wall protrusions involving the plasmodesmata have been reported for several unrelated viruses. The protrusions from the plasmodesmata into cells may have one or more canals. They may be quite short or of considerable length. They appear to be due to deposition of new wall material induced by the virus, and they may be lined inside and out with plasma membrane (e.g. Bassi et al., 1974). 3. Depositions of electron-dense material between the cell wall and the plasma membrane may extend over substantial areas of the cell wall (as with ONMV: Gill, 1974) or may be limited in extent and occur in association with plasmodesmata (as with BSMV: McMullen et al., 1977). They have been called paramural bodies. The major cytopathic effect of CEVd is the induction of numerous small membrane-bound bodies near the cell wall with an electron density similar to that of the plasma membrane (Semancik and Vanderwoude, 1976). They are found in all cell types.
5. Bacteroidal cells The first phase in the infection of soybean root cells by Rhizobium (i.e. development of an infection thread and release of rhizobia into the cytoplasm) appears not to be affected by infection of the plant with SMV. In the second stage a membrane envelope forms around the bacterial cell to form a bacteroid. Structural differences such as decreased vesiculation of this membrane envelope were observed in virusinfected roots (Tu, 1977). 6. Myelin-like bodies Myelin-like bodies consisting of densely staining layers that may be close-packed in a concentric or irregular fashion have been described for several plant virus infections (e.g. Kim et al., 1974). They probably reflect degenerative changes in one or more of the cell's membrane systems. They may be associated with osmiophilic globules, which are thought to consist of the lipid component of cell membranes. Kim et al. suggested that myelin-like bodies in bean leaf cells infected with comoviruses may be formed from the osmiophilic globules. 7. Cell death Drastic cytological changes occur in cells as they approach death. These changes have been studied by both light and electron microscopy, but they do not tell us how virus infection actually kills the cell. C. V i r u s - i n d u c e d s t r u c t u r e s in the cytoplasm The specialized virus-induced regions in the cytoplasm that are, or appear likely to be, the sites of virus synthesis and assembly (viroplasms) are discussed in Chapter 8. Here other types of inclusions will be described. These are usually either crystalline inclusions consisting mainly of virus, or the pinwheel inclusions characteristic of the potyviruses. Light and electron microscopy of these inclusions is reviewed in Christie and Edwardson (1977). 1. Crystalline inclusions Virus particles may accumulate in an infected cell in sufficient numbers and exist under
IV. CYTOLOGICAL EFFECTS
suitable conditions to form three-dimensional crystalline arrays. These may grow into crystals large enough to be seen with the light microscope, or they may remain as small arrays that can be detected only by electron microscopy. The ability to form crystals within the host cell depends on properties of the virus itself and is not related to the overall concentration reached in the tissue or to the ability of the purified virus to form crystals. For example, TYMV can readily crystallize in vitro. It reaches high concentration in infected tissue but does not normally form crystals there. By contrast, sTNV occurring in much lower concentrations frequently forms intracellular crystals. a. TMV In tobacco leaves s h o w i n g typical mosaic symptoms caused by TMV, leaf-hair and epidermal cells over y e l l o w - g r e e n areas may almost all contain crystalline inclusions, while those in fully dark green areas contain none. The junction between yellow-green and dark green tissue may be quite sharp, so that there is a zone where neighboring leaf hairs have either no crystals, or almost every cell has crystals. Warmke and Edwardson (1966) followed the development of crystals in leaf-hair cells of tobacco. Virus particles were first seen free in the cytoplasm as small aggregates of parallel rods with ends aligned. These aggregates increase in size. The growing crystals are not bounded by a membrane, and as they become multilayered they may sometimes incorporate endoplasmic reticulum, mitochondria, and even chloroplasts between the layers. The plate-like crystalline inclusions are very unstable and are disrupted by pricking or otherwise damaging living cells. They are birefringent when viewed edge-on, but not when seen on the flat face. They contain about 60% water and otherwise consist mainly of successive layers of closely packed parallel rods oriented not quite perpendicularly to the plane of the layers. Rods in successive layers are tilted with respect to one another. This herringbone effect can be visualized in freeze-fractured preparations for some strains of TMV (Fig. 3.7). Sometimes long, curved, fibrous inclusions, or
63
Fig. 3.7 Crystal of TMV rods in a freeze-etched preparation. Part of TMV crystal lying within a mesophyll cell that has been penetrated by glycerol. The crystal has retained its herringbone structure and the lattice spacing of about 24 nm, despite the fact that the tonoplast was ruptured. Bar = 1 gm. From Willison (1976), with permission. spike-like or spindle-shaped inclusions made up largely of virus particles, can be seen by light microscopy. Different strains of the virus may form different kinds of paracrystalline arrays. Most crystalline inclusions have been found only in the cytoplasm but some have been detected in nuclei (e.g. Esau and Cronshaw, 1967). b. BYV The inclusions found in plants infected with BYV occur in phloem cells and also appear in other tissues, for example, the mesophyll. By light microscopy, the inclusions are frequently spindle-shaped and may show banding (Fig. 3.8A). Electron microscopy reveals layers of flexuous virus rods (Fig. 3.8B). Most of the viral inclusions occur in the cytoplasm. Smaller aggregates of virus-like particles were seen in nuclei and chloroplasts (Cronshaw et al., 1966). c. O t h e r helical viruses
Rod-shaped viruses belonging to other groups m a y aggregate in the cell into more or less ordered arrays. These can frequently be observed only by electron microscopy, and other material besides virus rods may be present in the arrays. As well as such aggregates of
64
3 DISEASE S Y M P T O M S A N D H O S T R A N G E
Fig. 3.8 Inclusion bodies caused by BYV in parenchyma cells of small veins of Beta vulgaris. The banded form of these inclusions is shown (A) by light microscopy and (B) by electron microscopy. Bands are made up of flexuous virus particles in more or less orderly array. From Esau et al. (1966), with permission. virus rods, RCVMV induces the appearance of large crystals in the cytoplasm (Khan et al., ] 977). These contain RNA and protein and consist of a crystalline array of polyhedral particles about 10 n m in diameter. No virus rods were present. The significance of these unusual crystals in u n k n o w n . d. Small icosahedral viruses
M a n y small icosahedral viruses form crystalline arrays in infected cells (see Fig. 3.9D). Sometimes these are large e n o u g h to be seen by light microscopy (e.g. TNV: Kassanis et aI., 1970). lcosahedral viruses that do not normally form regular arrays may be induced to do so by heating or plasmolyzing the tissue to remove some of the water (Milne, 1967; Hatta, 1976). Many strains of BBWV induce cylindrical tubules of virus particles (Fig. 3.9A, B). Russo et al. (1979) described a strain that formed unusual hollow tubules that are square or rectangular in section. The walls are m a d e up of two parallel rows of virus particles.
e. Reoviruses and rhabdoviruses
Plant cells infected with viruses belonging to the Reoviridae or Rhabdoviridae frequently contain masses of virus particles in regular arrays in the cytoplasm. With some rhabdoviruses the bullet-shaped particles accumulate in more or less regular arrays in the perinuclear space (see Fig. 8.19) (see also Francki et al., 1985c). 2. Pinwheel inclusions Potyviruses induce the formation of characteristic cylindrical inclusions in the cytoplasm of infected cells (e.g. Hiebert and McDonald, 1973). The most striking feature of these inclusions viewed in thin cross-section is the presence of a central tubule from which radiate c u r v e d ' a r m s ' to give a p i n w h e e l effect. Reconstruction from serial sections shows that the inclusions consist of a series of plates and curved scrolls with a finely striated substructure with a periodicity of about 5 nm. The bundles, cylinders, tubes and pinwheels seen in section are aspects of geometrically complex
IV. CYTOLOGICAL EFFECTS
65
Fig. 3.9 Electron micrographs of thin sections of Nicotiana clevelandii leaves infected with BBWV-1 (then named petunia ringspot virus). (A,B) Longitudinal and transverse sections of tubes composed of virus particles. (C) Virus particles arranged between mitochondria. (D) Crystals of virus particles. Markers = 250 nm. From Hull and Plaskitt (1974), with kind permission of the copyright holder, 9 Karger, Basel.
structures. The general structure has been confirmed by examination of freeze-etched preparations (McDonald and Hiebert, 1974) and by the use of a tilting stage together with computer-assisted analytical geometry (Mernaugh et al., 1980) (Fig. 3.10). For some potyviruses, the pinwheels are tightly curved. For others they are more open. The inclusions induced by members of the group may differ consistently in various details. Studies using confocal laser scanning microscopy show that the three-dimensional structure of ZYMV cytoplasmic inclusions is a linear filamentous structure of varying thickness and length (Lim et aI., 1996). The average length varied from 9.4 ___0.3 to 20.1 ___0.4 ~m and the average w i d t h from 2.1 + 0.1 to 3.7 _ 0.1 gm. Pinwheel inclusions originate and develop in association with the plasma membrane at sites lying over plasmodesmata (Lawson et al., 1971; Andrews and Shalla, 1974). The central tubule of the pinwheel is located directly over the plasmodesmata and it is possible that the membranes may be continuous from one cell to the
next. The core and the sheets extend out into the cytoplasm as the inclusion grows. Later in infection, they may become dissociated from the plasmodesmata and come to lie free in the cytoplasm. Virus particles may be intimately associated with the pinwheel arms at all times and particularly at early stages of infection (Andrews and Shalla, 1974). The virus-coded protein that is found in these inclusions is discussed in Chapter 6 (Section VIII.C). 3. Caulimovirus inclusions Two forms of inclusion bodies (also termed viroplasms) have been recognized in the cytoplasms of plants infected with CaMV and other 'caulimoviruses' (see Fig. 8.23); they are not seen in cells of plants infected by the closelyrelated 'badnaviruses'. Both forms contain virus particles. Electron-dense inclusions are made up of ORF VI product and are considered to be the sites of virus synthesis and assembly (see Chapter 8, Section VII.B.2). Electron-lucent inclusion bodies are made up of ORF II product, one of the proteins involved in aphid transmission (see Chapter 11, Section III.F).
66
3 DISEASE S Y M P T O M S A N D H O S T R A N G E
Fig. 3.10 Morphological modifications of cytoplasmic inclusions according to the subdivisions of Edwardson (1974a) and Edwardson et al. (1984). (A) Scrolls induced in Nicotiana clevelandii by an unidentified isolate from eggplant in Nigeria. (B) Laminated aggregates induced by statice virus Y in Chenopodium quinoa. (C) Scrolls and laminated aggregates induced by strain O of TuMV in N. clevelandii. (D) Scrolls and short curved laminated aggregates in N. tabacum infected by PVY. Bars = 200 nm. From Lesemann (1988), with permission.
D. Cytological structures resembling those induced by viruses Some n o r m a l structures in cells could be mistaken for v i r u s - i n d u c e d effects--for example, crystalline or m e m b r a n e - b o u n d inclusions in plastids (e.g. N e w c o m b , 1967). P r o l a m e l l a r
bodies in chloroplasts give the a p p e a r a n c e in section of a regular array of tubes. These bodies m a y be i n d u c e d by certain chemical treatments (Wrischer, 1973). P h o s p h o r u s - d e f i c i e n t b e a n leaves ( T h o m p s o n et al., 1964) s h o w e d d e g e n e r a t i v e changes in the chloroplasts like those seen in some virus
v. THE HOST RXNOE OF V~RUSES
infections. Similarly, in sulfur-deficient Zea mays, chloroplasts contained many osmiophilic granules and small vesicles (Hall et al., 1972). Bundle sheath chloroplasts of C4 plants contain numerous small vacuoles near their periphery and vacuoles also have been described for chloroplasts in certain tissues of C3 plants (e.g. Marinos, 1967). In a spontaneous plastid m u t a n t of Epilobium hirsutum, the degenerate grana and vacuolation of the mutant plastids seen by electron microscopy (Anton-Lamprecht, 1966) bear some resemblance to the pathological changes induced by TYMV infection in Chinese cabbage. Nuclei of healthy cells sometimes contain crystalline structures that might be mistaken for viral inclusions (e.g. Lawson et al., 1971). Such virus-induced effects as disorganization of membrane systems, presence of numerous osmiophilic granules, and disintegration of organelles are similar to normal degenerative processes associated with aging or degeneration induced by other agents.
E. Discussion There are three main cytological effects of virus infection. Firstly, cellular components are frequently affected. Sometimes the impact of infection is major, inducing marked changes in the structures of organelles such as chloroplasts or even breakdown of internal components leading to death of the cell. These frequently play a role in the symptoms that infection produces. There are also many minor and transitory effects that virus infection has on cells. ! will describe in Chapter 13 some of the dynamism involved in virus replication and expression. Secondly, there are often aggregates of virus particles. As will be described in subsequent chapters, viruses can go through n u m e r o u s rounds of replication, which if allowed to happen in an uncontrolled manner could result in cell death. Thus, it is possible that aggregation of a virus is a mechanism for removing it from the replication pool. Thirdly, there are aggregates of virus gene products. The ORF VI inclusion bodies of CaMV are the sites of virus expression and
67
replication. However, the transactivation properties of the ORF VI product, described in Chapter 7 (Section V.B.7), could be deleterious to the normal function of the cell. Thus, it is possible that the aggregation of the ORF VI protein is a mechanism of sequestering it away from the cell components. Similarly, the potyviral genome is expressed from a polyprotein which gives rise to as many copies of protease and replicase molecules as of coat protein molecules. It is likely that the proteases and replicases could be deleterious to the cell and hence they are sequestered into inclusion bodies.
V. THE H O S T R A N G E OF VIRUSES Since the early years of the twentieth century, plant virologists have used host range as a criterion for attempting to identify and classify viruses. In a typical experiment, the virus under study w o u l d be inoculated by mechanical means to a range of plant species. These would then be observed for the development of virus-like disease symptoms. Back-inoculation to a host k n o w n to develop disease might be used to check for symptomless infections. In retrospect, it can be seen that reliance on such a procedure gives an oversimplified view of the problem of virus host ranges. Over the past few years, our ideas of what we might mean by 'infection' have been considerably refined, and some possible molecular m e c h a n i s m s that might make a plant a host or a non-host for a particular virus have emerged. The term 'host' is sometimes used rather loosely. Technically, it is defined as 'an organism or cell culture in which a given virus can replicate'. This w o u l d mean that a plant species in which the virus can replicate in the initially infected cell (subliminal infection) is a host. However, this is impractical and for this book I will use the terms local host for a species in which the virus is restrained to the inoculated leaf and systemic host for a species in which the virus spreads from the inoculated leaf to other, but not necessarily all, parts of the plant. More detailed aspects of the interaction of viruses with plants are discussed in Chapter 10.
68
~ DISEASE S Y M P T O M S AND H O S T R A N G E
A. L i m i t a t i o n s in h o s t r a n g e s t u d i e s 1. General Almost all the plant viruses so far described have been found infecting species among the angiosperms. Only a minute proportion of the possible h o s t - v i r u s combinations has been tested experimentally. The following arithmetic indicates the scale of our ignorance. Horvath (1983) tested the host range of 24 viruses on 456 a n g i o s p e r m species. He found 1312 new host-virus combinations, that is, 12% of those he tested. There may be about 250 000 species of angiosperms (Heywood, 1978) and over 900 plant viruses have been recorded. If the 12% rate applied on average to all these plants and viruses, then there may be more than 25 • 106 new compatible h o s t - v i r u s combinations awaiting discovery. In relation to this figure, the number of combinations already tested must be almost negligible. Our present knowledge of the occurrence and distribution of viruses among the various groups of plants is both fragmentary and biased. There are four probable reasons for this. First, plant virologists working on diseases as they occur in the field have been primarily concerned with viruses causing economic losses in cultivated plants. They have usually been interested in other plant species only to the extent that they might be acting as reservoirs of a virus or its vector affecting a cultivated species. Thus, until fairly recently all the k n o w n plant viruses were confined to the angiosperms. Within this group, most of the known virus hosts are plants used in agriculture or horticulture or are weed species that grow in cultivated areas. Most of the world's population is fed by 12 species of plant (three cereals: rice, wheat and corn; two sugar plants: beet and cane; three root crops: potato, sweet potato and cassava; two legumes: bean and soybean; and two tree crops:-coconut and banana). Eighty-seven of the 977 viruses recognized by the ICTV have names derived from these 12 host plants. The number of angiosperm species is commonly reported to be about 250 000. Thus, about 9% of the viruses have been described from about 0.005% of the k n o w n species. Tobacco and
tomato provide another example. Both these annual crops are of high value commercially on a per-hectare basis. Fifty-five viruses are listed with tobacco and tomato as part of the vernacular name. Not many of the species in lower phyla are used commercially. It is not surprising, therefore, that the first virus found in these groups was one associated with cultivated mushrooms. The second reason for our incomplete knowledge is a more speculative one. It seems likely that widespread and severe disease in plants due to virus infection is largely a consequence of h u m a n agricultural manipulations. Under natural conditions, viruses are probably closely adapted to their hosts and cause very little in the way of obvious disease. Thus, casual inspection of plants growing in their natural habitat may give little indication of the viruses that might be present. Adequate testing of a significant number of such species by means of inoculation tests, both by mechanical transmission and with possible invertebrate vectors, would be laborious and time-consuming. Very little systematic testing of this sort has been carried out. Thirdly, the selection of 'standard' test plants for viruses is to a great extent governed by those species that are easy to grow in glasshouses and to handle for mechanical and insect vector inoculation. Fourthly, the genera and species chosen for a host range study may not form a taxonomically balanced selection. Watson and Gibbs (1974) pointed out that most virologists, working in the north temperate zone, use mainly festucoid grasses in host range studies; whereas in other parts of the world non-festucoid groups predominate in the flora and in agricultural importance. 2. Technical limitations There are a number of technical difficulties and pitfalls in attempting to establish the host range of a virus. These will be discussed in relation to diagnosis in Chapter 15 (Section II). Another kind of potential limitation in the use of mechanical inoculation is illustrated by experiments with the viroid PSTVd. Based on mechanical inoculation tests, a particular
V. TIlE HOST RANGE OF VIRUSES
accession of Sotanum acaule has been considered immune to PSTVd. However, Agrobacteriummediated inoculation of this plant led to the systemic replication of the viroid, as did grafting with viroid-infected tomato (Salazar et al., 1988b).
69
advanced groups of plants to contain a higher proportion of species susceptible to the viruses examined than the lower phylogenetic groups. Generally speaking, viruses can infect a higher proportion of species in the family of the common field host and closely related families than in distantly related families.
B. Patterns of host range In spite of the limitations alluded to above, some general points can be made. Different viruses may vary widely in their host range. At one extreme BSMV is virtually confined to barley as a host in nature (Timian, 1974). At the other extreme, CMV, AMV, TSWV, TMV and TRSV have very wide host ranges. For instance, TSWV has a host range of over 925 species belonging to 70 botanical families, and CMV can infect more than 1000 species in more than 85 botanical families. The host range of one virus may sometimes fall completely within the host range of another a p p a r e n t l y unrelated virus. Thus, Holmes (1946) found that among 310 species tested, 83 were susceptible to both TMV and TEV, 116 were susceptible to TMV but not TEV, none to infection by TEV but not TMV, and 111 were not susceptible to either virus. There were also great differences in s y m p t o m response (Table 3.3). Bald and Tinsley (1970) showed that eight cocoa virus isolates could be placed in a host range containment and divergence series. The most virulent isolate infected 21 of the 26 species examined, and this host range contained the host ranges of the other seven isolates. Bald and Tinsley (1967) re-examined data of earlier workers by statistical procedures. There was a tendency for phylogenetically more TABLE 3.3 Host ranges of, and s y m p t o m responses to, TMV and TEV TMV
Localization
Symptoms?
TEV
111 ' 100 15 27 57
Immune/subliminal Local Systemic Local Systemic
No s y m p t o m s No s y m p t o m s No s y m p t o m s Symptoms Symptoms
227 15 8 7 53
N u m b e r of species. Data from Holmes (1946), with permission.
C. The determinants of host range Biological and statistical studies of the sort described in the previous section cannot lead us to an u n d e r s t a n d i n g of reasons w h y a virus infects one plant species and not another. For this we need biochemical, molecular biological and genetic information of the sort that is beginning to become available. There is no doubt that very small changes in the viral genome can affect host range. For example, Evans (1985) described a nitrous acid m u t a n t of CPMV that was unable to grow in cowpea but that could grow in Phaseolus vulgaris. The mutation was in the B-RNA. On the basis of present knowledge there are four possible stages where a virus might be blocked from infecting a plant and causing systemic disease: (1) during initial e v e n t s - - t h e uncoating stage; (2) during attempted replication in the initially infected cell; (3) during m o v e m e n t from the first cell in which the virus replicated; and (4) by stimulation of the host's cellular defenses in the region of the initial infection. These stages will be considered in turn. Molecular aspects of host range determination are discussed in Chapters 9 and 10. 1. Initial events a. Recognition of a suitable host cell or organelle
Bacterial viruses and most of those infecting vertebrates have specific proteins on their surface that act to recognize a protein receptor on the surface of a susceptible host cell. The surface proteins on plant rhabdo-, reo- and tospo-viruses may have such a cell recognition function. Such a function is unlikely to be of use in plants. For instance, virus particles of LNYV, whose outer membranes have been removed with detergent, are infectious to plants (Randles and Francki, 1972). However, the G
70
3 DISEASE SYMPTOMS AND HOST RANGE
protein spikes or projections (see Fig. 5.40) are likely to be important in recognizing membranes of insect vectors. Surface proteins on plant reoviruses almost certainly have a recognition role in their insect vectors. There is no evidence for plant cell recognition receptors on the surface of any of the ssRNA plant viruses; however, there are receptors on the surface of RNA viruses that have a circulative interaction with their biological vector (see Chapter 11, Section III.H.l.a). The evidence available for these small viruses suggests that host range is usually a property of the RNA rather than the protein coat. When it has been tested, the host range of a plant virus is the same whether intact virus or the RNA is used as inoculum. Attempts to extend host ranges by using infectious RNA rather than virus have generally been unsuccessful. Hiebert et al. (1968) showed that artificial hybrid Bromovirus particles, consisting of CCMV RNA in a coat of BMV protein, could still infect cowpea, which is immune to BMV. The hybrid could not infect barley, the normal host of BMV. These and similar tests showed that the host range of a viral RNA cannot be extended by coating the RNA in the protein of a virus that can attack the host. Atabekov and colleagues (Atabekov, 1975) have studied the host ranges of many 'hybrid' viruses and found examples where a heterologous coat protein limited host range. Thus, BMV RNA in a coat of TMV was unable to infect its normal hosts (Atabekov et al., 1970a). On balance, it appears that viral coat proteins play little if any positive part in cell recognition. This view is supported by the fact that viral uncoating following inoculation appears not to be host-specific (see below). Surface recognition proteins may be of little use to a virus in the process of infecting a plant because of the requirement that they enter cells through w o u n d s on the plant surface. Leaf-hair cells have been infected with various viruses by introducing the virus directly into the cell with a microneedle (see Chapter 9, Section II.B), thus presumably bypassing any virus-cell surface interaction. Similarly, intact virus particles or other infectious materials are able to pass from cell to cell through the plasmodesmata and cause infection while remain-
ing within the plasma membrane (see Chapter 9, Section II). From experiments using model membranes and plant viral coat proteins in various aggregation states, Datema ef al. (1987) concluded that hydrophobic lipid-coat protein interactions do not occur. Experiments on the binding and uptake of CCMV by cowpea protoplasts gave no evidence to support an endocytic uptake of virus mediated by specific receptors (Roenhorst et al., 1988). As discussed in Chapter 8, stages in the replication of many viruses take place in association with particular cell organdies. Recognition of a particular organdie or site within the cell by a virus (or by some subviral c o m p o n e n t or product) must be a frequent occurrence. Plant viruses may have evolved a recognition system basically different from that of viruses that normally encounter and recognize their host cells in a liquid m e d i u m or at a plasma membrane surface. b. Lack of specificity in the uncoating process
Various lines of evidence suggest that there is little or no host specificity in the uncoating process. Thus, for TMV (Kiho et al., 1972) and TYMV (Matthews and Witz, 1985), virus was uncoated as readily in non-hosts as in host species. Gallie et al. (1987c) showed that, when mRNA coding for the enzyme chloramphenicol acetyltransferase was packaged into rods with TMV coat protein and inoculated to protoplasts or plant leaves, the RNA became uncoated. Francki et al. (1986b) showed that, when VTMoV and the viroid PSTVd were inoculated together on to a susceptible host, the viroid was incorporated into virus particles. When such virus was inoculated to a species that was a host only for the viroid, viroid infection occurred, indicating that the nucleic acids had been uncoated. However, these experiments did not eliminate the possibility that VTMoV replicated only in the initially infected cells of the presumed nonhost species. 2. Replication Following inoculation of TMV to plant species considered not to be hosts for the virus, viral
v. THE HOST ~AXC~EOF VIRUSES RNA has been found in polyribosomes (Kiho et al., 1972). Furthermore, TMV particles uncoat and express their RNA in Xenopus oocytes (P.C. Turner et al., 1987b). However, there is some evidence that specific viral genes involved in replication may also be involved as host range determinants. Various host proteins have been found in replication complexes of several viruses (see Chapter 8, Sections IV.E.5 and IV.H.5). Mouches et al. (1984) and Candresse et al. (1986) obtained evidence that the replicase of TYMV consists of a 115-kDa viral-coded polypeptide and possibly a 45-kDa subunit of plant origin. The 115-kDa polypeptide from different tymoviruses showed great serological variability. They suggested that this variability might be a consequence of the need for the viral-coded polypeptide specifically to recognize the host subunit in different host species in order to form a functional replicase. Thus, the virus-coded peptide might be directly involved in defining host specificity. Most strains of CaMV infect only Brassicaceae. The few strains that also infect some solanaceous species can be divided into three types according to which species they infect. To determine which CaMV genes determine this host range, Schoelz et al. (1986) made recombinant viruses by exchanging DNA segments b e t w e e n the cloned strains. The resulting hybrids were then tested on the relevant solanaceous species. These experiments indicated that the first half of gene VI, which codes for the inclusion body protein (see Chapter 6, Section IV.A.I.b), determines whether the virus can systemically infect Datura stramonium and Nicotiana bigelovii. Squash leaf curl disease in the United States is caused by two distinct but highly homologous bipartite geminiviruses. The host range of one virus is a subset of the other (Lazarowitz, 1991). Analysis of agroinfected leaf disks indicated that virus replication was involved in the host restriction of one of the viruses. Replication of the restricted virus was rescued in trans by coinfection with the non-restricted virus. Sequence analysis revealed that the restricted virus had a 13-nucleotide deletion in the c o m m o n region. In other respects the sequences of the two common regions were
71
almost identical. Lazarowitz (1991) suggested that this deletion may have been involved in the host range restriction. As discussed in Chapter 7 (Section V.B.9), some viruses synthesize 'read-through' proteins. Successful read through depends on the presence of an appropriate suppressor tRNA in the host plant. The presence of such a tRNA may, in principle at least, be a factor determining host range for viruses depending on the read-through process. 3. Cell-to-cell movement Two lines of evidence strongly support the view that possession of a compatible and functional cell-to-cell m o v e m e n t protein is one of the factors determining whether a particular virus can give rise to readily detectable virus replication in a given host species or cultivar. First, the experiments s u m m a r i z e d in Chapter 9 show that many viruses contain a gene coding for a cell-to-cell m o v e m e n t protein. Viruses cannot usually infect leaves through the cut end of petioles or stems. However, Taliansky e, al. (1982a,b,c) showed that, if the upper leaves were already infected with a helper virus, a virus in solution could pass into the upper leaves and infect them. This happened with mutants defective in the transport function, and also with viruses that were not normally able to infect the species of plant involved. For example, they showed that BMV can be transported from conducting tissues in tomato plants preinfected with TMV. Tomato is not normally a host for BMV. The same occurred with TMV introduced into wheat plants already infected with BSMV. A similar situation appears to exist in field infections in sweet potato, where prior infection with SPFMV allows CMV to infect this host (Cohen e, al., 1988). Thus, when the resistance of an apparent non-host species is due to a block in the transport function, this block can be overcome by a preinfection with a virus that has a transport protein compatible with the particular host. This appears to be a fairly general phenomenon. However, in interpreting these phenomena involving the joint infection with two or more viruses, the possibility that one of the viruses can suppress the plant
72
3 DISEASE SYMPTOMS AND H O S T R A N G E
defense system must be borne in mind (see Chapter 10, Section IV.H, for suppression of host defense systems). The second line of evidence concerning the importance of cell-to-cell m o v e m e n t in host range is that a number of viruses have been shown to be able to infect a n d replicate in protoplasts derived from species in which they show no macroscopically detectable sign of infection following mechanical inoculation of intact leaves. For example, Beier et al. (1977) attempted to infect 1031 lines of cowpea by mechanical inoculation with CPMV. Sixty-five lines were defined as operationally immune because no disease developed and no virus could be recovered. Protoplasts could be prepared from 55 of the 'immune' lines. Fifty-four of these could be infected with CPMV. Similar results have been obtained with other hosts and viruses. Sulzinski and Zaitlin (1982) mechanically inoculated cowpea and cotton plants with TMV. They isolated protoplasts at intervals and determined the proportion of infected cells at various times using fluorescent antibody. Only about i in 5-15 • 1 0 4 protoplasts were infected and this number remained unchanged for at least 11 days. These results show that subliminally infected plants can support virus replication in individual cells, but that the virus cannot move out of the initially infected cells. Viruses that can be induced to invade and replicate systemically by a helper virus, as described in the first part of this section, can presumably infect the host concerned on their own in a similar subliminal fashion. This idea is s u p p o r t e d by the observation that the Geminivirus BGMV, normally confined to phloem tissue, moved to cells of many types in bean leaves doubly infected with BGMV and TMV (Carr and Kim, 1983). However, not all sequence differences in the movement proteins can be involved in defining host range. Solis and Garcia-Arenal (1990) showed that there were major differences in the C-terminal region of the m o v e m e n t proteins of TMV, another Tobamovirus, TMGMV, and TRV, and yet all of these viruses share the same range of natural hosts in the Solanaceae. A functional cell-to-cell m o v e m e n t protein
may not be the only requirement for systemic movement of a virus. CCMV infects dicotyledons (cowpea) and the related BMV replicates in monocotyledons (barley). Both viruses replicate in protoplasts of both hosts. If the movement protein was the only factor controlling systemic m o v e m e n t , transferring RNA3 between the two viruses should switch the host range. This does not occur. Therefore, systemic infection must also involve factors coded for by RNAs 1 or 2 or both (Allison et al., 1988). Since BMV must be encapsulated for systemic movement, it is probable that all four genes are involved in systemic m o v e m e n t and host specificity. 4. Stimulation of host-cell defenses In some host-virus combinations, the virus stimulates host-cell defense mechanisms that may be a factor limiting virus replication and movement to cells near the first infected cell, giving rise to local lesions, without subsequent systemic spread. This p h e n o m e n o n , which makes the plant resistant to the virus from a practical point of view, is discussed in Chapter 10 (Section III). Other aspects of host defense systems are discussed in other sections in Chapter 10. 5. Host genes affecting host range Host genes must be involved in all interactions between a virus and the plant cell following inoculation, whatever the outcome may be. The effects of certain host genes that limit virus infection to local lesions have been studied in some detail. These are discussed in Chapter 10 (Section III). 6. Summary Concerning the molecular basis for viral host ranges, present evidence suggests the following. 1. There is usually no receptor recognition system for the host plasma membrane and little host specificity in the initial uncoating process. 2. In some plant-virus combinations, molecular mechanisms may exist for blocking virus at some stage in the replication process in the cells where virus first gained entry.
VI. DISCUSSION AN[-)SUMMARY
This forms the basis for true i m m u n i t y to infection. 3. In m a n y host-virus combinations, replication can occur in the first infected cells but movement out of these cells is not possible because of a mismatch between a viral m o v e m e n t protein and some host cell structure. This situation gives rise to subliminal infection and, for practical purposes, a resistant species or cultivar. 4. In other v i r u s - h o s t combinations, m o v e m e n t may be limited by host responses to cells in the vicinity of the initially infected cell, giving rise to a 'local lesion host', which from a practical point of view is 'field resistant'. Using these ideas, the different kinds of host response to inoculation with a virus are defined in Table 10.1.
VI. DISC U SSIO N A N D SUMMARY The ability to be transferred to healthy host plant individuals is crucial for the survival of all plant viruses. Viruses cannot, on their own, penetrate the u n d a m a g e d plant surface. For this reason, each kind of virus has evolved ways to bypass or overcome this barrier. Viruses that are transmitted from generation to generation with high frequency via pollen and seed or in plant parts in vegetatively reproducing species can avoid the need to penetrate the plant surface. Many groups of plant viruses are transmitted by either invertebrate or fungal vectors, which penetrate the plant surface during the process of feeding or infection, carrying infectious virus into the plant cells at the same time. These are described in Chapter 11. Mechanical t r a n s m i s s i o n is a process whereby small w o u n d s are m a d e on the plant surface in the presence of infectious virus. For m a n y g r o u p s of viruses, if conditions are favorable, infection follows. Mechanical inoculation is a very important procedure in experimental virology. In the field, it is not considered to be of great importance except for a few g r o u p s such as t o b a m o v i r u s e s and potexviruses, which appear to have no other means of transmission. However, there are several viruses for which a biological vector is not
73
k n o w n and which may also be transmitted in this manner. These viruses are, however, well a d a p t e d to this form of transmission as they are relatively stable and occur in high concentration in infected leaves. These features make transmission possible w h e n leaves of neighboring plants abrade one another or roots come in contact. It is probable that successful entry of a single virus particle into a cell is sufficient to infect a plant but, in the normal course of events, m a n y m a y enter a single cell. Following initial replication in the first infected cell, virus moves to neighboring cells via the plasmodesmata, this being a relatively slow process. Virus then enters the vascular tissue, usually the phloem, m o v e m e n t in this tissue being m u c h more rapid. Most viruses code for specific gene products that are necessary for the virus to move out of the initially infected cell and t h r o u g h the plant. The final distribution of virus t h r o u g h the plant m a y be quite uneven. In some host-virus combinations virus m o v e m e n t is limited to local lesions. In others, some leaves m a y escape infection, while in mosaic diseases dark green islands of tissue m a y contain little or no virus. Some viruses are confined to certain tissues. For example, luteoviruses are confined mainly to the phloem. Some viruses penetrate to the d i v i d i n g cells of apical meristems. Others appear not to do so. There is clear evidence that both virus particles and infectious viral RNA can m o v e from cell to cell, but this may not be the form in which all viruses move. Evidence suggests that TMV m a y move in the form of nucleoprotein complexes m a d e up from viral RNA, viruscoded proteins, and some host components. Virus-coded m o v e m e n t proteins m a y exert their effect by altering the properties of the p l a s m o d e s m a t a to facilitate viral passage, but this has not been proven (virus m o v e m e n t is discussed in Chapter 9). The molecular basis for the host range of viruses is not fully understood. Most plant viruses do not have a m e c h a n i s m by which intact particles can recognize a host cell that is suitable for replication. Uncoating of the viral nucleic acid appears to occur in hosts and
74
3 DISEASE SYMPTOMS AND H O S T R A N G E
non-hosts alike. No particular step in the replication of a virus has yet been implicated in limiting host range. However, there is good evidence that in 'non-hosts' of some viruses the virus can replicate in the first infected cell but cannot move t o neighboring cells. This is because the viral-coded cell-to-cell movement protein does not function in the particular plant. For practical purposes, such a plant is
resistant to the virus in question. As described in Chapter 10 (Section IV), plants have a general defense system against 'foreign' nucleic acids that can be suppressed by viruses that infect that host. It is likely that the interaction between the host defense system and the viral defense suppression system is the major determinant of both host range and s y m p t o m expression.
C
H
A
P
T
E
R
Purification and Composition of Plant Viruses I. I N T R O D U C T I O N
a preparation contained absolutely no low- or high-molecular-weight host constituents (which is most unlikely), there are other factors to be considered:
To study the basic properties of virus particles it is usually necessary to separate them from the cellular components. Since the classic studies of Stanley (1935), Bawden and Pirie (1936,1937) and others in the 1930s, a great deal of effort has been put into devising methods for the isolation and purification of plant viruses. To ascertain the basic properties of a virus it is essential to be able to obtain purified preparations that still retain infectivity. It is not surprising that the first viruses to be isolated and studied effectively (TMV, PVX and TBSV) were among those that are fairly stable and occur in relatively high concentration in the host plant. Today, interest has extended to a range of viruses that vary widely in concentration in the host and in their stability towards various physical and chemical procedures. In this chapter, I will describe factors to be considered in virus purification and then the basic components of virus particles.
1. Most preparations almost certainly consist of a mixture of infective and non-infective virus particles. The latter will probably have one or more breaks in their nucleic acid molecules, most of which will have occurred at different places in different particles. 2. Most virus preparations almost certainly consist of a mixture of mutants even though the parent strain greatly predominates. Such mutants will differ in the base sequence of their RNA or DNA in at least one place. If the mutation is in the cistron specifying the coat protein, then this may also differ from the parent strain. 3. Purified preparations of m a n y viruses can be shown to contain one or more classes of incomplete, non-infective particles. 4. The charged groups on viral proteins and nucleic acids will have ions associated with them. The inorganic and small organic cations found in the purified virus preparation will depend very much on the nature of the buffers and other chemicals used during isolation. 5. Some of the larger viruses appear to cover a range of particle sizes having infectivity. 6. A variable proportion of the virus particles may be altered in some w a y during isolation. Enzymes may attack the coat protein. For example, extracts of bean (Phaseolus vulgaris) contain a carboxypeptidase-like enzyme that removes the terminal threonine from TMV (Rees and Short, 1965). Proteolysis at defined sites may give rise, during isolation of the
II. I S O L A T I O N There are no generally applicable rules for virus purification. Procedures that are effective for one virus may not work with another apparently similar virus. Even different strains of the same virus may require different procedures for effective isolation. A great deal has been written about purity and h o m o g e n e i t y as they a p p l y to plant viruses. In a chemical sense, there is no such thing as a pure plant virus preparation. Even if Virus acronyms are given in Appendix 1.
75
76
4 P U R I F I C A T I O N A N D C O M P O S I T I O N OF P L A N T VIRUSES
virus, to a series of coat protein molecules of less than full size (e.g. with SNMoV: Chu and Francki, 1983; and BaYMV: Ehlers and Paul, 1986). Coat proteins may undergo chemical modification when leaf phenols are oxidized (Pierpoint et al., 1977). More complex viruses, such as the reoviruses, may lose part of their structure during isolation (e.g. Hatta and Francki, 1977). Thus, for plant viruses, purity and homogeneity are operational terms defined by the virus and the methods used. A virus preparation is pure for a particular purpose if the impurities, or variations in the particles present, do not affect the particular properties being studied or can be taken account of in the experiment. Effective isolation procedures have now been developed for many plant viruses. Rather than describe these in detail, I shall consider, in general terms, the problems involved in virus isolation. Detailed protocols for isolation procedures for a number of viruses are given in numerous publications, including Hull (1985), Walkey (1991), Stace-Smith and Martin (1993), Dijkstra and de Jager (1998), and in various chapters in Foster and Taylor (1998).
A. C h o i c e of p l a n t m a t e r i a l 1. Assay host During the development of an isolation procedure, it is useful, but not always possible, to be able to assay fractions for infectivity. Of course, this is best done with a local lesion host. Great accuracy usually is not necessary in the preliminary assays, but reliability and rapid development of lesions are a great advantage. If no local lesion host is available, then assays must be done using a systemic host. Assays by the injection of insect vectors sometimes have been used where mechanical transmission is impossible. Electron microscopy or, if antisera or probes are available, dot blot ELISA or dot blot nucleic acid hybridization, can be used to follow the progress of purification. 2. Propagation host The choice of host plant for propagating a virus may be of critical importance for its successful
isolation. Various points have to be considered in the choice of a propagating host: 1. The host plant should be easy to grow, preferably from seed. However, care should be taken that there is no seed-transmitted virus such as SoMV, which is highly transmitted in seed of Chenopodium spp. 2. The virus should reach a high concentration in the host. 3. The host should not contain high amounts of substances such as phenolic materials, organic acids, mucilages and gums, certain proteins, and enzymes, particularly ribonucleases that can inhibit or irreversibly precipitate the virus. For example, many stone fruit viruses were very difficult to isolate from their natural hosts, since most members of the Rosaceae contain high concentrations of tannins in their leaves. Discovery of alternative non-rosaceaous hosts, for example cucumber (Cucumis sativus L.), has allowed the isolation of several such viruses. 4. The host plant constituents should be easily separable from the virus during purification. For instance, certain legumes (e.g. Canavalia ensiformis) contain large amounts of Rubisco (fraction 1 protein), often in an aggregated form that can co-purify with rod-shaped viruses. In practice, species of Chenopodium, Cucumis, Nicotiana, Petunia, Phaseolus and Vigna have been found suitable for the propagation of a large number of viruses. The plant species used, the conditions under which it is grown, and the time at which it is harvested should be chosen to maximize the starting concentration of infectious virus. For many viruses, concentration rises to a peak after a few days or weeks and then falls quite rapidly (Fig. 4.1). Sometimes the distribution of virus within the plant is so uneven that it is worthwhile to dissect out and use only those tissues showing prominent symptoms. Viruses frequently occur in much lower concentration in the midrib than in the lamina of the leaf. If the midrib and petiole are large, it may pay to discard them. In special situations, dissection of tissue is almost essential, for example with WTV and other
II.
ISOLATION
77
32
,_
4
=
0
-6 .~_
,,'" -
-
2
1
1
1
4
.5
8-tO
1
i2
Gr~ 0
B 32 m
15A
16 8 4
0
1
2
1
4
1
6
8
I0
18 t'X
",--
12 14 16 18 20 22 24 26 28 30 32 34 36 38 -
= Leof
number
Fig. 4.1 Change in concentration of PVA with age in leaves of a single Samsun tobacco plant inoculated on leaf number 1. Virus concentration was measured serologically. From week 15 onward, older leaves had died. Numbers beside graphs show weeks after infection. From Bartels (1954), with kind permission of the copyright holders, 9 Blackwell Science Ltd.
viruses causing t u m o r s where the virus is f o u n d associated w i t h the t u m o r tissue. Another reason for harvesting only certain parts of the infected plant may be to avoid high concentrations of inhibitory substances or materials that adsorb to the virus and are later difficult to remove. Such materials frequently occur in lower concentration in new y o u n g growth. In certain hosts, virus can only be isolated from such tissue. Similarly, root tissue may sometimes provide more favorable starting material than leaves (e.g. Ford, 1973), although virus concentration is usually lower in roots. For some viruses, flowers m a y provide a suitable source of virus at high concentration The possibility that the host used to culture a virus m a y already harbor another virus or become infected with one m u s t always be borne in mind. Contamination of greenhousegrown plants with u n w a n t e d viruses is not at all u n c o m m o n . Strains of TMV, PVX and TNV may be particularly prevalent, especially in greenhouses that have been used for virus w o r k for some time. It is not necessarily sufficient to use a host that is only a local lesion host for such contaminating viruses (e.g.N. glutinosa for TMV). Very small a m o u n t s of the contaminating virus m a y become differentially concentrated d u r i n g isolation of a second virus.
find themselves in an e n v i r o n m e n t that is abnormal. Thus, it is often necessary to use an artificial extraction m e d i u m designed to preserve the virus particles in an infectious, intact and u n a g g r e g a t e d state d u r i n g the various stages of isolation. The conditions that favor stability of purified virus preparations may be different from those n e e d e d in crude extracts or partially purified preparations (e.g. Brakke, 1963). Moreover, different factors m a y interact strongly in the extent to which they affect virus stability. The main factors to be considered in developing a suitable m e d i u m are as follows.
3. Extraction medium
b. Metal ions and ionic strength
Once infected plant cells are broken, and the contents released and mixed, the virus particles
Some viruses require the presence of divalent metal ions (Ca 2+ or Mg 2+) for the preservation
a. pH and buffer system M a n y viruses are stable over a rather narrow p H range, and the extract must be maintained within this range. As most viruses have an acid isoelectric point, buffer solutions with p H values in the range of 7-8 should prevent them from precipitating. However, the structural integrity of some viruses (e.g. bromoviruses) is lost on exposure to p H s in this range. The use of a buffer of p H 5 will precipitate m a n y normal plant proteins and is a d v a n t a g e o u s if the virus itself is not precipitated. Buffers comm o n l y used for virus extraction include borate, citrate, phosphate and Tris.
78
4 P U R I F I C A T I O N AND C O M P O S I T I O N OF PLANT VIRUSES
of infectivity and even for the maintenance of structural integrity. Ionic strength may be important. Some viruses fall apart in media of ionic strength below about 0.2 M, while others are unstable in media above this molarity. AMV particles may be precipitated by Mg 2+ concentrations above I mM and degraded by concentrations above 0.1 M (Hull and Johnson, 1968). For some viruses, NaEDTA may be included to minimize aggregation by divalent metals. On the other hand, NaEDTA disrupts certain viruses. Decisions on the pH, presence of chelating agent and ionic strength should take account of the factors stabilizing the virus particle (see Chapter 5). c. Reducing agents and substances protecting against phenolic compounds Reducing agents such as ascorbic acid, cysteine hydrochloride, 2-mercaptoethanol, sodium sulfite, or sodium thioglycollate are frequently added to extraction media. Dithiothreitol (Cleland's reagent) is a useful reducing agent as it has little tendency to be oxidized by air. These materials assist in preservation of viruses that readily lose infectivity through association with products resulting from oxidation of plant extracts and also may reduce adsorption of host constituents to the virus. Phenolic materials especially may cause serious difficulties in the isolation and preservation of viruses. Several methods have been used more or less successfully to minimize the effects of phenols on plant viruses during isolation: 1. Cysteine or sodium sulfite added to the extraction medium both probably act by inhibiting the phenol oxidase and by combining with the quinone (Pierpoint, 1966). 2. Polyphenoloxidase is a copper-containing enzyme. Two chelating agents with specificity for copper, diethyldithiocarbamate, and potassium ethyl xanthate, have been used to obtain infectious preparations of several viruses (e.g. PNRSV: Barnett and Fulton, 1971); the former is also a reducing agent. 3. Materials that compete with the virus for phenols have sometimes been used. For example, Brunt and Kenten (1963) used
various soluble proteins and hide powder to obtain infective preparations of CSSV from cocoa leaves. Synthetic polymers containing the amide link required for complex formation with tannins have been used effectively to bind these materials. The most important of these is polyvinyl pyrrolidone (PVP).
d. Additives that remove plant proteins and ribosomes Many viruses lose infectivity fairly rapidly in vitro. One reason for this loss may be the presence of leaf ribonucleases in extracts or partly purified preparations. Dunn and Hitchborn (1965) made a careful study of the use of magnesium bentonite as an additive in the isolation of various viruses. They found that, under appropriate conditions, contamination of the final virus product with nucleases was reduced or eliminated. In addition, ribosomes, 19S protein, and green particulate material from fragmented chloroplasts were readily adsorbed by bentonite, provided Mg 2. concentration was 10-3M o r greater. However, variation occurs in the activity of different batches of bentonite, and the material must be used with caution as some viruses are degraded in its presence. Charcoal may be used to adsorb and remove host materials, particularly pigments. Subsequent filtration to remove charcoal may lead to substantial losses of virus adsorbed in the filter cake. NaEDTA at 0.01 M in pH 7.4 buffer will cause the disruption of most ribosomes, preventing their co-sedimentation with the virus. This substance can be used only for viruses that do not require divalent metal ions for stability. e. Enzymes
Enzymes have been added to the initial extract for various purposes. Thus, Adomako et al. (1983) used pectinase to degrade mucilage in extracted sap of cocoa leaves prior to precipitation of CSSV. Improved yield of a virus limited to phloem tissue was obtained when fibrous residues were incubated with Driselase and other enzymes (Singh et al., 1984). This
II. ISOLATION
material contains pectinase and cellulose and presumably aids in the release of virus that would otherwise remain in the fibre fraction. The enzymes also digest materials that would otherwise co-precipitate with the virus. Jones et al. (1991) found that digestion of rice tissue with Celluclast gave more reliable yields of the two rice tungro viruses than did digestion with Driselase. Treatment with trypsin at an optimum concentration can markedly improve the purification of TuMV (Thompson et al., 1988).
f. Detergents and other additives Non-ionic detergents such as Triton X-IO0 or Tween 80 are often used in the initial extraction m e d i u m to assist in release of virus particles from insoluble cell components and to dissociate cellular membranes that may contaminate or occlude virus particles. However, detergents should not be used with enveloped viruses. The particles of some viruses, such as caulimoviruses, may be contained within inclusion bodies that have to be disrupted. Hull et al. (1976) showed that caulimovirus inclusion bodies were disrupted on treatment with urea and Triton X-100. 4. Extraction procedure Freezing of the plant tissue, say by liquid N2, before extraction enables disruption of vascular tissues releasing phloem-limited viruses and facilitates subsequent removal of host materials. For some viruses, however, freezing may have a deleterious effect. A variety of procedures are used to crush or homogenize the virus-infected tissue. These include (1) a pestle and mortar, which are useful for small-scale preparations, (2) various batch-type food blenders and juice extractors, which are useful on an intermediate scale, and (3) roller mills, colloid mills and commercial meat mincers, which can cope with kilograms of tissue. For long, fragile, rod-shaped viruses, grinding in a pestle and mortar may be the safest procedure to minimize damage. For instance, BYV particles were broken on using a blender (Bar-Joseph and Hull, 1974). The addition of acid-washed sand greatly improves
79
the efficiency of extraction. If an extraction m e d i u m is used, it is often necessary to ensure immediate contact of broken cells with the medium. The crushed tissue is usually expressed through muslin or cheesecloth. 5. Preliminary isolation of the virus a. Clarification of the extract
In the crude extract, the virus is mixed with a variety of cell constituents that lie in the same broad size range as the virus and that may have properties that are similar in some respects. These particles include ribosomes, Rubisco (fraction I) protein from chloroplasts, which has a tendency to aggregate, phytoferritin, membrane fragments, and fragments of broken chloroplasts. Also present are unbroken cells, cell wall fragments, all the smaller soluble proteins of the cell, and low-molecular-weight solutes. The first step in virus isolation is usually designed to remove as much of the macromolecular host material as possible, leaving the virus in solution. The extraction m e d i u m m a y be designed to precipitate ribosomes, or to disintegrate them. The extract may be subject to some treatment such as heating to 50-60~ for a few minutes, freezing and thawing, acidification to a p H of less than 5 or the addition of K2HPO 4, to coagulate much host material. The treatment has to be chosen on a case-by-case basis as it m a y d a m a g e the virus. For some viruses, organic solvents such as chloroform give very effective precipitation of host components. For others, the extract can be shaken with n-butanol-chloroform, which denatures much host material. The treated extract is then subjected to centrifugation at fairly low speed (e.g. 10-20 minutes at 5000-10 000 g). This treatment sediments cell debris and coagulated host material. With the butanol-chloroform system, centrifugation separates the two phases, leaving virus in the aqueous phase and m u c h denatured protein at the interface. It should be noted that although some viruses can withstand the butanol-chloroform treatment, quite severe losses m a y occur with others. Chloroform alone gives a milder treatment than does a butanol-chloroform mixture. Of
80
4 PURIFICATION AND COMPOSITION OF PLANT VIRUSES
course, solvents and detergents cannot be used for viruses that have a membrane. It must be remembered that there are safety considerations to be taken into account in using large volumes of these organic solvents. For many viruses, it pays to carry out the isolation procedure at 4~ and as fast as possible once the leaves have been extracted. On the other hand, some viruses occur in membranebound structures or other structures within the cell. It may take time for the virus to be released from these after the leaf extract is made. A lowspeed centrifugation soon after extraction may then result in much virus being lost in the first pellet but, on the other hand, this can be used to concentrate the virus.
b. Concentration of the virus and removal of low-molecular-weight materials i. High-speed sedimentation
Centrifugation at high speed for a sufficient time will sediment the virus. Provided the particular virus is not denatured by the sedimentation, it can be brought back into solution in an active form. This is a very useful step, as it serves the double purpose of concentrating the virus and leaving behind low-molecular-weight materials. However, high-speed sedimentation is a physically severe process that may damage some particles (e.g. some reoviruses; Long et al., 1976). Following high-speed sedimentation, some viruses remain as characteristic aggregates when the pellets are redissolved. The virus particles in these aggregates may be quite firmly bound together (Tremaine et al., 1976). If host membranes are involved in the binding, a non-ionic detergent may help to release virus particles. Sedimentation of viruses occurring in very low concentration may result in very poor recoveries. The major process causing losses appears to be the dissolving and redistribution of the small pellet of virus as the rotor comes to rest (McNaughton and Matthews, 1971). Redissolving particles from the surface of the pellet can lead to preferential losses of more slowly sedimenting components (Matthews, 1981).
ii. Precipitation with polyethylene glycol
Hebert (1963) showed that certain plant viruses could be preferentially precipitated in a singlephase polyethylene glycol (PEG) system, although some host DNA may also be precipitated. Since that time, precipitation with PEG has become one of the commonest procedures used in virus isolation. The exact conditions for precipitation depend on pH, ionic strength, and concentration of macromolecules. Its application to the isolation of any particular virus is empirical, although attempts have been made to develop a theory for the procedure (Juckes, 1971). PEG precipitation is applicable to many viruses, even fragile ones. For example, the procedure gave a good yield of intact particles of CTV, a fragile rod-shaped virus (R. F. Lee et aI., 1987). It has the advantage that expensive ultracentrifuges are not required, although differential centrifugation is often used as a second step in purification procedures. iii. Density-gradient centrifugation
Many viruses, particularly rod-shaped ones, may form pellets that are very difficult to resuspend. Density gradient centrifugation offers the possibility of concentrating such viruses without pelleting. A density gradient is illustrated in Fig. 4.2. The following modification of the density gradient procedure may sometimes be used even with angle rotors for initial concentration of a virus without pelleting at the bottom of the tube. A cushion of a few millilitres of dense sucrose is placed in the bottom of the tube, this being overlaid with a column of lowdensity sucrose about 2 cm deep, and the rest of the tube is filled with clarified virus extract. Under appropriate conditions, virus can be collected from the region of the interface between the two sucrose layers. In a similar application, a cushion of medium-density sucrose is placed in the bottom of the tube before centrifugation; this prevents contamination of pellets with chlorophyll material. Density-gradient centrifugation is used in the isolation procedure for many viruses. iv. Salt precipitation or crystallization
Salt precipitation was commonly employed before high-speed centrifuges became generally
Ii. ISOLATION
81
6. Further purification of the virus preparation
A
TMV
B
83 S
Fig. 4.2 Density-gradient centrifugation for the assay of viruses in crude extracts. An extract of 30 mg of epidermis (A) or the underlying tissue (B) from a tobacco leaf inoculated with TMV was sedimented in a 10-40% sucrose gradient at 3 5 0 0 0 r p m for 2 h . Absorbancy at 254 nm t h r o u g h the g r a d i e n t was measured with an automatic scanning device. Note absence of detectable 68S ribosomes in the epidermal extracts.
available. It is still a valuable method for viruses that are not inactivated by strong salt solutions. A m m o n i u m sulfate at concentrations up to about one-third saturation is most commonly used, although many other salts will precipitate viruses or give crystalline preparations. After standing for some hours or days, the virus is centrifuged d o w n at low speed and redissolved in a small volume of a suitable medium. v. Precipitation at the isoelectric point
Many proteins have low solubility at or near their isoelectric points. Isoelectric precipitation can be used for viruses that are stable under the conditions involved. The precipitate is collected by centrifugation or filtration and is redissolved in a suitable medium. vi. Dialysis
Dialysis through cellulose membranes can be used to remove low-molecular-weight materials from an initial extract and to change the m e d i u m . It is more usually e m p l o y e d to remove salt following salt precipitation or crystallization, or following density gradient fractionation in salt or sucrose solutions, as discussed in a subsequent section.
Virus preparations taken through one step of purification and concentration will still contain some low- and high-molecular-weight host materials. Further purification steps can remove more of these. The procedure to be used will depend very much on the stability of the virus, the scale of the preparation, and the purpose for which the preparation is required. Sometimes highly purified preparations can be obtained by repeated application of the same procedure. For example, a preparation may be subjected to repeated crystallization or precipitation from a m m o n i u m sulfate, or may be given several cycles of high- and low-speed sedimentation. The latter procedure leads to the preferential removal of host macromolecules because they remain insoluble w h e n the pellets from a high-speed sedimentation are resuspended. Losses of virus often occur with this procedure, either because some, or all, of the virus itself also becomes insoluble or because insufficient time is allowed for the virus to resuspend. This is a particular difficulty with some rod-shaped viruses. Losses may also occur from virus resuspending before the supernatant fluid is removed. Such losses may be severe with very small pellets, where the surface-to-volume ratio is high. Generally speaking, during an isolation it is useful to apply at least two procedures that depend on different properties of the virus. This is likely to be more effective in removing host constituents than repeated application of the same procedure.
a. Density-gradient centrifugation One of the most useful procedures for further purification and for assay, particularly of less stable viruses, is density-gradient centrifugation which was developed by Brakke (1951, 1960). This technique has proved to be highly versatile and has been widely used in the fields of virology and molecular biology. It has largely replaced the analytical ultracentrifuge in studies on viruses. A centrifuge tube is partially filled with a solution having a decreasing density from the bottom to the top of the tube. For plant viruses, sucrose is commonly used to
82
4 PURIFICATION AND COMPOSITION OF PLANT VIRUSES
form the gradient, and the virus solution is layered on top of the gradient. With gradients formed with cesium salts, the virus particles may be distributed throughout the solution at the start of the sedimentation or they may be layered on top of the density gradient. Brakke (1960) defined three ways in which density gradients may be used. Isopycnic gradient centrifugation occurs when centrifugation continues until all the particles in the gradient have reached a position where their density is equal to that of the medium. This type of centrifugation separates different particles based on their different densities. Sucrose alone may not provide sufficient density for isopycnic banding of many viruses. In rate zonal sedimentation the virus is layered over a preformed gradient before centrifugation. Each kind of particle sediments as a zone or band through the gradient, at a rate dependent on its size, shape and density. The centrifugation is stopped while the particles are still sedimenting. Equilibrium zonal sedimentation is like rate zonal sedimentation except that sedimentation is continued until most of the particles have reached an approximate isopycnic position. The role of the density gradient in these techniques is to prevent convectional stirring and to keep different molecular species in localized zones. The theories of density gradient centrifugation are complex. In practice, this technique is a simple and elegant method that has found widespread use in plant virology. A high-speed preparative ultracentrifuge and appropriate swing-out or angle rotors are required. Following centrifugation, virus bands may be visualized due to their light scattering. The contents of the tube are removed in some suitable way prior to assay. The bottom of the tube can be punctured and the contents allowed to drip into a series of test tubes. Fractionating devices based on upward displacement of the contents of the tube with a dense sucrose solution are available commercially. The UV absorption of the liquid column is measured and recorded, and fractions of various sizes can be collected as required. Fig. 4.2 illustrates the sensitivity of this procedure. Since successive fractions from a gradient can be collected, a variety of procedures can be
used to identify the virus, non-infectious viruslike components and host materials. These include infectivity, UV absorption spectra, and examination in the electron microscope. Using appropriate procedures, very small differences in sedimentation rate can be detected (Matthews and Witz, 1985). With rate zonal sedimentation, if the sedimentation coefficients of some components in a mixture are known, approximate values for other components can be estimated. If antisera are available, serological tests can be applied to the fractions, or antiserum can be mixed with the sample before application to the gradient. Components reacting with the antiserum will disappear from the sedimentation pattern. The various forms of density gradient centrifugation can give some indication of the purity of the preparation. They also allow a correlation between particles and infectivity to be made, and frequently reveal the presence of non-infective virus-like particles or multiparticle viruses. Bands of a single component may spread more widely in the gradient than is apparent from a trace of optical density. Several cycles of density gradient centrifugation may be necessary to obtain components reasonably free of mutual contamination. Work with multicomponent viruses has shown that it may be extremely difficult to obtain one component completely free of another, even by repeated density-gradient fractionation. Density gradients prepared from the non-ionic m e d i u m Nycodenz were used effectively for the further purification of several viruses (Gugerli, 1984). Brakke and Dayly (1965) showed that zone spreading of a major component by non-ideal sedimentation can cause zone spreading of a minor component that the major component overlaps. The way in which zones are removed from the gradient may have a marked effect on the extent of cross-contamination. Strong solutions of salts such as CsC1 are also effective gradient materials for viruses that are sufficiently stable. Successive fractionation in two different gradients may sometimes give useful results. The effective buoyant density and the stability of a virus in strong CsC1 solutions may depend markedly on pH and on other ions present (Matthews, 1974). Viruses
II. ISOLATION
that are unstable in CsC1 may be stable in Cs2SO4 (Hull, 1976b). When a virus preparation is subjected to density-gradient centrifugation in CsC1 or Cs2SO4, multiple bands may be formed. These may be due to the presence of components containing differing amounts of RNA, as is found with TYMV (Keeling et al., 1979). Virus particles containing uniform amounts of nucleic acid may sometimes form multiple bands in CsC1 or other salt gradients because of such factors as differential binding of ions (e.g. Hull, 1977a; Noort et al., 1982). b. Gel filtration Filtration through agar gel or Sephadex may offer a useful step for the further purification of viruses that are unstable to the pelleting involved in high-speed centrifugation. However, such a step will dilute the virus. c. Immunoa~nity columns Monoclonal antiviral antibodies can be bound to a support matrix such as agarose to form a column that will specifically bind the virus from a solution passed through the column. Virus can then be eluted by lowering the pH. Clarified plant sap may destroy the reactivity of such columns (Ronald et al., 1986), while low pH may damage the virus. To avoid such treatment, De Bortoli and Roggero (1985) developed an electrophoretic elution technique. The use of such columns is probably justified only in special circumstances. d. Chromatography Chromatographic procedures have occasionally been used to give an effective purification step for partially purified preparations. McLean and Francki (1967) used a column of calcium phosphate gel in phosphate buffer to purify LNYV. OBDV was purified using cellulose column chromatography (Banttari and Zeyen, 1969), while Smith (1987) used fast protein liquid chromatography to separate the two electrophoretic forms of CPMV.
83
e. Concentration of the virus and removal of lowmolecular.weight materials At various stages in the isolation of a virus, it is necessary to concentrate virus and remove salts or sucrose. For viruses that are stable to pelleting, high-speed centrifugation is commonly employed for the concentration of virus and the reduction of the amount of low-molecularweight material. Ordinary dialysis is used for removal or exchange of salts. For unstable viruses, some other procedures are available. A dry gel powder (e.g. Sephadex) can be added to the preparation, or the preparation in a dialysis bag can be packed in the dry gel powder and placed in the refrigerator for a period of hours, while water and ions are absorbed through the tubing by the Sephadex. Ultrafiltration under pressure through a membrane can be used to concentrate larger volumes and to remove salts. Pervaporation, in which virus solution in a dialysis bag is hung in a draft of air, may lead to loss of virus due to local drying on the walls of the tube, and to the concentration of any salts that are present. 7. Storage of purified viruses Storage of purified preparations of many plant viruses for more than a few days may present a problem. It is often best to avoid long-term storage by using the preparations as soon as possible after they are made. Under the best of conditions, most viruses except TMV lose infectivity on storage at 4~ in solution or as crystalline preparations under a m m o n i u m sulfate. Such storage allows fungi and bacteria to grow and contaminate preparations with extraneous antigens and enzymes. Addition of low concentrations of sodium azide, thymol or EDTA will prevent growth of micro-organisms, but EDTA may affect virus structure. Preparations may be stored in liquid form at low temperature by the addition of an equal volume of glycerol. By careful attention to the additives used in the medium (a suitable buffer plus some protective protein, sugar or polysaccharide), it may be possible to retain infectivity in deep-frozen solutions or as frozen dried powders for fairly long periods (e.g. Fukumoto and Tochihara, 1984). Preparations to be used
84
4 PURIFICATION AND COMPOSITION
OF P L A N T V I R U S E S
for analytical studies on protein or nucleic acid are best stored as frozen solutions, after the components have been separated. In solutions containing more than about 10 m g / m L of viruses such as TYMV, virus particles interact quite strongly and spontaneously degrade, especially if the ionic strength is low. 8. Identification of the infective particles and criteria of purity The best methods for determining the purity of a virus preparation depend on the purpose of the experiments and the conclusions to be drawn. The three most common purposes for which virus preparations are made are: (1) to establish the general properties of a newly isolated virusmthis aspect is discussed in the following sections; (2) to aid in the diagnosis of a particular disease--diagnosis is discussed in Chapter 15; and (3) to prepare specific antisera or diagnostic nucleic acid probes. a. Identification of the characteristic virus particle or particles With newly isolated viruses, or d u r i n g attempts to isolate such viruses, the main objective is to purify the virus sufficiently to identify the infectious particles and characterize them, at least in a preliminary way. The most suitable method is sucrose density-gradient centrifugation of the purified preparation. Fractions taken from the gradient can be assayed for infectivity singly, and in various combinations if a multiparticle system is involved. Ultraviolet absorption spectra can be obtained. Samples can be examined by electron microscopy for characteristic particles. Further a d v a n t a g e s of the sucrose gradient system are that it requires relatively small quantities of virus, and that an approximate estimate of sedimentation coefficients can be obtained. b. Criteria of purity There is much discussion about the purity of virus preparations, but essentially a preparation should be as pure as is necessary for its final use. Thus, if the preparation is to be used for raising an antiserum, it should not contain normal host proteins; and if it is to be used for
making a virus-specific hybridization probe, it should be free from host nucleic acids. On the latter point, if the probe is to be made using a cloning procedure, it should be recognized that cloning is a purification method itself. Many of the criteria for purity that are applied to virus preparations involve the application of the assay m e t h o d s discussed in Chapter 15. Crystallization used to be taken as an important criterion of purity, but it does not distinguish components that co-crystallize with the infectious virus. A nucleoprotein-type spectrum is often put forward as evidence of purity. This is a very poor measure since contamination with large amounts of host nucleoprotein, or polysaccharides, or significant amounts of host proteins can easily go undetected. Sucrose density gradient analysis or sedimentation in the analytical ultracentrifuge will often reveal the presence of non-viral material, provided the contaminants do not sediment together with the virus and provided they are present in high enough concentrations. Sedimentation analysis of purified preparations of some spherical viruses may reveal the presence of dimers or trimers of the virus (Markham, 1962). These might be confused with host contaminants unless other tests are applied. Likewise, preparations of rod-shaped viruses often contain a range of particles of various lengths shorter than the virus. These shorter particles and the virus may aggregate to give a range of sizes that sediment over a broad band, or as two or more discrete peaks. Sedimentation analysis is less useful for detecting the presence of non-viral materials with multiparticle viruses since more of the sedimentation profile is occupied by virus particles. Equilibrium density-gradient sedimentation in CsC1 or Cs2SO4 is a much more sensitive criterion of h o m o g e n e i t y than sedimentation velocity analysis for viruses or viral components that are stable in the strong salt solution. For example, a purified preparation of the B0 noninfectious nucleoprotein from a TYMV preparation appeared as a single component on sedimentation velocity analysis, but after equilibrium sedimentation in CsC1 it was clearly shown to be a mixture of more than one species (Matthews, 1981).
II. ISOLATION
Electron microscopy is often useful in a qualitative way for revealing the presence of extraneous material, provided it is of sufficient size and differs in appearance from the virus itself. Rubisco, phytoferritin, or pieces of host m e m b r a n e structures often can be detected in purified preparations. Small proteins and low-molecular-weight materials would not be detected unless they crystallized on the grids or were present in relatively large amounts. Serological methods, such as immunodiffusion and immunoelectrophoresis, provide very sensitive methods for detecting diffusible host impurities that are antigenic, as do nucleic acid hybridization procedures for detecting nucleic acid contamination. Where the gross composition of a particular virus such as TMV is well established, chemical analysis of major constituents, for example phosphorus nitrogen ration, would show up a nitrogen- or phosphorus-containing i m p u r i t y if it were present in sufficient amount. More sophisticated chemical methods might also be used in certain circumstances, for example end-group amino acid analysis on the protein, which should show only one amino acid. Fingerprinting of the peptides from tryptic digests of well-characterized virus proteins might also be used, but again the sensitivity as a test for purity would not be very high. Radiochemical methods can be used in various ways to test for purity. For example, a healthy plant can be labeled with 32p o r t h o p h o s p h a t e or 3~S sulfate for several days. This material is then mixed with equivalent unlabeled virus-infected tissue before the leaf is extracted. If millicurie (10 7 bq) amounts of radioactivity are used, the sensitivity of this method can be rather depressing. Even so, it is not a complete check on purity with respect to compounds containing these two elements. For example, it is possible that virus infection leads to a kind of contaminant not found in healthy tissues. At present, the only general rule for deciding what criteria of purity to apply is that they should ensure that the virus preparation is adequate for the desired purpose.
85
9. Virus concentration in plants and yields of purified virus a. M e a s u r e m e n t
of yield
There are many difficulties and ambiguities in obtaining estimates of virus concentration in a plant. i. Non-extracted virus
The method used to extract virus from the tissue may frequently leave a variable but quite large proportion of the virus bound to cell debris or retained in vesicles or organdies. Some viruses, particularly those rod-shaped viruses that occur as large fibrous inclusions in the cell, may remain aggregated during the initial stages of isolation. The losses during virus purification can be large and variable and the extent to which various factors lead to virus loss often have not been adequately assessed. ii. Method of measurement
The m e t h o d of m e a s u r i n g virus yield depends upon the purpose for which the virus is being purified. If it is for serology, the criterion is the a m o u n t of virus particles, whether they are infectious or not. If it is for infectivity studies or, say, for making fulllength clones of the genome, the integrity of the nucleic acid is of p a r a m o u n t importance. Infectivity m e a s u r e m e n t s cannot give an absolute estimate of a m o u n t of virus and are influenced by many variables. For a reasonably stable virus, an estimate of such losses can be obtained by adding a small, k n o w n a m o u n t of radioactively labeled virus to the starting material. The proportion of label recovered in the final preparation gives an estimate of virus loss. This method would not account for virus b o u n d within cell structures, however. For viruses occurring in high enough concentrations, sucrose density-gradient analysis on clarified extracts is probably one of the most direct methods for estimating virus concentration in plant extracts. Virus concentration estimates can also be obtained from the ultraviolet spectrum of the preparation (with the caveat noted above on impurities) or by i m m u n o - d o t blots or hybridization dot blots.
86
4 P U R I F I C A T I O N A N D C O M P O S I T I O N OF PLANT VIRUSES
iii. Tissue sampled
As discussed in Chapter 9 (Section II.J), the concentration of virus in different parts of the plant may vary w i d e l y - - e v e n in different parts of a leaf showing mosaic symptoms. iv. Basis for expressing results
Virus concentration or virus yield is usually expressed as weight per unit of fresh weight of tissue or weight per millilitre of extract. However, different tissues and different leaves, even in the same plant, vary widely in their water content u n d e r the same conditions, reflecting, for example, the size of vacuoles or the amount of fibrous material.
b. Reported yields of virus There is no entirely satisfactory answer to many of the preceding problems. Most workers, in reporting the isolation of a virus, if they give estimates of yield at all, express their results as weight of purified virus obtained from a given weight of starting material. This is usually whole leaves or leaf laminae without midribs. Reported yields for the same virus from different laboratories may vary quite widely because of such factors as host species, growing conditions and isolation procedure. Yields vary from several milligrams of virus per gram of fresh weight of tissue for viruses like TMV and TYMV, d o w n to 1-2 ~ g / g for rice tungro viruses (Jones et al., 1991) or even fractions of btg/g for luteoviruses (e.g. Matsubara et al., 1985). 10. Discussion and summary Different viruses vary over a 10 000-fold range in the amount of virus that can be extracted from infected tissue (from 0.4 or less to 4000 ~tg/g fresh weight). They also vary widely in their stability to various physical, chemical and enzymatic agents that may be encountered during isolation and storage. For these reasons, isolation procedures have to be optimized for each virus, or even each strain of a virus. Important factors for the successful isolation of a virus are: (1) choice of host plant species and conditions for propagation that will maximize virus replication and minimize the formation of interfering substances; and (2) an extraction m e d i u m that will protect the virus from inactivation or irreversible aggregation.
Most viruses can be isolated by a combination of two or more of the following procedures: high-speed sedimentation, densitygradient fractionation, precipitation using polyethylene glycol, salt precipitation or crystallization, gel filtration, and dialysis. Density-gradient fractionation of the purified preparation, with combined physical and biological examination of particles in fractions from the gradient, can usually best achieve positive identification of the infectious virus particle, or particles. When a new virus or strain is being isolated, it is essential to back-inoculate the isolated virus to the original host to demonstrate that it, and it alone, is in fact the cause of the original disease. Attention must also be paid to the precise conditions under which purified viruses are stored, as many viruses may lose infectivity and undergo other changes quite rapidly following their isolation. As noted earlier, in a chemical sense there is no such thing as a pure virus preparation. Purity and homogeneity are operational terms. A virus preparation is pure for a particular purpose if the impurities or inhomogeneities in it do not interfere with the objectives of the experiment.
III. C O M P O N E N T S Most plant viruses have their genomes enclosed in either a tube-shaped or an isometric shell made up of many small protein molecules. The majority of viral genomes are RNA (Table 4.1) with most of these being messenger or (+)-sense RNA. In these small geometric viruses, there is usually only one kind of protein molecule but some have two. Some viruses have (-)-sense or dsRNA genomes while others have ssDNA or dsDNA. A few plant viruses have an outer envelope of lipoprotein. Some of the larger viruses contain several viral-coded proteins, including e n z y m e s involved in nucleic acid synthesis. Thus, the major components of all viruses are protein and nucleic acid. A few contain lipids, while some contain small molecules such as polyamines and metal ions. This section deals with the isolation and some properties of these various components.
III. COMPONENTS
87
TABLE 4.1 Viral genomes in host from different kingdoms
Genome nucleic acid
Plants No.
dsDNA ssDNA RT dsRNA (-)ssRNA ( + )ssRNA Total
0 166 31 45 100 635 977
% 0 17.0 3.2 4.6 10.2 65.0
Fungi No. 1 0 0 27 0 1 29
% 3.5 0 0 93.0 0 3.5
Animal No. 606 58 112 383 604 525 2288
% 26.5 2.5 4.9 16.7 26.5 22.9
Bacteria No. % 445 88 0 1 0 57 591
75.4 14.9 0 0.2 0 9.6
ds, double-stranded; ss, single-stranded; RT, viruses that replicate by reverse transcription- they may have either DNA or RNA genomes. The genomes of, at least, one m e m b e r of each virus genus has n o w been sequenced or will be sequenced soon. There is great interest in comparing the nucleotide and derived amino acid sequences of different viruses, from the point of view of virus evolution, virus classification, and the functional roles of viral genes and controlling elements. Reeck et al. (1987) h a v e d r a w n attention to the terminological m u d d l e that has arisen t h r o u g h the imprecise use in the literature of the term ' h o m o l o g y ' in c o m p a r i n g two sequences w h e n in fact 'similarity' is w h a t is meant. Similarity is a fact, a m e a s u r a b l e property of two sequences, to which measures of statistical significance can be a p p l i e d . H o m o l o g y describes a relationship b e t w e e n two things. The w o r d has a precise m e a n i n g in biology of ' h a v i n g a c o m m o n evolutionary origin'. Thus, a sequence is h o m o l o g o u s or it is not. Sequences that are h o m o l o g o u s in an evolutionary sense m a y range from being highly similar to h a v i n g no significant sequence similarity. In the interest of clarity, I will a t t e m p t to maintain this distinction.
A. Nucleic acids For m a n y viruses containing ssRNA, the R N A can act directly as an m R N A u p o n infection. Such molecules are called positive sense or plus- (+)-strand RNAs. The c o m p l e m e n t a r y sequences are called negative sense or m i n u s ( - ) - s t r a n d RNAs. The types of nucleic acid found in plant viruses are listed in Table 4.1. In the following sections, most attention is given to ssRNAs since most plant viruses contain this type of nucleic acid. The RNAs of plant viruses contain adenylic, guanylic, cytidylic
and uridylic a c i d s m a s are found in cellular RNAs. Except for the m e t h y l a t e d guanine in 5' cap structures (Section III.A.3.c), no minor bases have been d e m o n s t r a t e d to be present in plant viral RNAs. The plant reoviruses, containing dsRNA, have regularities in base composition like those found in dsDNA, n a m e l y guanine = cytosine and adenine - uracil. For the viruses containing ssRNA there is no expectation that guanine should equal cytosine or that adenine should equal uracil, and m a n y viruses show a w i d e deviation from any base-pairing rule. In plant viruses containing DNA, deoxyribose replaces ribose, and t h y m i n e is present instead of uracil as with other DNAs. If the D N A is ds, guanine = cytosine and adenine - thymine. A d a m s et al. (1981) in general gives a good account of the properties of nucleic acids. 1. Isolation The usual aim in nucleic acid isolation is to obtain an u n d e g r a d e d and u n d e n a t u r e d product in a state as close as possible to that existing in the virus particle. A variety of physical and chemical agents can be used to remove the protein from viruses and give infectious nucleic acid, p r o v i d e d that (1) the nucleic acid was infectious w i t h i n the intact particle, (2) extremes of p H are avoided, and (3) the nucleic acid is protected against attack by nucleases. M u c h of the pioneering w o r k on viral RNA was carried out w i t h TMV. However, the RNA of this virus while it is w i t h i n the protein coat has exceptional stability. For most other viruses the RNA within the intact virus particle is probably subject, in v a r y i n g degrees, to some degradation, w i t h loss of infectivity. This occurs both
88
4 PURIFICATION AND COMPOSITION OF PLANT VIRUSES
in the intact plant and during isolation and storage of the virus. Best conditions for maintaining intact RNA within the virus particle differ from one virus to another. Isolation of viral RNA often involves the use of phenol. Gierer and Schramm (1956) used a phenol procedure to isolate infectious TMV RNA. Phenol is an effective protein denaturant and nuclease inhibitor. There are many variations of the basic procedure. Particular variations that work well for one virus may not be effective for another. However, phenol has been reported to be unsatisfactory for some viruses (e.g. WSMV: Brakke and van Pelt, 1970; BYDV: Brakke and Rochow, 1974). Destruction of nucleases and effective release of the viral nucleic acid may be achieved by a preliminary incubation with pronase in the presence of 1% SDS. Deoxyribonucleases require Mg 2+ and thus are inhibited by the use of a chelating agent such as EDTA. Phenol extraction is then used to remove pronase and protein digestion products. In most currently used RNA isolation methods, SDS is included during the phenol extraction. If this is done, good nucleic acid preparations can be obtained with most viruses. If the virus preparation is at a high enough concentration the particles can be disrupted by treatment with SDS, and if necessary with heat a n d / o r pronase, and the nucleic acid directly fractionated on an agarose or acrylamide gel. Various procedures can be used to fractionate or to purify further the nucleic acids isolated from viruses or from infected cells. These include sucrose density gradients; equilibrium density-gradient centrifugation in solutions of CsC1 or Cs2SO4; fractionation on columns of hydroxylapatite (Bernardi, 1971); cellulose chromatography (e.g. Jackson et al., 1971); and electrophoresis in polyacrylamide or agarose gels (e.g. Symons, 1978). This last is the method of choice for many purposes since the procedure can be fast and simple, requires very small amounts of nucleic acids, and can detect heterogeneity that would not be seen by other methods (e.g. Fowlks and Young, 1970). In summary, for any particular virus, a protocol for nucleic acid isolation has to be devised that maximizes the required qualities in the final product. These qualities may include yield, purity, structural integrity, infectivity, or
the ability to act with fidelity in in vitro translation systems (e.g. Brisco et al., 1985). 2. Methods for determining size The size of a viral nucleic acid is perhaps the most important property of the virus that can be expressed as a single number. There is some uncertainty for any method of measurement of nucleic acid size except with a full sequence analysis. There is often confusion about the terminology for describing the size of a nucleic acid (or protein) molecule. The term molecular weight (MW) is the sum of the atomic weights of all the atoms in the molecule. Molecular radius (Mr) is the radius of space occupied by a (macro)molecule and is usually expressed as relative to the M r of a h y d r o g e n atom; this is generally regarded as the more correct term for a nucleic acid or protein molecule. The unit dalton (Da) is frequently used for a molecular weight. The availability of sequence data, especially for nucleic acids, enables these terms to be avoided in most cases and units such as bases (b), kilobases (kb) and kilobase pairs (kbp) are much more informative. However, for protein, the unit kilodaltons (kDa) is in widespread use.
a. Sequence analysis The nucleotide sequences for many viral RNAs and DNAs and viroid RNAs have been determined. These allow chemically precise determinations of molecular weight.
b. Gel electrophoresis Electrophoresis in either a polyacrylamide (PAGE) or agarose gel is now a widely used procedure for estimating the M r of viral nucleic acids, by reference to the mobilities of standard nucleic acids of k n o w n M r. Standard 'RNA or DNA ladders' are available commercially. Any secondary structure in ss nucleic acids will affect the mobility and, therefore, the M r estimate. Even formaldehyde treatment may not eliminate all secondary Mructure. If the analysis is performed in a Tris-EDTA buffer at pH 7.5 containing 8 M urea at 60~ more reliable estimates of M r may be obtained (Reijnders et al., 1974). Glyoxal has also been used as an effective denaturant (Murant et at., 1981). Estimates
IiI. COMPONENTS
of M r using gels will be approximate whatever method and markers are used. c. Renaturation kinetics
Where it appears necessary to check whether a virus is monopartite or whether it might have, say, two RNAs or DNAs of different base sequence but of the same size, then renaturation of the nucleic acid with complementary sequences can be useful. The sequence complexity of the nucleic acid can give a clear indication as to whether one, two or more different molecules exist (e.g. Gould, et al., 1978). Nucleic acid hybridization is discussed in more detail in Chapter 15 (Section V.C). d. Electron microscopy
The size of ds nucleic acids has been estimated from length measurements made on electron micrographs of individual molecules. This method may give erroneous results with ss nucleic acids because of doubt as to the internucleotide distances (Reijnders et al., 1973). Errors are considerable and reliable marker molecules must be used. e. Physicochemical methods
Gierer (1957, 1958) estimated the size of isolated TMV RNA by measuring the sedimentation coefficient and the intrinsic viscosity of the RNA. He found the general relationship between molecular weight (m) and sedimentation coefficient (s) for this RNA to be m = 1 1 0 0 s 22. With the s equal to 31 he determined that the molecular weight of the isolated intact TMV RNA was 2.1 • 106. Empirical relationships between MW (m) and S20w of the general form m = ks d have been published for RNAs by several workers (e.g. Spirin, 1961; Hull et al., 1969). Such a relationship is very useful but it should be remembered that methods depending on the determination of s are much more laborious than those using PAGE and are no more accurate. 3. ssRNA genomes a. Heterogeneity When RNA is prepared from a purified virus preparation and subjected to some fractiona-
89
tion procedure that separates RNA species on the basis of size, RNAs of more than one size are usually found. Apart from degradation during RNA isolation and storage, there are several possible reasons for heterogeneity. i. Muhipartite genomes
Of the 70 genera of plane viruses, 28 have their genome in two or more separate pieces of different size, as discussed in detail in Chapter 6 (see Appendix 2). ii. Other sources of heterogeneity
1. Some degradation of the viral RNA may have occurred inside the virus before RNA isolation. 2. Less than full-length copies of the viral RNA (subgenomic pieces) synthesized as such in infected cells may be encapsulated (Palukaitis, 1984). 3. Some host RNA species may become accidentally encapsulated. For example, the empty protein shells found in preparation of EMV and several other tymoviruses contain an average of two to three small RNA molecules with tRNA activity (Bouley et al., 1976). These are very probably encapsulated host tRNAs. Similarly, host RNAs m a y become encapsulated in TMV coat protein (Siegel, 1971). 4. A number of viruses contain small satellite RNAs unrelated to the viral genome in nucleotide sequence (see C h a p t e r 14, Section II). 5. Small defective interfering RNAs made up of viral sequences are found in some virus infections (see Chapter 8, Section IX.C). 6. As noted in the next section (III.A.3.b), ssRNAs in solution may have a substantial degree of secondary and tertiary structure. A single RNA species may assume two or more distinct configurations under appropriate ionic conditions. These 'conformers' may be separable by electrophoresis or centrifugation. Thus, they give rise to additional heterogeneity that is not based on differences in length (Dickerson and Trim, 1978). 7. The RNA may migrate as a transient dimer under certain conditions (e.g. Asselin and Zai tlin, 1978).
90
4 P U R I F I C A T I O N A N D C O M P O S I T I O N OF PLANT VIRUSES
b. Some physical properties of ssRNAs i. Ultraviolet absorption
Like other nucleic acids, plant viral RNAs have an absorption spectrum in the ultraviolet region between 230 and 290 nm that is largely due to absorption by the purine and pyrimidine bases. The absorption spectra of the individual bases average out to give a strong peak of absorption near 260 nm with a trough near 235 nm (see Fig. 15.4). Ultraviolet absorption spectra are usually of little use for distinguishing between one viral RNA and another. The absorbance of an RNA solution at 260 nm measured in a cell of 1 cm path length is a convenient measure of concentration. The absorbance per unit weight varies somewhat with base composition and also with secondary structure. In general, for ssRNA at I m g / m L in 0.1 M NaC1, A26 o = 25; at 1 m g / m L for TMV RNA, 326 o = 29; for TYMV RNA, A2~,~~ = 23 (in 0.01 M NaC1 at pH 7) (Matthews, 1991). ii. Secondary structures
In the intact virus particle, the three-dimensional arrangement of the RNA is partly or entirely determined by its association with the virus protein or proteins (see Chapter 5). Here I shall consider briefly what is known about the conformation of viral RNAs in solution. dsDNA has a well-defined secondary structure imposed by base-pairing and base-stacking in the double helix. Single-stranded viral RNAs have no such regular structure. However, it has been shown by a variety of physical methods that an RNA such as that of TMV in solution near pH 7 at room temperature in, say, 0.1 M NaC1 does not exist as an extended thread. Under appropriate conditions, ssRNAs contain numerous short helical regions of intrastrand hydrogen-bonded base-pairing interspersed with ss regions. They behave under these conditions as more or less compact molecules. The degree of secondary structure in the molecule under standard conditions will depend to some extent on the base sequence and base composition of the RNA. The helical regions in the RNA can be abolished, making the molecule into a random, extended, disorganized coil by a variety of changes in the environment (Spirin, 1961).
These include heating, raising or lowering the EH, or lowering the concentrations or changing the nature of the counterions present (e.g. Na+, Mg 2+) (Boedtker, 1960; Eecen et al., 1985). The change from helical to random coil alters a number of measurable physical properties of the viral RNA. The absorbance at 260 nm and viscosity are increased, while the sedimentation rate is decreased. A polyribonucleotide lacking secondary structure has about 90% the absorbance of the constituent nucleotides found on hydrolysis. A fully base-paired structure (e.g. a synthetic helical polyribonucleotide composed of poly(A) plus poly(U) has about 60% of the UV absorbency of the constituent nucleotides. The changing absorption characteristics of RNA with changing pH are due in part to changes in the extent of the base-paired helical regions in the molecule, and in part to shifts in the absorption spectrum of individual bases due to changes in their ionization state. Methods for probing the structure of RNAs in solution are discussed by Ehresmann et al. (1987). Viral RNAs with amino-acid accepting activity almost certainly have a three-dimensional tRNA-like configuration near the 3' terminus when they are in solution under appropriate conditions (see Section III.A.3.c). iii. Pseudoknots (reviewed by Deiman and Pleij, 1997; Herman and Patel, 1999) RNA molecules can fold into complex threedimensional shapes and structures in order to perform their diverse biological functions. The most prevalent of these forms is the pseudoknot, which in its simplest manifestation involves the loop of a stem-loop structure basepairing with a sequence some distance away (Fig. 4.3). The involvement of pseudoknots has been recognized in a great variety of functions of viral RNAs including control of translation by - 1 frameshifting (see Chapter 7, Section V.B.10) (Giedroc et al., 2000), by read-through of stop codons, by internal ribosome entry sites (see Chapter 7, Section V.B.4), and by translational enhancers (see Chapter 7, Section V.C). The tRNA-like structures at the 3' termini of a variety of plant viral RNAs also form pseudoknots (Figs 4.4 and 4.5). These structures are
III. COMPONENTS
!' : . . . . .
A
$1::, I
,,, ' '
5.
.
.
.
'' '' I
I
' '
' '
.
involved in the regulation of RNA replication (see Chapter 8, Section IV.H.3). The structure of the classical pseudoknot of TYMV has been solved by NMR (Kolk et al., 1998). The structure displays internal mobility, which may be a general feature of RNA pseudoknots that regulates their interactions with other RNA molecules or with proteins.
L2
.... ' '
//o llmll II II
I(~1
I1~i ~ 9
II'--II II,,,,II
.
.
.
.
.
.
v=.
.
3'
L1
iv. Effective buoyant density in solutions of cesium salts
L1 5'
B L2
L1 Sl
C s=
L2
5' D
91
3'
When a dense solution of CsC1 or Cs2SO 4 is subjected to centrifugationunderappropriateconditions, the dense Cs ions redistribute by sedimentation and diffusion to form a density gradient in the tube. If a nucleic acid is present, it will band at a particular position in the density gradient. This density, k n o w n as the effective buoyant density, provides a useful criterion for characterizing virus nucleic acids and for distinguishing between RNAs and DNAs. dsDNAforms a band ataboutl.69-1.71 g/cm3inasolutionofCsC1; the exact banding position depends on the G + C content of the DNA. Thus, the method can be used to discriminate between DNAs with differing base composition. Under the same conditions, RNA is pelleted from the gradient. Cs2SO4 solutions form gradients in which all the various kinds of nucleic acid can form bands: d s D N A b a n d i n g at 1.42-1.44 g/cm3; ssDNA at about 1.49 g/cm3; D N A - R N A hybrids at about 1.56g/cm3; dsRNAs at about 1.60 g/cm3; and ssRNAs at about 1.65 g / c m 3. If a fluorescent dye, such as e t h i d i u m bromide, is intercalated into the nucleic acid to enable it to be located more easily, the nucleic acid will band at a lower density. c.
Fig. 4.3 The H(airpin) type of classical RNA pseudoknot. (A) Secondary structure. The dotted lines indicate the base-pair formation of the nucleotides from the hairpin loop with the complementary region at the 5' side of the hairpin. (B,C) Schematic folding. (D) Three-dimensional folding, showing the quasi-continuous double-stranded helix. The stem regions ($1 and $2) and the loop regions (L1 and L2) are indicated. L1 crosses a deep groove and L2 a shallow groove respectively. From Deiman and Pleij (1997), with permission.
End-group
structures
Many plant viral ssRNA genomes contain specialized structures at their 5' and 3' termini. This section summarizes the nature of these structures, while their biological functions are discussed in Chapter 6 (Section II.D). i. The 5' cap
Many m a m m a l i a n cellular messenger RNAs and animal virus messenger RNAs have a methylated blocked 5' terminal group of the form: m7G 5 pppS X(m)pyIm)p
where Xim) and y(m) a r e two methylated bases.
92
4 PURIFICATION AND COMPOSITION OF PLANT VIRUSES
A
D 5'
3'
G U U lOO c U U U C UAGC U V I II I AAUC_ G AA U U A 80 G G A CCGCU C ulV llt~ u
3' A76
A, C C ,0 A A C AGGCU i i i i i I U CC GA U 2~ G G C U c C c ao C C U A GC C C
~G~&GU_AU A AU~a~a ,o U G C "
C-G U-A
DG
C G
D
13
D
20
i
U6G(~
DA2~
CD mG
A
U - AAC C-G U-A
20:1
u-A
3oG-C40
- G
uC-G C
uACA I
A
E
L2 AC A
II
4,
,~i
i , i i
. GGGC GGG,~I~GCC ,o I ~J UCU uG U 2, U IV A A LI CGU
A
G-C ~C-Go
A
C
G-C A-U
/
U
,~U
CSO
C
T
U U
G-C G-U C-G10,
o,
? 9 1 A A Ti Ai? uiq A C C A
GCUUUGG 7
1
5'
76
3'
C
D
mG U C ~m
3oG- C4o C-G U C UACA A
F
-"
49
C
U- A~" C-G U-A
i
C
GACC-~
GGc D 20D D G 20.1D A 21A
U-A C-G U-A G-C U-A ~C -13 C C C III A CAC
A
T cu999
G
ao UCA~o UCCCG & C UCGGA/~CCA, 3' , , i,
6O
C U loUG-CGGGUC mA A mG ,, ,,, I G mCCCAG UCUG T~C
c c c III A C C
B
A
1G-C G U U-A7o U-A U-A C-G
G-CSO
6~
C C
5'
~------
C
Fig. 4.4 C o m p a r i s o n b e t w e e n the tRNA-like structure from TYMV R N A (A,B,C) and canonical t R N A (D,E,F). (A) Secondary structure of the 106 3 ' - t e r m i n a l nucleotides of TYMV RNA. (B) Scheme of folding of the tRNA-like structure (86 nucleotides). (C) Three-dimensional wire m o d e l of the t R N A q i k e structure (86 nucleotides). (D) Secondary structure of yeast t R N A w'l. (E) Scheme of folding of the yeast t R N A va] (to be c o m p a r e d with (B)). (F) Three-dimensional m o d e l of canonical tRNA. From D u m a s et al. (1987), with permission.
I ll.
c ~, ;o T
cccG~ccc l~li Ill
A
u &
G--C
qc.~ r
\
z'o
c
?O-C-G G-C-~O A-U
9 O"U U
o '~Ged-,' ,"o
Ar
6-t, ' / u iiill,i, 1651~r
u
u 1-/Ai
ae/I/c
& I/
j /
-
Ge,U 1 2 8 ~ " 0--C" - / G-C t c-~
,,.
uu
".~b"*/Veuu
M.
]~,-4 4
z zo .,~-..--~,
zzo
G-z6s
~_i, ~:~
U-A tl A
~ &/C
g:l
G A ,& A
30"
T
G
tltl
"-.-C&CGR3A ~P-t I I I I t l | 5 ' - GC/GUCU I
~o,-
21-uu ~ o =
cccc4~ccG
\ Y
~0-G-C C-G
.
7, the icosahedra are skew, and r i g h t - h a n d e d and l e f t - h a n d e d versions are possible. The physical m e a n i n g of h and k in a virus structure is illustrated in the following section. Since each of the triangles formed w i t h the P p a r a m e t e r can be further s u b d i v i d e d into f2 smaller triangles, T gives the total n u m b e r of subdivisions of the original faces and 20T the total n u m b e r of triangles. Fig. 5.17 gives some examples. Thus, the n u m b e r of structural subunits in an icosahedral shell is 20 • 3 • T = 60T.
D. Clustering of subunits The actual detailed structure of the virus surface will d e p e n d on h o w the physical subunits are packed together. For example, three clustering possibilities for the basic icosahedron are s h o w n in d i a g r a m iv.
9
137
I
0
Diagram iv In fact m a n y smaller plant viruses are based on the P = 3, f = 1, T = 3 icosahedron. In this structure, the structural subunits are c o m m o n l y clustered about the vertices to give p e n t a m e r s and hexamers of the subunits. These are the morphological subunits seen in electron micrographs of negatively stained particles (e.g. Fig. 5.30B). Since there are always 12 vertices w i t h 5-fold s y m m e t r y in icosahedra, we can calculate the n u m b e r of morphological subunits (M) (assuming clustering into p e n t a m e r s and hexamers) as follows: M-
[ ( 6 0 T - 60)/6] hexamers + (60/5) p e n t a m e r s = 1 0 ( T - 1) h e x a m e r s + 12 p e n t a m e r s
In p h o t o g r a p h s of virus particles where the p e n t a m e r s and hexamers can be u n a m biguously recognized (e.g. one-sided images of negatively stained particles or freeze-fracture replicas of the outer faces of larger viruses), the p a r a m e t e r s h and k m i g h t be used to establish the icosahedral class of the particle. This procedure w o u l d be particularly useful for shells containing large n u m b e r s of hexamers, h and k represent the n u m b e r s of hexamers that
Fig. 5.17 Ways of sub-triangulation of the basic icosahedron shown in Fig, 5.16 to give a series of deltahedra with icosahedral symmetry (icosadeltahedra). (A) The basic icosahedron, with T = 1 (P = 1, f = 1). (B) With T = 4 (P = 1, f = 2). (C) With T = 3 (P = 3,f= 1). (D) With T = 12 (P = 3, f = 2). From Caspar and Klug (1962), with permission.
138
5 ARCHITECTURE
A N D ASSEMBLY OF V I R U S P A R T I C L E S
must be traversed to move by the shortest route from one pentamer to the next. Thus, we must identify two adjacent pentamers. For example, the structure s h o w n in diagram v was observed on the surface of freeze-etch replicas of phage )~ (Bayer and Bocharov, 1973).
C2 //
h = 4 , k= 1, T = 21 Diagram v This indicates a skew 'right-handed' skewness.
icosahedron
with
E. 'True' and 'quasi' symmetries In the basic icosahedron (Fig. 5.16), a feature located half w a y along any edge of a triangular face is positioned on an axis of rotation for the whole solid. Thus, it is on a 'true' or icosahedral s y m m e t r y axis. This is a true dyad. In any more complex icosahedron (T > 1) there is more than one kind of 2-fold symmetry. For example, in a P = 3, T = 3 shell (Fig. 5.17C), the center of one edge on each of the 60 triangular faces is on a true d y a d axis relating to the solid as a whole. The center positions of the other two edges of a face have only local 2-fold symmetry. These are called 'quasi' dyads. On the 3' axis of the 'quasi' symmetry, the three chemically identical but structurally independent subunits in an icosahedral asymmetric unit are designated A, B and C (Harrison et al., 1978) (Fig. 5.18).
E Bacilliform particles Some virus particles, for instance those of AMV and b a d n a v i r u s e s , are bacilliform w i t h r o u n d e d ends separated by a tubular section. Hull (1976a) suggested that the structure of these particles is based on icosahedral s y m m e try. The r o u n d e d ends w o u l d have constraints
Fig. 5.18 Arrangement of protein subunits as found in several T = 3 plant and animal viruses. The nomenclature for the chemically identical subunits follows Harrison et al. (1978). From Krishna et al. (1999), with permission. of icosahedra with the three-dimensional curvature determined by 12 pentamers, six at each end. The t w o - d i m e n s i o n a l structure of the tubular section w o u l d be made up of hexamers. Hull derived various hexamer structures from icosahedra cut across the 2-fold, 3-fold, 5-fold and interlattice axes.
VI. SMALL ICOSAHEDRAL VIRUSES A. Subunit structure The coat protein subunits of most small icosahedral viruses are in the range of 20-40 kDa; some are larger but fold to give effective 'pseudomolecules' within this range (Section VI.B.6). In contrast to rod-shaped viruses, the subunits of most small icosahedral viruses have a relatively high proportion of [3-sheet structure and a low proportion of R-helix (Denloye et al., 1978; O d u m o s u et al., 1981) and have the same basic structure. This comprises an eightstranded antiparallel [3 sandwich, often termed a [3 barrel or 'jellyroll', which is shown schematically in Fig. 5.19.
VI. SMALL ICOSAHEDRAL VIRUSES
139
Fig. 5.19 (see Plate 5.1) (A) The icosahedral capsid contains 60 identical copies of the protein subunit- in blue, labeled A. These are related by 5-fold symmetry elements (pentagons at vertices), 3-fold symmetry elements (triangles in faces) and 2-fold symmetry elements (ellipses at edges) - shown in yellow. For a given-sized subunit, this point group symmetry generates the largest possible assembly (60 subunits) in which every protein lies in an identical environment. (B) Schematic of the subunit building block found in many RNA, and some DNA, viral structures. Such subunits have complementary interfacial surfaces which, when they repeatedly interact, lead to the symmetry of the icosahedron. The tertiary structure of the subunit is an eight-stranded [3-barrel with the topology of the jellyroll - see part (C). In this diagram the {3-strand and helix coding are identical to that in (C). Subunit sizes generally range between 20 and 40 kDa with variation between different viruses occurring at the N and C termini and in the size of insertions between strands of [3-sheet. These insertions generally do not occur at the narrow end of the wedge (B-C, H-I, D-E and F-G turns). (C) The topology of viral {3-barrel showing the connections between strands of the sheets (represented by fat arrows) and positions of the insertions between strands. The green cylinders represent helices that are usually conserved. The C-D, E-F and G-H loops often contain large insertions. From Johnson and Spier (1999), with permission.
The o v e r a l l s h a p e is a t h r e e - d i m e n s i o n a l w e d g e w i t h t h e B-C, H - I , D - E a n d F - G t u r n s b e i n g at the n a r r o w (interior) end. M o s t v a r i a tion b e t w e e n the s u b u n i t sizes o c c u r s at the N a n d C t e r m i n i a n d b e t w e e n the s t r a n d s of [3 s h e e t at the b r o a d e n d of the s u b u n i t . It is the d e t a i l e d p o s i t i o n i n g of the e l e m e n t s of the [3 b a r r e l a n d of the N a n d C t e r m i n i t h a t g i v e the flexibility to o v e r c o m e the q u a s i - e q u i v a l e n c e p r o b l e m s . This is i l l u s t r a t e d in t h e d e t a i l e d s t r u c t u r e s d e s c r i b e d below. The coat p r o t e i n s u b u n i t s f o r m one, t w o or t h r e e s t r u c t u r a l d o m a i n s , the S (shell) d o m a i n , the R d o m a i n ( r a n d o m - b i n d i n g ) a n d the P (pro-
t r u d i n g ) d o m a i n . All v i r u s e s h a v e t h e S d o m a i n . T h e R d o m a i n is s o m e w h a t of a m i s n o m e r b u t d e f i n e s a n N - t e r m i n a l r e g i o n of the p o l y p e p t i d e c h a i n t h a t a s s o c i a t e s w i t h the viral RNA. As it is r a n d o m , n o s t r u c t u r e c a n be d e t e r m i n e d by X-ray c r y s t a l l o g r a p h y . T h e P d o m a i n gives surface protuberances on some viruses.
B. Virion structure At p r e s e n t , w e can d i s t i n g u i s h s e v e n k i n d s of s t r u c t u r e a m o n g the p r o t e i n shells of s m a l l i c o s a h e d r a l or i c o s a h e d r a - b a s e d p l a n t v i r u s e s
140
5
ARCHITECTURE
AND
ASSEMBLY
OF
VIRUS
PARTICLES
whose architecture has been studied in sufficient detail. These are: T = 1 particles; bacilliform particles based on T = 1; geminate particles based on T = 1; T = 3 particles; bacilliform particles based on T = 3; p s e u d o T = 3 particles; and T = 7 particles. 1. T = 1 particles (satellite viruses) The satellite viruses are the smallest k n o w n plant viruses having a particle diameter of about 17 n m and capsids m a d e up of 17-21kDa polypeptides (see Chapter 14; Section II.A, for satellite viruses). The structure of STNV was the first to be solved and s h o w n to be m a d e of 60 protein subunits of 21.3 kDa arranged in a T = 1 icosahedral surface lattice. The structure of the protein subunit has been determined crystallographically with refinement to 2.5 ,~ resolution (Jones and Liljas, 1984). The general topology of the polypeptide chain is like that of the S domains of TBSV and SBMV (Fig. 5.20) but the packing of the subunits in the T = 1 icosahedral structure is clearly different (Rossmann et al., 1983) and there is no P domain. In the amino terminus of STNV, there are only 11 disordered residues followed by an ordered helical section (residues 12-22) buried in the RNA. Three different sets of metal ion b i n d i n g sites (probably Ca 2+) have been located. These link the protein s u b u n i t s together. A more detailed structure for the protein shell has been proposed by Montelius et al. (1988). The crystal structure of STNV has been studied at 16 A resolution using neutron diffraction in H20/D20 (Bentley et at., 1987). At 40% D20, scattering arises largely from the RNA component. The two main RNA motifs are s h o w n in Fig. 5.21. These are in fact connected by regions of weaker RNA density. Each spherical motif (II) is connected to five symmetry-related extended motifs (I). They form a continuous n e t w o r k of RNA density on the inside surface of the protein coat. The I motifs form the edges of an icosahedron, leaving triangular holes centerd on the 3-fold axes. It is into these holes that triple-helical arms of amino-terminal regions of the protein subunits penetrate, and make close
A
TBSV 2 a
Q6
8
SBMV
aO
2
\ c
STNV -
"
2 o81 aB2
5
3
S Fig. 5.20 Diagrammatic representation of the backbone folding of the coat protein of (A) TBSV, (B) SBMV, and (C) STNV shown in roughly comparable orientations. From Rossmann et al. (1983), with permission. contact with the RNA. Basic amino acids are well placed to m a k e contact with the RNA. Except for this p r o t e i n m R N A interface, the inner face of the protein shell is separated from
VI. SMALL ICOSAHEDRAL VIRUSES
interactions are organized by Ca 2+ in STNV, an anion in STMV and apparently neither of these in SPMV. Finally, nucleic acid was visible only in electron density maps of STMV and s h o w e d as double-helical RNA segments associated with each coat protein dimer.
It
- (.v
1~
141
..j
2. Bacilliform particles based on T = 1 symmetry As noted in Section V.F, it has been p r o p o s e d that the structure of bacilliform particles is based on icosahedral symmetry. i;_ ".%
,@
a. Alfamovirus a n d Ilarvirus genera
r r
Fig. 5.21 Low-resolution structure of the RNA within STNV determined by neutron diffraction. Positive density at 40% D20 looking down the 5-fold axis from the center of the virus. I is the RNA density motif, which lies along the edges of each triangle and forms the edges of an icosahedron. Its length is about 45 A and its diameter is 22-25 A. The RNA density of motif II lies along each 5-fold axis at a distance of 67 A from the virus center. Minor regions of density (III) at higher radii correspond to positive fluctuations of protein density. See Bentley et al. (1987) for details of I, II and III. From Bentley et al. (1987), with permission.
the RNA by a thin layer of solvent. The crosssection of motif I fits well with the idea that it represents a double helix in the RNA. If this is so, then 72% of the total RNA w o u l d be in double-helical form. The fact that STNV RNA has thermal denaturation kinetics like that of a transfer R N A m i n d i c a t i n g a high degree of secondary structure (Mossop and Francki, 1979a)confirms this. The structures of two other satellite viruses have been resolved, STMV and SPMV (Larson et al., 1993a,b; Ban and McPherson, 1995) and have been c o m p a r e d to that of STNV (Ban et al., 1995). In spite of all having the [3-barrel structure in the subunit that the 5-fold contacts at the narrow end, the three viruses were remarkably different in the arrangements of the secondary structural elements. Also, the 5-fold protein
Purified preparations of AMV contain four nucleoprotein c o m p o n e n t s present in major a m o u n t s (bottom, B; middle, M; top b, Tb; and top a, Ta). They each contain an RNA species of definite length. The genome is split between the Tb, M and B RNAs. Three of the four major c o m p o n e n t s are bacilliform particles, 19 n m in diameter. The fourth (Ta) is normally spheroidal with a diameter slightly larger than 19 n m (Fig. 5.22). However, two forms of Ta c o m p o n e n t have been recognized (Heijtink and Jaspars, 1976). Ta t is spheroidal and soluble in 0.3 M MgSO 4. T a b is a rodlet and insoluble in 0.3 M MgSO 4. These two particles appear identical in other properties. The Ta c o m p o n e n t s h a v e 120 protein subunits, and Cusack et al. (1983) raised the possibility that these m a y have a nonicosahedral structure. F r o m a careful s t u d y of the m o l e c u l a r weights of the RNAs, the protein subunit and the virus particles, Heijtink et al. (1977) conc l u d e d that the n u m b e r of coat p r o t e i n m o n o m e r s in the four major c o m p o n e n t s is equal to 60 + (n • 18), n being 10, 7, 5 or 4. As noted above (Section V.F), it has been proposed that such particles are based on icosahedra cut across various axes with the tubular portion m a d e up of hexamer subunits. From optical diffraction studies on electron micrographs of negatively stained AMV particles, Hull et al. (1969b) suggested that the tubular structure of the bacilliform particles was based on a T = 1 icosahedron cut across its 3-fold axis with rings of three hexamers (18 coat protein
142
5 ARCHITECTURE ANt) ASSEMBLYOF VIRUS PARTICLES
monomers) forming the tubular portion (Fig 5.23). Proton magnetic resonance studies provide a low-resolution model in which the coat protein consists of a rigid core and a flexible aminoterminal part of about 36 amino acid residues (Kan et al., 1982). The protein behaves as a water-soluble dimer stabilized by hydrophobic interactions between the two molecules. This dimer is the morphological unit out of which the viral shells are constructed. Under appropriate conditions of ionic strength, ionic species, pH, temperature and protein concentration, the protein dimer forms a T = 1 icosahedral structure built from 30 dimers (Fig. 5.24) (Driedonks et al., 1977). This structure has been confirmed by X-ray crystallographic analysis (Fukuyama et al., 1983) at 4.5 A resolution that was further refined to 4.0 A resolution together with cryoelectron microscopy and image reconstruction (Kumar
~-19-~
-,~-19-~
et al., 1997). This showed that the subunit structure and dimer association is structurally similar to CCMV. Large holes are observed at the pentamer axes, giving a porous particle structure. The virus is unstable with regard to high ionic strength and SDS, and is sensitive to RNase, which might be explained by the holes in the protein coat. Conformational changes occur at mildly alkaline pH (Verhagen et al., 1976). AMV nucleoproteins are readily dissociated into protein and RNA at high salt concentrations, and bacilliform particles can be reformed under appropriate conditions. Thus, the virus is mainly stabilized by protein-RNA interactions (van Vloten-Doting and Jaspars, 1977). Under a wide range of solvent conditions, AMV particles do not show the phenomenon of swelling described below for bromoviruses (Oostergetel et al., 1981). Compared with the small isometric viruses,
.,i--19-~
..q.-19-~ I
top a (Ta)
~ top b (Tb)
I I
48 ,~
' I I
58
I
I I I I I
',..L/
middle (M)
bottom (B)
T
m 9~4nm
Fig. 5.22 AMV particles. Top: Sizes (in nm) of the four main classes of particle. Bottom: Schematic of the distribution of protein and RNA in AMV bottom component. RNA is indicated by the tinted area and the protein molecules are represented by ellipsoids. The model is derived from the analysis of both the 30S and the bottom component by small-angle neutron scattering. Bottom part reprinted from Cusack et al. (1981), with permission.
VI. SMALL ICOSAHEDRAL VIRUSES
A
B
2 ~.'Qd
C D Fig. 5.23 Geodestix models showing the proposed structure of the components of AMV. (A) Topz, (53S) component; (B) tOPa component; (C) toPb component; (D) middle component; (E) bottom component. From Hull et al. (1969b), with permission.
'y,e .,~
Fig. 5.24 T = 1 of the 30S AMV particle composed of 30 dimers of coat protein. The 2-fold symmetry axes of the dimers coincide with the dyad positions of an icosahedron leaving large holes at 5- and 3-fold positions. For clarity, the rear-facing subunits are omitted from the model. From Driedonks et al. (1977), with permission.
143
AMV may be regarded as being in a permanently swollen state and thus resembles CMV. Purified preparations of some AMV strains can be fractionated by polyacrylamide gel electrophoresis to reveal the presence of at least 17 nucleoprotein components (Bol and LakKaashoek, 1974). These are all made up from the single viral coat protein (van Beynum et al., 1977). Besides the four major RNA species, there are at least 10 minor RNA species of different lengths. The nucleoproteins occurring in minor amounts contain both major and minor species of RNA. Particles of somewhat different size may contain the same RNA, while particles of the same size may contain different RNAs. Thus, a small variation is possible in the amount of RNA encapsulated by a given amount of viral protein (Bol and Lak-Kaashoek, 1974). Although AMV is labile in gradients of CsC1, it is stable and bands isopycnically in gradients of CSRSO4 and metrizamide (Hull, 1976b). In these gradients it forms two bands at very similar densities, the major band containing mainly bottom component and the minor band predominantly the other components. This indicates that the components have very similar, but not identical, protein:RNA ratios. Some AMV strains (e.g. 15/64 and VRU) form unusually long particles (Fig 5.25) (Heijtink and Jaspars, 1974; Hull, 1970a) with particles up to 1/zm long; these particles do not contain any abnormally long RNA molecules and most likely contain several molecules of the genome segments. Reconstitution experiments indicated that the tendency to form long particles is coat protein-directed (Hull, 1970a). In a comparison of the coat proteins of normal-length particle strains with those of long particle strains, Thole et al. (1998) identified two amino acid substitutions, Ser 66 and L e u 175, that were associated with the long particles. When these amino acid alterations were introduced into a normallength particle strain, long particles were formed. Ilarviruses such as TSV have quasi-isometric or occasionally bacilliform particles of four different size classes (van Vloten-Doting, 1975) that appear to share many properties in common with the AMV group (van Vloten-Doting,
144
5 ARCHITECTURE
A N D ASSEMBLY OF V I R U S P A R T I C L E S
1976). The top c o m p o n e n t of TSV has been crystallized to give a hexagonal space group, but the crystals were not amenable to X-ray diffraction as they were disordered (Senke and Johnson, 1993). b.
Ourmiaviruses
M e m b e r s of the Ourmiavirus genus have bacilliform virions of 18 n m diameter and of three lengths, 30, 37 and 45.5 n m (Lisa et al., 1988). The protein subunits cluster into dimers or trimers. The pointed ends m a y be formed from icosahedra cut through 3-fold axes for a d i m e r or 2-fold axes for a trimer. The tubular b o d y does not form a continuous geometrical net as with AMV, but contains discontinuities m a r k e d by fissures between double disks of the protein (Fig. 5.26). Particles consisting of two, three, four or six double disks have been observed, with four- and six-disk particles being rare. Figure 5.26 also illustrates these structures dia g r a m m a tic ally.
3. Other particles based on T = 1 symmetry (geminiviruses) Geminiviruses contain ssDNA and one type of coat polypeptide. Particles in purified preparations consist of twinned or geminate icosahedra (Fig. 5.27). From a study of negatively stained particles together with models of possible structures, Francki et al. (1980b) suggested that CSMV consists of two T = 1 icosahedra joined together at a site w h e r e one morphological subunit is missing from each, giving a total of 22 morphological units in the geminate particle. This structure received s u p p o r t from a study of the size of the D N A and protein subunit of this virus, indicating that each geminate particle shell, consisting of 110 polypeptides of 28 kDa arranged in 22 morphological units, contains one molecule of ssDNA with molecular weight 7.1 • 105 (about 2.1 kb). The structure of the coat protein of MSV has been d e d u c e d by modeling it on the atomic coordinates of STNV coat protein as a template
Fig. 5.25 Electron micrograph of unfractionated preparation of VRU strain of AMV negatively stained in saturated uranyl acetate. Bar - 100 nm. From Hull (1970a), with permission.
VI. SMALL ICOSAHEDRAL VIRUSES
145
motif (Zhang et al., 2001) (Fig. 5.28). The fine structure of MSV particles was determined by cryoelectron m i c r o s c o p y and t h r e e - d i m e n sional image reconstruction (Zhang et al., 2001) (Fig. 5.28) and confirmed the model suggested by Francki et al. (1980b). 4. T = 3 particles a. Tymovirus genus The particles of tymoviruses are about 30 n m in diameter and m a d e up of subunits of about 20 kDa.
i. Classes of particle Purified TYMV preparations can be fractionated on CsC1 density gradients into a large n u m b e r of components. There are three classes of particle.
Fig. 5.26 Structure of Ourmia melon mosaic virus. Top: Averaged images of negatively stained particles. Each image was built up photographically by equal
superimposed exposures of 10 original particle images. The top row has two double disks and the bottom row has three. Bottom: Sketches showing the suggested arrangement of double disks in particles of different length. Particles of type D have not yet been observed. (Courtesy of R. G. Milne.)
Fig. 5.27 Purified geminivirus virus from Digitaria negatively stained in 2% aqueous uranyl acetate. al. (1986), with Bar = 5 0 n m . From Dollet et
permission.
1. The empty protein shell. About one-third to one-fifth of the particles found in a TYMV preparation isolated from infected leaves are e m p t y protein shells (the top or T component). These contain no RNA but otherwise are identical in structure to the protein shell of the infectious virus (B1 component). Full particles can be converted to e m p t y ones by freeze-thaw or by alkali treatment. W h e n TYMV B 1 is taken to p H 11.6 in 1 M KC1, the particles swell from 14.6 to 1 5 . 2 n m radius within 30 seconds (Keeling et al., 1979). RNA escapes from these particles in 3-10 minutes in a partially d e g r a d e d state. In addition, an a m o u n t of protein is lost within 1-3 minutes that is equivalent to the loss of one p e n t a m e r or hexamer of subunits from each particle. No such loss of protein occurs with the minor nucleoproteins (Keeling and Matthews, 1982). On return to p H 7.0, the radius of the resultant e m p t y shells returns to normal. The nucleoproteins containing less than the full genome RNA do not swell at p H 11.6 u n d e r the same conditions, and their RNA does not escape, although it is also d e g r a d e d within the particle. 2. Infectious virus nucleoprotein (B or bottom components and particles derived from them). The infectious virus fractionates to form two density classes in CsC1 gradients (Bla and Bib) that are equally infectious (Matthews, 1974). A third B1 fraction, B~c, more dense than B~b, has been
146
5 ARCHITECTURE AND ASSEMBLY OF VIRUS PARTICLES
Fig. 5.28 (see Plate 5.2) Models of MSV-N coat protein and complete capsid. Ribbon drawings of polypeptide chain of (A) STNV coat protein and (B) MSV-N coat protein. In these drawings ]3-strands are labeled according to convention. (C) Fit of 110 copies of the MSV-N pseudo-atomic model, shown as Ca tracings (in yellow), the two apical capsomeres, 10 peripentonal capsomeres, and 10 equatorial capsomeres. The rectangular boxed region indicates the position of close-up views of the models shown in (D) and (E) and the double-ended arrows the cross-sectional views shown in (F) and (G). (D) Close-up view of the 2- and 3-fold icosahedral and (E) equatorial capsomere interactions. Cross-sectional slices of the MSV-N particle (F) through the equatorial region and (G) peripentonal capsomeres of the particle. The icosahedral 2-, 3and 5-fold axes are indicated as a black ellipsoid, triangle and pentamer respectively. From Zhang et al. (2001), with permission.
VI.
characterized. Bib and Blc both contain copies of the coat protein mRNA as well as a molecule of genome RNA. These B1 components can be converted in strong solutions of CsC1 to a B 2 series with higher densities, especially if the pH is above 6.5. These are designated B2a, B2B and B2c. Their formation is prevented by the presence of 0.1 M MgC12 in the CsC1. 3. Nucleoprotein particles containing subgenomic R N A s and having densities in CsCI intermediate between that of the T and B components. Mellema et at. (1979) and Keeling et al. (1979) isolated a series of eight minor components. The coat mRNA and a series of eight other subgenomic RNAs of discrete size have been isolated from these particles. These eight subgenomic RNAs have not yet been firmly allocated to particular nucleoprotein species. If the coat mRNA and the eight others are encapsidated in various combinations, and numbers of copies per particle, there may in fact be a very large number of particles of slightly differing density. Some properties of the minor noninfective nucleoprotein fractions have been determined by Mellema et al. (1979) and Keeling et al. (1979). Their RNA content ranges from about 5% for fraction I to 28% for fraction 8. The proportion of total minor nucleoproteins relative to the infectious nucleoproteins is about 5% on a particle number basis. The coat protein cistron is found in most of the minor nucleoproteins along with other subgenomic RNAs (Pleij et al., 1977; Higgins et al., 1978). ii. The protein shell Using negative-staining procedures, Huxley and Zubay (1960) and Nixon and Gibbs (1960) showed that the protein shell of TYMV is made up of 32 protuberances, occupying two structurally distinct sites in the shell. Klug and colleagues used TYMV extensively in developing X-ray diffraction and electron microscopy as tools for the study of smaller isometric viruses. Klug et al. (1966) and Finch and Klug (1966) concluded that the protein shell has 180 scattering centers lying at a radius of about 14.5 nm. These points were identified with protuberances of the protein structure units at the surface of the particle. A higher resolution X-ray
S M A I . I. I C ' O S A H E 1 ) R A L
VIRUSES
147
analysis to 3.2 A was made by Canady et al. (1995, 1996). Each individual protein subunit is somewhat banana-shaped. Within the intact virus, each protein subunit is made up of about 9% ~-helix and 43 % ~-sheet. About 48% of the polypeptide is in an irregular conformation (Hartman et al., 1978). This conclusion was broadly confirmed by Tamburro et al. (1978). The polypeptide chain forms an eight-stranded antiparallel sandwich (]3 barrel) (Fig. 5.29) (Canady et al., 1996). The X-ray data on the virus gave good agreement for a model of the protein shell with 32 scattering centers lying at a radius of about 121 A and extending to a radius of 159 A from the center of the particle. Figure 5.30A summarizes the knowledge of the external arrangement of the protein subunits, while Fig. 5.30B shows views of these particles in three different orientations obtained by the three-dimensional image reconstruction technique. Thus, the surface of the particle has the subunits in a T = 3 arrangement with the pentameric and hexameric protein aggregates protruding from the surface and forming deep valleys at the quasi 3fold axes. The N-terminal 26 residues of the A subunit are disordered, whereas those of the B and C subunits interact around the interior of the quasi 6-fold cluster where they form an annulus (Fig. 5.31). There are extensive internal contacts between the A, B and C subunits (Fig. 5.32). The three histidine residues of each protein subunit are positioned to the interior and are accessible for interaction with the RNA genome. The appearance of the interior surface of the virus capsid suggests that a pentameric subunit is lost during virus disassembly. The structure of the tymovirus PhyMV has been resolved to 3.0 A (Krishna et al. 1999). The basic structure is similar to that of TYMV but there are some differences. The N-terminal 17 residues of the A subunits making up the 12 pentamers show order that is not seen in TYMV and have a conformation that is very different from that observed in the B and C subunits constituting the hexameric capsomeres. Analysis of interfacial contacts indicates that
148
5 ARCHITECTURE
A N D ASSEMBLY OF V I R U S P A R T I C L E S
Fig. 5.29 Ribbon diagram of the TYMV coat protein with orientations of the quasi-sixfold, quasi-threefold and quasitwofold axes indicated. The B-subunit is shown; the A- and C-subunits are virtually the same except that the N-terminal 26 residues are disordered in the A-subunit. The eight strands of the [3-barrel are labeled, along with the N and C termini. Helix CD, which is a regular r is indicated. The EF helix, appearing behind, is irregular, with some qualities of a 310-helix. Two small segments of ~-sheet are also formed by residues 6-8 and 140-142. From Canady et al. (1996), with permission. hexamers are held together more strongly than pentamers and that h e x a m e r - h e x a m e r contacts are stronger than p e n t a m e r - h e x a m e r contacts. These observations suggested an explanation for the formation of e m p t y capsids which might be initiated by a change in the conformation of the A-subunit N-terminal arm. iii. Location of the R N A
Finch and Klug (1966) considered that folds of the RNA in TYMV were intimately associated with each of the 32 morphological protein units. They thought that it was the presence of this RNA in and around these positions that enhanced the appearance of the 32 morphological subunits seen in electron micrographs of this virus, as compared with empty protein shells. However, neutron small-angle scattering shows that there is very little penetration of the RNA into the protein shell and that the protein subunits are densely packed. C o m p a r i s o n of cryoelectron microscopy images of full and empty capsids of TYMV identified strong inner features a r o u n d the 3-fold axes of the full, but
not the empty, particles (Bottcher and Crowther, 1996). This suggested that substantial parts of the RNA are icosahedrally ordered. The X-ray analysis indicates density attributable to RNA to form a core or radius ---40 _Awith a hollow at its center (radius ---25/~) and pronounced projections extending to a radius of ---75/~ along the 5-fold directions (Canady et al., 1996). b. Bromovirus g e n u s The protein shell of these viruses is 25-28 n m in diameter, made up of 180 protein subunits of 19.4 kDa. The three genomic RNAs (1, 2 and 3) are packaged separately; the subgenomic RNA for coat protein (RNA4) is packaged with RNA3. i. Stability of the virus
A l t h o u g h b r o m o v i r u s e s are very similar to tymoviruses in the arrangement of their viral RNA and protein shells, they have much less stable particles. As the p H is raised between 6.0 and 7.0, particles of BMV swell from a radius of
lI. MOVEMENT AND FINAl. DISTRIBU
III Furo
III Alfamo/ilar.
RT Badna _
I I daeo ~ "
~
III Tobamo ~ _
li Diantho ~
Begomo
NepoA I
~ 1O0
~
III Bromo III Cucumo
, 100 "
88 ~
~
99
Tubuleformers
A
Tenui
100
III
Tobra
I
T ombusll Umbra II
Phloem Tospo
\A
Rhabdo N
Capilio/Tdcho Caulimo
~
RT
NepoB Como I
III
I
Fig. 9.12 Relationships a m o n g putative 30-kDa MP superfamily members determined by bootstrappir
Branches with less than 80% support (100 bootstrap replicates) were collapsed. Labels 0, RT, N, A, I, II an( the types of polymerase encoded by the viruses, which are respectively: 0; RNA-dependent DNA polymer strand virus; ambisense-strand virus; and positive-strand supergroups I, II and III RNA-dependent RNA (see Chapter 8, Section IV.B.1). The thin-lined polygon encloses those MPs known to form virion-bearing 1 Melcher (2000), with permission. A model for the two stages, intra- and intercellular transport, is s h o w n in Fig. 9.13. At the early stage, the virus replicates on the endoplasmic reticulum, leading to viral RNA accumulating in large irregular bodies. P30 colocalizes with this viral RNA, suggesting that viral replication and translation occur at the same subcellular site. The viral RNA and P30 co-localize w i t h m i c r o t u b u l e s w h i c h are thought to direct the R N A - P 3 0 complex to the plasmodesmata, which the P30 gates open to allow the complex to pass through (the intercellular transport stage). The mechanisms by which the MP targets and gates plasmodesmata are u n k n o w n , but it is interesting to note that P30 binds specifically to the cell wall enzyme, pectin methylesterase, which is also an RNA binding protein (Dorokhov et al., 1999). This gating is only t e m p o r a r y at the infection front, and behind the front p l a s m o d e s m a t a
a p p e a r to return to their n o r m a l S virus infection of protoplasts, h i g h weight forms of P30 have been d( accumulation of which was enhan inhibition of the 26S proteasome (I Beachy, 2000). It was s u g g e s t e d ubiquinated P30 is d e g r a d e d by the somes. It has also been suggeste phosphorylation could act as a me sequester P30 in cell walls and in function (Citovsky et al., 1993); in thatiana P30 is N-terminally processe t h o u g h t to also inactivate it (Hu I 1995). The similarities b e t w e e n the R C N M V noted above, and the fac TMV, R C N M V does not require vir. tein for cell-to-cell m o v e m e n t (Xi 1993), suggests that these two viruse ilar local m o v e m e n t strategies.
150
5 ARCHITECTURE
A N [ ) ASSEMBLY OF V I R U S P A R T I C L E S
13.5 n m to over 15 n m (Zulauf, 1977). The consequences of this swelling d e p e n d on m a n y factors and particularly on the ionic conditions. The particles are readily disrupted in I M NaC1 at p H 7.0. W h e n swollen virus is dissociated, the protein subunits can reassemble u n d e r a variety of conditions to form a range of products, described by Bancroft and Horne (1977) (see Section IX.A). Of particular interest is the fact that, in the presence of trypsin, the coat protein of BMV loses 63 amino acids from the amino terminus, and can then self-assemble into a T = 1 empty shell (Cuillel et al., 1981). BMV can be fully dissociated into subunits by high pressures, the formation of T = 1 particles being a step in the disassembly process (Silva and Weber, 1988). Only swollen particles disassemble under high pressure (Leimkt~hler et al., 2000). The structural p o l y m o r p h i s m of BMV has been investigated using neutron small-angle scattering (Jacrot et al., 1976; Chauvin et al., 1978). At low p H (around 5.0) the virus is in a compact state u n d e r a range of ionic condi-
tions. Near p H 7.0 at moderate ionic strength, the virus particle swells and the RNA penetrates more deeply between the protein subunits; Mg 2+ suppresses this effect. Titration studies have s h o w n that BMV contains several (probably two) cation-binding sites per subunit. Both sites titrate together at about p H 6.7. It is probable that both carboxylate groups of the protein and RNA phosphates contribute to this b i n d i n g (Pfeiffer and Durham, 1977). When both sites bind H § Ca R+ or Mg 2§ the virus is compact. As these ions are released, the virus swells. The N-terminal 25 amino acids of the coat protein are rich in basic amino acids, and structure p r e d i c t i o n m e t h o d s indicate that this sequence may interact in a helical form with the RNA (Argos, 1981) thus forming an R domain. ii. Particle structure Finch et al. (1967b) found that the morphologi-
cal units of BBMV protrude at least 1.5 n m from the body of the particle. The negative stain
Fig. 5.32 (see Plate 5.4) Hydrogen bonding, intersubunit contacts, and accessibility diagram for the B-subunit of TYMV. Atoms were assumed in contact if they were 4.11 A apart for van der Waals interactions, 3.3 :~ for hydrogen bonds, and 3.8/k for salt bridges. Residues were considered inaccessible if they had less than 5 ./~2 accessible surface area. From Canady et al. (1996), with kind permission of the copyright holder, 9 John Wiley and Sons Ltd.
V1. SMALl. I('.OSAHEDRAL VIRUSES
151
appeared to penetrate into the center of the virus particles, suggesting the presence of an appreciable central hole, which is not found in TYMV. The presence of this hole, about 5.5 nm in radius, was confirmed by X-ray diffraction studies (Finch et al., 1967b), and for BMV by small-angle neutron scattering. Finch and Klug (1967) suggested that the absence of a central hole in TYMV may be due to the need to pack a higher proportion of RNA into the particle. A detailed analysis has been made of the structure of native and swollen particles of CCMV using X-ray crystallography (to 3.2 A) and cryoelectron microscopy (Speir et al., 1993, 1995). The polypeptide chains of the coat protein subunits are arranged in [3 barrels with the C-terminal regions of adjacent subunits being interwoven. Additional particle stability is provided by contacts between metal ion (primarily Ca2§ carboxyl cages on each subunit and by protein interactions with regions of ordered RNA. Swelling of the particle results in a 29 A radial expansion due to electrostatic repulsion at the carboxylate cages. Complete disassembly of the particle is prevented by preservation of the interwoven C termini and by the protein-RNA interactions.
There is substantial penetration of the RNA into the shell of protein (Jacrot et al., 1977). The packing of the protein subunits is such that about 15% of the surface (at a radius of 11.7 nm) could be made up of holes, which would expose the RNA to inactivating agents and could explain the sensitivity of this virus to RNase. Cryoelectron microscopy and reconstruction to 23 A resolution shows that CMV is structurally similar to CCMV (Wikoff et al., 1997). The CMV structure was confirmed by X-ray crystallography at 8 A resolution, which also showed that the coat protein subunits have a [3-barrel structure. Thus, CMV and CCMV particle structures are similar in (1) particle morphology, (2) size and orientation of their [3-barrels, (3) stabilizing interactions for hexamer formation, and (4) subunit primary sequence. However, CMV has relatively unstable particles w h e n compared with those of bromoviruses. CMV particles do not swell at pH 7.0 under conditions where those of bromoviruses d o m o r in reality, they do not 'shrink' at pHs below 7. Thus, they behave essentially as swollen particles stabilized primarily by protein:RNA interactions.
iii. Location of the R N A
d. Tombusviridae family
In a cryoelectron microscopy study of CCMV, Fox et al. (1998) concluded that RNAs 1, 2 and 3 +4 were packaged in a similar manner against the interior surface of the virion shell. The viral RNA appeared to have an ordered conformation at each of the quasi-threefold axes. c. Cucumovirus genus
Purified preparations of cucumoviruses contain four RNA species housed in three particles of 3 0 n m diameter in the same arrangement as the bromoviruses. Some isolates of cucumoviruses have associated with them a small satellite RNA (see Chapter 14, Section II.B.2). Using electron microscope methods similar to those employed in their study of BBMV, Finch et al. (1967a) showed that CMV resembled bromoviruses, both in surface structure and in the fact that there is a central hole in the particle.
The 30-nm diameter particles of tombusviruses are composed of subunits of 38-43 kDa encapsidating a single species of genomic RNA. The structure of TBSV has been determined crystallographically to a resolution of 2.9 (Harrison et al., 1978; Olson et al., 1983) and those of TCV and CarMV to 3.2 A (Hogle et al., 1986; Morgunova et al., 1994). Each virus contains 180 protein subunits arranged to form a T = 3 icosahedral surface lattice, with prominent dimer clustering at the outside of the particle, the clusters extending to a radius of about 17 nm. TBSV, CarMV and TCV have very similar structures. The essential features of the structure are summarized in Fig. 5.33. The two distinct globular P and S domains of the protein subunit are connected by a flexible hinge involving five amino acids (Fig. 5.33B).
152
5 ARCHITECTURE
I
A N D ASSEMBLY OF V I R U S P A R T I C L E S
')
N -
R
o j~e
S
P
66
3,5
167
113
B
s2
q2
Fig. 5.33 Architecture of TBSV particle. (A) Order of domains in polypeptide chain from N terminus to C terminus. The number of residues in each segment is indicated below the line. The letters indicate R-domain (possible RNA binding region), arm 'a' (the connector that forms the ~-annulus and extended arm structure on C-subunits and that remains disordered on A- and B-subunits), S-domain, hinge, and P-domain. (B) Schematic of folded polypeptide chain, showing P-, S- and R-domains. (C) Arrangement of subunits in particle. Here, labels A, B and C denote distinct packing environments for the subunit (see Fig. 5.18). S-domains of A-subunits pack around 5-fold axes; S-domains of B- and C-subunits alternate around 3-fold axes. The differences in local curvature can be seen at the two places where the shell has been cut away to reveal S-domain packing near strict (top) and quasi (bottom) dyads. (D) The two states of the TBSV subunit found in this structure, viewed as dimers about the strict (s2) and local (q2) 2-fold axes. Subunits in C positions have the interdomain hinge 'up' and a cleft between 2-fold-related S-domains into which fold parts of the N-terminal arms. Subunits in the quasi-twofold-related A and B positions have hinge 'down', S-domains abutting, and a disordered arm. Parts (A)-(C) from Olson et al. (1983), with permission; (D) from Harrison ct al. (1978), with kind permission of the copyright holder, (-~ Pr3' and replication 3'>5'; therefore there must be controls to prevent interference between the two processes. Most of the R N A elements involved in replication operate in cis, showing that the template RNA is an integral part of the replication corn-
333
plex. The elements at the 3'-end of the template RNA that initiate (-)-strand synthesis appear to be well defined. This is in contrast to those at the 5'-end of the genomic RNA, which initiate (+)-strand synthesis. It seems likely that ( - ) strand and (+)-strand synthesis are highly co-ordinated and that, once the (+)-strand template has been 'captured' by the in vivo replication complex, the full round of RNA replication will occur. The lack of (+)-strand synthesis in most in vitro replication systems indicates either that an important factor is lost during extraction or that there are conformational constraints imposed by the location of the complex in vivo. The observations on TBSV pX (see Section IV.L.2) indicate that there might be host-specific cis-acting elements. It will be of interest to see whether this is a widespread feature among (+)-strand viruses, and what is the nature of the host specificity.
V. REPLICATION OF NEGATIVE. SENSE S I N G L E . S T R A N D E D R N A VIRUSES A. P l a n t R h a b d o v i r i d a e Jackson et al., 1999)
(reviewed by
Rhabdoviruses have large membrane-bound particles containing a single species of ( - ) sense ssRNA; see Chapter 2 (Section III.G) for a general description, and Chapter 6 (Section VII.A) for genome organization. Basically, the virion RNA is associated with the nucleocapsid protein (N) to form coiled nucleocapsid; a large protein (L), considered to be the replicase, is also associated with the nucleocapsid. The nucleocapsid is encased in the matrix (M) protein which, in turn, is enveloped in a membrane to form the bacilliform particle. Virusencoded glycoproteins (G) extend through this membrane. It has been suggested (Cartwright et al., 1972) that there are structural interactions between the N, M and G proteins. The overall structure and replication resembles that of animal rhabdoviruses but there are some differences. For instance, all vertebrate rhabdoviruses replicate and assemble in the cytoplasm, as do some plant rhabdoviruses,
334
s virus REPLICATIO~
but other plant rhabdoviruses replicate in the nucleus. The ( - ) - s t r a n d genome of rhabdoviruses has two functions, as the template for transcription of m R N A s for individual genes (described in Chapter 7, Section IV.A.1) and as the template for replication via a full-length (+) strand. The polymerase complex undertakes both functions but the switch mechanism is not fully understood, even for the m u c h studied animal-infecting vesicular stomatitis virus (Rodriguez and Nichol, 1999). 1. Cytological observations on replication Because of their large size and distinctive morphology, the r h a b d o v i r u s e s are particularly amenable to study in thin sections of infected cells. Morphologically they appear to fall into three groups. First there are the nucleorhabdoviruses that accumulate in the perinuclear space with some particles scattered in the cytoplasm. With some viruses of this group, structures resembling the inner nucleoprotein cores have been seen within the nucleus. The envelopes of some particles in the perinuclear space can be seen to be continuous with the inner lamella of the nuclear membrane. Fig. 8.18 illustrates this group. I m m u n o g o l d labeling with an a n t i s e r u m against the five structural proteins of PYDV s h o w e d that viral proteins accumulate mainly in the nucleus (Lin et al., 1987). In situ h y b r i d i z a t i o n d e m o n s t r a t e d that ( - ) - s t r a n d genomic RNAs are found only in nuclei of infected plants, whereas the (+)-strand RNA sequences are in both nuclei and cytoplasm (Martins e, aI., 1998). Immunofluorescence and i m m u n o g o l d labeling s h o w e d that N and L proteins were in viroplasms in the nucleus and the M2 protein was more generally distributed within the nucleus. In the second group, the cytorhabdoviruses (e.g. LNYV), maturation of virus particles occurs in association with the endoplasmic reticulum (ER), and particles accumulate in vesicles in the reticulum. Biochemical evidence suggests that the nucleus might be involved in the early stages of infection by members of this group. In the third group, there are structures that appear to be rhabdovirus nucleocapsid cores
Fig. 8.18 Electron micrograph of a thin section of a maize leaf infected with a Hawaiian isolate of MMV. Virus particles apparently budding through the inner nuclear membrane (INM) and through intracytoplasmic extensions (double arrows) of the outer nuclear membrane (ONM); single arrows indicate constriction of the INM. Alignment of particles at P1 and P2 suggests budding on the ONM. Cy, cytoplasm; N, nucleus. Bar = 0.3 ~m. From McDaniel el a[. (1985), with kind permission of the copyright holder, 9 The American Phytopathological Society.
lacking the s u r r o u n d i n g m e m b r a n e (Francki et al., 1985a). When examining rhabdoviruses in the cell it must be remembered that the outer nuclear m e m b r a n e is contiguous with the ER. Thus, n u c l e o r h a b d o v i r u s e s b u d d i n g t h r o u g h the inner nuclear m e m b r a n e into the perinuclear space m a y further be included in vesicles derived from the outer m e m b r a n e and be found in the cytoplasm. Similarly, cytorhabdoviruses that associate with the ER may affect
V.
R E P L I C A T I O N OF N E G A T I V E - S E N S E
the outer nuclear membrane giving an appearance of a nuclear involvement. 2. Nucleorhabdoviruses Most of the studies have been performed on SYNV for which a reasonably detailed picture of its replication has been developed. a. In vitro studies A salt extraction procedure proved effective in isolating an active polymerase complex from SYNV-infected Nicotiana edwardsonii leaf tissue (Wagner et al., 1996). The products of the in vitro polymerase reactions included fulllength, polyadenylated N and M2 mRNAs and (+)-strand leader RNA. Animal rhabdoviruses do not polyadenylate their (+)-strand leader transcript, and it is suggested that this feature of SYNV may reflect its replication in the nucleus. The polymerase complex comprises the N, M2 and L proteins (Wagner and Jackson, 1997) and addition of antibodies to L protein inhibited the in vitro system. The reaction condition of this system favors sgRNA transcription over (-)-sense replication, possibly owing to depletion of N protein during extraction. Some small virus-sense RNAs were formed and it is suggested that the formation of genomic (-) strand is inhibited by specific signal sequences in the (+)-strand RNA (Wagner and Jackson, 1997).
b. Replication The following steps have been described for the replication of nucleorhabdoviruses (Fig 8.19A) (Jackson et al., 1999): 1. On entry into the cell, virus particles associate with the ER and release the nucleocapsid cores into the cytoplasm. 2. It is thought that the nucleocapsid cores enter the nucleus through the nuclear pore complexes. 3. Primary transcription takes place using the L protein incorporated in the nucleocapsid core to give mRNAs that are transported to the cytoplasm and translated. 4. The core polymerase proteins, N, M2 and L, are transported back to the nucleus where
SINOLE-STRANDED
RNA VIRUSES
335
they initiate genomic RNA replication and further mRNA synthesis. 5. Granular electron-dense viroplasms, containing N, M2 and L proteins, form near the periphery of the nucleus and are the site of viral replication 6. In the late stages of replication, M protein associates with the newly synthesized nucleocapsid cores coiling them. This complex then associates with G protein that is concentrated at sites on the inner nuclear membrane. 7. Newly synthesized virus particles bud into the perinuclear space. 3. Cytorhabdoviruses Few molecular details are available on cytorhabdoviruses, most of the information coming from cytological studies. Features of the infection cycle are shown in Fig. 8.19B and are summarized by Jackson et al. (1999). Early in infection, some cytorhabdoviruses (e.g. LNYV) induce blisters on the outer nuclear membrane that develop into small vesicles containing some virus particles. These are not found with other viruses (e.g. BYSMV). In both cases, masses of fibrillar viroplasms are found in the cytoplasm closely associated with dense networks of ER. Newly synthesized virus particles bud into vesicles derived from the ER. The three major structural proteins of FLSV, the G, N and M proteins, were found at the periphery of viroplasms, but only the N protein was detected in the granular matrix of the viroplasm (Lundsgaard, 1992). In a similar immunocytological study, the FLSV G protein was detected in ER and the perinuclear membrane but not in Golgi apparatus (Lundsgaard, 1995). B. T o s p o v i r u s e s The tospovirus genome consists of three RNA segments, L, M and S, enclosed in a membranebound particle (see Fig. 6.11 for genome organization). RNA L is (-)-sense and M and S have an ambisense strategy (see Chapter 7, Section IV.A.2). There are four structural polypeptides (Fig. 5.42). Two are glycosylated (G1 and G2) and are at the surface. The N protein binds to
336
s
VIRUS REPLICATION
uncoating
~r ~,
aooara,us/
golgi apparatus
.f'''-,~"~'--~ plasma
nucl
nucleus
~
,,.~i~~,~.~.,,,.. i jr viral proteins
translation
A r
*
'
/
endr~ umm -"~"
/
I nucleocapsid core 'oplasm :: .:. , Ant~genome
+~ Genome~1_...- proteins viral
space
-9-
"-E
cell wall
cell-to-cell movement complexes "! ~
'lplasmodesmata I'
unc~
~ ~ nucleus )
":":....
IAntigenome
mRNAs
.~"
m
mRNAs
~toplasmic~
il
d,~
"
membranes
"-~~anslation cell-to-cell movement complexes
* ~" ,.., I~ y~plasmodesmata
Fig. 8.19 Model for the replication cycle of (A) plant nucleorhabdoviruses, and (B) cytorhabdoviruses. From Jackson et al. (1999), with permission.
the RNA, and there is a large protein, the RdRp, occurring in a minor amount that is encoded by RNA L. These and other properties place tospoviruses in the Bunyaviridae, a large family of viruses replicating in vertebrates and invertebrates. Little is known about the details of tospovirus replication except that it occurs in the cytoplasm in association with the Golgic stack membranes. Double enveloped particles in the ER cisternae may result from fusion of Golgi-based particles with the ER (Kitajima et al., 1992; Kibbert et al., 1997). It is thought that tospovirus replication is similar to that of other bunyaviruses with the concentration of N protein regulating the switch from the production of viral mRNAs to replication of the genome
(Storms, 1998). For an account of bunyavirus replication, see Elliott (1999).
VI. R E P L I C A T I O N OF D O U B L E S T R A N D E D R N A VIRUSES
A. Plant Reoviridae Plant members of the Reoviridae family are placed in three genera: Phytoreovirus with 12 dsRNA genome segments, the type member being WTV; Fijivirus, with 10 dsRNA genome segments, the type member being FDV; and Oryzavirus with 10 dsRNA genome segments, the type member being RRSV (see Chapter 6, Section VI.A, for details of genome organiza-
Vl.
tion). Little is known about the molecular aspects of plant reovirus replication but it is likely to be similar to that of animal reoviruses (described by Joklik, 1999). 1. Intracellular site of replication Plant reoviruses replicate in the cytoplasm as do those infecting mammals (Wood, 1973). Following infection, densely staining viroplasms appear in the cytoplasm (Fig. 8.20). Viroplasms were present in cells of various tissues of leafhopper vectors infected with WTV as well as infected plant cells (Shikata and Maramorosch, 1967). Immunofluorescence demonstrated the presence of viral antigen in the cytoplasm of cultured leafhopper cells (Chiu et al., 1970). It is not yet possible to relate the in vitro studies on the replication of WTV to the structures seen cytologically. Enzyme digestion experiments and radioautographic assay of the incorporation of 3Hlabeled uridine into maize cells infected with the fijivirus, MRDV, indicate that much of the viroplasm is made up of p r o t e i n - p r o b a b l y viral proteins. Viral RNA appears to be synthesized in the viroplasm, where the mature particles are assembled. The mature particles then
Fig. 8.20 Leaf vein tumor cells of maize experi-mentally infected with MRDV. The three different kinds of inclusions caused by MRDV are easily recognizable: viroplasm (arrows), cytoplasmic tubules along and inside which the virus particles are aligned (double arrows), and part of a virus crystal (top right). From Bassi and Favali (1972), with permission.
R E P L I C A T I O N OF I ) O U B L E - S T R A N D E D R N A V I R U S E S
337
migrate into the cytoplasm where they may (1) remain as scattered particles, (2) form crystalline arrays, or (3) become enclosed in or associated with tube-like proteinaceous structures (Fig. 8.20) (Bassi and Favali, 1972; Favali et al., 1974). The autoradiographic studies failed to implicate the nucleus, mitochondria or chloroplasts in virus replication. A detailed electron microscopic study supported the view that the viroplasms caused by FDV in sugarcane are the sites of virus component synthesis and assembly (Hatta and Francki, 1981c). The viroplasms are composed mostly of protein and dsRNA. Some areas contained numerous isometric particles 50-60 nm in diameter. Some appeared to be empty shells while others contained densely staining centers of dsRNA. These particle types appeared to be incomplete virus particles or cores. Complete virus particles were seen only in the cytoplasm. 2. Packaging of RNAs A problem faced by reoviruses is the selection of progeny for packaging. Each particle contains the genomic complement of one copy each of 10 or 12 dsRNA species. The evidence that every WTV particle appears to contain one copy of each genome segment is: (1) RNA isolated from virus has equimolar amounts of each segment (Reddy and Black, 1973); and (2) an infection can be initiated by a single particle (Kimura and Black, 1972). Thus, what are the macromolecular recognition signals that allow one, and one only, of each of 10 or 12 genome segments to appear in each particle during virus assembly? For example, the packaging of the 12 segments of WTV presumably involves 12 different and specific protein-RNA a n d / o r RNA-RNA interactions. The first evidence that may be relevant to this problem came from the work of Anzola e* al. (1987) with WTV. They determined the structure of a defective (DI) genomic segment 5, which was only one-fifth the length of the functional $5 RNA because of a large internal deletion. However, this DI RNA was packaged at one copy per particle, as for the normal sequence. Thus, they established that the sequence(s) involved in packaging must reside
.338 S
s VIRUS REPLICATION
1 5'
S2
~UCUUU
5' ~ C U C a'
rr162
S7 5' ~ U U C C GCUU 3' ~ C A A G % CGAA=
=
3' ~ G A A A G G A A
=
~ ~,,~ ~
S8
$3 5' ~ U U C C C 3' ~ A c A A G G G
$4
5' 3'
$9 5' ~ U U U C U C C U 3' ~ A C AAAGAGGA =
~aauCca ~ A C
~
UA GGU =,
cA
S 10
s' ~ u u u a u u G a A ~ ~ A'~'AcA'AccU
$11
5' ~ ~ U u u c U A c c = 3 ' ~ ] ~ AC AAAAGGUG G ~
U
$5
5
,
j
,
|
~UUAGGACU
3' ~ u c
~UcCUGA~
s' ~ u U c u c c
S6 3'
~ A
5' ~ U U U CUCUUUA~ 3' ~ m ~ A C AAAGGAGAAAU ~
-~
c GAGG ~
s 12 s' ~II~]G~cUGcuu.~ 3'~AC
ACUU%CGJ~ ~
Fig. 8.21 Terminal sequence domains of the positivesense strands of the 12 genome segments of WTV. The segment-specific inverted repeats near the 5' and 3' termini are oriented to indicate potential base-pairing interactions. The conserved 5' hexanucleotide and 3' tetranucleotide shared by all 12 genome segments are shown in white on black. From Anzola et al. (1987), with kind permission of the copyright holder, 9 The National Academy of Sciences, USA. within 319 base-pairs of the 5'-end of the (+) strand and 205 from the 3'-end. Reddy and Black (1974, 1977) showed that an increase in the DI RNA content in a virus population led to a corresponding decrease in the molar proportion of the normal fragment; that is, the D[ fragment competes only with its parent molecule. Thus, there must be two recognition signals, one that specifies a genome segment as viral rather than host, and one that specifies each of the 12 segments. Anzola et al. (1987) sequenced the 5'- and 3'terminal domains of all 12 genome segments of WTV (Fig. 8.21). They suggested that a fully conserved hexanucleotide sequence at the 5' terminus and a fully c o n s e r v e d t e t r a n u c l e o t i d e sequence at the 3' terminus might form the recognition signals for viral as opposed to host RNA. They also f o u n d segment-specific inverted repeats of variable length just inside the conserved segments (Fig. 8.21). They suggested that these might represent the specific recognition sequence for each i n d i v i d u a l genome segment. A similar inverted repeat was found for segment 9 of RDV (Uyeda et al., 1989).
Other m e m b e r s of the plant reoviruses family each have conserved 5' and 3' sequences (Kudo et al., 1991; M e r t e n s at al., 2000). I n f l u e n z a viruses, w h i c h h a v e s e g m e n t e d ssRNA genomes, also have comparable conserved sequences at the 5' and 3' termini of each genome segment (Stoeckle et al., 1987). These similarities strengthen the idea that the 5'- and 3'-terminal sequences have a role in the packaging of these segmented R N A genomes. However, the problem is by no means solved. If only R N A - R N A recognition is involved, how is this brought about to give a set of 12 dsRNAs for packaging? If p r o t e i n - R N A recognition is important, how are 12 specific sites constructed out of the three proteins k n o w n to be in the nucleoprotein viral core? 3. Replication In Reovirus the ( - ) - s e n s e strands are synthesized by the viral replicase on a (+)-sense template that is associated with a particulate fraction (Acs et al., 1971). These and related results led to the proposal that dsRNA is formed within the nascent cores of developing virus particles, and the d s R N A remains within these particles. If true, this mechanism a l m o s t certainly applies to the plant reoviruses. It implies that the mechanism that leads to selection of a correct set of 12 genomic RNAs could involve the ss (+) strand. Thus, the base-paired inverted repeats illustrated in Fig. 8.21 could be the recognition signals. It m a y be that other virus-coded 'scaffold' proteins transiently present in the developing core are involved in RNA recognition rather than, or as well as, the three proteins found in mature particles. Xu et al. (1989) constructed a series of transcription vectors that allowed production of an exact transcript of $8 RNA and of four analogs that differed only in the immediate 3' terminus. Their experiments provided three lines of evidence supporting the view that the 5'- and 3'-terminal domains interact in a functional way: Nuclease T1 sensitivity assays showed that even a slight c h a n g e in the 3 ' - t e r m i n a l sequence can affect the conformation of the 5' terminus.
VII.
Translation in vitro is slightly decreased by alterations in the 3' terminus, which extends the potential for 3'-5' terminal base-pairing, and is increased by changes that reduce potential base-pairing. Computer modeling for minimal energy structures for six WTV transcripts predicted a conformation in which the terminal inverted repeats were base-paired. Dall et al. (1990) developed a gel retardation assay with which they demonstrated selective binding of WTV transcripts by a component of extracts from infected leafhopper cell cultures. Using terminally modified and internally deleted transcripts, they established that the segment-specific inverted repeats present in the terminal domains were necessary but not sufficient for optimal binding. Some involvement of internal sequences was also necessary. There was no evidence for discrimination in binding between transcripts from different segments. The binding component or components present in extracts of infected cells, which are not present in those of healthy cells, have not yet been characterized.
VII. REPLICATION OF REVERSE T R A N S C R I B I N G VIRUSES The Caulimoviridae is the only family of plant viruses with dsDNA genomes; see Chapter 2 (Section III.A) for a description of the family. In 1979 very little was known about the replication of this group, but since then progress has been very rapid. There have been two main motivating factors. First, it was hoped that these viruses, because of their dsDNA. genomes, might be effective gene vectors in plants. This aspect is discussed in Chapter 16 (Section IX). Second, the realization that the DNA is replicated by a process of reverse transcription made their study a matter of wide interest. The family comprises six genera that form two groups, the 'caulimoviruses' and the 'badnaviruses'. These two groups differ in genome organization but have essentially the same replication methods. Most experimental work has been carried out on the 'caulimovirus'
R E P L I C A T I O N OF REVERSE T R A N S C R I B I N O VIRUSES
339
CaMV and the 'badnavirus' RTBV. Reviews include Hohn et al. (1985), Hull et al. (1987), Pfeiffer et al. (1987), Mason et al. (1987), Hull (1996) and Hohn and Ftitterer (1997). Although the replication of members of the Caulimoviridae is by reverse transcription and, in many respects, is similar to that of retroviruses, it does differ from that of retroviruses in several important points: The replication does not involve integration into the host genome for transcription of the RNA. This is done from an episomal minichromosome. The virus does not encode an integrase gene. The virion DNA is circular dsDNA and not the linear DNA with long terminal repeats characteristic of retroviruses. (This and the previous point relate to the lack of integration.) The DNA phase of the replication cycle is encapsidated rather than the RNA phase, which is encapsidated in retroviruses. Thus, the Caulimoviridae are known as pararetroviruses. As with retroviruses, the replication cycle of pararetroviruses has two phases, a nuclear phase where the viral DNA is transcribed by host DNA-dependent RNA polymerase II; and a cytoplasmic phase where the RNA product of transcription is reverse-transcribed by virusencoded RNA-dependent DNA polymerase or reverse transcriptase (RT) to give DNA. In retroviruses, the RT activity is part of the pol gene which also includes the RNase H activity that removes the RNA moiety of the RNA:DNA intermediate of replication. The pol gene in turn is part of the gag-pol polyprotein that is cleaved by an aspartate proteinase, the gag being analogous to coat protein. In pararetroviruses, the reverse transcriptase and RNaseH activities are closely associated. In badnaviruses the coat protein and pol are expressed from the same ORF but in caulimoviruses they are expressed from separate ORFs (see genome maps in Figs 6.1, 6.3 and 6.4). All plant pararetroviruses encode an aspartate proteinase. A. R e v e r s e t r a n s c r i p t a s e Most studies have been performed on retrovirus pol. The reverse transcriptase has a characteristic motif of tyrosine-isoleucine-aspartic
340
s VIRUSREPLICATION
acid-aspartic acid (YIDD) and several amino acid motifs identify the RNase H domain. Processing of the 66-kDa pol region of h u m a n immunodeficiency virus (HIV) by the aspartate protease removes the RNase H domain giving a heterodimer of 66- and 51-kDa proteins. This enzyme complex has three activities for the conversion of ssRNA to dsDNA, RNA-dependent DNA polymerase, DNA-dependent DNA polymerase and RNase H. The crystal structure of HIV RT has been determined at 3.5 (Kohlstaedt et al., 1992) and it was shown to have a structure resembling a right hand, as described above for RdRp (Section IV.B.1). The first indication that the product of CaMV ORF V is analogous to the retrovirus pol gene came from sequence comparison (Toh et al., 1983). The N-terminal domain of ORF V has an aspartate protease motif (Toruella et al., 1989) that autocatalytically cleaved an N-terminal doublet of polypeptides (20 and 22 kDa) from in vitro translated ORF V transcript. Mutants of the protease active site were not processed. This fitted with other features of the replication cycle as will be described below. ORF V was expressed in yeast as a 60-kDa protein that had RT activity on a synthetic template (Takatsuji et al., 1986). In contrast, expression of ORF V in E. coli gave a protein of 78 kDa, the size expected from that ORE but it did not have any RT activity. An activity gel analysis revealed that RT activity associated with CaMV particles is also 60-kDa (Takatsuji et al., 1992). Deletion analysis showed that removal of between 143 and 185 N-terminal amino acids from the E. coliexpressed protein gave RT activity similar to that of the yeast-expressed protein. This suggests that CaMV RT is translated as an inactive precursor form that is converted to the active form by proteolytic processing. It is presumed that this is by the N-terminal aspartate protease activity. The pol motifs are in the C-terminal part of ORF III of 'badnaviruses' and that of RTBV has been studied in insect cells (Laco and Beachy, 1994). The predicted 87-kDa product was detected and was processed to give a 62-kDa and 55-kDa protein. Sequencing showed that these proteins were N co-terminal. Both proteins exhibited RT and DNA polymerase
activities but only the 55-kDa protein had RNase H activity. The precise weights of the 62and 55-kDa proteins were determined by mass spectrometry (Laco et al., 1995) enabling the C termini to be identified. Mutagenesis of the putative active site of the aspartate protease prevented the 87-kDa protein being processed in insect cells. Using antisera to specific fragments of the RTBV ORF III product, Hay et al. (1994) detected a 13.5-kDa protein corresponding to the aspartate protease in extracts from infected plants. The protease antibody labeled the surface of the virus particle. Antibodies against the RT domain identified proteins of 68, 65 and 56 kDa, the latter two probably corresponding to the 62- and 55-kDa proteins found on expression of this ORF in insect cells (Laco and Beachy, 1994). Mutation of the y1339, D1341 o r D 1342 residues of the RT core motif abolished RT activity, whereas that of the 1134~did not (S.-C. Lee and R. Hull, unpublished observation).
B. Replication of 'caulimoviruses' There are a large number of publications concerned with CaMV nucleic acid replication and the p h e n o m e n o n of reverse transcription. Many of these are referred to in the review articles noted at the beginning of this section. Here, only a few key or recent references will be given. By 1983, various aspects of CaMV nucleic acid replication led three groups to propose that CaMV DNA is replicated by a process of reverse transcription involving an RNA intermediate (Guilley et al., 1983; Hull and Covey, 1983b; Pfeiffer and Hohn, 1983). Here are some of the observations that led to the model: A full-length RNA transcript is produced that has terminal repeats (Covey and Hull, 1981). DNA in virus particles has discontinuities (see Fig. 6.1) while that found in the nucleus does not, but is supercoiled and is associated with histones as a minichromosome (M~nissier et al., 1982; Olszewski et al., 1982). dsDNA exists in knotted forms (see Fig. 4.10). Other forms of CaMV DNA in the cell are not encapsulated, such as an ss molecule of 625 nucleotides with the same polarity as the
VII. REPLICATION OF REVERSETRANSCRIBING VIRUSES
(z-strand covalently linked to about 100 ribonucleotides (Covey et al., 1983). Since 1983, a detailed picture of the replication of CaMV has been built up. 1. Replication pathway The replication pathway is outlined in Fig. 8.22. Essentially, the replication has two phases: transcription of an RNA template from the virion DNA and then reverse transcription of the RNA template to give dsDNA. The transcription phase occurs in the nucleus and the reverse transcription phase in the cytoplasm. In the first phase of replication, the dsDNA of the infecting particle moves to the cell nucleus, where the overlapping nucleotides at the gaps
9
90 ~
90 ~
9
9
341
are removed, and the gaps are covalently closed to form a fully ds DNA. The covalently closed DNA associates with host histones to form minichromosomes that are the template used by the host enzyme, D N A - d e p e n d e n t RNA polymerase II, to transcribe two RNAs of 19S and 35S, as indicated in Fig. 6.1. This is described in more detail in Chapter 7 (Section IV.C.1). The two p o l y a d e n y l a t e d RNA species migrate to the cytoplasm for the second phase of the replication cycle that takes place in the viroplasms (e.g. Mazzolini et al., 1985). The 19S RNA is the mRNA for gene VI that is translated in large amounts to produce the viroplasm protein. Gene VI is the only Caulimovirus gene to be
CYTOPLASM
tRNA
9
4._._.~
SE
,,
o
....
/
,
35S RNA --
mRNA
1
TRANSLATION
0 O0
,, ./
~
INCLUSION BODY ,,
Fig. 8.22 Diagram of the replication cycle of CaMV.
342
s VIRUSREPLICATION
transcribed as a separate transcript from its own promoter, suggesting that it may have an important role at an early stage following infection (Gowda et at., 1989). Mutagenesis of the coding part of gene VI showed that it was the protein product rather than the mRNA that was responsible for transactivation which is described in more detail in Chapter 7 (Section V.B.7). To commence viral DNA synthesis on the 35S RNA template, a plant methionyl tRNA molecule forms base-pairs over 14 nucleotides at its 3'-end with a site on the 35S RNA corresponding to a position immediately d o w n s t r e a m from the D1 discontinuity in the R-strand DNA (see below). The viral reverse transcriptase commences synthesis of a DNA ( - ) strand and continues until it reaches the 5'-end of the 35S RNA with the RNase H activity removing the RNA moiety of the RNA:DNA duplex giving what is termed 'strong-stop DNA'. At this point, a switch of the enzyme to the 3'-end of the 35S RNA is needed to continue the copying. The switch is made possible by the 180nucleotide direct repeat sequence at each end of the 35S RNA which enables the 3'-end of the strong-stop DNA to hybridize with the 3'-end of the 35S RNA. When the template switch is completed, reverse transcription of the 35S RNA continues up to the site of the tRNA primer, which is displaced and degraded to give the D1 discontinuity in the newly synthesized DNA. The rest of the used 35S template is removed by an RNase H activity. In this process, two polypurine tracts (PPT) of the RNA are left near the position of discontinuities D2 and D3 in the second DNA strand (+ strand). Synthesis of the second (+) strand of the DNA then occurs, initiating at these two RNA primers. The growing (+) strand has to pass the D1 gap in the ( - ) strand, which again involves a template switch. There are several observations that support and enhance this model for CaMV replication (reviewed in Hohn and Ftitterer, 1997): Both ( - ) - and (+)-strand DNA synthesis are resistant to aphidocolin, an inhibitor of DNA -+ DNA synthesis. RT activity is associated with viral inclusion bodies and virus particles.
Various unencapsidated nucleic acid molecules interpreted as being replication intermediates have been isolated. These include 'strong-stop' DNAs which have ribonucleotides at the 5'-end, DNA molecules that are partially double- and partially single-stranded, compatible with being products of defective replication and hairpin structures (Turner and Covey, 1988). Replication intermediates are associated with apparently incomplete virus particles (Thomas et al., 1985; Marsh and Guifoyle, 1987; Ftitterer and Hohn, 1987). The finding of RT activity in inclusion bodies and virus particles, and of replication intermediates in virus particles, indicates that, as with retroviruses, the reverse transcription of CaMV occurs in particle-like proviral structures. 2. Inclusion bodies As noted in an earlier section, CaMV (and other caulimoviruses) induce characteristic inclusion bodies or viroplasms in the cytoplasm of their host cells (Fig. 8.23). There are two forms of inclusion bodies, electron-dense ones that are made up of ORF VI product, and electron-lucent ones that are made up Of ORF II product (Espinoza et al., 1991). Virus particles are found in both types of inclusion body. The electron-dense inclusion bodies are the site for progeny viral DNA synthesis and for the assembly of virus particles; it is not known whether virus replication takes place in the electron-lucent inclusion bodies. Viral coat protein appears to be confined to the inclusion bodies and most virus particles are retained within them. At an early stage in their development, the ORF VI product inclusion bodies appear as very small patches of electron-dense matrix material in the cytoplasm, surrounded by numerous ribosomes. Larger inclusion bodies are probably formed by the growth and coalescence of the smaller ones, leading to mature inclusion bodies that vary quite widely in size from about 0.2 to 20 ~m in diameter. They are usually spherical and are not membrane-bound. They often have ribosomes at the periphery and consist of a fine granular matrix with some electron-lucent areas not bounded by membranes. Virus particles are
vii. REPI,ICATION OF REVERSE TRANSCRIBING VIRUSES
343
Fig. 8.23 Electron micrographs of the inclusion bodies of CaMV Cabb B-JI in infected turnip leaves immunogold labeled with anti-P62 (ORF VI product) antiserum. (A) Electron-dense inclusion body with gold particles preferentially labeling the inclusion body matrix (bar = 200 nm). (B) Cell showing an electron-dense inclusion body (filled-in star) heavily labeled and an electron-lucent inclusion body (open star) without gold particles contained within the same cell (bar = 1 ~tm). (C) An electron-lucent inclusion body showing the lack of gold particles (bar = 500 nm). From Espinoza et al. (1991), with permission.
344
8 VIRUS REPLICATION
present in scattered or irregular clusters in the lucent areas and the matrix. Little is k n o w n about the w a y CaMV particles are assembled. No e m p t y virus shells are found in infected tissue. These observations suggest that encapsulation m a y be closely linked to D N A synthesis. The role of glycosylation and p h o s p h o r y l a t i o n of the coat protein remains to be determined. 3. Discontinuities The D N A of 'caulimoviruses' has gaps or discontinuities at specific sites, one (D1) in one strand (the + or a strand) and one or more in the other strand. Those of CaMV have been studied in detail and s h o w n to comprise an overlapping sequence with the 5'-end being in a fixed position and the 3'-end being in a variable position giving an overlap varying b e t w e e n 8 and 40 nucleotides (Fig. 8.24) (Richards et al., 1981). As explained above (Section VII.B.1), these discontinuities arise from replication where the advancing D N A strand reaches the priming site. Thus, D1 is at the tRNA priming site
8010 G1
G2
5,_ 3i___
C. Replication of 'badnaviruses' The replication of RTBV, the most studied of the 'badnaviruses', is similar to that of CaMV in most respects and is s u p p o r t e d by the detection
1
20 9 9 HoTTTTTTAACCATAGTCTCGGTACTTAGCC ---5i
AG CdCC C GCTTAAAAAATTGGTATCAGAG C CATGAATCG G ---3i
TCGGGGGCGAATTTTTTA5i 4200
4220
5 i--- TATTC TTTC AGAG G G GAG GAG GAG G TTATC AGAAG AAAAA OH 3i---ATAAGAAAGTCTCCCCTCCTCCTCCAATAGTCTTCTTTTTGA---5i 5iAGGTTATCAGAAGAAAAACT---3i
1620
G3
for ( - ) - s t r a n d synthesis and the discontinuities in the other strand are at the PPTs of RNA, generated by RNase H cleavage that give (+)strand priming. In an analysis of the PPTassociated (+)-strand priming, N o a d ei al. (1998) s h o w e d that altering the length of the 13-base-pair PPT by +25 To significantly reduced priming efficiency but did not affect the site of the 5'-end of the new (+)-strand D N A which is 3 nucleotides from the PPT 3'end. There is a short pyrimidine tract 5' to the PPT that plays an important role in PET recognition in vivo. N o a d et al. (1998) propose a model for pararetroviral (+)-strand priming in which the pyrimidines enhance the PPT recognition d u r i n g RNase H cleavage, and suggest that the fidelity of primer maturation involves PPT length m e a s u r e m e n t and 3'-end recognition by the RNase H.
1640
5 i--- C C ATTTTTAAG AG TG G G G G G G TTG ATTAC TGCA OH 3 i--- G G TAAAAATTC TCAC C C C C C CAAC TAATG AC G TCGG TTG A--- 5 i 5 iG GTTGATTACTG CAG C CAACT---3 i
Fig. 8.24 Structure of CaMV gaps or discontinuities. G1 (also referred to as D1) is on the transcribed ~ strand and G2 (D2) and G3 (D3) are on the complementary strand. For each of the gaps the upper and lower sequences are those of discontinuous strand and the middle sequence the unbroken strand. These sequences are the most common found, but for each the 5' terminus is at a fixed position; the 3' terminus may vary from the shown position. The numbers above each sequence are the positions on the CaMV sequence. From Richards et al. (1981), with permission.
VIII.
of characteristic replication intermediates (Bao and Hull, 1994). The 'strong-stop' DNA was shown to have methionine initiator tRNA at its 5'-end (Bao and Hull, 1993b). However, the discontinuity at the (+)-strand priming site did not map to the PPT site predicted from the sequence but to a site about 1400 base-pairs away (Bao and Hull, 1992). Furthermore, both the 5' and 3' termini at this discontinuity were heterogeneous in position giving structures varying from a gap of 10 nucleotides to an overlap of 103 nucleotides.
VIII. REPLICATION OF SINGLE. STRANDED D N A VIRUSES There are two families of plant viruses that have ssDNA genomes, the Geminiviridae and the Circoviridae. The replication of members of these two families is ssDNA ~ ssDNA via a dsDNA stage. Many of the features of replication of the two families are similar but there are some differences. Most is known about replication of the Geminiviridae. There are four genera in the Geminiviridae, the genome organizations of which are described in Chapter 6 (Section V.A). Three of the genera, the mastreviruses, the curtoviruses and the topocuroviruses, have monopartite genomes whereas many of the begomoviruses have bipartite genomes. However, DNA A of bipartite begomoviruses contains all the information necessary for virus replication, the genes on DNA B encoding proteins involved in movement to the nucleus and between cells (as described in Chapter 9, Sections II.C and II.D.5). Therefore, with the caveat that the nuclear localization properties of DNA B B V 1 0 R F are required for nuclear shuttling, for discussing replication, DNA A of these bipartite viruses can be considered to be comparable to the DNAs of the monopartite viruses.
R E P L I C A T I O N OF S I N G L E - S T R A N I ) E I ) I ) N A VIRUSES
345
agroinfection of whole plants, and transfection of protoplasts. Grimsley et al. (1987) showed that Agrobacterium containing tandem repeats of MSV DNA inoculated to whole maize plants led to symptoms of MSV infection. Since MSV DNA is not mechanically transmissible, and intact virus can infect only by means of an insect vector, this experiment provided a very sensitive demonstration that Agrobacterium could interact with a monocotyledon. Elmer et al. (1988b) adapted Grimsley et al.'s agroinfection procedure to provide a simple and efficient assay for TGMV replication. They produced transgenic Nicotiana benthamiana plants containing multiple tandem copies of TGMV B-DNA. They found that an inoculum containing as few as 2000 Agrobacterium cells TGMV A-DNA could produce 100% virus infection. Agroinfection is also a highly efficient way of introducing DNAs A and B together into N. benthamiana (Hayes et al., 1988e). The technique has also been extended to other genomes including Digitaria streak virus (Donson et al., 1988), MSV in various species of Poaceae (Boulton et al., 1989), and ACMV (B. A. M. Morris et al., 1988a; Klinkenberg et al., 1989). The use of protoplasts to study plant viruses is described in Section III.A.5. Both protoplasts and agroinfection allow the effects of mutations on the viral genome to be explored.
B. In vivo observations on geminiviruses Geminivirus particles usually accumulate in the nucleus, and with some, such as MSV, large amounts of virus accumulate there. In some infections, fibrillar rings, which must be part of a spherical structure, appear in the nucleus (Francki et al., 1985a) but their composition and significance are not known. Nuclei isolated from Nicotiana tissue infected with TGMV synthesized variable amounts of (+) and ( - ) strands of both DNAs A and B (Coutts and Buck, 1985).
A. Methods for studying geminivirus replication
Novik, 1998)
Two methods have been of great use in elucidating the details of geminivirus replication:
Two initial observations pointed to geminiviruses using the rolling-circle mechanism of
C. Rolling-circle replication (reviewedby
346
s VIRUSREPLICATION
replication. One observation came from twodimensional electrophoresis of extracts of ACMV-infected plants that revealed five putative replication intermediates (Saunders et al., 1991): 9 subgenomic (-)-strand DNA associated with genomic (+)-strand DNA; 9 unit-length (-)-strand DNA; 9 virion DNA ranging from one to two genome lengths-concatemeric virion ss DNA, ds DNA, and partially ss DNA. The other observation came from the replicational release of the BCTV genome that had beeen agro-inoculated to Nicotiana benthamiana as a tandem construct (Stenger et al., 1991). Rolling-circle replication (see also Chapter 14, Section I.D) is common in the replication of bacterial viruses and plasmids. It is a two-step process, in the first phase of which the ss (+) strand is the template for the synthesis of ( - ) strand to generate a ds, replicative form (RF). This RF has two functions. It is the template for transcription as described in Chapter 7 (Section IV.E), and it is the template for (+)-strand synthesis generating free ssDNA. The priming of (-)-strand synthesis is usually by an RNA molecule that is generated through RNA polymerase or DNA primase activity. (+)-strand synthesis is primed by a site-specific nick in the (+) strand of the RF. D. G e m i n i v i r u s replication (reviewed by Hanley-Bowdoin et al., 1999; Guticrrcz, 1999, 2000a) Many features of the geminivirus replication cycle have recently been elucidated, though there are still several points that are poorly understood. The elucidation of the replication cycle has revealed several aspects of the normal cell cycle as geminivirus replication depends upon many host functions. Of especial interest is that geminiviruses replicate in differentiated cells that are in the G phase and have shut down most of their DNA replication activities. Thus, geminiviruses reactivate the replication activities that they require and convert the cell back to S phase. The geminivirus replication cycle is outlined in Fig. 8.25.
1. Minus-strand synthesis A small oligonucleotide complementary to the 3' intergenic region has been isolated from several mastreviruses (Donson et al., 1984; Hayes ef al., 1988c; Morris et al., 1992). This oligonucleotide can be extended by DNA polymerase in vitro and may be the in vivo (-)strand primer. This is supported by the finding that sequences in the 3' intergenic region of WDV have been implicated as being involved in replication (MacDonald et al., 1988b; Kammann et al., 1991). No analogous molecules have been found in curtoviruses or begomoviruses. The two-dimensional electrophoretic analysis of ACMV replication intermediates (Saunders et al., 1991) indicated that (-)-strand synthesis is primed within the 5' intergenic region. The primers found in the mastreviruses and begomoviruses contained ribonucleotides. Little is known about the proteins and mechanisms of (-)-strand synthesis. It is most likely that the virus particles are targeted to the nucleus (see Chapter 9, Section II.C) where this stage of the replication occurs. As no viral proteins other than the coat protein have been detected in virus particles, and as the coat protein is not required for replication (Elmer et at., 1988a; Woolston et al., 1989), it is generally believed that (-)-strand synthesis is effected by host factors. 2. Plus-strand synthesis The priming of geminivirus (+)-strand synthesis is through a DNA cleavage at a specific site in vivo. The progeny of plants infected with heterodimers of different strains of BCTV, WDV or ACMV (Stenger el al., 1991; Heyraud et al., 1993a,b; Stanley, 1995) were shown to comprise predominant genotypes dependent on the arrangement of the parental genomes. The sequences of the progeny were consistent with (+)-strand DNA synthesis by rolling circle initiating or terminating within the conserved hairpin sequence in the 5' intergenic region (see Chapter 6, Section V.A.2, for details of this common region). The initiation site was mapped to the conserved nonanucleotide sequence, TAATAATTSAC, in the loop of the hairpin (Stanley, 1995). The geminivirus Rep protein is
VIII. REPLICATIONOF SIN(3LE-STRANDEI)DNAVIRUSES 347 complement~w-sen=e strandsynthesis
7
(3
virus particle
~
RNA primer / synthesis
@
~t-t,-OJmt tr~m,~/8~ion
(coat protein)
ceil-to-celland va~.d=r =pread (BC1/BVI?)
J continuedstranddisplacement to produceconcatemericD ~
primerdisplacement and gap sealing
~ ~ . release Of ' ViruS-SenSestrand
( J
)
0
nicking of virus-sensestrand (ACI?)
transcriptionally active rninichrornosorne (AC2/AC3control?)
stranddisplacement
Fig. 8.25 Diagram of replication of a geminivirus. Kindly provided by J. Stanley. a site-specific endonuclease that nicks and ligates (+)-strand viral DNA at the same position in vitro (Laufs et al., 1995; Orozco and Hanley Bowdoin, 1996). There are two major categories of molecular organization of the (+)-strand DNA replication origin. The replication origin for mastreviruses consists of a large cis-acting region where the Rep protein forms multiple complexes (Castellano et al., 1999) and that for the begomoviruses contains one binding site for Rep (Fontes et al., 1992, 1994b; Lazarowitz et al., 1992; Orozco and Hanley-Bowdoin, 1998). 3. Plus-strand origin The origin of TGMV (+)-strand synthesis has been studied in detail (reviewed in HanleyBowdoin et al., 1999) and has been compared with those of other begomoviruses and of curtoviruses. The features of the TGMV (+)-strand origin is illustrated in Fig 8.26. The origin is in the left-hand side of the common region and
overlaps the promoter for AC61 (also termed AL61) (described in Chapter 7, Section IV.C.2). Six cis elements have been identified in this region. 1. The hairpin element is common to all geminivirus genomes (Arguello-Astorga et al., 1994). This comprises a GC-rich stem and an AT-rich loop, and mutagenesis demonstrates that it is the structure of the stem rather than the sequence that is essential for its activity. The stems of mastreviruses are much longer than those of members of the other two genera. The 5'-TAATATTAC loop is conserved in all geminiviruses and is found in the (+)strand origins of other nucleic acids that replicate by rolling circle. There is some sequence flexibility in the loop sequence but the cleavage is between the TTSAC as noted above. 2. The binding site for Rep has several features: (i) It is virus-specific (Table 8.5) but has some conserved sequence. (ii) The GGAT repeat is an absolute requirement. (iii) The spacing
348
s v i r u s REPLICATION
conserved sequence
T......
T
1Ti
j A / " I ~ (+) DNA TIC J CG CG TA
A T hairpin co c oc
AL61 RNA ~ iC~'AAA~ATA~OAAT~OTAOTAA~O>+ 7.4 the predominant form was a monomer. Preformed oligomers interacted very poorly with DNA. The authors suggested that there was a stepwise assembly of the protein-DNA complex with the monomers interacting with the DNA and then with other monomers to assemble the oligomeric structure. Secondly, Rep binds to the (A)C3 protein. The (A)C3 protein locates to the nucleus and enhances DNA accumulation of begomo- and curtoviruses. It is thought that this enhancement is through the binding to Rep (HanleyBowdoin et al., 1999). Thirdly, Rep binds to retinoblastoma factor from both plants and animals (see below). Finally, RepA binds to GRAB proteins (geminivirus RepA-binding proteins) (Xie et al., 1999). Using the WDV RepA protein as a bait in the yeast two-hybrid system (Fig. 8.5), a family of GRAB proteins was isolated from wheat suspension cultured cells. The 37 amino acids at the C terminus of RepA, a region conserved among other mastreviruses, is involved in the interaction with an N-terminal domain of the GRAB proteins. The expression of GRAB1 or GRAB2 proteins in wheat cells inhibits WDV replication. GRAB proteins have significant
amino acid homology with the NAC domain of proteins involved in plant development and senescence. 5. Geminivirus control of the cell cycle (reviewed by Gutierrez, 2000b) As noted above, geminiviruses replicate in differentiated plant cells in which host DNA replication has ceased. The viral replication is dependent upon host DNA replication factors, and thus the cell cycle has to be modified. There are several lines of evidence suggesting that the Rep protein is involved in this modification (reviewed in Hanley-Bowdoin et aI., 1999). Many Rep proteins are recalcitrant to stable constitutive expression in transgenic plants. In those plants in which expression does occur, and in plants infected with TGMV, the nucleus becomes round and migrates to the cell center, features associated with de-differentiation. Rep proteins (or RepA of mastreviruses) bind to retinoblastoma (Rb) proteins from a variety of sources including plants (Horvath e, al., 1998; Liu et al., 1999b). Animal Rb proteins regulate cell growth most probably through control of the transition of the G0/G1 into S phase of the cell cycle (reviewed by Gutierrez, 1998; de Jager and Murray, 1999). It is thought that the plant analogs of Rb proteins have a similar function. Various animal DNA viruses control their host cell cycle through the binding of a virusencoded protein with the host Rb protein through a LxCxE motif (reviewed by HanleyBowdoin et al., 1999; Gutierrez, 2000a). Curtovirus and begomovirus Rep proteins have this LxCxE motif as does the RepA/Rep protein of mastreviruses. However, CSMV does not have a LxCxE motif (Horvath et al., 1998) and, although mutagenesis of this motif in BYDV reduces the Rep binding capacity, it does not abolish the replication of the virus (Liu et al., 1999b). As noted above, the curtovirus and begomovirus Rep proteins and the mastrevirus RepA protein bind Rb proteins from various sources, and it is likely that this binding is in a similar manner to that of animal viral proteins to their host Rb protein. Thus, the suggestion is that the binding inhibits the Rb protein control that maintains the host cell in the G
VIII.
phase of the cell cycle, enabling it to return to S phase and produce the factors required for viral replication. However, for this to occur Rep must be expressed from the incoming virus. Therefore there must be enough capability in the newly infected cell to initiate ( - ) - s t r a n d synthesis to give the dsDNA for transcription of the mRNA for Rep. Details of this initial event have not yet been determined, but a model is shown in Fig. 8.28.
E. Nanovirus replication As described in Chapter 6 (Section V.B.1), the genomes of nanoviruses are distributed over at least six small circular ssDNA species, each of which (with one exception) has a single ORF. Although they have not been studied in as much detail as geminiviruses, nanoviruses have several features in c o m m o n with gemiCDK J Cyc E2F
.1. (.:+)
~i:: E2F
.........,
Geminivirus RBR - binding protein
RepA (LxCxE)
( Mastrevirus )
G~ iS
Transition
S-pha~e
@
Rep ( non-LxCxE )
( Begomoviru$ ) ( Curtovirus ? )
Fig. 8.28 A proposal for the two mechanisms that geminiviruses may use to interfere with the retinoblastoma-related (RBR)/E2F pathway. The G1/S transition is normally regulated by phosphorylation of RBR by CDK/cyclin complexes. RBR phosphorylation releases the RBR-bound E2F (the question mark indicates that a DP-like protein has not been identified yet in plants) transcription factors required for G1/S transition and S-phase functions. Geminivirus proteins (RepA in mastreviruses and Rep in begomoviruses) are proposed to sequester RBR and release RBR-bound factors, thus bypassing the normal cellular control. Mastrevirus RepA protein interacts with RBR through its LxCxE motif while begomovirus, and most likely curtovirus, Rep does so using a different motif. From Gutierrez (2000b), with kind permission of the copyright holder, 9 Kluwer Academic Publishers.
REPLICATION
OF SINGLE-STRANDED
DNA
VIRUSES
351
niviruses that suggest that their replication mechanisms are very similar. Each of the nanovirus DNA species has a c o m m o n region that is predicted to form a stem-loop, the loop containing the sequence 5'TANTATTAC-3' (see Burns et al., 1995) which is found in the origin of geminivirus (+)-strand synthesis (see Section VIII.D.2). At least one of the DNA species is inferred from the amino acid sequence to encode a Rep protein (see Table 6.1) (see Harding et al., 1993; Wu et at., 1994b; Sano et al., 1998; Timchenko et al., 1999). In vitro tests with E. coli-expressed protein showed that the BBTV Rep protein has site-specific cleaving and joining activity (Hafner et at., 1997a). In FBNYV, five of the DNA segments appear to encode a Rep protein (Timchenko et al., 1999). Site-specific D N A cleavage and nucleotide transfer activities have been shown in vitro for those from DNAs I and 2 (Rep 1 and 2) and the essential tyrosine residue that catalyzes these reactions has been identified by mutagenesis. Rep I and 2 proteins hydrolyze ATP and this activity is essential for multiplication of the viral DNA. Each of the five Rep proteins initiated replication of the DNA species by which it was encoded, but only Rep2 was capable of replication of all the six DNAs that did not encode a Rep protein. Thus, only one of the Reps is a master Rep, and this is capable of triggering replication of heterologous nanovirus DNAs (Timchenko et al., 2000). Nanoviruses face the same problem as geminiviruses in that they need to start replicating their DNA in cells that are not transcriptionally active. The DNA of SCSV is able to self prime ( - ) - s t r a n d synthesis (Chu and Helms, 1988). Using self-primed extension with a DNAd e p e n d e n t D N A polymerase, Hafner et at. (1997b) showed that all six DNAs of BBTV had endogenous primers bound to the genomic DNA. These primers were heterogeneous in size and appeared to be derived from DNA5, and that DNA self-primed more efficiently than the other DNAs. It is suggested that the function of the protein encoded by DNA5 is important early in the infection process. The product of BBTV DNA5 contains the LxCxE motif characteristic of the retinoblastoma (RB)-binding protein described above
352
s VIRUSREPLICATION
(Section VIII.D.5) (Wanitchakorn et al., 2000). The yeast two-hybrid system showed that this protein has RB-binding activity, and the activity is dependent upon the LxCxE motif. None of the five Rep proteins of FBNYV contains the LxCxE motif. However, the 20-kDa protein encoded by DNA10 does contain this motif and also an F-box associated with binding to a ubiquitin-ligase (a plant SKP1 homolog) (Aronson et al., 2000). The protein from DNA10, named Clink (cell cycle link), binds to RB and stimulates viral replication; the product of BBTV DNA5, described above, is a homolog of Clink. However, Clink is not an absolute requirement for infection of Nicotiana benthamiana and it is likely that this virus encodes one or more cell cycle-modulating activity. From its association with a constituent of the ubiquitin-protein turnover pathway, it is suggested that, as well as blocking the action of the RB protein, it targets that protein for processing.
IX. M U T A T I O N A N D RECOMBINATION The main two ways by which faults arise in replication is by mutation and recombination. | shall discuss here the mechanisms leading to these faults, and in Chapter 17 the impact that these faults in replication have on the variation and evolution of plant viruses. Recombination is also involved in various other viral phenomena, which will be described here. A. M u t a t i o n (reviewed by Domingo and Holland, 1997; Drake et al., 1998) Replication mutations can be either base substitutions, base additions or base deletions. In discussing mutations, one has to distinguish between mutation frequency and mutation rate (see Domingo, 1999). Mutation frequency is the proportion of mutants (averaged for an entire sequence or specific for a defined site) in a genome population. Mutation rate is the frequency of occurrence of a mutation event during genome replication. The relationship between mutation frequency and mutation rate is discussed by Drake and Holland (1999). Here
I will discuss mutation rate, but the frequency is important in t h e a n a l y s i s of variation and evolution. The rate of mutational errors depends on the mode of replication, the nucleotide sequence context and environmental factors. As shown in Fig 8.29, nucleic acids that replicate DNA --* DNA have much lower mutation rates than those that replicate by other pathways. This is because DNA-dependent DNA polymerase has a proofreading ability that checks that the correct nucleotide has been added, whereas the other polymerases (DNAdependent RNA polymerase, RdRp and RT) do not (Steinhauer et al., 1992). The crystal structures of RNA replicases and RT do not reveal the 5' to 3' exonucleolytic proofreading domain present in D N A - d e p e n d e n t DNA polymerases (Kohlstaedt et al., 1992; Joyce and Steitz, 1994; Hansen et al., 1997). Most of the studies on error rate have been undertaken in vitro and these have shown that parameters such as ionic composition of the medium and relative concentration of nucleoside triphosphate substrates can have significant effects (Domingo and Holland, 1997). Similarly, in vitro studies show that the sequence context being copied can have an effect with some regions being hypermutagenized. There is little evidence on which to judge the significance of these effects on plant viruses in vivo. In an attempt to minimize selection pressure, Kearney et al. (1993) measured the drift in two foreign sequences (dihydrofolate reductase and neomycinphosphotransferase II) cloned into a TMV vector. It was suggested that these foreign sequences would not be subject to the selection constraints of the -3 -4
10 -
DNA
10
=
RNA T 10 -3 -
10"' 1 DNA
....
-3
10 -
10
-4
10 -4
~ RNA
Fig. 8.29 Error rates of transcription within and between RNA and DNA.
IX. MUTATION AN[) RECOMBINATION
viral sequence, and it was. determined that the accumulated rate w a s ,,o.
>e-, 0
z z z ~ ,~ J:l ,,Q OOO
CP
t-
oo
_J ..J ..I
II amino acid 692
MP 1 .....
!11
I!
CD
I
9
\amino acid 1116
.... -"
Conserved NTP Bind~'ngand HeticaseMotiis ....
20"C
22"C
24~
26~
TMV
+t,
+
+
+
+
O b N L - 1,1
+
+
+
+.
ObNL-3.1
+
+
+
+
ObNL-2.1 ObYW-2-2
+ +
+ +
. .
Ob
.
.
150
elicitation and virus localization
HR 18"'C
.
. .
.
.
.
.
.
.
. .
CD
_h
.
.
_ .
.
LRR
216-224 297-301 320-325
590
928
1144
28"C
-
.
NBS
_1
" AI' 8
B Virus"
Toll protein a n d the interleukin-1 receptor in m a m m a l s , a n u c l e o t i d e b i n d i n g site a n d four imperfect leucine-rich regions (Fig. 10.4). Thus, it belongs to the TIR-NB-LRR class of R genes (see Fig. 10.2). A deletion analysis s u g g e s t e d that the TIR, NB a n d LRR d o m a i n s all play an i n d i s p e n s a b l e role in the i n d u c t i o n of H R ( D i n e s h - K u m a r et al., 2000). The N gene is e x p r e s s e d from t w o transcripts, N s a n d N c, via alternative splicing p a t h w a y s ( D i n e s h - K u m a r a n d Baker, 2000). The N s transcript codes for the full-length N protein a n d is m o r e p r e v a l e n t before, a n d for 3 hours after, TMV infection. The N L transcript codes for a t r u n c a t e d N protein (Ntr), lacking 13 of the 14 leucine-rich repeats, a n d is m o r e p r e v a l e n t 4-8 h o u r s after infection. A TMV-sensitive tobacco variety t r a n s f o r m e d to express the N protein but not the N tr p r o t e i n is susceptible to TMV, w h e r e a s transgenic plants expressing both N s a n d N L transcripts are completely resistant.
.
.
" V i r u s e s w e r e t e s t e d for i n d u c t i o n o f the h y p e r s e n s i t i v e r e s p o n s e ( H R ) and l o c a l i z a t i o n to i n o c u l a t e d l e a v e s o f t o b a c c o cv, X a n t h i N N .
8
150
NBS
i
iill
216-224 297-301 320-325
LRR
I
I
(+). FIR a c c o m p a n i e d b y v t n i s I o c a l i z a t i o n s ( - ) , s y s t e m i c v i r u s s p r e a d ,
Fig. 10.3 Identification of HR-inducing region in TMV. (A) Mutations of the 126/183-kDa gene of TMV strain Ob that lead to induction of the HR. Hatched region is the genomic map representing the portion of the TMV replicase (amino acids 692-1116) that is required for induction of the N gene-mediated HR. At the bottom is an expanded view of this region of the Ob replicase, indicating the positions in Ob that result in HR. Shaded bars represent the nucleotide triphosphate-binding domain and helicase consensus motifs. Asterisk denotes the amber termination codon in the 126/183-kDa protein. (B) Temperature sensitivity of N against TMV Ob and ObNL mutants. From Padgett et al. (1997), with kind permission of the copyright holder, @ The American Phytopathological Society.
Fig. 10.4 Schematic diagram of the N and N tr proteins. An analysis of N protein amino acid sequence identified three domains of possible functional significance. These domains are indicated and shown to scale within the full-length N protein. CD, putative cytoplasmic domain of N with sequence similarity to the cytoplasmic domains of Toll, interleukin-lR and MyD88; NBS, putative nucleotide-binding site comprising three motifs from amino acid 216 to 325; LRR, leucine-rich repeat region consisting of 14 imperfect tandem leucinerich repeats. Alternative splicing yields a truncated protein, N tr. N tr is identical to the N-terminus of N except for the C-terminal 36 amino acids, indicated by the black box next to the LRR on the right. From Dinesh-Kumar and Baker (2000), with kind permission of the copyright holder, 9 The National Academy of Sciences, USA.
II1.
However, the ratio of N s t o N L mRNAs before and after TMV infection is critical as the expression of either one mRNA alone or the two at a 1:1 ratio gives incomplete resistance. It is suggested that the relative ratio of the two N messages is regulated by TMV signals (DineshKumar and Baker, 2000). The N gene has been transferred to tomato, where it confers resistance to TMV (Whitham et al., 1996). E. O t h e r v i r a l - h o s t h y p e r s e n s i t i v e responses 1. TMV and the N' gene (reviewed in Culver et al., 1991) The N' gene, originating from Nicotiana sylvestris, controls the HR directed against most tobamoviruses, except U1 (vulgare) and OM strains that move systemically and produce mosaic symptoms in N'-containing plants. However, mutants that induce necrosis can easily be isolated as spontaneous mutants of virus strains causing systemic symptoms. The TMV coat protein gene is involved in the induction of the N' gene HR (Saito et al., 1987a). Five independent amino acid substitutions have been identified as being involved in the HR elicitation (Culver and Dawson, 1989; Culver et al., 1991), but the HR response varied for each mutation, from strong elicitors producing visible necrosis in 2-3 days to weak elicitors requiring at least 6 days for necrosis to appear. This suggested that the structure of the coat protein might influence the response. To investigate this, Culver et al. (1994) made a set of amino acid substitutions that would have predicted structural effects on the coat protein, and examined their HR. Substitutions eliciting the HR were within, or would predictably interfere with, interface regions between adjacent subunits in ordered aggregates of coat protein (see Chapter 5, Section III.B.5, for TMV structure). Substitutions that did not elicit the HR were either conservative or located outside the interface regions. The HR-inducing substitutions formed rod-shaped particles with reduced quaternary stability, and the strength with which a coat protein elicited the HR correlated with the degree of destabilization of quaternary
S T E P S IN T H E I N D U C T I O N
OF D I S E A S E
445
structure. However, mutations that affected the tertiary structure of the coat protein did not elicit the HR. It was suggested that, to elicit the HR, the weakened quaternary structure exposed a receptor binding site (Culver et al., 1994). It is likely that the size distribution a n d / o r lifetime of small coat protein aggregates in elicitors allow the N' gene to recognize the invading virus (Toedt et al., 1999). To identify the elicitor site further, Taraporewala and Culver (1996) studied various amino acid substitutions in relation to the known tertiary structure of the coat protein (see Chapter 5, Section III.B.3, for details of TMV coat protein tertiary structure). They showed that substitutions that disrupted the right face of the coat protein s-helical bundle interfered with N' gene recognition. The elicitor active site covers approximately 600 y~2 and comprises 30% polar, 50% non-polar and 20% charged residues. Comparison of the coat proteins of various tobamoviruses and the effects of substitutions in these coat proteins on the N' gene HR revealed the presence of a conserved central hydrophobic cavity surrounded by surface features that were less conserved (Taraporewala and Culver, 1997). These findings suggested that the N'-gene specificity is dependent upon the three-dimensional fold of the coat protein as well as upon specific surface features within the elicitor active site (Fig. 10.5). 2. TMV and the L genes of Capsicum Capsicum spp. carry genes, L 1 (C. annuum), L ~ (C. frutescens) and L 3 (C. chinense) that confer HR resistance to TMV; there is also a genetically uncharacterized HR in eggplant, Solanum melongena (Berzal-Herranz et al., 1995; Dardick and Culver, 1997; de la Cruz et al., 1997; Dardick et al., 1999). The HR of each of these genes is induced by TMV coat protein. Chimeric constructs in which the coat protein gene of TMV U1 was substituted by those of other TMV strains showed that U1, U2, ORSV and CGMMV coat proteins elicited phenotypically similar HRs in pepper (L 1 gene). U1 and CGMMV coat proteins did not elicit the HR in tobacco and eggplant respectively (Dardick and Culver 1997; Taraporewala and Culver, 1997). In a comparison of the elicitation of the
446
,0
I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
No HR
HR
Non-elicitor CP
O Receptors
O
Elicitor CP
O
O
O
O
Receptors Signal transduction
Defence mechanisms
Nucleus
Nucleus
Prime gene host cells HR in tobacco, pepper and eggplant, Dardick et al. (1999) showed that the (x-helical bundle was essential in all cases. Differences in recognition were considered to result from how these hosts perceived the coat protein surface features a n d / o r quaternary configurations. This suggests that these resistance genes are functionally related and may be structurally homologous. 3. TMV and the Tm-2 genes of tomato (reviewed in Culver et al., 1991) There are two allelic genes in tomato, Tin-2 and Tin-2 e, that give an HR to certain strains of ToMV (see Table 10.3). As well as being determined by virus strain, the HR is also dependent on genotype of tomato and on environmental conditions, especially temperature. The response can vary from a very mild necrotic lesion giving apparent subliminal infection, through the normal necrotic local lesion to systemic necrosis. By comparison of sequences of HR-inducing and non-inducing isolates of TMV, and use of
Fig. 10.5 Model explaining the effects of elicitor and non-elicitor TMV coat proteins (CPs) on induction of the N'gene hypersensitive response. From Culver (1997), with permission.
mutagenesis and chimeric viruses, the 30-kDa movement protein (MP) has been identified as being the inducer of the HR in both Tin-2 and Tin-2'- plants (Culver et al., 1991; Calder and Palukaitis, 1992; Weber et al., 1993). Sequence comparison of Tin-2 resistance-breaking isolates (Ltbl and C32) with wild-type isolates identified two amino acid differences for Ltbl (Cys to Phe at position 68 and Glu to Lys at position 133) and also two in C32 (Glu to Lys at position 52 and Glu to Lys at position 133) (see Fig. 9.6). In Tin-2'- resistance-breaking isolates, amino acid substitutions are at the C-terminus of the MP (Ser to Arg at position 238 and Lys to G|u at position 244) (Weber et al., 1993). The C-terminal region of the MP, including the sites of the two substitutions, is not required for cellto-cell movement (Weber and Pfitzner, 1998), although the N-terminus is (see Fig. 9.6). Even so, it now appears that both these resistance genes operate via an HR rather than inhibition of movement out of the initially infected cell due to incompatibility of the MP (Weber and Pfitzner, 1998).
III.
S T E P S IN T H E I N D U C T I O N
OF D I S E A S E
447
TABLE 10.3 Genetic interactions between ToMV-resistant tomato plants and strains of the virus
0
1
Virus genotype 2 22
1.2
1.2 2
Wild type
M
M
M
M
M
M
Tin-1 Tin-2* Tin-22. Tm-1/Tm-2 Tm-1/Tm-2 ~ Tm-2/Tm-2 ~Tm-1/Tm-2/Tm-22
R R R R R R R
M R R R R R R
R M R R R R R
R R M R R R R
M M R M R R R
M R M R M R R
Host genotype"
Plants with genotype marked with an (*) may show local and variable systemic necrosis rather than mosaic when inoculated with virulent strains. M, systemic mosaic; R, resistance. " Modified from Fraser (1985), with permission. A sequence-characterized amplified region (SCAR) marker linked to the Tm-22 resistance gene in tomato has been produced (Dax et al., 1998).
t h r o u g h convergent evolution rather than through recombination (Malcuit et al., 2000). 5. TCV and the HRT and rrt genes of Arabidopsis
4. PVX and the N genes of potato HR in potato to PVX is controlled by two genes, N b and N x . N b has been m a p p e d to a resistance gene cluster in the upper arm of chromosome V (De Jong et al., 1997; Rouppe van der Voort et al., 1998) and N x to a region of chromosome IX similar to that containing the gene S w - 5 for resistance to TSWV (Tommiska et al., 1998). To identify the viral elicitor of the N b HR, hybrid viruses were constructed between avirulent PVX strains and virulent strains (Malcuit et al., 1999). The N b avirulence determinant was m a p p e d to the PVX 25-kDa gene encoding the ME The isoleucine at position 6 in this protein was shown to be involved in the elicitor function. However, this amino acid is present in the resistance-breaking strain HB and may act as a determinant of the three-dimensional structure of the avirulence domain that is specifically recognized by the N b gene product (Malcuit et al., 1999). The Nx-mediated resistance is elicited by the PVX coat protein gene (Kavanagh et al., 1992), a single amino acid at position 78 being an i m p o r t a n t d e t e r m i n a n t (Santa Cruz and Baulcombe, 1993). It is suggested that the multiple virulenceavirulence determinants were acquired by PVX
Inoculation of the Arabidopsis ecotypes Di-0 or Di-17 with TCV results in an HR on the inoculated leaves (Simon et al., 1992; Dempsey et aI., 1993, 1997). The HR development is conferred by a single d o m i n a n t gene t e r m e d H R T (Dempsey et al., 1997), located on chromosome 5 and encoding a classical leucine zipper-NBSLRR protein (Cooley et al., 2000). H R T shares extensive homology with the R P P 8 gene family that confers resistance to the oomycete, Peronospora parasitica. The TCV coat protein is the Avr factor recognized by HRT and mutations in the very N-terminus of the coat protein produced hypervirulent strains of TCV that failed to induce an HR (Oh et al., 1995; Wang and Simon, 1999; Cooley et al., 2000). H R T may not be sufficient for complete resistance as m a n y of the H R + p r o g e n y become fully infected (Kachroo et al., 2000). A recessive allele, rrt, which regulates resistance to TCV, has n o w been identified (Kachroo et al., 2000). 6. C a M V and solanaceous hosts
The D4 and W260 strains of CaMV induce chlorotic local lesions and a systemic mosaic in Datura s t r a m o n i u m , Nicotiana bigelovii and N. edwardsonii, whereas strain CM1841 induces necrotic local lesions limiting it to the inoculated leaf (Schoelz et al., 1986; Schoelz and
448
lO I N I ) U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
Shepherd, 1988). Using chimeric viruses constructed between these two strains it was shown that the HR was elicited by gene VI (see Chapter 6, Section IV.A.1, for details of the CaMV genome). This deduction was confirmed by infiltration of Agrobacterium tumefaciens containing a binary vector expressing gene VI of WD260 strain into leaves (Palanichelvam et at., 2000). Mutational analysis and chimeras identified the N-terminal one third of gene VI as being involved (Wintermantel et al., 1993; Broglio, 1995) and gene II, the aphid transmission factor, also has a light-dependent influence (Qiu et al., 1997). A point mutant at position 1628 (amino acid 94 in gene II) of strain W260 could systemically infect N. bigelovii at low but not at high light intensity, causing necrotic local lesions at under both conditions; wild-type W260 gave systemic infection at both light intensities. Thus, the HR and containment of the virus within the HR region is conditioned by host, virus strain and growing conditions. The translational transactivation of ORF VI can be uncoupled from its ability to elicit HR (Palanichelvam and Schoelz, 2001). 7. Other viruses By expression of individual genes of TBSV, Scholthof et al. (1995a) showed that the MP, p22, was responsible for the HR on N. edwardsonii. The p22 residues responsible for cell-to-cell movement were separable from those eliciting the HR (Chu et al., 1999). Three mutants of p22 impaired in cell-to-cell movement elicited necrotic local lesions on N. edwardsonii, whereas two mutants, capable of cell-to-cell movement, gave chlorotic lesions and systemic infection. The resistance to infection of cowpea by strains (e.g. Fny) of CMV involves an HR and a localization of infection (Kim and Palukaitis, 1997), responses that can be separated by mutation at two sites (nucleotides 1978 and 2007, codons 631 and 641) in the viral 2a polymerase gene. Changes to both sites of Fny strain allowed systemic infection without an HR and increase of viral synthesis in protoplasts, whereas changing position 1978 alone resulted in systemic infection, systemic HR and an increase of viral RNA accumulation in protoplasts. It is suggested that the inhibition of
RNA accumulation in protoplasts, where an HR does not occur, leads to localization of infection in whole plants and that different plant genes are involved in eliciting the HR and the localization response (Kim and Palukaitis, 1997). An R gene, Cry, has been recognized in cowpea cultivar Kurodane-Sanjaku which confers an HR resistance to CMV strain Y. This resistance is overcome by a legume strain of CMV (CMV-L) and its elicitation was shown by point mutation to be associated with amino acid 631 in the 2b replicase gene ((Karasawa et al., 1999). The HR caused by CMV in Chenopodium amaranticolor is affected by mutation of both the 3a movement protein and the 3b coat protein. It was considered that movement of the virus from the initially infected cell was required to elicit the HR (Canto and Palukaitis, 1999b). The begomoviruses, BDMV and BGYMV, differentially infect certain bean (Phaseolus vulgaris) varieties, with some cultivars giving an HR to BDMV. A series of hybrid DNA-B components, containing BDMV and BGYMV sequences, were co-inoculated with BDMV DNA-A or BDMVA-green fluorescent protein into seedlings of cv. Topcrop (susceptible to BDMV and BGYMV) and cvs Othello and Black Turtle Soup T-39 (BDMV resistant) (GarridoRamirez et al., 2000). The BDMV avirulence determinant was mapped to B V 1 0 R F and most likely BV1 protein. The product of BV1 is the nuclear shuttle protein (Chapter 9, Section II.C), and thus represents a new class of viral avirulence determinant. F. H o s t p r o t e i n c h a n g e s in the hypersensitive response The HR involves a series of complex biochemical changes at and near the infection site that include the accumulation of cytotoxic phytoalexins, the deposition of callose and lignin in the cell walls and the rapid death of plant cells forming the necrotic lesion (Dixon and Harrison, 1990). The regulation of HR is equally complex, involving the interplay of many potential signal-transducing molecules including reactive oxygen species, ion fluxes, G proteins, jasmonic and salicylic acids, protein phosphorylation cascades, activation of
III. GTEPS IN THE INDUCTION OF DISEASE
transcription factors and protein recycling by the polyubiquitin system (Dangl et al., 1996; Hammond-Kosack and Jones, 1996). To identify some of the tobacco genes eliciting the HR, Karrer et al. (1998) made a cDNA library from tobacco leaves undergoing HR and cloned the cDNAs into a TMV-based expression vector. Infectious transcripts from this TMV vector were inoculated to Xanthi nn tobacco (lacking the N gene) and those giving an HR were characterized. One of the 12 unique clones that were sequenced encoded ubiquitin, which was considered to elicit the necrotic response by a co-suppression mechanism. Increases in the activity of a variety of enzymes other than those designated as PR proteins have been observed during the HR to viruses. Peroxidase, polyphenoloxidase and ribonuclease activities are increased (e.g. Wagih and Coutts, 1982). The metabolism of phenylpropanoid compounds is strongly activated by infection with various pathogens, including viruses that induce an HR. This activation leads to the accumulation of compounds derived from phenylalanine, such as flavonoids and lignin. Activation of the pathway involves de novo enzyme synthesis, for example O-methyltransferase (Collendavelloo et al., 1982). Van Kan et al. (1988) identified two genes in Samsun NN tobacco coding for a glycine-rich protein that is strongly induced during the HR to TMV infection. Among other host proteins that are affected by the HR in tobacco, and which may be important in the reaction, are the myb oncogene homolog which is increased (Yang and Klessig, 1996), catalase which is decreased (Yi et al., 1999) and a glycine-rich RNAbinding protein which decreases in the early stages of HR but increases in the later stages (Naqvi et al., 1998). One of the earliest detectable events in the interaction between a plant host and a pathogen that induces necrosis is a rapid increase in the production of ethylene, which is a gaseous plant stress hormone. In the HR to viruses, there is an increased release of ethylene from leaves (e.g. Gfiborjfinyi et al., 1971). The fact that ethepon (a substance releasing ethylene), introduced into leaves with a needle,
449
can mimic the changes associated with the response of Samsun NN to TMV is good evidence that ethylene is involved in the initiation of this HR (van Loon, 1977). An early burst of ethylene production is associated with the viruslocalizing reaction, although the increase in ethylene production is not determined by the onset of necrosis but by a much earlier event (De Laat and van Loon, 1983). The outcome of the interaction between a virus and a host resistance gene may be controlled by a quantitative interaction between the viral avirulence determinant(s) and the host resistance determinant(s) (Collmer et al., 2000). The resistance to BCMV conferred by the I allele in cultivars of P. vulgaris varied according to I allele dosage and temperature, giving a range or responses from immunity, to hypersensitive resistance, to systemic phloem necrosis leading to plant death. In assessing the role(s) of all these changes in host proteins during the HR, it should be remembered that some of the changes are actually involved in the response and some are secondary to the response. At present, it is difficult to distinguish between these roles.
G. Other biochemical changes during the hypersensitive response Many other biochemical changes have been observed during the HR. For example, Uegaki et al. (1988) detected 19 sesquiterpenoids that were considered to be stressinduced compounds in Nicotiana undulata inoculated with TMV. An enhanced NADPHd e p e n d e n t oxygen-generating system was found in a membrane-rich cellular fraction from tobacco leaves reacting hypersensitively to TMV infection. Early electrolyte leakage occurred from cells of cowpea leaves during an HR to virus infection and other stress stimuli (Pennazio and Sapetti, 1981). Free abscisic acid concentration was raised up to 18-fold of normal in tissue near or within necrotic local lesions caused by TMV in tobacco (Whenham and Fraser, 1981). The cytokinin content of Xanthi N N tobacco leaves with systemic acquired resistance was increased (Balfizs et al., 1977).
450
10 I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
H. Systemic necrosis 1. Specific necrotic response On occasions the necrosis induced by virus infection is not limited to local lesions but spreads. This is usually from expanding local lesions that reach veins and result in systemic cell death. The systemic necrosis can range from necrosis in a few areas of upper leaves or sporadic necrotic spots mixed with mosaic symptoms to widespread necrosis leading to death of the plant. The systemic necrotic symptoms are dependent on host genotype, virus strain and environmental conditions. Some Nicotiana spp. that contain the N gene may develop systemic cell death in response to TMV infection, even when the plants are maintained at 17-20~ (Dijkstra et al., 1977). A single amino acid substitution may alter the coat protein of TMV from being a strong elicitor of the HR in N' gene-containing tobacco to being a weak elicitor resulting in systemic spread of necrosis (Culver and Dawson, 1989). ToMV-22 infection of tomato variety GCR 267, which is homozygous for the Tm-22 gene, results in systemic necrosis 2-3 weeks after infection. Infection with ToMV-30.22, in which the wild-type ToMV 30-kDa gene had been replaced by the Tm-2 e 30-kDa gene, led to much milder symptoms and later development of systemic necrosis (Weber et al., 1993). Infection of N. clevelandii with CaMV strains D4 and W260 leads to systemic cell death elicited by the CaMV gene VI (Kir~ily et al., 1999). The F1 generation of crosses between N. ctevelandii with N. bigelovii, which does not give systemic necrosis with these CaMV strains, developed systemic mosaic symptoms on inoculation with W260, whereas the F2 plants segregated 3:1 for systemic mosaic versus systemic death. The plant gene responsible for cell death has been named ccdl. Thus, the systemic death phenotype is induced by the interaction between ccdl and CaMV gene VI (Kiraly et al., 1999). 2. Non-specific necrosis Some other types of necrotic response are less specific and can occur in hosts that are different genera and apparently not in a gene-for-gene
manner. Several deletion mutants of TMV coat protein cause an apparent non-specific necrosis (Dawson et al., 1988; Dawson and Bubrick, 1989). The 5' non-coding region of GCMV RNA2, cloned into a PVX vector, induces a systemic necrotic response in N. benthamiana, N. clevelandii and N. tabacum (Fernandez et al., 1999). GCMV itself does not infect N. ctevelandii and is symptomless or gives very mild symptoms in the other two Nicotiana spp. Joint infection of tomato with TMV and PVX can cause systemic necrosis (Plate 3.20). Ringspot symptoms, in which necrotic rings spread in an apparently diurnal manner, are described in Chapter 3 (Section II.B.5).
I. Programed cell death and plant viruses Multicellular organisms have mechanisms for eliminating developmentally misplaced or unwanted cells, or for sacrificing cells to prevent the spread of pathogens. This is termed programed cell death (PCD) or apoptosis; apoptosis is a specific case of PCD with a distinct set of physiological and morphological features (reviewed in Mittler and Lam, 1996; Gilchrist, 1998). Although much of the work on PCD has been done in animal systems, there is increasing interest in this process in plants. The HR to plant pathogens has various features in common with PCD (see Mittler and Lam, 1996; Mitsuhara et al., 1999; Pontier et al., 1999). Certain animal viruses can inhibit PCD (Osborne and Schwartz, 1994). It will be of interest to determine whether plant viruses have similar properties.
J. Local acquired resistance Most experimental work on local acquired resistance has been carried out with tobacco varieties containing the N gene. Ross (1961a) showed that a high degree of resistance to TMV developed in a 1-2-mm zone surrounding TMV local lesions in Samsun NN tobacco (Fig. 10.6). The zone increased in size and resistance for about 6 days after inoculation. Greatest resist-
III.
Fig. 10.6 Acquired resistance to infection. Upper: Disc cut from a Samsun NN tobacco leaf inoculated first with TMV (large lesion in center) and 7 days later given a challenge inoculation with a concentrated TMV inoculum. Note absence of lesions from the second inoculation in a zone around the original lesion. Lower: Similar experiment in which PVX was the challenge virus. No zone free of lesions is present. From Ross (1961a), with permission. ance developed in plants g r o w n at 20-24~ Resistance was not found in plants g r o w n at 30~ There appears to be no local acquired resistance without the HR (Costet et al., 1999). K. S y s t e m i c a c q u i r e d r e s i s t a n c e (see Ryals et al., 1996 for review) Our initial k n o w l e d g e about systemic acquired resistance came mainly from the w o r k of Ross and colleagues (e.g. Ross, 1961b, 1966) on TMV in the tobacco variety Samsun N N and on TNV in pinto bean. In tests with tobacco, lower leaves are inoculated with TMV and then some days later the same leaves or u p p e r leaves m a y be challenged by a second inoculation with TMV (Fig. 10.7). Acquired resistance is measured by the reduction in diameter of the lesions (and with some viruses, reduction in number). With bean, one p r i m a r y leaf is inoculated and
S T E P S IN WEE I N D U C T I O N
OF D I S E A S E
451
the opposite primary leaf challenged by inoculation some days later. Lesions were about one-fifth to one-third the size found in control leaves, but lesion n u m b e r was not reduced with TMV in S a m s u n N N tobacco. Resistance was detectable in 2-3 days, rose to a m a x i m u m in about 7 days, and persisted for about 20 days. Leaves that developed resistance were free of virus before the challenge inoculation. No conditions have been found that w o u l d give complete resistance. In plants held at 30~ no resistance developed. Mechanical or chemical injury that killed cells did not lead to resistance, nor did infection with viruses that do not cause necrotic local lesions. On the other hand, m a n y other nonspecific agents applied to leaves will induce the p h e n o m e n o n (e.g. Gupta ef al., 1974). In such experiments it is not possible to be sure that the same p h e n o m e n o n is being studied because m a n y treatments affect lesions size. The resistance induced by TMV was not specific for TMV, but was effective for TNV and several other viruses. A similar lack of specificity in the resistance acquired following the d e v e l o p m e n t of necrotic local lesions was found with various other h o s t - v i r u s combinations giving the HR. However, virus-specific factors m a y regulate the extent of the resistance (van Loon and Dijkstra, 1976). Ross (1966) pointed out that effects on lesion n u m b e r tend to be more variable and develop later. He considered that a fall in lesion n u m b e r merely means that in highly resistant leaves lesions do not become large e n o u g h to be countable. 1. Pathogenesis-related (PR) proteins (reviewed by van Loon and van Strien, 1999) Gianinazzi et al. (1970) and van Loon and van K a m m e n (1970) s h o w e d that changes in the pattern of soluble leaf proteins occurred in tobacco leaves r e s p o n d i n g hypersensitively to infection with TMV. They were termed pathogenesisrelated proteins, or 'PR proteins', by A n t o n i w et al. (1980). These proteins have been studied extensively and their activities characterized (see Linthorst, 1991; Shewry and Lucas, 1997; D e m p s e y e, al., 1999). There are 14 families of PR protein (PR1-14), five of which are listed in
452
lO INDUCTION OF DISEASE 2: VIRUS-PLANT INTERACTIONS
Fig. 10.7 Resistance acquired at a distance from the site of inoculation. Right: Samsun NN leaf inoculated first on the apical half with TMV and 7 days later given a challenge inoculation over its whole surface with TMV. Left: Control leaf given only the second inoculation. From Ross (1961b), with permission. TABLE 10.4 Some types of PR proteins Type
Biological activity
Inhibit
1 2 3 4 5
Not known [3-1,3-Gluconases Chitinases Chitin binding Membrane permeabilization
Oomycetes Fungi Fungi Fungi Fungi
Table 10.4, and for m a n y there are two classes: acidic PR proteins and their basic homologs. They h a v e been identified in a range of plant species a n d s h o w n to be i n d u c e d by a variety of microbial infections (viruses, viroids, bacteria and fungi) and by t r e a t m e n t with certain chemical elicitors, notably salicylic acid (SA) and acetyl salicylic acid (aspirin) (White, 1979). These various PR proteins are not i n d u c e d foll o w i n g infection with a virus that does not cause necrotic local lesions in the host used (Fig. 10.8, lane 4). The kinetics of the H R - i n d u c e d expression of a tobacco peroxidase gene differ from those of the i n d u c t i o n of PR proteins especially in being insensitive to inducers of PR genes such as SA, m e t h y l j a s m o n a t e a n d e t h e p h r o n (Hiraga et al., 2000). It is not yet established w h e t h e r any of the PR proteins play a role in the limitation of virus
Fig. 10.8 Accumulation of PR proteins in the intercellular fluid of tobacco after various treatments. Plants were sprayed with water (H), p-coumaric acid (C) or salicylic acid (S), or inoculated with AMV (A) or TMV (T). Samples of the intercellular fluid were electrophoresed in non-denaturing polyacrylamide gels. The position of the major PR proteins is indicated in the margin. From Bol and van Kan (1988), with kind permission of the copyright holder, 9 Blackwell Science Ltd.
III.
spread and acquired resistance. If any of the PR proteins are, in fact, active in the antiviral response, they may be found among those proteins for which a function has not yet been established. However, when expressed to high levels in transgenic tobacco plants with the NN constitution, individual PR proteins did not affect the necrotic response to inoculation with TMV, for example the lb gene (Cutt et al., 1989) and the la and s genes (Linthorst et al., 1989). 2. Effect of salicylic acid on viruses Application of SA suppresses the replication of TMV and PVX in inoculated tissue (Chivasa et al., 1997; Naylor et al., 1998) by not only decreasing the accumulation of virus, but also changing the ratio of viral genomic RNA to mRNA accumulation. It also affects the replication of AMV in cowpea protoplasts but not in tobacco leaves (Hooft van Huijsduijnen et al., 1986; Murphy et al., 1999a). SA treatment appears to have no effect on the replication of CMV but causes a significant delay in the development of systemic symptoms, suggesting that this treatment affects the entry of the virus into the vasculature (Naylor et al., 1998). It also affects the long-distance movement of AMV (Murphy et al., 1999a). At least in the case of AMV, this treatment appears to have different effects in different hosts. For other viruses, for example PVY, SA appears to have no effect on virus infection (Pennazio et al., 1985). 3. Pathway for systemic acquired resistance to viruses The characterization of PR proteins presented a conundrum in that, although they were induced by viral infection and associated with systemic acquired resistance, their activities, as far as is known (Table 10.4), do not appear to be related to virus inhibition but function in fungal and bacterial resistance. However, the HR induced by TMV infection of N-gene tobacco has many features in common with HRs caused by fungi and bacteria (reviewed in Murphy et al., 1999a). The reaction is mediated by a sustained burst of reactive oxygen (ROS) followed first by local and then by systemic accumulation of SA. Since the
STEPS IN THE INDUCTIC)N OF DISEASE
453
production of SA induces metabolic heating, its local accumulation can be detected very early on by thermography (Chaerle et al., 1999). SA is essential for the localization of the virus to the vicinity of the necrotic lesion and for the establishment of systemic acquired resistance. The SA-induced resistance to TMV and PVX replication and to CMV movement in tobacco can be inhibited by salicylhydroxamic acid (SHAM) (Chivasa et al., 1997; Naylor et al., 1998). However, SHAM does not inhibit the SA-induced synthesis of PR proteins (Chivasa et al., 1997; Chivasa and Carr, 1998), which suggests that there are two branches in the pathway to SA-induced resistance (reviewed by Murphy et al., 1999a) (Fig. 10.9). One branch leads to the production of PR proteins that confer resistance to fungi and bacteria, and the other induces resistance to viral replication and movement. SHAM is a relatively selective competitive inhibitor of the alternative oxidase (AOX) in the mitochondrial electron flow pathway in plants (reviewed in Murphy et al., 1999a). The SHAMsensitive pathway, induced by SA and potentially by AOX, is critical in the early stages of N gene-mediated resistance to TMV in tobacco. However, it does not explain all the observations, and other antiviral mechanism(s) must also play a role in the HR. Various other chemicals are involved in the signal transduction p a t h w a y for defense response against pathogens including nitric oxide, mitogen-activated protein (MAP) kinases, jasmonic acid and ethylene (see Dempsey et al., 1999; Klessig et al., 2000; Zhang and Klessig, 2000). However, it is becoming recognized that, although there are many common features in the defense response, there is not just one single response pathway. For instance, the HR formation and T M V / N gene and T C V / H R T gene resistance are dependent upon SA but not on ethylene or jasmonic acid, and are unaffected by mutations in N R P 1 (Knoester et al., 1998; Murphy et al., 1999a; Kachroo et al., 2000), whereas they are all required for resistance to other pathogens (see Dong, 1998; Shah et al., 1999). As noted in Section II.A, there is not yet a detailed understanding of the signal transduction
454
10 I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
Triggering of R gene-mediated resistance
Salicylic acid
l_
i ,,, nahG(SAhydroxylase) CNand
SHAM
AOX 9 t i
Induction of PR proteins and resistance to fungi and bacteria
Induction of resistance to virus movement and replication
Fig. 10.9 Possible model to explain the induction of resistance to viruses and other pathogens in tobacco. Recognition of a pathogen by the product of a resistance (R) gene results in localized cell death and the activation of a defense signal transduction pathway which includes salicylic acid (SA) as one of its components. Steps subsequent to SA in the defense signal transduction pathway are prevented in transgenic plants that express SA hydroxylase, the product of the bacterial nahG gene. Downstream of SA, the defense signal transduction pathway appears to divide. One branch (right) leads to the induction of extracellular pathogenesis-related (PR) proteins and systemic acquired resistance (SAR) to bacteria and fungi. The other (left) branch leads to the induction of resistance to the replication or long-distance movement of viruses. The virus-specific branch can be activated by cyanide (CN-) and antinomycin A (AA), or be inhibited with salicylhydroxamic acid (SHAM), independently of the other branch. These observations have led to the suggestion that alternative oxidase activity (AOX) might play a role in the induction of SAR to viruses. This suggestion is consistent with the finding that AOX protein and transcript levels are raised in tobacco tissue expressing SAR. From Murphy et al. (1999a), with kind permission of the copyright holder, 9 Elsevier Science. p a t h w a y following the interaction of an A v r gene with an R gene. The action at a distance i n v o l v e d in SAR p r e s u m a b l y requires the translocation of some substance or substances. Ross (1966) has presented good evidence that transport of a resistance-inducing material is involved. For example, w h e n the midrib of an
u p p e r tobacco leaf was cut, resistance did not develop in the portion of the lamina distal to the cut. Similarly, killing sections of petiole of inoculated leaves w i t h boiling water, while allowing the leaf to r e m a i n turgid, prevented d e v e l o p m e n t of resistance in other leaves. O t h e r e x p e r i m e n t s s h o w e d that, in large tobacco plants, the material m o v e d equally well both up and d o w n the stems. The nature of the material that migrates is u n k n o w n , as is the actual m e c h a n i s m of resistance in the resistant uninfected leaves. The migrating material m i g h t involve SA, ethylene, jasmonic acid, nitrous oxide or even small peptides such as systemin (see H o w e and Ryan, 1999). This m e c h a n i s m m a y or m a y not be the same as that in the zone of tissue around necrotic lesions. Systemic acquired resistance can be i n d u c e d by non-necrotic localized viral infection (Roberts, 1982). Systemic acquired resistance is not effective w h e n the challenge virus is one that moves systemically (Pennazio and Roggero, 1988). Thus, one also has to consider an inherent host response (see Section IV).
L. Wound healing responses Wounds involving necrosis caused by mechanical injury, insects a n d v a r i o u s pathogens, including viruses, frequently result in a series of w o u n d healing responses by the plant. These responses m u s t involve the non-specific induction of m a n y host-coded proteins. The most complex response is the d e v e l o p m e n t of a w o u n d p e r i d e r m and cell wall changes, including lignification, suberization and the deposition of callose. Virus-induced necrosis m a y lead to such w o u n d responses. A p e r i d e r m was formed in y o u n g bean leaves inoculated with the VM strain of TMV, which gives very small lesions, but not in old leaves or in leaves inoculated with the U1 strain giving large lesions (Wu, 1973). Various workers have noted a deposition of callose in cells a r o u n d necrotic local lesions, leading to thickening of cell walls and probably blocking of p l a s m o d e s m a t a (e.g. Hiruki and Tu, 1972; Wu, 1973). Stobbs et aI. (1977) found that callose deposition in live cells extended b e y o n d the m a r g i n of detectable virus, while remaining
Ili.
within the zone of fluorescent metabolite accumulation in pinto bean leaves infected with TMV. Cell wall glycoproteins were determined chemically following extraction from leaves by Kimmins and Brown (1973). They found an accumulation of glycoproteins following inoculation of hypersensitive hosts with TMV or TNV. An identical response occurred in leaves that had been mock inoculated. Other observations suggest that cell wall modifications may not be a factor limiting spread. Appiano et al. (1977) considered that the conspicuous cell wall lignification seen in lesions caused by TBSV in Gomphrena leaves was not a barrier to spread of virus because lignification did not follow the whole cell perimeter, and because virus could be detected beyond the cells with modified walls. There may be several systemic signaling pathways for wound response (see Rhodes et al., 1999) but recent research suggests that the SAR and wound response pathways may not be completely independent (Maleck and Dietrich, 1999). M. A n t i v i r a l factors The lesion response in tobacco varieties containing the N gene has been associated with the presence of a protein with antiviral properties named 'inhibitor of virus replication' (IVR) (see Loebenstein and Gera, 1981; Loebenstein et al., 1990). IVR, released into the medium from protoplasts of tobacco Samsun NN, inhibits the replication of TMV in protoplasts from both local lesion-responding and systemically infectible tobacco varieties. A 23-kDa protein with antiviral properties has been purified from these extracts (Gera et al., 1990) and a cDNA to a protein recognized by anti-AVR antiserum has been isolated and cloned (Akad et al., 1999). It was suggested that the N-gene products serve as a signal for the induction of IVR-like proteins or that IVR is similar to PR proteins (Loebenstein et al., 1990; Akad et al., 1999). Sela and colleagues have partially purified and characterized 'antiviral factors' (AVF) from Nicotiana cultures with the N gene (reviewed by Fraser, 1987a). Parallels have been d r a w n between the AVFs and human interferons (e.g.
STEPS IN THE I N D U C T I O N OF DISEASE
455
Sela et al., 1987; Edelbaum et al., 1990). However, the significance of this work in relation to host plant resistance remains to be established, as does the relationship of the AVFs and the IVR described above. N . A b i l i t y of v i r u s to spread t h r o u g h various barriers The factors that control the ability of a virus to pass various tissue boundaries in reaching the sieve elements of an inoculated leaf, to spread through the vascular system and to exit in photosynthetic product sink leaves are discussed in Chapter 9 (Sections II.E and II.G). One form of resistance that is not similar to the gene-for-gene-type resistance (see Section II.A) for the gene-for-gene hypothesis) is shown by the Arabidopsis RTM1 gene (Chisholm et al., 2000). This gene is necessary for restriction of long-distance movement of TEV without causing an HR or SAR. The gene product is similar to the a-chain of the lectin, jacalin, from Artocarpus integrifolia; jacalin belongs to a family of proteins with members that are implicated in defense against insects and fungi. O. S y s t e m i c h o s t r e s p o n s e As described in Chapter 3 (Section II.B), there is a wide range of systemic symptoms induced by viruses, the most common and characteristic of which is the mosaic symptom. The mosaic symptom comprises areas of the leaf showing various degrees of chlorosis together with areas that remain green and are termed 'dark green islands'. The dark green, light green and even chlorotic patches that make up mosaics range from relatively large (e.g. TMV, TYMV and AbMV) to relatively small, giving a fine mosaic (e.g. CPMV in cowpea, AMV in tobacco). These areas are often delimited by the vein structure of the leaf, giving streak or stripe symptoms in monocotyledons. The development of mosaic symptoms is described in more detail in Chapter 9 (Section IV.D). Two approaches are used to study the mosaic symptomatology: mutagenesis of the viral genome and microscopic examination of the infected leaf. Most of
456
10 I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
the work has been performed on TMV and TYMV, although there is an increasing amount of information accruing from studies on other viral genomes. 1. Chlorosis In most mosaic symptoms there are regions (domains) with different levels of chlorosis which can vary from almost completely white to very pale green-yellow. The chlorotic response is obviously the effect of the virus either directly or indirectly on the chloroplasts in that region of the leaf causing loss of chlorophyll by perturbation of chloroplast structure and function. a. T Y M V and T M V
The effects of virus infection on the cytology of chloroplasts, especially in relation to TYMV, are described in Chapter 9 (Sections III.K.3 and IV.C). In understanding the effects that the virus has on causing the chlorotic element of a mosaic s y m p t o m it is important to consider both the developmental stage of the leaf at the time of infection and the possibility of variants of the virus in different chlorotic domains. If the leaf is almost fully grown when it becomes infected, chlorosis must result from the breakd o w n of mature chloroplasts. In systemic infections where leaves are very young when they become infected, it is likely that the virus affects chloroplast development. In some infections, such as TMV in tobacco, the disease in individual plants appears to be produced largely by a single strain of the virus. However, it has been known for many years that occasional bright yellow islands of tissue in the mosaic contain different strains of the virus. Such strains probably arise by mutation, and during leaf development come to exclude the original mild strain from a block of tissue. In Chinese cabbage plants infected with TYMV there may be many islands of tissue of slightly different color from which different strains of the virus can be isolated (Chalcroft and Matthews, 1967a,b). Specific amino acid substitutions and deletions in the coat protein of TMV affect the production of chlorotic symptoms (Dawson et al.,
1988). Mutants that retained the C-terminus of the TMV coat protein induced the strongest chlorotic symptoms in tobacco (Dawson et al., 1988) in both expanded and developing leaves (Lindbeck et al., 1991, 1992). The chlorotic s y m p t o m formation is related to the concentration of TMV capsid proteins forming aggregates in infected cells but not accumulating in chloroplasts. In contrast to this, in infections with YSI/1, a naturally occurring chlorotic mutant of the U1 strain of TMV, coat protein is found in the chloroplasts associated with the thylacoid membrane fraction (Banerjee et al., 1995). Reinero and Beachy (1986) detected TMV coat protein in both the stroma and membrane fractions of the chloroplasts of infected cells. The coat protein of YSI/1 differs from that of U1 in two nucleotides, one of which, giving an Asp to Val change at amino acid 19, is responsible for the chlorotic phenotype (Banerjee et al., 1995). The coat protein of another natural chlorotic mutant, flavum, also has a substitution at amino acid 19, this time Asp to Ala (Wittmann et al., 1965). However, these severe chlorotic symptoms are unusual in TMV infections and are associated only with mutants. In natural infections, the chlorotic element of the mosaic is usually light green and is not accompanied by the accumulation of coat protein bodies or with coat protein in chloroplasts. b. CMV Chlorosis is a strain-specific s y m p t o m of CMV with strains CMV-M and CMV-Y inducing severe systemic chlorosis in tobacco. Pseudorecombinants between CMV-M and a green mosaic-inducing strain, CMV-Fny, located the gene responsible on CMV-M RNA3. Further experiments with recombinant RNA3 transcribed from engineered cDNAs showed that the s y m p t o m in tobacco was controlled by the coat protein gene (Shintaku and Palukaitis, 1990). The determinant was further narrowed down to the amino acid at position 129 in the coat protein (Shintaku et al., 1992), which suggested that the local secondary structure influenced the s y m p t o m type. The induction of severe chlorotic spots on tobacco leaves inoculated with CMV-Y is also associated with amino
III.
STEPS IN THE I N D U C T I O N OF DISEASE
457
sis. Stratford a n d C o v e y (1989) c o n s t r u c t e d a series of h y b r i d C a M V g e n o m e s b e t w e e n t w o strains that cause severe (strain Cabb B-JI) or mild (strain Bari-1) disease in turnips. The determ i n a n t s for the d e g r e e of leaf chlorosis w e r e located in a d o m a i n consisting of part of gene VI together w i t h the large intergenic region a n d nucleotides 6103-6190 of gene VII (see Fig. 10.10). The t r a n s f o r m a t i o n of tobacco to e x p r e s s s e g m e n t s of C a M V D N A c o n t a i n i n g gene VI resulted in transgenic plants s h o w i n g virus-like s y m p t o m s ( B a u g h m a n et al., 1988; Balazs, 1990; G o l d b e r g et al., 1991). G e n e VI f r o m t w o different virus isolates p r o d u c e d different s y m p t o m s - e i t h e r mosaic-like or a bleaching of the leaves. S y m p t o m p r o d u c t i o n w a s b l o c k e d by deletions or f r a m e s h i f t m u t a t i o n s in gene VI. P r o d u c t i o n of s y m p t o m s w a s closely correlated w i t h the a p p e a r a n c e in the leaves of the 66-kDa gene p r o d u c t , as s h o w n by i m m u n o b l o t t i n g . H o w e v e r , it w a s e s t i m a t e d that the a m o u n t of
acid 129 a n d w i t h two n u c l e a r - c o d e d recessive host genes (Takahashi a n d Ehara, 1993). In m u t a g e n e s i s e x p e r i m e n t s w i t h CMV-Y a n d a green m o s a i c strain, CMV-O, S u z u k i et al. (1995) s h o w e d that the coat p r o t e i n a m i n o acid 129 w a s i n v o l v e d not only in the i n d u c t i o n of chlorosis b u t also in severe veinal necrosis a n d necrotic local lesion p r o d u c t i o n (Table 10.5). They c o n c l u d e d that the s y m p t o m a t o l o g y w a s associated w i t h the tertiary structure of the coat protein molecule as well as the local s e c o n d a r y structure s u r r o u n d i n g a m i n o acid 129. c. CaMV Two a p p r o a c h e s h a v e been u s e d to dissect the s y m p t o m d e t e r m i n a n t s of CaMV: i n f e c t i n g plants w i t h chimeric viral g e n o m e s a n d transf o r m i n g plants w i t h viral genes. Both of these a p p r o a c h e s i d e n t i f y g e n e VI, e n c o d i n g the protein P6 as the major factor in i n d u c i n g chloro-
TABLE 10.5 Coat protein involvement in symptom production by CMV
Base Inoculum a
Position ~
Amino Acid Change(s)
Position
Substitution
Symptoms on tobacco plants
None None Ser to Pro Pro to Ser Ser to Leu Pro to Leu Ser to Phe Pro to Phe Ser to Gly
Chlorosis Green mosaic Green mosaic Chlorosis NLL, veinal necrosis NLL, veinal necrosis NLL NLL Green mosaic
None Phe to Ser Ala to Ser Ala to Ser Ala to Glu Ala to Asp
NLL Chlorosis Chlorosis Chlorosis Green mosaic Slight vein necrosis
A. Summary of coat protein mutants
Y3 03 Y(SP) O(PS) Y(SL) O(PL) Y(SF) O(PF) Y(SG)
1644 1645 1644, 1645 1646 1645 1645, 1646 1645 to 1646
None None U to C C to U UC to CU C to U C to U CC to UU UCU to GGC
B. Summary of Y(SF) revertants
Y(SF) Y(SF)R1 Y(SF)R2 Y(SF)R3 Y(SF)R4 Y(SF)R5
1645 1698 1698 1690 1672
None U to C G to U G to U C to A C to A
129 147 147 144 138
a Designation refers to transcripts from wild-type and mutant RNA3 cDNA. Each transcript was inoculated with transcripts from cDNAs of CMV-Y RNAs I and 2. bNucleotide number (CMV-Y refer to Nitta et at., 1988; CMV-O, Hayakawa et al., 1989) of altered base(s). NLL, necrotic local lesions on inoculated leaves. From Suzuki et al. (1995), with permission.
458
10 INDUCTION OF DISEASE 2: VIRUS-PLANT INTERACTIONS
the 66-kDa protein produced in transgenic tobacco plants was only about one-twentieth of that found in infected turnips. Furthermore, the fact that tobacco is not a host of CaMV and that transformation with gene VI of FMV, which infects tobacco, gave no symptoms (Goldberg et al., 1991) raised the possibility that this was a non-host effect. To investigate this, Cecchini et al. (1997a) constructed a collection of transgenic Arabidopsis lines expressing gene VI sequences from CaMV isolates Cabb B-JI, Bari-1 and a recombinant, Baji-31, which causes very severe infections of Arabidopsis (Cecchini et al., 1997b). They showed that the symptom character elicited in the gene VI-expressing host plants was dependent on the level of P6 expression and also upon the P6 sequence itself. Using differential display PCR, the changes in abundance of mRNA species in P6 transgenic and CaMVinfected Arabidopsis plants were similar when compared with those in uninfected plants (Geri et al., 1999). mRNA species that were downregulated in transformed and infected plants included those for a phenol-like sulfotransferase and a glycine-rich RNA-binding protein; upregulated species included a myb protein, glycine-rich and stress-inducible proteins. In further studies on the interactions between CaMV and Arabidopsis, Cecchini et al. (1998) inoculated Arabidopsis ecotype Col-O gll with 30 CaMV isolates. Thirteen isolates failed to cause symptoms, the remainder inducing symptoms that varied between mild and very severe. Some of the symptoms differed markedly from those produced by that isolate in turnip (Brassica rapa). A greater variety of symptoms was observed in a single Arabidopsis ecotype infected by a range of CaMV isolates than was produced by a single isolate in a range of Arabidopsis ecotypes. An EMS-mutagenized Arabidopsis (Col-O dvl) gave altered symptoms to CaMV, including an uncharacteristic necrotic response. With the detailed molecular understanding of the Arabidopsis genome, this approach should reveal plant genes involved in interactions leading to symptom production (see Schenk et al., 2000). The ability of CaMV isolate, W260, to overcome resistance in A. thaliana ecotype Tsu-O is associated with ORF VI (Leisner et al., 2001).
d. MSV Immuno-histochemical and in situ hybridization techniques have been used to localize MSV in infected maize plants (Lucy et al., 1996). The virus did not invade the apical meristem and was present only in areas of leaves displaying the characteristic chlorotic streak symptoms. Strains of MSV differ in the width of the streaks and amount of chlorosis that they cause. Strain Ns produces earlier symptoms with broader and more chlorotic streaks than does strain Nm (Boulton et al., 1991a); it also has a broader host range. Site-directed mutagenesis and construction of chimeric viruses showed that genome nucleotide 40 affected the streak width and that nucleotide 2473 determined the severity of chlorosis, the length of the streaks, the latency and host range (Table 10.6). Nucleotide 40 is in gene V1, which controls cell-to-cell movement; nucleotide 2473 is in the large intergenic region and the nucleotide change alters a potential promoter sequence 101 nucleotides upstream of the initiation codon of the C1 gene (see Chapter 7, Section IV.E.3 for mastrevirus promoters). Mutagenesis of promoters downstream of that at-101 showed that the -101 promoter alone conferred the chlorosis, streak length, latency and host range phenotype (Boulton et al., 1991b). 2. Vein clearing An early systemic symptom of mosaic-causing viruses is vein clearing, which occurs temporarily when the first flush of virus reaches the young leaves. Dawson and colleagues have been trying to explain this phenomenon (reviewed in Dawson, 1999). In TMV infections vein clearing occurs at temperatures above 25~ and its intensity increases with increasing temperature up to 40~ It is light dependent and can be manifest within 5 minutes of permissive conditions being reached. It precedes most of the virus replication in the leaf and is not associated with any chloroplast abnormalities. The conclusion is that it is an optical illusion. 3. Vein banding As well as patches of chlorosis and of light and dark green, some mosaic-inducing viruses cause characteristic chlorotic vein banding.
III.
STEPS IN T H E I N D U C T I O N OF D I S E A S E
459
TABLE 10.6 Pathogenicity characteristics of wild-type and mutant MSV genomes in infected maize plants Virus or hybrid Experiments 1 and 2 MSV-Nm a MSV-Ns MSV-Nm (s40) MSV-Nm (s2473) MSV-Nm (s40, 2473) Experiments 3 and 4 MSV-Nm MSV-Ns MSV-Nm (TATA, 101) MSV-Ns (TATA, 101)
Symptoms (days pi) b
Streak morphology d
Chlorosis ~'
Viral DNA concentration r
Host rangeg
9 6 9 6 6
N, S W, L W, S N, L W, L
Mild Severe Mild Severe Severe
33 100 76 95 108
R B R B B
8 6 6c 6
N, S W, L N, L W, L
Mild Severe Severe Severe
ND ND ND ND
R B ND B
MSV-Ns, Nigerian strain of MSV inducing wide, long, severely chlorotic streaks; MSV-Nm, variant of MSV-Ns giving narrow, discontinuous, mildly chlorotic streaks; (s40) and (s2473), substitutions at nucleotides 40 and 2473 respectively; (TATA, 101), TATA sequences at -57 and -62 modified to GATA and TACA. b Earliest day of symptom appearance following agroinoculation of at least 50 plants with each construct. c In experiment 4, symptoms were first seen on MSV-Nm (TATA, 101)-infected plants 8 days following inoculation. d N , narrow; W, wide; L, long; S, short. e Assessed visually as degree of streak chlorosis. Relative concentration, percentage values expressed relative to MSV-Ns-infected tissue = 100; mean value from two experiments. g B, broad host range; R, restricted host range. ND, not determined; pi, post-inoculation. From Boulton et al. (1991b), with permission. This is especially m a r k e d in m a n y strains of CaMV. In a s t u d y on the distribution of CaMV, A1-Kaff a n d C o v e y (1996) f o u n d that there w a s more virus in the vein b a n d s in systemically i n v a d e d e x p a n d e d leaves t h a n in the interv e i n a l areas, b u t in y o u n g e r s y s t e m i c a l l y infected leaves there w a s less virus in chlorotic vein b o r d e r s t h a n in the i n t e r v e i n a l g r e e n islands. 4. Dark green islands D a r k green islands in the mosaic p a t t e r n are cytologically a n d biochemically n o r m a l as far as has b e e n tested. In m o s t cases, for e x a m p l e t o b a m o v i r u s e s (Atkinson a n d M a t t h e w s , 1970), c u c u m o v i r u s e s (Loebenstein et al., 1977) a n d p o t y v i r u s e s (Suzuki et al., 1989), they contain low or zero a m o u n t s of infectious virus a n d no detectable viral protein or viral d s R N A . A b o u t 50% of the plants r e g e n e r a t e d f r o m d a r k green island material w e r e virus free (Murakishi a n d Carlson, 1976). H o w e v e r , in C a M V infections, the d a r k green islands contain u n u s u a l accum u l a t i o n s of virus (A1-Kaff a n d Covey, 1996). The interveinal areas of CaMV-infected turnip,
w h i c h d e v e l o p e d as d a r k green islands, h a d s o m e of the highest virus c o n c e n t r a t i o n s in the leaf. In a detailed s t u d y of the d a r k green islands in T M V - i n f e c t e d tobacco, A t k i n s o n a n d M a t t h e w s (1970) d i s t i n g u i s h e d t w o types of island: the 'true' island w h i c h r e m a i n e d d a r k green for the life of the leaf, a n d the ' p s e u d o ' island in w h i c h virus p r o d u c t i o n b e g a n at a late stage of leaf d e v e l o p m e n t . The p s e u d o islands h a d f u z z y b o u n d a r i e s w h e r e a s those a r o u n d true islands w e r e sharp. These a u t h o r s m a d e a close analysis of the virus content of cells on either side of the c h l o r o t i c - d a r k green b o u n d ary (see Fig. 9.21). The chlorotic cells c o n t a i n e d large n u m b e r s of crystalline T M V inclusions w h e r e a s n o n e of the d a r k green cells h a d a n y inclusions. TMV particles w e r e o b s e r v e d in d a r k green cells close to the b o u n d a r y but d e c r e a s e d in n u m b e r a w a y f r o m the b o u n d a r y until no virus particles w e r e o b s e r v e d in the center of the d a r k green island. Several lines of evidence s h o w that d a r k green islands are resistant to s u p e r i n f e c t i o n w i t h the s a m e virus or closely related viruses
460
10 I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T I N T E R A C T I O N S
(e.g. Chalcroft and Matthews, 1967a,b; Atkinson and Matthews, 1970; Loebenstein et al., 1977). Various factors can influence the proportion of leaf tissue that develops as green islands in a mosaic. These include leaf age, strain of virus, season of the year, and removal of the lower leaves on the plant (Crosbie and Matthews, 1974b). The dark green islands of tissue may not persist in an essentially virus-free state for the life of the leaf. 'Breakdown' leading to virus replication usually takes place after a period of weeks, as in the case of pseudo islands, or after a sudden increase in temperature (Atkinson and Matthews, 1970; Matthews, 1973; Loebenstein et al., 1977). Dark green islands may be part of a more general p h e n o m e n o n in plant virus infections. In local lesions caused by TYMV in Chinese cabbage, a proportion of cells show no evidence of virus infection and remain in that state for considerable periods even though neighboring cells are fully infected. It may be that these cells are in a resistant state, like that found in dark green islands of the mosaic. Perhaps a proportion of cells in all infected leaves develop the resistant state, but only cells still retaining the potential to divide can give rise to microscopic or macroscopic islands of dark green tissue. Dark green islands may also represent domains within the leaf into which virus had not entered before host defense mechanisms became activated (see below). P. D e v e l o p m e n t of m o s a i c disease The structure of mosaic s y m p t o m s is described in Chapter 9 (Section IV.D). Various hypotheses have been proposed as to how these patterns develop. As noted previously, they develop only in sink leaves to which the virus is transported systemically and are most obvious in leaves that develop at or after the time of systemic invasion. When the third edition of this book was written (Matthews, 1991), it was thought that mosaic patterns were laid d o w n in the shoot apex. It was assumed that, in a plant infected with a mixture of virus strains, the first virus particle to establish itself in a dividing cell pre-empted that cell and all or almost all its
progeny, giving rise in the mature leaf to a macroscopic or microscopic island of tissue occupied by the initial strain. The main observations to support this hypothesis were: 9 Mosaics caused by TMV and TYMV are most apparent in leaves that were in the developmental stage at the time of systemic invasion; in the case of TMV, this is tobacco leaves less than 1.5 cm long. 9 There is often a gradient in sizes of the patches making up the mosaic, with younger leaves having larger islands. 9 In some cases, one half of the leaf would be dark green and the other half light green or chlorotic. 9 Detailed examination showed microscopic mosaics of cells with different levels of virus infection. In these, different horizontal layers of mesophyll had different type of virusinduced chloroplast damage (see Fig. 9.27). However, there are several points that do not support this hypothesis: 1. Only some viruses reach the meristematic region (see Chapter 9, Section II.J.8) where they would have to be to infect cells during the division stage. 2. The recent understanding of long-distance m o v e m e n t of viruses in relation to sourcesink translocation of photoassimilates (see Chapter 9, Section II.G) shows that systemically moving viruses unload from the vascular system some distance from the meristematic region. They would then have to move from cell to cell to reach dividing leaf primordial. 3. The recent studies on cell-to-cell communication indicate that there are symplastic d o m a i n s within which there is good communication and between which the communication is poor. 4. There is an increasing realization that the gene-silencing host defense system (see Section IV.A) plays a very important role in the overall picture of virus infection. Ratcliff et al. (1997) suggested that dark green islands might be the result of gene silencing. Thus, the picture that is emerging is that there is a complex of factors involved in the development of mosaic symptoms including systemic and local movement of the virus, the
Iii.
ability of the virus to invade meristematic tissues, the strain of the virus and its propensity to mutate and, probably above all, the conflict between the invasiveness of the virus and the response of the host defense system(s). It is quite likely that there is a different balance between these, and other, factors in the develo p m e n t of mosaic s y m p t o m s in different virus-host combinations.
Q. Symptom severity A combination of factors leads to differences in the severity of symptoms caused by different isolates or strains of a virus. Some of these have been noted above in the discussion of chlorosis (Section III.O). The factors controlling s y m p t o m severity are both viral and host. For instance, Martin et al. (1997a) examined the responses of 116 ecotypes of Arabidopsis to inoculation with YoMV. They defined five s y m p t o m groups of ecotypes based on stunting, abnormal flower or seed formation, and plant death. Lee et al. (1996) have identified a single locus, TTR1, in Arabidopsis that confers tolerance to TRSV. In some cases, differences in s y m p t o m severity have been attributable to individual viral genes. For instance, c o m p a r i s o n of the nucleotide sequences of two almost symptomless mutants ofTMV (LII and LIIA) with the parent-type strain showed that a change from Cys to Tyr at amino acid position 348 of the 126-kDa and 183-kDa protein(s) was involved in loss of s y m p t o m production. Two other amino acid changes in these proteins may also have been involved (Nishiguchi et al., 1985). The differential symptom determinants of the Holmes' masked (M) and U1 strains of TMV have been m a p p e d to the 126/183-kDa protein, this time to the N-terminal region (Shintaku et al., 1996). Single or multiple substitutions between eight nucleotides in nonconserved domains in the two strains altered the symptomatology but did not always induce complementary visual symptoms. In some cases, there were spontaneous second-site mutations that also influenced the s y m p t o m phenotype. There was no correlation between the severity of systemic s y m p t o m s and virus accumulation. Pseudorecombinants and recombinants between CMV and TAV identified the 3' region of RNA3 (encoding the coat protein) as the
STEPS IN THE I N D U C T I O N OF DISEASE
461
major d e t e r m i n a n t of s y m p t o m severity (Salanki et al., 1997). One combination of RNA1 and RNA2 from TAV and RNA3 from CMV gave more severe symptoms in N. benthamiana than either of the parental viruses or any other pseudorecombinant. On the other hand, RNA2 of a subgroup I CMV strain was identified as being involved in s y m p t o m severity in tomato (Hellwald et al., 2000). In an increasing n u m b e r of cases the transgenic expression of viral genes has been used to explore s y m p t o m expression. For instance, Ghorbel et al. (2001) suggest that CTV p23 is responsible for some s y m p t o m expression in Mexican lime. However, transgene expression does not necessarily identify true s y m p t o m determinants. Not all severity determinants are proteins. 7873
90
109
I_ _ 3 _ 1 _
aal
spread 780 / q05
ehlorosis vil"~.
\N,
~
stunting
1785 6103 215o
v
timing
4438
3956
stunting Fig. 10.10 Location of CaMV genome domains containing strain-specific symptom determinants. Arrows with roman numerals represent viral genes. Domains containing sequences involved in determining different symptom characters between CaMV strains Cabb B-JI and Bari 1 infections are labeled: spread, rate of spread of systemic vein clearing symptoms on younger leaves; stunting, whole plant stunting; timing, period between inoculation and appearance of first systemic vein clearing symptoms; chlorosis, change in chlorophyll content of leaves. The dashed line represents a domain that could contain loci influencing stunting and timing (nucleotides 7873-109) and symptom spread (nucleotides 109-331), but its specific role has not been determined. From Stratford and Covey (1989), with permission.
462
10 I N D U C T I O N
OF D I S E A S E 2" V I R U S - P L A N T
INTERACTIONS
The 3'-terminal non-coding region of TVMV determines the s y m p t o m severity in N. tabacum (Rodriguez-Cerezo et al., 1991a). However, for many viruses, it is the combination of different viral genes (and possibly non-coding regions) that determines the severity of the symptoms. In their analysis of the effect of various CaMV genes on the symptomatology of the virus in turnip, Stratford and Covey (1988) detected a variety of loci affecting disease development (Fig. 10.10). As well as the effect of gene VI on chlorosis (see Section III.O.l.c), plant stunting was affected by at least two separate loci, one containing parts of genes I and II and the second within the reverse transcriptase gene (V). Thus, different aspects of the disease process can be assigned to specific parts of the genome, and much of the viral genome appears to be involved. An example of more than one RNA viral gene being responsible is shown in experiments with hybrids between the tombusviruses CymRSV and C1RV. Burgyan et al. (2000) demonstrated that the necrotic response of N. beilthamiana is associated with the produce of ORF 1 (p35) as well as p19. As members of the plant reovirus group cannot be cloned by single local lesion selection, isolation of mutants has been difficult. However, Kimura et al. (1987) injected dilute inoculum of the type strain (0) of RDV into insect vectors, which were then fed on rice plants. They repeatedly selected for rice plants that showed unusually severe symptoms. By these means, they obtained a severe strain (S) (see Fig. 10.11). The fourth largest genome segment of strain S had an apparent MW about 20 kDa larger than that of strain 0. The corresponding gene product in strain 0 had an M r of about 43 kDa, and in strain S the M was about 44 kDa. This protein is located in the outer envelope of the virus. The idea that this gene product is involved in producing the more severe symptoms receives support from the fact that neurovirulence in a reovirus infecting mice has been shown to be controlled by the outer envelope protein (Weiner et al., 1977). We have to distinguish between s y m p t o m severity and s y m p t o m phenotype. For instance, the common strain of the begomovirus, TGMV, induces extensive chlorosis in N. benthamiana,
Fig. 10.11 Stunting effect of rice dwarf Phytoreovirus infection in rice. Healthy plant on right. Plants infected with a standard strain (O) and a severe mutant (S). From Kimura et al. (1987), with permission. whereas the yellow strain produces vein chlorosis on systemically-infected leaves. The symptoms of these two strains also differ on Datura strallmJliunl. The difference in phenotype expression of these two strains is determined by a single nuc|eotide in the 3' region of the gene encoding the movement protein (Saunders et al., 2001a).
F
R. Recovery For several viruses, infected plants apparently recover after the initial outbreak of systemic s y m p t o m s . This is especially marked in nepoviruses (see Section IV.G) (Wingard, 1928; Lister and Murant, 1967), tobraviruses (Cadman and Harrison, 1959) and caulimoviruses (A1Kaff and Covey, 1995). Virus can be isolated from the symptomless young leaves of plants that have recovered from nepovirus infection
IV. I N H E R E N T H O S T RESPONSE
(Wingard, 1928) and also from symptomless seedlings in which nepoviruses are seed transmitted (Lister and Murant, 1967). AMV can also show recovery in N. tabacum but not in Chenopodium amaranticolor (Ross, 1941; Gibbs and Tinsley, 1961); virus can still be isolated from recovered tobacco leaves (Gibbs and Tinsley, 1961). The virus concentrations in tobacco plants rises and falls in a cyclical manner (reviewed in Hull, 1969) (Fig. 10.12). Ratcliff et al. (1999) note that m a n y of the viruses that appear to recover (except CaMV) have the ability to invade the meristem; this is likely to be associated with pollen transmissibility of the virus.
IV. INHERENT HOST RESPONSE A. Gene silencing The application of genetic engineering to confer resistance to viruses in plants described in [
I
..........
I
..........
]
II
150
i50
i,a,,l ;Z) t"-t~
I00
t00
z l.iJ I-0 el. 0 laJ _I
50
/
z (..9
50
I/
,,
,Z
II " __
0
\0 l
5
DAYS
,L
14 AFTER
....
t. ....
24
i
........
31'
INOCULATION
Fig. 10.12 Concentration and specific infectivity of AMV harvested from whole tobacco plants at different times following inoculation. Zero lesions are assumed for time 0. Curve A, amount of purified virus nucleoprotein (mg/kg total leaf wet weight); curve B, numbers of local lesions induced by sap inoculation; curve C, number of local lesions when purified virus samples were equalized spectrophotometrically. From Kuhn and Bancroft (1961), with permission.
463
Chapter 16 (Section VII) was directed primarily at the expression of viral sequences that would interfere with the normal functioning of the target virus. The initial approaches were to express wild-type or mutated viral genes in the expectation that the gene products would block crucial steps in viral replication a n d / o r propagation. However, various observations on several systems suggested that the mechanism(s) by which the transgenic plants were being protected against viral infection were, in some cases, not as predicted. These observations included: 9 In m a n y cases, it was the transcript and not the protein that was involved in effecting protection. For instance, some lines of potato plants, transgenic for the PLRV coat protein and showing high levels of resistance to the virus, did not contain detectable levels of coat protein ( K a w c h u k et al., 1991). Resistance to TSWV and TEV was induced by transforming tobacco plants with translationally deficient versions of the N gene and of the coat protein respectively (de H a a n et al., 1992; Lindbo and Dougherty, 1992b). 9 Often plants expressing a viral transgene at low level were more resistant than those expressing the transgene at high levels. Thus, in plants transgenic for the TSWV N gene, the highest levels of resistance were found in plants accumulating the lowest levels of transcript (Pang et al., 1993). 9 Some tobacco lines transformed with either the complete or truncated coat protein gene of TEV were initially susceptible to TEV infection but the plants 'recovered' from the infection after about 3-5 weeks. The recovered tissue could not be reinfected with TEV, nor could protoplasts from it. There was a significant decline in the steady-state level of the transgene RNA in the recovered tissue, although nuclear run-off studies showed no differences in transgene transcription rates between recovered and uninoculated tissue (Lindbo et al., 1993). These and other observations suggested that the resistance might be due to homologydependent gene silencing. To test this possibility, crosses were m a d e between lines of tobacco
464
lO I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T I N T E R A C T I O N S
Fig. 10.13 Analysis of the hybrid progeny of tobacco lines, transgenic in PVX replicase, 3.2 (PVX susceptible) and 3.3 (PVX resistant). Leaves were harvested for RNA analysis when the plants were at the five-leaf stage. The probe for gel blot analysis of total RNA (8 ~g) samples was a 1.6-kb fragment of the PVX RdRp DNA (a} or a rubisco small subunit probe (c) to confirm equal loading of the RNA samples. The genotype of each line was determined by DNA gel blot analysis of EcoRV-digested DNA (8 ~g per track) (b). After removal of the leaf sample for RNA analysis the plants were inoculated with PVX and assigned as either resistant (r) or susceptible (s) depending on the absence or presence of systemic mosaic symptoms at 14 days post-inoculation. Samples of the lines 3.2.2 and 3.3.4 and non-transformed tobacco were included at each stage of the analysis for comparison and labeled as 3.2, 3.3 and nt respectively. From Mueller et al. (1995), with kind permission of the copyright holder, 9 Blackwell Science Ltd. t r a n s f o r m e d w i t h the replicase e n z y m e gene of PVX (Mueller et al., 1995) (Fig. 10.13). Some of the lines (e.g. line 3.3) were virus resistant and expressed the transgene at low level; other lines (e.g. 3.2) were susceptible a n d expressed the transgene at relatively high levels. All the p r o g e n y from crosses b e t w e e n these two lines that were virus resistant expressed the transgene at low levels. The p r o g e n y of crosses with non-transgenic tobaccos s h o w e d the transgenic p a r e n t p h e n o t y p e . These o b s e r v a t i o n s were in accord with an increasing n u m b e r of cases in w h i c h t r a n s f o r m a t i o n w i t h h o m o l o g s of e n d o g e n o u s plant genes led to both the transgene a n d e n d o g e n o u s gene expression being c o - s u p p r e s s e d (e.g. Napoli et al., 1990; v a n der Krol et al., 1990). This co-suppression is d u e to either transcriptional gene silencing (TGS) or post-transcriptional gene silencing (PTGS) or, possibly, a c o m b i n a t i o n of the two (for reviews
see Depicker a n d van M o n t a g u , 1997; Vaucheret et al., 1998; M a t z k e et al., 2000).
B. Transcriptional and posttranscriptional gene silencing The m a i n features of TGS a n d PTGS are given in Table 10.7. As most plant viruses have R N A g e n o m e s that replicate in the cytoplasm, it is likely that the transgenic resistance operates by a PTGS m e c h a n i s m involving RNA. In an i m p o r t a n t experiment, English et al. (1996) s h o w e d that PTGS operates against the w h o l e RNA in which a target sequence is located. PVX vectors in w h i c h the [3-glucuronidase (GUS) or green fluorescent protein (GFP) gene h a d been inserted b e t w e e n the triple gene block and the coat protein gene (see Fig. 6.39 for PVX g e n o m e organization) were inoculated to tobacco plants
IV. INHERENT HOST RESPONSE
465
TABLE 10.7 General features of gene silencing mechanisms Feature
Position-dependent gene silencing
Homology-dependent TGS
Homology-dependent PTGS
Homology of silencing sequence and silenced genes
No
Yes
Yes
Properties of the silenced target locus
Cis-located
Homology with silencing locus in the promoter
Homology with silencing locus in the transcribed region
Properties of the silencing locus
Not applicable
Frequently direct or inverted repeated sequences
Frequently direct or inverted repeated sequences
Silencing step
Transcriptional
Transcriptional
Post-transcriptional
Localization
Nuclear
Nuclear
.Cytoplasmic
T r a n s - i n a c t i v a ti o n
No
Yes; paramutagenic
Yes
Phenotype
Recessive
Dominant, sometimes epistatic
Dominant and epistatic
Protein levels
Strongly reduced
Strongly reduced
Reduced
RNA levels
Strongly reduced
Strongly reduced
Reduced
Influence of position effects
Not applicable
Strong
Variable
Influence of sequence organization
Not applicable
Strong
Strong
Influence of position effects
Strong
Variable
No
Influence of locus organization
Variable
Variable
No
Silencing inducer signals
Developmental regulation by thresholds of non-coding RNAs and silencing proteins
In cis
and in trans inactivation by DNA-DNA pairing
RNA thresholds
Silencing effectors
Silencing proteins and/or non-coding RNAs mediating heterochromatinization
Methylation and packing of the chromatin
Degradosomes and anti-sense RNA
Methylation of gene
Present in promoter or complete locus
Present in promoter or complete locus
When present, found in transcribed and downstream regions
Stability/persistence
Meiotic resetting Developmentally controlled
Persistent in different degrees
Meiotic resetting Infrequent somatic resetting Developmentally controlled
Environmental susceptibility
Yes (gene regulation) No (genetic programs)
Variable
Yes (weak silencing loci) No (strong silencing loci)
Examples
C i s - r e g u l a t e d elements and enhancers Position-effect variegation X-chromosome dosage compensation
Natural paramutation Homology-dependent silencing of transgenes, endogens and transposons
Natural in trans silencing Variation of transgene expression Co-suppression Virus-induced gene silencing
sequence
Silencing capacity of a locus
Silencing susceptibility of a locus
From Depicker and van Montagu (1997), with kind permission of the copyright holder, 9 Elsevier Science.
466
10
INDUCTION
OF D I S E A S E 2: V I R U S - P L A N T
line 3.3/-
I '
PVX RdRp hemizygous low transgene RdRp RNA PVX resistant
INTERACTIONS
line3.2/-
I
PVX RdRp hemizygous | high transgene RdRp RNt PVX susceptible /
x non trans I
PVX RdRp homozygous low transgene RdRp RNA PVX resistant
PVX RdRp homozygous| high transgene RdRp RN1 PVX susceptible |
III
crossed
line 3.3/3.2 hemizygous at each transgene locus
low transgene RdRp RNA PVX resistant Fig. 10.14 Experimental testing of the relationship of virus resistance and PTGS. Lines 3.3 and 3.2 are tobacco lines carrying a PVX RdRp transgene. The diagram illustrates the phenotypes of plants carrying these genes either individually or in combination, and shows how it was concluded that the PVX RdRp transgenes conferring PVX resistance were also able to confer PTGS. The primary data were presented by Mueller et al. (1995). The crosses with non-transformed plants (upper panel) revealed that the transgene phenotype with these lines was not affected by transgene dosage. The crosses between the two lines showed that the transgene in line 3.3 could suppress expression of the transgene in line 3.2. The results also show that PVX resistance and low transgene expression were epistatic to high-level expression and PVX susceptibility. From Baulcombe (1999a), with kind permission of the copyright holder, 9 Springer-Verlag GmbH and Co. KG. that were transformed with, and expressed, the GUS gene at low and high levels; the low-level expressers showed PTGS whereas the highlevel expressers did not. The low-level expressers were resistant to both the GUS and to PVX in the PVX:GUS construct but not to the PVX:GFP construct. Plants expressing high levels of GUS transgene were susceptible to both constructs (Fig. 10.14). Various experiments have s h o w n that PTGS is highly nucleotide sequence specific. For
instance, the above example demonstrated that it targeted GUS but not GFP inserts in the PVX vector inoculated to GUS transgenic plants. A PTGS transgene based on the TEV coat protein confers resistance against TEV but not against other potyviruses (Lindbo et al., 1993).
C. Genes involved in posttranscriptional gene silencing PTGS has been f o u n d in other eukaryotic organisms and has been studied in detail in fungi where it is termed 'quelling', and in nematodes and Drosophila where it is termed RNA interference (RNAi) (Gura, 2000; Marx, 2000). By s t u d y i n g mutants of these organisms and of Arabidopisis that were defective in PTGS, several genes t h o u g h t to be involved in the defense p a t h w a y have been identified (see Morel and Vaucheret, 2000). An R N A - d e p e n d e n t RNA p o l y m e r a s e (RdRp) (QDE-1) was s h o w n to be required for PTGS in the fungus Neurospora crassa (Cogoni and Macino, 1999a) and in Arabidopsis (Dalmay et aI., 2000). Mutagenesis of Arabidopsis identified four silencing defective (sde) loci, one of which, sdel, is a h o m o l o g of the RdRP, QDE1, from N. crassa. Although sde mutations affect transgene silencing, they do not have any effect on virus-induced silencing (Dalmay et al., 2000). In a similar study, Mourrain et al. (2000) isolated two Arabidopsis mutants, sgs2 and sgs3, impaired in PTGS. SGS2 protein is similar to N. crassa QDE-1 and to the nematode, Caenorhabditis elegans EGO-l, both RdRps; EGO-1 functions in RNAi (Smardon et a/., 2000). The functions of SGS3 and of the other sde proteins are u n k n o w n but these proteins are likely to be involved in the PTGS p a t h w a y in plants. Analysis of a further Arabidopsis m u t a n t impaired in PTGS, AGO-l, revealed one amino acid essential for PTGS that is also present in QDE-2 and RDE-1 in a highly conserved motif (Fagard et al., 2000). A m u t a n t of the alga, Chlamydomonas reinhardtii, Mut6, required for the silencing of a transgene and two transposon families, is h o m o l o g o u s to the D E A H - b o x family of RNA helicases (Wu-Scharf et al., 2000). A different helicase, encoded by the gene smg-2, is involved in PTGS in C. elegans (Domeier
Iv. I~ m J
Resistant
Recovery
mRNA from 1-2 transgene copies Susceptible
Fig. 10.16 Initiation of RNA-mediated virus resistance (RMVR) according to the threshold model. A threshold of RNA containing virus sequences is required to initiate RMVR. Plants with between three and eight transgenes meet this level and show resistance. Some plants with one or two transgene copies exceed this level only in conjunction with RNA from the virus, and resistance develops some time after infection. Other plants with one or two transgene copies express insufficient mRNA to exceed the threshold even in conjunction with viral RNA, and show no resistance. From Waterhouse et al. (1999), with kind permission of the copyright holder, 9 Elsevier Science.
IV. I N H E R E N T H O S T R E S P O N S E
469
and by Dougherty and Parks (1995), for the induction of PTGS is illustrated in Fig. 10.16. Multiple copies of transgenes can induce their methylation, which, it is suggested, might lead to the production of short aberrant RNA species that induce PTGS (see Waterhouse et aI., 1999; Mette et al., 2000), although not all cases of silencing involve DNA methylation (Wang and Waterhouse, 2000).
Staining tobacco plants showing PTGS of the ~-glucuronidase (uidA) transgene for Gus expression revealed no GUS staining of leaves or cross-sections of stems, but dark blue GUS staining of the shoot apical and axillary meristems (Fig. 10.17) (B6clin et al., 1998). This indicates that silencing does not affect meristems and thus takes place during the development of each leaf.
E. P T G S s y s t e m i c signaling (reviewed by Fagard and Vaucheret, 2000)
E Induction and maintenance
In grafting experiments, plants showing PTGS transmit this character 100% from silenced stocks to non-silenced scions expressing the corresponding transgene (Palauqui et al., 1997), even when the stock and scion are separated by 30cm of stem of non-target wild-type plant. This suggested that a transgene-specific diffusible messenger mediates the propagation of de novo PTGS through the plant. Systemic silencing has also been demonstrated by Voinnet et al. (1998), who infiltrated lower leaves of N. benthamiana, transgenic with a GFP construct, with A. tumefaciens carrying a 35S:GFP construct. After 7-14 days the GFP expression was lost from the upper leaves, especially around the veins. The identity of the signal molecule is unknown but all the evidence (sequence specificity, systemic translocation) suggests a small antisense RNA.
In a virus-induced gene silencing (VIGS) approach, Ruiz et al. (1998) inoculated N. benthamiana plants with PVX vectors carrying the exons of the endogenous phytoene desaturase (PDS) gene. The PDS mRNA was affected in all the green tissue. When PVX:GFP was inoculated to GFP transgenic plants, the PVX:GFP was silenced. From their analysis of the silencing in different parts of these plants, Ruiz et al. (1998) suggested that there are two stages in PTGS: initiation of silencing and maintenance of silencing. The initiation of silencing is an RNA-mediated defense reaction as a transgene is not necessary, and the maintenance requires the presence of the transgene and probably involves methylation of the transgene dependent on an R N A - D N A interaction (Jones et al., 1999). Thus, for PTGS in transgenic plants there are three stages: initiation, systemic signaling and maintenance.
Fig. 10.17 (see Plate 10.2) Histochemical analysis of uidA expression in the meristems of silenced and non-silenced tobacco plants. Histochemical GUS assays were performed on stems and shoot tips of homozygous 6b5 tobacco plants showing uidA PTGS and hemizygous 23b9 plants that did not show uidA PTGS. (A) Staining of stem cross-section of non-silenced 23b9 plant. (B) Staining of stem cross-section of silenced 6b5 plant. A non-silenced axillary meristem is visible as a dark blue spot. (C} Staining of the shoot apical meristem of a silenced 6b5 plant. From B6clin et al. (1998), with permission.
470
10 I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
G. PTGS in virus-infected plants
Systemic accumulation ofTMV::GFP and PVX::GUS
(reviewed by Carrington and Whitham, 1998; Marathe et al., 2000a; Carrington et al., 2001) The finding that viruses can both initiate and be the targets of gene silencing in transgenic plants raised the speculation that PTGS is part of a defense system (also termed RNAmediated defense, RMD; Voinnet et al., 1999) in plants against viruses and other 'foreign' nucleic acids (Baulcombe, 1996a; Pruss et al., 1997). The similarities between plant defense against viruses and gene silencing were e n h a n c e d by several f u r t h e r o b s e r v a t i o n s including: (1) various experiments similar to that s h o w n in Fig. 10.18 indicated that homologous genes carried by virus infection of a transgenic plant silenced the expression of the transgene. (2) In nepovirus infection of nontransgenic plants there are severe viral symptoms on the inoculated and first systemically infected leaves; however, u p p e r leaves developing after systemic infection are s y m p t o m free and contain lesser a m o u n t s of virus than do the systemic leaves (Wingard, 1928) (see Section III.R). These recovered leaves are resistant to reinoculation of the virus. When the recovered leaves of N. clevelandii plants inoculated with the TBRV strain W22 were challenged with a PVX:W22 construct, the PVX:W22 RNA could not be detected, in contrast to the high levels of this RNA is control plants (Ratcliff et at., 1997). Thus, the PVX:W22 was suppressed by the infection of the plants by TBRV W22 in a m a n n e r similar to PTGS in transgenic plants. (3) Kohlrabi (Brassica oleracea gongylodes) plants inoculated with CaMV initially develop systemic s y m p t o m s from which they recover completely. This recovery coincides with marked changes in intermediates of viral replication (see Chapter 8, Section VII.B, for CaMV replication cycle). The viral minichromosome accumulates but the levels of both of the main viral polyadenylated RNAs decline rapidly, a l t h o u g h n o n - p o l y a d e n y l a t e d fragments persist (Covey et al., 1997). These changes in the levels of viral replication intermediates are consistent with gene silencing halting the replication of the virus. However, the difference between these two examples of host defense against viral infection
Co-inoculate TMV::GFP and PVX::GUS Systemic accumulation PVX::GUSGF only
Co-inoculate TMV::GFP and PVX::GUSGF Fig. 10.18 RNA sequence specificity in cross-protection (PTGS). NicotiaJla bet~thatniaJm plants were inoculated with mixed TMV and PVX vector constructs. The TMV vector was TMV:GFP. The PVX vectors were either PVX:GUS or PVX:GUSGF in which the GF component represents the 5' part of the GFP reporter gene. Crossprotection between the two vector constructs was assessed ill systemically infected tissue (Ratcliff et al., 1999). In the plants inoculated with PVX:GUSGF and TMV:GFP there was an interaction involving the GFP sequence in the TMV construct and the GEP-derived GF in PVX:GUSGF. The consequence of this interaction was cross-protection and subsequent suppression of TMV:GFP in the systemic parts of the plant, which is suggested to involve a PTGS-like mechanism. From Baulcombe (1999a), with kind permission of the copyright holder, 9 Springer-Verlag GmbH and Co. KG. is that the replication of the RNA of TBRV is cytoplasmic whereas that of the DNA of the p a r a r e t r o v i r u s CaMV occurs in both the nucleus and cytoplasm. The silencing of TBRV is consistent with the PTGS mechanism, whereas there is evidence that silencing of CaMV involves both PTGS and TGS (A1-Kaff et al., 1998; Covey and A1-Kaff, 2000). Thus, PTGS can be considered to be a generalized response to infection of plants by viruses. This raises the questions of how this defense is initiated and maintained. As RNA viruses replicate via a complementary strand, the formation of d s R N A is an integral part of the replication cycle. Furthermore, host RdRp
~v. ~ . E R E N T , O S T RESPONSE
activity is induced in virus-infected plants (Xie et al., 2001). The replication intermediate, or
dsRNA formed by transcription by the host RdRp, would be a target for the pathway giving the 22-25 nucleotide characteristic degradation products (see Ruiz et al., 1998). Presumably the systemic movement of the signal would be similar to that described above for the transgenic situation. If PTGS operated more rapidly than the virus spread, the virus would not be able to move from the site of infection. If the virus moved more rapidly than the PTGS, it would establish systemic infection, the effectiveness of which would depend upon the 'aggressiveness' of the virus and the 'responsiveness' of the PTGS. Factors involved in virus aggressiveness are discussed in the next section.
H. Suppression of gene silencing (reviewed in Carrington et al., 2001) Obviously, if PTGS is a normal defense system in plants against 'foreign' nucleic acids, for a virus to establish infection successfully this defense has to be overcome. In studying the
471
mediator of synergistic interactions (Section V.F.4) between PVX and PVY, Pruss et al. (1997) identified that the TEV HC-Pro gene potentiated the synergism and suggested that this might be due to it interfering with a host defense system. Experiments reported by Anandalakshmi et al. (1998) and by Kasschau and Carrington (1998) confirmed that the potyviral HC-Pro did indeed suppress the action of PTGS. Both groups essentially used the same approach of inoculating transgenic plants with virus vectors containing various inserts. The experiments are summarized in Fig. 10.19 and Table 10.8. Since this initial demonstration of viral suppression of gene silencing, the phenomenon has been shown for several other viruses. Voinnet et al. (1999) examined 16 viruses using the PVX vector system and showed suppression of PTGS in 12 of them (Table 10.9). The experiments of Brigneti et al. (1998) (Fig. 10.20), which confirmed the suppression of gene silencing by TEV and CMV (B6clin et al., 1998), also showed that it was the 2b gene of CMV that was involved and that the suppression was
TABLE 10.8 Experiment demonstrating suppression of gene silencing by TEV P1/HC-Pro Transgene genotype Plants Non-transgenic 407 F3 generation U-6B • 407#34 U-6B-407#7 U-6B • 407#17 P1/HC-Pro (+) P1/HC-Pro ( - ) 407 • U-6B#13 407 • U-6B#9 407 x U-6B#25 P l / H C - P r o (+) Pl/HC-Pro (-)
Gus
TEV-GFP
TEV-GUS +++
2n
+++ ++ +++ +++
+++
2n 2n 2n
P1/HC-Pro
Infection
ln.2n
+++
+++
-
+ + +
-
-
+ + +
+ + +
2n
-
+ + +
=
2n 2n
ln/2n =
++ +++
+++
Experimental system: Transgenic Nicotiana benthamiana expressing a non-translatable GUS gene (line 407). Two infectious cDNAs to TEV, TEV-GUS containing the GUS gene and TEV-GFP containing the GFP gene (inserted between P1 and HC-Pro). 407 is immune to TEV-GUS and susceptible to TEV-GFP. N. benthamiana line U-6B expressing TEV P1/HC-Pro genes. 407 crossed with U-6B, F1 selfed to give F2; F3 generation as in Table. There were three phenotypes: homozygous GUS:hemizygous P1/HC-Pro (#17, #25); homozygous GUS:nul P1/HC-Pro (#7, #9); nul GUS:nul P1/HC-Pro (#34, #13). From Kasschau and Carrington (1998), with kind permission of the copyright holder, 9 Elsevier Science.
472
10 INDUCTION OF DISEASE 2: VIRUS-PLANT INTERACTIONS
Fig. 10.19 See caption on the following page. V o i n n e t et al. (1999) i d e n t i f i e d t h r e e f u r t h e r v i r a l g e n e p r o d u c t s t h a t a r e i n v o l v e d in g e n e silencing suppression (Table 10.9) and observed that there was no common feature between them except that they were frequently i d e n t i f i e d as ' p a t h o g e n i c i t y determinants'.
mediated by the protein rather than a nucleic acid. T h i s p r o t e i n l o c a l i z e s to t h e n u c l e u s v i a a n arginine-rich nuclear localization signal (22KRRRRR27) ( L u c y et al., 2000). T h e n u c l e a r targeting PTGS.
is r e q u i r e d
for t h e s u p p r e s s i o n
of
TABLE 10.9 Suppression of PTGS of GFP mRNA caused by various plant viruses Protein
Other known functions
. . Vein centric Whole leaf
~ 2b
OL and NL
Whole leaf
AC2
Host-specific longdistance movement Virion-sense gene expression transactivator
0/6 0/9 0/9 8/9 7/9 7/9 10/10
. . . OL OL OL OL
. . . Whole Whole Whole Whole
~ ~ ~ HC-Pro
RYMV
-
-
-
Pl
TMV TRV TBSV
4/6 7/9 7/9
OL and NL OL and NL NL only
Vein centric Whole leaf Vein centric
~ ~ 19 kDa
Virus genus
Virus
Suppression of PTGS
Old or new leaves
Whole leaf or vein centric
Atfamovirus Comovirus Cucumovirus
AMV CPMV CMV
0/9 5/6 20/20
. . OL and NL NL only
Geminivirus
ACMV
6/6
Nepovirus Potexvirus
TBRV PVX FoMV NMV NVX VMV PVY/TEV TEV
Sobemovirus Tobamovirus Tobravirus Tombusvirus
Potyvirus
and and and and
. . . NL NL NL NL
. . . leaf leaf leaf leaf
Genome amplification Viral synergy Long-distance movement Polyprotein processing Aphid transmission Virus accumulation Long-distance movement Host-specific spread and s y m p t o m determinant
PTGS of the GFP was induced in transgenic N. benthamiana by Agrobacterium infiltration as described in Voinnet et al. (1999), with kind permission of the copyright holder, 9 The National Academy of Sciences, USA. OL, old leaves; NL, new leaves.
IV. I N H E R E N T H O S T RESPONSE
473
Fig. 10.19 (see Plate 10.3) Virus-induced gene silencing (VIGS) of transgene-encoded GFP in mixed virus infection. (A) Schematic diagram of the PVX vector constructs carrying the GFPt coding sequence (PVX-GFPt) or the TEV P1/HC-Pro sequence and GFPt. GFPt is the $65T version of GFP. (B} Northern blot analysis of GFP transgene mRNA levels in mockinoculated GFP transgenic plants (lane 1) or plants infected with either PVX-GFPt (lane 2) or PVX-P1/HC-Pro/GFPt (lanes 3 and 4). Lane 4 shows a 3-fold longer exposure of lane 3. Total RNA was extracted from upper leaves of the plants 20 days after inoculation; a GFP-specific probe was used. Arrows indicate the positions of PVX-P1/HC-Pro/GFPt genomic RNA, PVX-GFPt genomic RNA and GFP transgene mRNA (from top to bottom respectively). (C-F) Transgenic GFP Nicotiana benthamiana plants coinoculated with (C) PVX-GFP and PVX-5'TEV, showing complete suppression of virus-induced gene silencing (VIGS) of GFP; (D) PVX-GFP and PVX-HC, showing an almost complete suppression of VIGS of GFP. (E) PVX-GFP and PVX-HC, showing partial suppression of VIGS of G F P - t h e inset is a closer view of a partially silenced leaf showing silencing in the vicinity of the veins; (F} PVX-GFP and PVX-noHC, showing complete VIGS of GFP- the red fluorescence is that of chlorophyll. From Anandalakshmi et al. (1998), with kind permission of the copyright holder, 9 The National Academy of Sciences, USA. There w e r e also differences in the degrees a n d spatial details of the s u p p r e s s i o n , r a n g i n g from s u p p r e s s i o n in all tissues of all infected leaves to s u p p r e s s i o n only in the veins of n e w l y e m e r g e d leaves. This suggests that suppressors m i g h t be targeted to different parts of the gene silencing m e c h a n i s m . For instance, if the s u p p r e s s o r blocks the initiation stage of
silencing it w o u l d s h o w in n e w l y e m e r g i n g leaves w h e r e virus w a s just starting to be synthesized, w h e r e a s if it s u p p r e s s e d the maintenance stage it w o u l d s h o w in both old a n d y o u n g leaves (Brigneti et al., 1998; Voinnet et al., 1999). The e v i d e n c e points to p o t y v i r u s H C - P r o s u p p r e s s i n g at the m a i n t e n a n c e stage a n d the c u c u m o v i r u s 2b protein blocking at
474
lO I N D U C T I O N OF DISEASE 2" V I R U S - P L A N T
INTERACTIONS
Fig. 10.20 (see Plate 10.4) PTGS of a GFP transgene induced by infiltration with A. tumefaciet~s and subsequent suppression by PVY. (A) Schematic representation of the experimental system. GFP silencing was induced in transgenic N. benthamiana line by infiltration with a hypervirulent strain of A. tumeJiwieJts carrying a binary Ti plasmid shown in (B). Complete GFP silencing was achieved at 10 days post-infiltration and plants appeared uniformly red (chlorophyll fluorescence) under UV illumination. (B) Structure of the binary Ti plasmid cassette used to generate transgenic N. benthamiana plants expressing GFP and to induce GFP silencing in this line by A. tltmefaciens (strain cor308) infiltration. The right and left borders of the T-DNA (RB and LB) flank a kanamycin resistance gene (KanR) in a nos promoter (pnos) and a nos terminator (nos) cassette. (C) Transgenic N. bepttha,liatu~ plant showing high levels of GFP expression under UV illumination. (D) Transgenic N. benthamiana after induction of gene silencing by infiltration with A. tumefaciens carrying the binary Ti plasmid shown in (B). (E-G) Suppression of PTGS by PVY. GFP-silenced transgenic N. benthamiana, as in (C), 15 days after infection with PVY viewed under UV illumination. (E) Whole plant; (B) close-up of leaf; (C) close-up of stem. From Brigneti et al. (1998), with kind permission of the copyright holder, 9 The Oxford University Press.
the initiation stage (Brigneti et al., 1998). Furthermore, there were differences within suppression at the maintenance stage, with TMV, CPMV and TBSV suppressing around the veins and the other viruses suppressing over all the old and y o u n g leaves (Table 10.9). This is taken as suggesting that, in the veinoriented phenotype, the suppression is targeted against the systemic signal of silencing (Voinnet et al., 1999). Virus-specific spatial differences with the interference of silencing have also been reported by Teycheney and Tepfor (2001). To analyze the point of potyviral HC-Promediated suppression of the PTGS pathway, Llave et al. (2000) transiently delivered HC-Pro by Agrobacterium injection into tissue of a plant with a silenced GUS transgene. They confirmed that HC-Pro suppresses one or more maintenance steps by targeting a factor that is required on a continual basis or is relatively labile. They s h o w e d that HC-Pro inhibits a step required for the accumulation of the small RNAs and that it reduces the level of cytosine methylation of the t r a n s g e n e sequence. It is possible that the methylation of the PTGS transgene locus is guided by small RNAs diffusing from the cytoplasm and interacting with chromosomal DNA. HC-Pro a p p e a r s to p r e v e n t plants from responding to the mobile silencing signal but does not eliminate the ability to produce or send the signal (Mallory et al., 2001). A suppressor of gene silencing in one host may evoke a different reaction in another. For instance, the 2b gene of TAV, placed in a TMV vector, functions as a virulence determinant suppressing PTGS in N. benthamiana (Li et al., 1999). However, the same gene evokes an HR in N. tabacum.
The plant viral suppressors of PTGS, P 1 / H C Pro and the 2b gene do not suppress transcriptional gene silencing (Marathe et al., 2000b).
I. Other mechanisms of avoiding PTGS Table 10.9 shows that not all viruses appear to be able to suppress gene silencing, at least in the vector system and host used. Thus, there may be other m e c h a n i s m s by which some viruses can avoid the host defense systems. Brigneti et al. (1998) suggest that PVX, which does not appear to have a suppression mecha-
V.
nism, might out-compete the defense system by very rapid replication and spread, or it might avoid the defense by compartmentalization.
I N F L U E N C E OF O T H E R A G E N T S
475
specific suppression of gene expression (virusinduced gene silencing, or VIGS) is leading to a high t h r o u g h p u t procedure for functional genomics in plants (Baulcombe, 1999b).
J. Discussion The realization that plants have a general defense system against viruses has given potential answers to many aspects of the interactions of viruses with plants, but has also raised many questions. The common features of the defenses in plants (PTGS), fungi (quelling) (Cogoni and Macino, 1999c) and animals (RNAi) (Sharp, 2000) are indicating that this is likely to be an ancient system to control 'foreign' nucleic acids and is probably involved in the regulation of gene expression during development (see Fagard et al., 2000; Matzke et al., 2000). The picture that is beginning to appear in relation to PTGS and plant virus infection is one of balance and counter-balance between the virus establishing itself and moving through the plant and the plant system responding to the invasion of a foreign nucleic acid. The results of these interactions can explain why some cell regions become infected and others do not, as in the case of the common mosaic symptom. External factors, such as environmental conditions, and internal factors, such as the plant genome, also play a significant part in the outcome of the initial infection of the plant by the virus. The mechanism(s) of PTGS are beginning to be unraveled. It is becoming more apparent that PTGS and TGS (through methylation) are interlinked (Wang et al., 2001). The overall outline and the similarities to the centre of expression of transgenes are shown in Fig. 10.21. It is likely that there are several pathways by which the final event of silencing occurs. Observations such as the difference in suppression of PTGS by potyviruses and cucumoviruses suggest that there are several routes to the RNA degradation stage. Mutagenesis of model plants such as Arabidopsis will reveal the pathway(s) and gene products involved in PTGS. However, it should be recognized that there may be differences in plants from other families. Similarly, it appears that there are two mechanisms by which viruses can suppress PTGS and it is likely that others will be found. The ability to use viruses to induce sequence-
V. I N F L U E N C E OF O T H E R AGENTS Most viruses mutate frequently to give new strains that may have a marked effect on the disease produced (see Chapter 17). The disease produced by a particular virus in a given host and the extent to which it replicates are sometimes influenced markedly by the presence of a second independent and unrelated virus or by infection with cellular parasites. These latter effects are discussed below.
A. Viroids and satellite R N A s The molecular basis for disease induction by viroids and satellite RNAs is discussed in Sections I.E and II.13.5 of Chapter 14.
B. Defective interfering nucleic acids Virus-dependent defective nucleic acids are biologically similar to satellite nucleic acids in that they require a helper virus for replication and encapsidation, but they differ from satellites in being deletion a n d / o r recombination mutants of the helper virus (reviewed in Bruening, 2000): They are found associated with both RNA and DNA viruses including the following genera: RNA viruses -- Alfamovirus, B romovirus, Closterovirus, Carmovirus, Cucumovirus, Potexvirus, Rhabdovirus, Tobamovirus, Tombusvirus and Tospovirus; DNA viruses-Begomovirus and Curtovirus. As they frequently interfere with the replication of the helper virus, they are termed defective interfering (DI) nucleic acids. The properties of DI particles and nucleic acids are discussed as a manifestation of faults in virus replication (Chapter 8, Section IX.C). The presence of DI nucleic acids in the plant tends to make disease symptoms milder. Thus Hillman et al. (1987) described a DI RNA derived from a culture of TBSV (see Fig. 8.40). In N. clevelandii, TBSV alone causes fatal necrosis, but addition of increasing amounts of the DI RNA to
476
10
INDUCTION
OF D I S E A S E 2: V I R U S - P L A N T
INTERACTIONS
Fig. 10.21 Model of pathways and genes involved in PTGS in plants, dsRNA can be produced in three ways; as a hairpin RNA produced by read-through transcription of two transgenes integrated into the genome as all inverted repeat; as the replicative form of an RNA virus; or as a product of a host-encoded RdRp acting off a ssRNA that is in some way aberrant (and hence recognized as a template). A complex of proteins binds to and then cleaves the dsRNA into short 22-25nucleotide segments. These dsRNA-protein complexes cleave ssRNA molecules containing the same sequences as the oligomers. The same or modified versions of complexes travel to the nucleus, where they direct methylation of DNA sequences homologous to the oligomers, and also travel to other cells where they propagate the RNA degradation mechanism and direct sequence-specific DNA methylation in their nuclei. From Finnegan et al. (2001), with kind permission of the copyright holder, 9 The American Association for the Advancement of Science. the i n o c u l u m resulted in increasingly m i l d e r s y m p t o m s , w h i c h w a s a c c o m p a n i e d by a r e d u c e d p r o d u c t i o n of TBSV. Ismail a n d Milner (1988) isolated a DI particle of SYNV w h o s e RNA w a s 77% as long as that of s t a n d a r d virus. Particles containing this R N A w e r e only a b o u t 80% as long as the s t a n d a r d virus. They were non-infectious on their o w n but, w h e n m i x e d with s t a n d a r d virus, the resulting s y m p t o m s in N. edwardsonii w e r e chlorotic m o t t l i n g instead of the n o r m a l vein-clearing. The DI D N A of A C M V interferes w i t h the replication of the p a r e n t virus in N. benthamiana causing a r e d u c t i o n in infected plants, a delay in s y m p t o m d e v e l o p m e n t a n d s y m p t o m attenuation (Stanley a n d T o w n s e n d , 1985).
C.
Other associated nucleic acids
E x a m i n a t i o n of the causal agents for the whiteflyt r a n s m i t t e d diseases i n d u c i n g yellow veins in Ageratum conyzoides a n d leafcurl in cotton has r e v e a l e d three D N A species. O n e of these r e s e m b l e d D N A A of b e g o m o v i r u s e s , one h a d the characteristics of a satellite (DNA ]3) a n d one h a d features suggestive of a r e c o m b i n a n t b e t w e e n D N A A a n d a n a n o v i r u s (Stanley et al., 1997; M a n s o o r et al., 1999; S a u n d e r s a n d stanley, 1999). A l t h o u g h the D N A A of AYVV systemically infects A. conyzoides a n d N. benthamiana, it causes no s y m p t o m s in either host. D N A A + [3 gives the w i l d - t y p e s y m p t o m s , w h i c h are not affected by the presence of the
V. INFLUENCE OF OTttER A(3ENTS
477
nanovirus-like D N A (J. Stanley, u n p u b l i s h e d data). It is thought possible that other apparently single-component begomoviruses might have a similar complex etiology. D.
Cross-protection
Cross-protection is the protection conferred on a host by infection with one strain of a virus that prevents infection by a closely related strain of that virus. It was demonstrated by McKinney (1929), who s h o w e d that tobacco plants infected with a green mosaic virus (TMV) developed no further s y m p t o m s w h e n inoculated with a yellow mosaic virus. Salaman (1933) found that tobaccos inoculated with a mild strain of PVX were i m m u n e from subsequent inoculation with severe strains of the virus, even if inoculated after only 5 days. They were not i m m u n e to infection with the unrelated viruses, TMV and PVY. This p h e n o m e n o n , which has also been called antagonism, or interference, was soon s h o w n to occur very c o m m o n l y a m o n g related virus strains. It is most readily demonstrated w h e n the first strain inoculated causes a fairly mild systemic disease and the second strain causes necrotic local lesions or a severe disease. D e v e l o p m e n t of such lesions can be readily observed and a quantitative assessment made. Interference between related strains can also be demonstrated by mixing the two viruses in the same inoculum and inoculating to a host that gives distinctive lesions for one or both of the two viruses or strains. Interference by type TMV with the formation of yellow local lesions by another strain is s h o w n in Fig. 10.22. Strains of AMV differ in the aggregation bodies that their particles form in infected cells (see Chapter 17, Section II.C.l.b). Hull and Plaskitt (1970) used electron microscopy to investigate the cross-protection interactions between two strains of AMV. W h e n the challenging strain was inoculated at the same time as the protecting strain or a short time after (about 4 hours), the two aggregation types were found side by side, merging into each other in the same cell. W h e n there was a longer interval between inoculation of the protecting and challenging strains (about 7 hours) the two strains were found in separate parts of the cytoplasm of the same cell and, after an interval of about 10 hours, in separate cells. Only w h e n
Fig. 10.22 Tobacco leaf inoculated with a mutant strain of TMV that produces large slowly developing yellow local lesions. The isolate is not completely freed of the type strain (causing no obvious local lesions). The sharp boundaries to the yellow (light) areas in the lower left quarter of the leaf delineate areas in which type TMV has multiplied and prevented invasion by the yellow strain. (Courtesy of B. Kassanis.) cross-protection was complete as assessed by back inoculation to an indicator host could the aggregation bodies of the challenging strain not be found. There have been several mechanisms prop o s e d for cross-protection, w h i c h were reviewed in the previous edition of this book (Matthews, 1991). However, the recent recognition of gene silencing m e c h a n i s m s being induced by plant virus infection (see Section IV.G) gives a m u c h more rational explanation of this p h e n o m e n o n . Ratcliff et aI. (1999) have p r o d u c e d convincing evidence that PTGS is responsible for cross-protection in a tobravirus and a potexvirus system. Cross-protection with mild virus strains is
478
10 I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
used as a control measure (see Chapter 16, Section IV.A).
E. Concurrent protection Concurrent protection is the reduction in challenging virus infection rate a n d / o r titer due to the co-inoculation with a protecting virus that does not accumulate or induce symptoms in that host plant (Ponz and Bruening, 1986). The cowpea line Arlington shows no symptoms after inoculation with CPMV and no infectivity or accumulation of capsid antigen can be detected (Beier et al., 1977). Co-inoculation of Arlington with CPMV and CPSMV reduced the n u m b e r s of CPSMV-induced lesions. Inoculation of an isogenic line derived from Arlington with CPMV also protected it against infection with CLRV and SCPMV (Bruening et al., 2000). This protection was elicited by CPMV RNA1.
F. Interactions between unrelated viruses 1. Complete dependence for disease For some virus combinations, there is complete dependence of one virus on the other for its replication. The dependent virus is termed a satellite virus and is described in fully Chapter 14 (Section II.A). 2. Incomplete dependence f~)r disease This situation exists b e t w e e n two viruses where both are normally associated with a recognized disease in the field. For example, the important tungro disease of rice is normally caused by a mixture of RTBV, a reverse transcribing DNA virus, and RTSV, an RNA virus (reviewed by Anjaneyulu et al., 1994). RTSV on its own is transmitted by the rice green leafhopper, but causes no symptoms. RTBV causes severe symptoms in rice but no vectors are known for this virus on its own; it requires the presence of RTSV for transmission (see Chapter 2, Section III.A.6). Thus, in the disease complex RTSV gives the transmission and RTBV most of the symptoms. Most, if not all, umbraviruses are associated with luteoviruses, which provide their insect transmission (see Chapter 11, Section III.H.l.a).
An even more complex system is that of groundnut rosette disease which involves three agents: an u m b r a v i r u s (GRV), a luteovirus (GRAV) and a satellite RNA (GRV sat-RNA). The sat-RNA is the major cause of symptoms (Murant et al., 1988a; Murant and Kumar, 1990) and both it and GRAV are required for aphid transmission (Murant, 1990). The GRAV coat protein encapsidates its own genome and also that of GRV and sat-RNA, the presence of sat-RNA being essential for encapsidation (Robinson et al., 1999). 3. Effects on numbers of local lesions Thomson (1961) found that, when PVX was inoculated on to tobacco leaves together with TMV or PVY, the PVX lesions were more numerous. The increase was highly variable, ranging from 2- to 80-fold depending on conditions. From studies on the effect of dilution and of changing ratios of the two viruses, Close (1962) concluded that the m a x i m u m stimulation by TMV of PVX lesion formation occurs at a definite concentration of TMV in the inoculum and does not depend on the ratio of amounts of the two viruses. This can possibly be explained by virus suppression of PTGS (see Section IV.H). 4. Synergistic effects ~m virtls replicati~n Joint infection of tobacco plants with PVX and PVY is characterized by severe veinal necrosis in the first systemically infected leaves. Leaves showing this synergistic reaction contain up to 10 times as much PVX as with single infections but only the same amount of PVY (e.g. Stouffer and Ross, 1961). Ultrastructural studies and fluorescent antibody staining showed that both viruses were replicating in the same cells and that the increased production of PVX was due to an increase in virus production per cell rather than an increase in the number of cells supporting PVX replication (Goodman and Ross, 1974a). The level of PVX (-)-strand RNA increased disproportionately to that of (+)strand RNA in doubly infected tissues, suggesting that the synergism involves an alteration in the normal regulation of the relative levels of the two RNA strand polarities during viral replication (Vance, 1991). TVMV, TEV and PepMoV also give a synergistic reaction with
V.
PVX (Vance et al., 1.995) and various other combinations of potyviruses and unrelated viruses have synergistic interactions. The joint infection of V1CMV and CMV causes cowpea stunt (Anderson et al., 1996) and corn lethal necrosis is due to joint infections of the machlovirus, MCMV, with the potyvirus, M D M V or SCMV (see Scheets, 1998). Joint infections of potyvirus SPFMV and the crinivirus SPCSV lead to severe s y m p t o m s in sweet potato (Ipomoea batatas), w h e r e a s the i n d i v i d u a l viruses cause mild s y m p t o m s (Karyeija et al., 2000). In all these cases, the concentration of the potyvirus is similar to that in a single infection, whereas that of the other virus is increased markedly; in the sweet p o t a t o disease complex, SPCSV remained limited to the phloem. However, corn lethal necrosis is also caused by joint infection of MCMV w i t h the r y m o v i r u s WSMV (Scheets, 1998). In this case, the concentrations of both viruses were enhanced by the dual infection. To determine which region of the potyviral genome enhanced PVX multiplication, Vance et al. (1995) inoculated tobacco plants transgenically expressing various regions of the TEV genome (Fig. 10.23). The synergistic reaction was only in plants expressing PI-(HC-Pro)-P3. This region also enhanced the replication of CMV and TMV and, for all three enhanced viruses, it w a s the expression of the HC-Pro protein that was required for synergism (Pruss et al., 1997). The recognition that potyviral HCPro suppresses the host gene silencing defense m e c h a n i s m (see Section IV.H) suggests that the synergy m i g h t be due to reduction of the host inhibition of the replication of the other virus. A striking synergistic effect was found with BRNV in herbaceous hosts. In mixed infections with the sobemovirus, SNMoV, the normally very low concentration of the raspberry virus was increased 1000-fold (Jones and Mitchell, 1986). In rice tungro disease, the presence of RTSV e n h a n c e s the s y m p t o m s c a u s e d by RTBV. Unlike the potyviral synergisms, there was no significant increase of either RTBV or RTSV in joint infection w h e n c o m p a r e d w i t h single infections (F. Sta Cruz, O. A z z a m and R. Hull, u n p u b l i s h e d data). The presence of a satellite nucleic acid can either reduce or enhance the helper virus repli-
I N F L U E N ( ; E OF O T H E R A G E N T S
479
VpG P1
' HC-pro
P3
CI
'
Nla
NIb
f CP
6K Line
.
T V M V gene
Km segregation
Protein level
1
coat 13rotein (CP)
98:1
0.30%
2
coat protein (CP)
65:22
0,25%
3
CI
66:26
0.15%
4
CI
64:22
0,10%
5
Nla
73:22
#
6
Nla-glnH
59:20
0.20%
8
Cl-6K-Nla-NID-CP
*
0,02%
9
CI-6K-Nla-NIb-CP
*
002%
10
PI-(HC-ProFP3
40:1
1%"
11
P 1-(HC-Pro)-P3
2:1
1%"
Synergism
Fig. 10.23 Diagrammatic representation of the TVMV polyprotein showing the localization of the individual genes and a listing of the TVMV transgenic lines indicating the genes included in the construct, the segregation characteristics of the transgene as measured by selection on kanamycin, and the protein level of the encoded gene product(s) in plant leaves. Km, kanamycin; CI, cytoplasmic inclusion body protein; NIa, nuclear inclusion body protein a; glnH, Escherichia coli glutamine-binding protein; NIb, nuclear inclusion body protein b. *Indicates that these seeds are homozygous and do not segregate as measured on kanamycin plates. #Indicates that no appropriate antibodies were available for this measurement. Expression levels for lines 1-9 refer to percentage of transgene-derived protein/total soluble protein. qndicates that expression levels for lines 10 and 11 refer to percentage of transgene expression compared with a wild-type virus infection. In all cases, plants used to determine the level of gene expression were grown from the same generation of seed used to produce plants for the studies on synergism with PVX. From Vance et al. (1995), with permission.
cation and s y m p t o m production (see Chapter 14, Section II.B.5). 5. Effects on virus movement As discussed in Chapter 9 (Section II.D.6), infection and systemic m o v e m e n t by one virus in a particular host m a y allow the cell-to-cell and systemic m o v e m e n t of an unrelated virus that n o r m a l l y w o u l d not move from the initially infected cells in that host. Similarly, a fully systemically infecting virus can c o m p l e m e n t the m o v e m e n t of a tissue-restricted virus (e.g. p h l o e m - l i m i t e d virus) out of that tissue.
480
10 I N D U C T I O N
OF D I S E A S E 2: V I R U S - P L A N T
INTERACTIONS
6. Effects on macroscopic disease symptoms Not uncommonly, a mixed infection with two viruses produces a more severe disease than either alone. The classical example of this synergistic effect is the mixture of PVX and PVY in tobacco described above, which produces a severe veinal necrosis instead of the milder mottling or vein-banding diseases seen with either virus separately (Smith, 1931). In m a n y potato varieties, these two viruses together produce a 'rugose mosaic' that is more severe than the disease produced by either virus alone. Many strains of ToMV produce only a mosaic disease in tomatoes but, in combination with PVX, a severe 'streak' disease ensues and usually kills the plant (Plate 3.20). Double infections with PVY and the viroid PSTVd caused severe necrotic disease in some potato cultivars in the field (Singh and Somerville, 1987). Isolated protoplasts have been used to study mixed infections. For example, Otsuki and Takebe (1976a) inoculated tobacco mesophyll protoplasts with TMV and CMV. Some 70-80% of protoplasts supported both viruses without any synergistic or antagonistic effects. On the other hand, in mixed infections of BMV and CCMV in tobacco protoplasts, both types of virus particle are produced, but only those of BMV are infectious. Only RNA3 of CCMV was synthesized in the mixed infections (Sakai et al., 1983). 7. Cytological effects in mixed infecticms Mixed infections leading to necrosis m u s t clearly have marked effects on individual cells. In mixed infections where cell death does not occur, two different viruses may replicate in the same cell producing their characteristic inclusion bodies or arrays of virus particles without any significant indications of mutual interference (e.g. Poolpol and Inouye, 1986; Langenberg, 1987).
G. Interactions between viruses and fungi 1. Effect of virus infection on fungal diseases a. Increased resistance
Observations in the field and glasshouse tests have shown that infection with Phytophthora
infestans developed less rapidly in potato plants infected with one of a n u m b e r of viruses (Mtiller and Munro, 1951). Numerous other examples are k n o w n where virus infection reduces susceptibility to, or development of, fungal and bacterial parasites. For example, infection of a hypersensitive tobacco cultivar with TMV induced systemic and long-lived resistance against Phytophthora parasitica, Peronospora parasitica and Pseudomonas tabaci (McIntyre et al., 1981). Similarly, systemic resistance to anthracnose in cucumber was induced by inoculation with TMV (Jenns and Kuc, 1980), as was resistance to Peronospora tabacina in tobacco (Ye et al., 1989) and Erysiphe chicoracearum (Marte et al., 1993). The protection against E. chicoracearum was related to an accumulation of hydroxyproline-rich glycoproteins induced by the TMV infection (Raggi, 1998). The development of resistance of this sort probably involves the PR proteins discussed in Section III.K.1. Indeed, fungicidal compounds have been isolated from plants reacting with necrosis to virus infection (e.g. Burden et al., 1985). Generalized defense reactions may not be involved in some other virus-fungus interactions. Thus, infection of faba bean with BYMV decreased pustule density on leaves subsequently inoculated with Uromyces viciafabae. Changes were most marked on leaves showing yellowing symptoms (Omar et al., 1986). b. Increased susceptibility
Russell (1966) showed that sugar beet plants in the field that were infected with BMYV had greatly increased susceptibility to Alternaria infection. BYV had no such effect. BMYV increased and BYV decreased susceptibility to another fungus, Erisphye polygoni. Plants infected with both viruses had about the same susceptibility as healthy plants. The precise extent of the interaction d e p e n d e d on the genetic constitution of the host plant and on the environmental conditions. Many other instances of increased susceptibility to fungi have been reported following virus infection. For example, prior infection of wheat or barley with BYDV predisposed the ears to infection by Cladosporium spp. and
VI.
Verticillum spp. (Ajayi and Dewar, 1983). Sporulation of Helminthosporium maydis on corn leaves began sooner and was more abundant in lesions formed on leaves infected with MDMV (Stevens and Gudauskas, 1983). Asparagus infected with AV-2 had increased susceptibility to Fusarium crown and root rot (Evans and Stephens, 1989). 2. Effects of fungal infection on susceptibility to viruses Infection by a rust fungus may greatly increase the susceptibility of leaves to several viruses (Yarwood, 1951). For example, pinto bean leaves were heavily inoculated with the uredinial stage of Uromyces phaseoli on one half-leaf and then later with TMV over the whole leaf. Subsequent estimations of the amounts of TMV showed the presence of up to 1000 times as much virus infectivity in the rusted as in the non-rusted half-leaves. Rust infection partially suppressed the development of visible necrotic lesions. This suppression made it impossible to determine whether the increase in virus content was due to an increase in the n u m b e r of successful entry points or to increased virus multiplication or to a combination of both of these factors. Other fungi may induce resistance or apparent resistance to viral infection. Xanthi tobacco plants that had been injected in the stem with a suspension of spores of Peronospora tabacina produced fewer and smaller necrotic local lesions w h e n inoculated with TMV 3 weeks later (Mandryk, 1963). The fungus Thielaviopsis basicola causes necrotic local lesions in tobacco. Hecht and Bateman (1964) found that, if tobacco plants were infected on the lower leaves with this fungus, upper leaves became resistant to infection with TNV or TMV, as judged by size of the necrotic viral lesions. Inoculation of cucumber with Colletotrichum lagenarium or Pseudomonas lachrymans induced systemic resistance to CMV in an upper leaf (BergstrOm et al., 1982). Some of these inhibitory effects at a distance may be due to SAR or possibly PR proteins discussed in Section III.K. On the other hand, a glucan preparation from Phytophthora megasperma protected a number of Nicotiana spp. from infection
I)ISCUSSION AND SUMMARY
481
by several viruses without the induction of pathogenesis-related proteins (Kopp et al., 1989). 3. Effects of virus infection on non-vector insects Infection by TMV improved the suitability of tomato for survival of Colorado potato beetle larvae (Hare and Dodds, 1987). It was suggested that virus infection may facilitate the adaptation of p h y t o p h a g o u s insects to 'marginal' host plant species. 4. Interactions between virus infection and air pollutants There is increasing concern about the effects of air pollutants such as ozone on plant growth. There have been several investigations of the effect of virus infection on the severity of ozone leaf damage. For most host-virus combinations, viral infection reduced damage. For example, infection with TMV reduced leaf damage in tobacco leaves due to ozone from 11% to 5% (Bisessar and Temple, 1977). The effect was seen in both glasshouse and field trials. Subacute doses of sulfur dioxide caused small but consistent increases in the content of SBMV in beans and MDMV in maize (Laurence et al., 1981).
VI. D I S C U S S I O N A N D SUMMARY The effects of a virus entering a cell in an uninfected plant depend on the interplay of two genomes, that of the virus and that of the plant. The activities of these genomes can be affected by influences such as the environment, the growth stage of the plant, the site of virus entry and the effects of other pathogens. The interplay between viral and host genes in disease induction is currently one of the most biologically interesting and practically important areas of plant virus research. Site-directed mutagenesis, transgenic plants, non-destructive fluorescent probes and other techniques together with in vitro or protoplast systems, and sometimes whole plants, can identify the main function or functions of a gene product, for example that it is a replicase or a protease. However, to study disease induction intact
482
lO I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T I N T E R A C T I O N S
plants or parts of plants must be used, and a full virus replication cycle studied. For these reasons there may be significant difficulties ahead. The in planta roles of some of the protein products of viral genes involved in disease induction may be very difficult to study for several reasons including: 1. The proteins may be present in very low concentration, as a very few molecules per cell of a virus-specific protein could block or derepress some host-cell functions. 2. It is quite possible that such proteins would be present in the infected cells for a short period relative to that required for the completion of virus synthesis. 3. The virus-specified polypeptide may form only part of the active molecule in the cell. 4. The virus-specified polypeptide may be biologically active only in situ, for example in the membrane of some particular organelle. As noted in Section III, many macroscopic disease symptoms may be due to quite unexpected side effects of virus replication. The following hypothetical example illustrates this kind of possibility. Consider the NIa proteinase coded for by TEV (Carrington and Dougherty, 1988). The amino acid sequences that function as substrate recognition signals have been identified. In the usual host under normal conditions, this proteinase can accumulate to high levels within infected cells without causing cell death. This must mean that the proteinase does not significantly deplete the amount of any vital host protein. Suppose that we change to another host species, or to different environmental conditions. In the new situation, some host-coded protein essential for cell function might be sensitive to cleavage, with a new pattern of disease developing as a consequence. In such circumstances, it might be difficult to identify the host protein involved, and its functions. In attempting to u n d e r s t a n d disease induction in molecular terms it may be most profitable to concentrate initially on effects that are k n o w n to be a direct consequence of and essential for virus replication, or for virus m o v e m e n t from cell to cell. A l t h o u g h molecular approaches will no
doubt continue to increase our understanding of the role of viral genes in the induction of disease, difficulties are emerging. For example, site-directed mutagenesis, which is in principle a powerful technique, has several limitations. (1) The n u m b e r of possible permutations and combinations of base sequence alteration is enormous. (2) It will be difficult to find changes that are not fatal for the virus when a full infectious cycle is required. (3) Because of the high mutation rate in RNA viruses and the existence of recombination (Chapter 17), it is often necessary to check the complete base sequence of the engineered mutant culture to ensure that spontaneous changes have not caused reversion to the original sequence or altered the sequence in some other way. (4) A more general difficulty may be that, as increasingly detailed experiments are carried out with a particular virus, most or all viral genes may be found to have interacting roles in various aspects of disease induction. The new techniques being developed in genomic research should provide powerful tools for unraveling details of the host responses to viral infections. Not only will they increase the understanding of virus disease resistance (see Michelmore, 2000), but also they should give information on the permissive situation. For instance, microarray analysis of infection of Arabidopsis with the fungal pathogen Alternaria brassicicola (albeit an incompatible situation) revealed substantial changes in the steady-state abundance of 705 host mRNAs (Schenk et aI., 2000). Depending on the response to inoculation with a virus, plants have been described as either immune or infectible. If a plant appeared immune it is considered a non-host for the virus, and the virus does not replicate in any cells of the intact plant or in isolated protoplasts. This extreme resistance is discussed in detail in Section III.A. In infectible species or cultivars, the virus can infect and replicate in isolated protoplasts. The plant may be either resistant or susceptible to infection. Until recently, two kinds of resistance were recognized. In resistance involving subliminal infection, virus multiplication is limited to the initially infected cells because the
VI.
viral-coded protein necessary for cell-to-cell movement of virus is non-functional in the particular host. In the past, m a n y examples of this type of resistance were described as immune. In the second kind of resistance, infection was considered to be limited by a host response to a zone of cells around the initially infected cell, usually resulting in necrotic local lesions. Uninfected tissue surrounding these lesions becomes resistant to infection. This was called acquired resistance. However, at least two forms of host resistance response have now been recognized. These can be termed a specific resistance response and a generalized resistance response. The specific resistance response is directed by one or more host genes, the products of which interact with certain viral determinants and limit the spread of the virus from the site of initial infection. The limitation is usually by an HR. It is likely that the containment by the specific response is not always by an HR and may sometimes be by a mechanism that does give a visible symptom. The generalized resistance is that mounted by the plant against 'foreign' nucleic acids, in the case of viruses usually by PTGS (see Section IV) (reviewed by Covey, 2000; Ding, 2000). Very recently, it has been demonstrated the begomovirus AC2 and curtovirus C2 genes supress a general stress defence response in plants (Sunter et al., 2001). This points to at least two forms of generalized resistance, both of which viruses can overcome. A virus that does not cause systemic disease in a particular plant has been termed as nonpathogenic for that plant. If a virus or virus strain causes systemic disease in a particular species or cultivar, it is pathogenic. A gene for resistance introduced into such a susceptible species or cultivar may make the virus avirulent. However, the virus may mutate and overcome the host resistance to become a virulent strain. Thus, both host and viral genes interact to determine the outcome of inoculation. The change from an avirulent to a virulent virus strain may involve no more than a single amino acid change in a virus-coded protein (for examples see Sections III.A-E). In species or cultivars that are susceptible, the virus replicates and moves systemically. In
DISCUSSION AND SUMMARY
483
a sensitive reaction, disease ensues. If the plant is tolerant, there is no obvious effect on the plant, giving rise to a latent infection. The consequences of infection for a susceptible plant are determined by both host and viral genes. For example, a single base change in the TMV coat protein gene may be sufficient to alter the nature of the resulting disease. In the light of these new findings on host responses to virus infection, it is necessary to modify the ideas of the events leading to the establishment of fully systemic virus infection. If the virus is not contained by extreme resistance or a local hypersensitive response, or if it overcomes the HR, the outcome of the infection is dependent on the 'aggressiveness' of the virus overcoming the generalized resistance response of the host. Several viruses have been found to encode genes that suppress the generalized PTGS defense (see Section IV.H), and two basic systems have been recognized that operate at different times in the infection process. It is likely that other viral genes with be found that suppress PTGS and that there will be other variants of the suppression process. However, for other viruses such as PVX, no PTGS suppression system has been detected and it is suggested that the rapid multiplication and spread of the virus outcompetes the host defense response. As noted earlier for viruses that overcome the defense mechanisms, m a n y environmental factors influence the course of infection and disease. These include light, t e m p e r a t u r e , water supply, nutrition and the interactions between these factors during the growing season. Complex interactions may occur w h e n plants are infected with two unrelated viruses or with a virus and a cellular pathogen. Factors such as these would affect the expression of either the plant or viral genome or both and thus, the outcome of an infection depends upon the individual circumstances. The actual processes involved in the induction of disease symptoms are not well understood. For strains of a given virus we should distinguish between the severity of a particular s y m p t o m and different s y m p t o m phenotypes. Many of the biochemical changes involved may not be directly connected with virus replication. Stunting probably involves changes in
484
10 I N D U C T I O N OF DISEASE 2: V I R U S - P L A N T
INTERACTIONS
the balance of growth hormones. The formation of mosaic patterns in virus-infected leaves involves events that occur in the early stages of
leaf ontogeny and the local balance between the aggressiveness of the virus and the host defense systems.
C
H
A
P
T
E
R
11
Transmission 1:
By Invertebrates, Nematodes and Fungi I. I N T R O D U C T I O N
cle and the cellulose cell wall. This problem is overcome either by avoiding the need to penetrate the intact outer surface (e.g. in seed transmission or by vegetative propagation) or by some method involving penetration through a w o u n d in the surface layers, such as in mechanical inoculation and transmission by insects. There is considerable specificity in the mechanism by which any one virus is naturally transmitted (see Appendix 2). In this chapter, I will discuss h o w their biological vectors transmit viruses. Mechanical and seed transmission of viruses are described in Chapter 12 together with how viruses spread through crops and natural populations.
One could envisage a virus surviving for hundreds of years in an individual tree of a long-lived species. However, being obligate parasites, viruses usually depend for survival on being able to spread from one susceptible individual to another fairly frequently. Knowledge of the ways in which viruses are transmitted from plant to plant is important for several reasons: 9 From the experimental point of view, we can recognize a particular disease as being caused by a virus only if we can transmit the virus to healthy individuals by some means and reproduce the disease. 9 Viruses are important economically only if they can spread from plant to plant fairly rapidly in relation to the normal commercial lifetime of the crop. 9 Knowledge of the ways in which a virus maintains itself and spreads in the field is usually essential for the development of satisfactory control measures. 9 The interactions between viruses and their invertebrate and fungal vectors are of considerable interest, both from the scientific point of view and in developing new approaches to the control of viruses. 9 Certain methods, particularly mechanical transmission, are very important for the effective laboratory study of viruses.
II. T R A N S M I S S I O N BY INVERTEBRATES Many plant viruses are transmitted from plant to plant in nature by invertebrate vectors (see Appendix 3). Of some 22 phyla in the invertebrates, only three have many members that feed on living green land plants. These are the Urinamia and Crustacea phyla of the Arthropoda and the Nematoda. Both of the Arthropoda phyla, and the Nematoda, contain vectors of plant viruses. Two additional phyla, the Annelida and Mollusca, have a few plant feeders, and it these that may contain potential vectors of a strictly mechanical sort.
A. Arthropoda
Viruses cannot penetrate the intact plant cuti-
I have followed the taxonomy used by Webb et aI. (1978) and by Richards and Davies (1994). Of the three subphyla of the Uniramia, one, the
Virus acronyms are given in Appendix 1.
485
486
11 TR A N SMISSI ON 1: BY INVERTEBRATES, NEMATODES AND FUNGI
Hexapoda (Insecta), contains members feeding on living green land plants; one class of the Crustacea, the Arachnida, has members feeding on green land plants. Both of these groups contain virus vectors.
1. Insecta A m o n g 29 orders in the living Insecta there are nine with members feeding on living green land plants and that might, therefore, be possible vectors. These are listed below with the approximate n u m b e r of vector species at present known:
1. Collembola--chewing insects; some feed on green plants (0)
2. Orthoptera--chewing insects; some feed on green plants (27)
3. Dermaptera--chewing insects; a few feed on green plants (1)
4. Coleoptera~chewing insects; m a n y feed on green plants; see Table 11.1 for vectors
5. Lepidoptera~chewing insects; larvae of many
are still uncertainties about the classification within the Homoptera (Campbell et al., 1995).
2. Arachnida Only one of 12 orders in this class, the Acari (mites and ticks), contains members feeding on living green land plants. The Acari have four families containing mites that are green plant feeders: the Tetranychidae, Tarsonemidae, Eriophyidae and Acaridae. Virus vectors are known in the third and possibly the first of these families.
B. Nematoda There are 10 orders in the Nematoda (Goodey, 1963). Most of the nematodes parasitic in living green plants belong to the Tylenchida, but none from this group has yet been found to be a virus vector. Vectors known so far are confined to the Dorylaimida group containing only a few plant parasites.
feed on green plants (4)
6. Diptera--larvae of a few feed on green plants
(2) 7. Hymenopteramlarvae of a few feed on green plants (0)
8. Thysanoptera (thrips)msome are rasping and sucking plant feeders (10)
9. Hemiptera--feed by sucking on green plants Suborder Heteroptera, families Myridae and Piesmatidae ( - 4 ) Suborder Homoptera, see Table 11.1. The first seven orders listed are all chewing insects, and representatives of these orders feed on living green plants as larvae or adults, or both. Vectors of a strictly mechanical sort have been found a m o n g the Orthoptera, Dermaptera and larvae of Lepidoptera and Diptera. Except for a few viruses, vectors in these orders are of minor importance. Important vectors occur in the Coleoptera. The Collembola and Hymenoptera contain relatively few species that are common pests of agricultural plants. There may be potential vectors among them. The Thysanoptera contains vector species for one i m p o r t a n t g r o u p of plant viruses. The Homoptera is numerically the most important suborder containing plant virus vectors. There
C. R e l a t i o n s h i p s b e t w e e n p l a n t v i r u s e s and invertebrates The transmission of viruses from plant to plant by invertebrate animals is of considerable interest from two points of view. First, such vectors provide the main method of spread in the field for many viruses that cause severe economic loss. Second, there is considerable biological and molecular interest in the relationships between vectors and viruses, especially as some viruses have been shown to multiply in the vector. Such viruses can be regarded as both plant and animal viruses. Even for those that do not multiply in the animal vector, the relationship is usually more than just a simple one involving passive transport of virus on some external surface of the animal. Transmission by invertebrate vectors is usually a complex phen o m e n o n involving specific interactions between the virus, the vector and the host plant coupled with the effects of environmental conditions. In this chapter I consider the groups involved as virus vectors and outline the kinds of relationship that have been found to exist between virus and vector. Vectors are considered in relation to the ecology of viruses in
III.
APHIDS
(APHIDIDAE) 487
TABLE 11.1 Distribution of plant virus vectors among selected Homoptera and Coleopterafamilies Order, suborder, family
Homoptera Auchenorrhyncha Cicadidae Membracidae Cercopidae Cicadellidae Fulgoroidea Sternorrhyncha Psyllidae Aleyroididae Aphididae Pseudococcidae Coleoptera Chrysomelidae Coccinellidae Cucurlionidae Meloidae
Common name of insect group
Cicada Treehopper Spittlebug Leafhopper Planthopper
Approx. no. species described
No. viruses transmitted
3 200 4 500 3 600 15 000 19 000
0 1 0 49 28
0 1 0 31 24
2 000 1 200 4 000 6 000
0 3 192 19
0 43 275 10
20 000 3 500 36 000 2 100
48 2 10 1
30 7 4 1
Psyllid Whitefly Aphid Mealybug Leaf beetle Ladybird beetle Weevil Blister beetle
No. vector species
From Nault (1997), with permission. Chapter 12 a n d in relation to disease control in Chapter 16. As a general rule, viruses that are transmitted by one type of vector are not transmitted by any of the others. This specificity is not only at the level of vector type, family, genus or species but can be even at the level of biotype. There are two basic interactions b e t w e e n viruses and their biological vector. They m a y be taken up internally w i t h i n the vector, t e r m e d internally borne or circulative, or they m a y not pass to the vector's interior, in which case they are t e r m e d externally borne or non-circulative (Hull, 1994a; Gray, 1996). Virus-vector relationships have been studied in most detail in insects, and especially aphids. The basic features of these relationships are described in Section III.D.
III. APHIDS (APHIDIDAE) A. Aphid life cycle and feeding habits A m o n g insects, the aphids have evolved to be the most successful exploiters of higher plants as a food source, particularly flora of t e m p e r a t e regions. It is therefore not surprising that they have also d e v e l o p e d into the m o s t i m p o r t a n t group of virus vectors.
1. Life cycle In t e m p e r a t e climates, aphids frequently alternate b e t w e e n a p r i m a r y and a secondary host. They are remarkable for the n u m b e r of forms produced. A complete cycle is s h o w n in Fig. 11.1. There are m a n y variations in this cycle, d e p e n d i n g on aphid species and on climate. For e x a m p l e , s o m e m a y o v e r w i n t e r as parthenogenetic viviparous forms. Some m a y pass t h r o u g h their life cycle on one host species or several species within one genus. Myzus persicae, an i m p o r t a n t vector aphid, has Prunus spp. as its p r i m a r y host, alternating w i t h seco n d a r y hosts in over 50 plant families. There are three kinds of variability in a p h i d s that m a y affect their ability to transmit a virus: 9 An a p h i d species m a y contain different clones or races (beotypes), with or w i t h o u t obvious morphological differences. 9 Aphids can exist in different forms, as noted earlier. 9 Successive m o u l t s by the developing insect define the n u m b e r of stages or instars. 2. Mouthparts The m o u t h p a r t s of aphids consist of two pairs of flexible stylets, held within a groove of the
488
11 TRANSMISSION
1: BY I N V E R T E B R A T E S , N E M A T O D E S A N D F U N G I
PRIMARY HOST WINTER /
fundatrices (apterae)
~.SPRING fundatrigeniae (apterae)
\
eggs
fundatrigeniae
~v~tPear~aae,,,!' e ~
I
gynoparae (alatae = r e m ~ sexuparae (apterae)
A UTUMN
!
males (alatae)
virginoparae (exulae apterae) virginoparae (alatae)
9 vi rgi nopa rae (apterae) ~"
HOLOCYCLE
SPRING
WINTER
(alatae = emigrants)
I
_ virginoparae (apterae)
virginoparae (alatae)
"
virginoparae (alienicolae alatae) I virginoparae (apterae)
#,
A UTUMN
vi rg i no pa rae (apterae)
virginoparae (alatae)
SUMMER
SECONDARY HOST ANHOLOCYCLE
Fig. 11.1 Two different types of aphid life cycle. In the holocycle, the aphid alternates between two hosts, overwintering usually as eggs on the primary host. In the anholocycle, the aphid overwinters as apterae, which usually find shelter. From Robert and Lemaire (1999b), with permission.
labium. They are extended from the labium during feeding. The maxillary stylets have a series of toothlike projections near their tips (Fig. 11.2). Details of the feeding mechanism such as the sucking pump and the esophageal valve have been described by Forbes (1977) and McLean and Kinsey (1984) (Fig. 11.3). 3. Feeding habits At the beginning of feeding a drop of gelling saliva is secreted. The stylets then rapidly penetrate the epidermis and, in exploratory probes, the aphid may feed there temporarily. Penetration usually continues into the deeper layers with a sheath of gelled saliva forming during penetration. The stylets usually move between ceils until they reach a phloem sieve tube (a process that may take minutes or hours) (Fig. 11.5A). Only the maxillary stylets enter the sieve tube. Compression by the cell wall causes the tips to open, exposing the end of the food and salivary canals. Electronic monitoring of insects while they are feeding can give information on their feeding behavior and sites of feeding (for further information see Ellsbury et al., 1994). It
should be recognized that there are two systems of electronic monitoring, AC and DC, which give different waveform patterns; also there have been differences in the terminology used by European and American scientists. Evidence from electronic monitoring of aphids (Fig. 11.5A) while they feed in the phloem suggests that they salivate, but it is the watery, enzyme-bearing, non-gelling saliva. Indeed, if aphids did not salivate in the phloem, the circulative and propagative viruses that are restricted to the phloem would have no access to this tissue from an aphid carrying the virus. No gelling sheath saliva is secreted during feeding in a sieve tube, but, on withdrawal, such saliva is used to seal the lumen in the salivary sheath that had been occupied by the stylets. This feeding process causes minimal damage to the sieve tube and to surrounding cells in the stylet path. Penetration by aphids has been reviewed by Pollard (1977). 4. Role of the host plant Both physical and chemical features of the plant may markedly affect aphid feeding behavior. For example, resistance of certain
III.
APHIDS
(APHIDIDAE) 489
Fig. 11.2 Mouthparts of
Myzus persicaerevealed by scanning electron microscopy. (A) Labium with joint area and bristles. Mandibular stylets protrude from the labium. (The aphid was frozen in liquid nitrogen immediately after it had withdrawn its stylets from a leaf.) (B) Tip of labium and mandibular stylets at higher magnification. (C) Tip of mandibular stylets showing ridges in both stylets, and the overlap of the tip of one stylet. From de Zoeten (1968), with permission. brassicas to Brevicoryne brassicae (L.) (but not to M. persicae) has been shown to depend on the physical state of the wax on the leaf surface (Jadot and Roland, 1971). The density of trichomes on soybean leaves influenced probing behavior by several aphid species. Spread of SMV in the field correlated negatively with density of pubescence (Gunasinghe et al., 1988). Specific chemicals may either attract or inhibit feeding by particular aphid species. For example, sinigrin, a mustard oil glucoside found in the Brassicaceae, stimulates feeding by aphids that normally feed on brassicas, but inhibits uptake by species that do not feed on members of this family (Nault and Styer, 1972). Several examples are known where infection of a plant with a virus makes the plant more suitable for the insect vector to grow and reproduce. Aphis fabae produced more young per mother on beet plants infected with BtMV than on healthy plants (Kennedy, 1951). Because overcrowding began sooner on virus-infected plants, emigration of aphids began sooner. The cumulative increase of the aphid numbers in a set of plants would be substantial. In other experiments with individual leaves, this differ-
ence appeared on leaves of all ages on the plant. Baker (1960) found somewhat similar effects with four species of aphids on beet. Myzus persicae preferentially selected plants infected with BYV for feeding, and subsequently bred more rapidly and lived longer than on normal green plants. Individuals of two aphid species excreted fewer droplets of honeydew when feeding on plants infected with BYDV compared with healthy plants (Ajayi and Dewar, 1982). Maturation of aphids as alatae on oats was favored by BYDV infection (85%) compared with that on healthy plants (35%) (Gildow, 1983). On the other hand, aphids reared on BYDV-infected wheat had a shorter lifespan (Araya and Foster, 1987). In transmission experiments, plants are involved in three ways: (1) for breeding the aphids, (2) for providing virus-infected material, and (3) for providing healthy plants to test the ability of aphids to infect. It is common practice to rear virus-free aphids on a plant species 'immune' to the virus under study. However, when aphids are placed on the virus-infected plant, the change of species
490
11 T R A N S M I S S I O N 1: BY INVERTEBRATES, N E M A T O D E S A N D FUNGI
Cl
9.
\Mx
..
......
.
I
pDm
Eph
Fig. 11.3 Stylized sagittal view of head of an adult pea aphid (Acyrthosiphon pisum), showing the primary structure of the anterior alimentary canal. The epipharyngeal pharynx protuberance (Pp) (actually offset 3 ~m from the midline), included here, demonstrates precisely its relative location with respect to the precibarial valve. An enlarged sagittal view, at the juncture of the precibarial canal and cibarium (Cb), is shown within the circle. The dashed line that scribes an arc between the precibarial valve piston (PcV) and the precibarial valve receptacle (PcVR) denotes a closed valve. Ac, antecylpeus; C1, clypeus; CpA, cybarial pump apodemes; CpD, cibarial pump diaphragm; CpDm, cibarial valve dilator muscle; Cv, cardiac valve, DPcS, distal precibarial sensilla; DPcSd, distal precibarial dendrites; Eph, epipharynx; Hph, hypopharynx; HphS, hypopharyngeal sensillum; Mg, midgut; Mx, maxillary stylets; Mxa, maxillary stylets apodemes; Oe, esophagus; Pc, precibarium; PcVA, precibarial valve apodemes; PcVdm, precibarial valve dilator muscle; PcVS, precibarial valve suture; PPcS, proximal precibarial sensilla; PPcSd, proximal precibarial dendrites; Sd, salivary duct; SP, salivary pump; Tb; tentorial bar. From McLean and Kinsey (1984), with permission. m a y influence their feeding behavior. The species and even the variety of plant used as a source of virus or as a test plant may affect the efficiency of transmission. Source plants m a y change in their efficiency with time after inoculation if virus concentration changes. As discussed in Chapter 9 (Section II.J), the concentration of virus in a systemically infected plant may vary widely even in adjacent areas of tissue. This can affect the efficiency with which an aphid acquires virus. Species and varieties may differ in their susceptibility to a given virus more w h e n one species of aphid is used as a vector than with another. 5. Environmental conditions Environmental conditions, particularly temperature, h u m i d i t y and w i n d , m a y have
marked effects on aphid m o v e m e n t and feeding. For example, both transmission and acquisition efficiency for SDV by an aphid vector were greater at 20-22~ than at 10-11~ or 29~ (Damsteegt and Hewings, 1987). High h u m i d i t y f a v o r e d the transmission of PVY and PLRV by various aphid species (Singh et al., 1988a). Light quality influenced the translocation of two luteoviruses in Lycopersicon and affected the recovery of virus by the aphid Myzus persicae from the plants (Thomas et al., 1988). These various effects are discussed in relation to virus ecology and control of virus diseases in Chapters 12 and 16. Environmental factors may also affect transmission t h r o u g h effects on plant susceptibility and on the concentration of virus in source plants.
Stylet sheath
Phloem
t
Xylem
Fig. 11.4 Stylet sheath laid down in the tissues of a plant by an aphid feeding on the phloem sieve tubes. With kind permission from Dixon (1973).
B. The vector groups of aphids There is no agreed classification for the 3700 aphid species that have been described. Eastop (1977) uses the superfamily Aphidoidea, which contains three families. The Adelgidae consists of about 45 species living only on Coniferae. They are of no k n o w n significance for viruses. The Phyloxeridae, with about 60 species, have been little studied with respect to viruses except for Phylloxera on vines. The Aphididae contain 10 subfamilies, of which the Aphidinae contain more than one-half of the species of aphids and most of the important virus vectors. About 50% of the approximately 600 or so viruses with invertebrate vectors are transmitted by aphids. Data concerning vector groups can be regarded only as indicative, because the numbers are biased in various ways. For example, in virus-vector studies there has been a marked preference for trials with certain aphid species such as M. persicae and Aphis gossipii that are widespread and easily reared on a range of
plant species. Negative results in transmission trials are often not reported. Only a very small proportion of the possible virus-vector combinations has actually been tested. Twenty-two viruses affecting the Solanaceae are transmitted by aphid species, most of which had not encountered potatoes until about 400 years ago (Eastop, 1977). Thus, it seems certain that many more actual and potential vectors exist. For example, in a 3-year study in Holland eight new vector species for PVY-N were discovered (Piron, 1986).
C. Aphid transmission by cell injury Unstable viruses occurring in low concentration may be readily transmitted by aphid feeding, but some stable viruses such as TMV, TYMV and SBMV are not. This curious fact has not yet been fully explained. Most experimental work on this problem has been done with TMV, with the following results: (1) aphids cannot transmit TMV via their stylets (Harris and Bradley, 1973); (2) under laboratory conditions they can transmit by making small w o u n d s when they claw the surface of the leaf (Harris and Bradley, 1973); (3) they can ingest TMV from infected plants and through membranes, and release virus again in an infectious state (Pirone, 1967); (4) aphid saliva does not inhibit TMV infection; and (5) when purified TMV is mixed with poly(L)-ornithine and potassium chloride, aphids can acquire TMV through a membrane and transmit it to plants via their stylets (e.g. Pirone, 1977). Perhaps the poly(L)ornithine in some way makes the cell penetrated by the stylets susceptible to infection or facilitates the retention of the virus in the stylets.
D. Types of aphid-virus relationship The basic concepts of virus-vector interactions were introduced by Watson and Roberts (1939), who coined the terms 'persistent' and 'non-persistent' to describe the length of time for which the aphid vector was able to transmit the virus after acquisition. Since then, the terminology has been modified and refined
492
11 T R A N S M I S S I O N
l: BY INVERTEBRATES, N E M A T O D E S A N D FUNGI
Fig. 11.5 Electronic monitoring of insect feeding. (A) Three composite probes made by a pea aphid (Act.trthosiphon pisum) on broad bean (Vicia faba) leaf. Lines above the waveform explain the hierarchical interpretation (from the top of the waveforms) of the terminology used by North American researchers. Line (a) defines feeding behavior encompassing all others, line (b) defines probing as the durations of all stylet insertions into a plant, whether as short test probes or as long ingestion (or exploratory) probes. Line (c) defines the categories or waveforms within the probes according to the classical AC interpretations for aphids (salivation or S, X-waveforms, and Ingestion or I). Salivation can also be split, as on line (d), into high-voltage salivation (H) and two types of low-voltage salivation (L1 and I~2), although the biological meanings of these patterns are not known. Lines below the waveform tracing explain the terminology used by the Europeans, where line (e) defines ingestion as synonymous with feeding, and line (f) defines the differences between the probes and stylet penetrations. European waveform interpretations are not listed because they are the same when an AC monitor is used; the DC monitor produced waveforms, which are not yet convertible into AC waveform terms. NP, nonprobing or non-penetrating. From Backus (1994). (B) Representative waveforms made during the electronic monitoring of the leafhopper, Cicadulina mbila, on a host, Di~itaria san~,uinalis (upper) and a non-host, rice (lower). IS, initial salivation; PI, phloem sieve element ingestion; P, probe (period from start of feed to end)" NP, non-probing activities; ?, beginning of probe. Strip chart to be read from left to right. Each section between two vertical lines represents 20 seconds. From Mesfin et al. (1995). (C) Strip-chart recording feeding of tile thrips, Frankliniella occidentalis, on broad bean. PR, probe initiation; S, salivation and stylet movement; I, ingestion; PO, probe termination and pull out spike; B, baseline. From Hunter et al. (1994), with permission. m a n y t i m e s to take a c c o u n t of n e w f i n d i n g s , b u t often c a u s i n g c o n t r o v e r s y ( r e v i e w e d in Hull, 1994a; N a u l t , 1997; G r a y a n d Banerjee, 1999; Blanc et al., 2001). The c u r r e n t , m o s t w i d e l y a c c e p t e d , t e r m i n o l o g y is g i v e n in Table 11.2, w h i c h d i f f e r e n t i a t e s b o t h b e t w e e n a n d w i t h i n e x t e r n a l l y a n d i n t e r n a l l y b o r n e interactions. This d i f f e r e n t i a t i o n is b a s e d o n the
region(s) of the v e c t o r in w h i c h the interaction(s) occurs a n d also takes into a c c o u n t the v i r u s g e n e p r o d u c t ( s ) i n v o l v e d in the interaction. Table 11.2 s u m m a r i z e s the m a i n p r o p e r t i e s of t h e d i f f e r e n t k i n d s of r e l a t i o n s h i p s . Essentially, there are t h r e e s t a g e s in the transm i s s i o n cycle:
III. APHIDS (APHIDIDAE)
493
T A B L E 11.2 R e l a t i o n s h i p s b e t w e e n p l a n t v i r u s e s a n d t h e i r v e c t o r s
Virus transmission group Virus Site in Type of vector transmission product interacting with vector
Transmission characteristics Transovarial Aquisition Retention Transtadial Virus in Latent Virus passage vector period multiples transmission time (max time hemoin vector dose) (half-life) lymph
Externally Nonborne persistently transmitted, stylet-borne
Capsid Helper factor
Seconds to minutes
Minutes
No
No
No
No
No
Nonpersistently transmitted, foregutborne (semipersistent)
Capsid Helper factor
Minutes to hours
Hours
No
No
No
No
No
Persistent, circulative
Hours to days
Days to weeks
Yes
Yes
Hours to days
No
No
Persistent, propagative
Hours to days
Weeks to months
Yes
Yes
Weeks
Yes
Often
Internally borne
1. The acquisition phase in which the vector feeds on the infected plant and acquires sufficient virus for it to be able to transmit it. 2. The latent period in which the vector has acquired sufficient virus but is not able to transmit it. For externally borne viruses there is little or no latent period. 3. The retention (transmission) period is the length of time during which the vector can transmit the virus to a healthy host. Some
further
definitions
are
needed.
Inoculativity is the ability of an aphid or other insect to deliver infectious virus into a healthy plant. The acquisition feed is the feeding process by which the insect acquires virus from an infected plant. The inoculative (transmission) feed is the feed during which virus is delivered into a healthy plant. These relationships have been studied in most detail with aphids but are applicable to most other viruses that are transmitted by arthropods with piercing-sucking mouthparts. The basic features of the interactions between viruses and their arthropod vectors are s h o w n diagrammatically in Fig. 11.6.
E. Non-persistent transmission Because of its role in the field transmission of m a n y economically i m p o r t a n t viruses, nonpersistent transmission by aphids has been s t u d i e d in m a n y laboratories over seven decades, and molecular details of the interactions are becoming more clearly understood. 1. The non-persistent viruses Of the approximately 290 or so k n o w n aphidborne viruses, m o s t are n o n - p e r s i s t e n t (see A p p e n d i x 2). The following virus genera have definite m e m b e r s transmitted in a non-persistent manner: Alfamovirus, Caulimovirus (by M.
persicae), Cucumovirus, Fabavirus, Macluravirus and Potyvirus. These genera include viruses with helical and isometric particles, and with D N A a n d R N A mono-, bi- and tri-partite genomes. 2. Acquisition time Non-persistently transmitted viruses are acquired rapidly from plants, usually in a matter of seconds. During this time, aphids stylets do not usually penetrate b e y o n d the epidermal cells and w h e n they penetrate b e y o n d the epidermis
494
11 TRANSMISSION 1: BY INVERTEBRATES, NEMATODESAND FUNGI
ES
" --~_
,,;-
ASG
,
.
OiO
.
.
(~
.
-
......... ,, __~.~Vv-.~~". _qy
t
I>AE) virus in the insect and transovarial transmission, and the factors that affect its efficiency, may be of considerable economic importance. i Age of vector when infected For several leafhopper vectors under experimental conditions, n y m p h s are more efficient vectors than adults, and adults decrease in efficiency as they age. For example, Sinha (1967) followed the appearance of WTV antigen in n y m p h s and adults of A. constricta after a 1-day acquisition feed. By 32 days after infection, 50% of the individuals infected as n y m p h s had antigen in their h e m o l y m p h and salivary glands. For individuals infected as adults, only 5% showed antigen at these sites. With the remaining insects, antigen was confined to the primary site of infection in the filter chamber of the intestine. Antigen had spread over a more limited region of intestine than with the nymphs. These experiments suggest that, as the insect ages, the intestine becomes more refractory to infection, and that virus present in the filter chamber of many adults may not be able to pass into the body cavity. ii Time after infection Slykhuis (1963) showed that the leafhopper Endria inimica lost its ability to transmit WASMV after a variable number of days. Some transmitted intermittently and none was infective after 72 days. iii Temperature Sinha (1967) examined the effect of increased temperature on the spread of WTV in A. constricta. Groups of n y m p h s were given an acquisition feed of 1 day at 27~ and then held at 27~ for three further days to allow virus infection to become established in the filter chamber. Virus antigen was found in the filter chamber at this stage. One group continued at 27~ and the other was held at 36~ High temperature prevented the spread of virus from the intestine to the hemolymph, salivary glands and other parts of the intestine. By day 6, antigen had disappeared from the filter chamber of 60% of the insects. Temperature had a marked effect on the rate of transovarial passage of RSV in its planthopper vector Laodelphax striatellus. At 17.5~ 83%
5 13
of viruliferous females passed virus to their progeny at a rate greater than 90%. At 32.5~ only 12.5 To of females reached this frequency of transovarial transmission (Raga et at., 1988). iv Genetic variation in the leafhopper Different lines or races within a vector species may vary widely in their efficiency as vectors. Thus, to demonstrate with a reasonable degree of confidence that a species is not a vector, it may be necessary to test populations from various regions where the virus and insect occur. v Change in properties of the virus Long-term culture of a virus in plants without recourse to leafhopper transmission may lead to loss of ability to be transmitted, as noted earlier for WTV and rhabdoviruses.
V. WHITEFLIES
(ALEYRODIDAE)
Three genera of viruses are transmitted by whiteflies, the begomoviruses of the Geminiviridae family and the criniviruses and some closteroviruses of the Closteroviridae. Most of these viruses are found in the tropics and subtropics where they can be of substantial importance.
A. Whiteflies Begomoviruses are transmitted exclusively by Bemisia tabaci whereas the whitefly-transmitted closteroviruses and the criniviruses are transmitted by the glasshouse whitefly, Trialeuroides vaporariorum (BPYV, TICV and CCSV), T. abutilonea (DVCV and AYV), B. tabaci (CYSDV and SPSVV) and B. argentifolia (LCV). The most studied vector is B. tabaci, which is important in the epidemiology of several major diseases. Only the first instar of the larva is mobile, and it does not move far. Adults are winged, and many generations may be produced in a year. The n y m p h s of B. tabaci are phloem feeders. The B type of B. tabaci, sometimes named B. argentifolia, is of increasing importance as a vector. The general m o r p h o l o g y of whiteflies is reviewed by Gill (1990) and details of the ultrastructure of the mouthparts of B. tabaci
514
11 T R A N S M I S S I O N 1: BY INVERTEBRATES, NEMATODES A N D FUNGI
have been studied by Rosell et al. (1995). Adult whitefly mouthparts are similar to those of other homopterans, especially aphids, and comprise the labrum, the lamium and the stylets. The stylet bundle, made up of two mandibular and two maxillary stylets, is the feeding organ. The feeding b e h a v i o r of whiteflies resembles that of aphids in being piercing and sucking and involving a salivary sheath (Janssen et al., 1989).
B. Begomoviruses Most, if not all, bipartite begomoviruses are transmitted in the persistent circulative manner. Virus particles have been observed in the gut epithelial cells and associated with salivary glands of whitefly vectors and are thought to follow a similar circulative route to those of luteoviruses (Cohen and Antignus, 1994; Hunter et al., 1998). As with circulative aphidtransmitted viruses (Section III.H.l.a), the GroEL homolog, symbionin, is implicated in the circulative transmission in whiteflies of TYLCV (Morin et al., 1999). However, the nontransmissibility of AbMV is not the result of lack of binding of the coat protein to symbionin (Morin et al., 2000). However, certain complexities have been described for some begomoviruses (reviewed in Gray and Banerjee, 1999). TYLCV appears to persist in its whitefly vector for longer than expected from its infectivity, and is reported to be transovarially transmitted. SCLV particles have been associated with cytopathological abnormalities in some vector tissues and this virus can have detrimental effects on the vector biology and reproduction. These observations have been taken to suggest that the viruses are propagative but no replication intermediates have yet been detected. Sexual transmission of TYLCV-Is from male to female and from female to male insects has been suggested (Ghanim and Czosnek, 2000). As with the persistent circulative transmission of luteoviruses (Section III.H.l.a), begomoviruses have to cross at least two barriers, the gut wall and membranes, to enter and exit the salivary glands; it is likely that these involve receptor-mediated strategies. 'Squash-
blot' experiments show that ACMV can be acquired by two non-vectors, the whitefly, T. vaporariorum, and the aphid, M. persicae (Liu et al., 1997b), which suggests that the virus can cross the gut wall but not enter the salivary glands. The coat protein is shown to be essential for the acquisition process because u n e n c a p s i d a t e d cloned nucleic acid is not acquired (Azzam et al., 1994). By making chimeras between the genomes of a whiteflytransmissible isolate of ACMV-NOg and a nontransmissible isolate ACMV-K, Liu et al. (1997b, 1998) demonstrated that the defects responsible for lack of transmissibility were in the coat protein and DNA-BC1 gene of ACMV-K. The BC1 gene is thought to mediate cell-to-cell spread of the virus (see Chapter 9, Section II.D.2.i) and thus may influence the virus distribution in plant tissues. Comparison of coat protein sequences of transmissible and non-transmissible isolates of TYLCV and then mutagenesis showed that the region of the coat protein between amino acids 129 and 134 was essential for the correct assembly of virion and for whitefly transmission (Norris et al., 1998). In contrast with these observations with ACMV, SLCV does not cross the gut wall barrier of non-vector B. tabaci (Rosell et al., 1999).
C. Closteroviruses and criniviruses Both the monopartite whitefly-transmitted closteroviruses and the bipartite criniviruses are transmitted in a foregut-borne, semipersistent manner. LIYV is retained in the vector for a maximum of 3 days whereas LCV and CYSDV persist for 4 and 9 days respectively. As with the aphid-transmitted closteroviruses, BYV and CTV, (see Section III.F.1), LIYV encodes two capsid proteins: the major protein (CP) and a minor protein (CPm) (see Chapter 6, Section VIII.F.2). Also as with the aphidtransmitted closteroviruses, CPm is found at one end of the particles (see Fig. 5.12) (Tian et al., 1999). Purified LIYV virions could be transmitted by B. tabaci after in vitro acquisition, and transmission was neutralized by antiserum to CPm but not by antiserum to CP. Thus, CPm is involved in the transmission of LIYV (Tian et al., 1999).
VI.
VI. THRIPS (THYSANOPTERA) (reviewed by Ullman et al., 1997) Of the 5000 or so species of thrips, only 10 species, all in the family Thripidae, are vectors of plant viruses (Table 11.5). Most of these vector species are extremely polyphagous and able to reproduce on a broad range of host plants. Thrips tabaci is cosmopolitan, feeding on at least 140 species from over 40 families of plants. It reproduces mainly parthenogenetically. The larvae are rather inactive but the adults are winged and very active. Thrips tabaci feeds by sucking the contents of the subepidermal cells of the host plant. Adults live up to about 20 days. Several generations can develop in a year. Viruses from four plant virus families or groups are transmitted by thrips (Table 11.5). The ilarviruses, sobemoviruses and carmoviruses are pollen transmitted, the thrips carrying the pollen and inoculating it by
THP, IPS
( THYSAN()PTERA )
mechanical damage during feeding. Tospoviruses are transmitted in a persistent propagative manner. A . T h r i p a n a t o m y (reviewed by Nagata, 1999) Although detailed studies have been made on the anatomy of Hercinothrips femoralis, most of the information on the internal anatomy is applicable to other members of the Thripidae. Thrip feeding a p p a r a t u s consists of one mandible that punches a hole in the leaf and two maxillae that are inserted into the plant cell and through which the cell contents are sucked (Fig. 11.16). The m o u t h p a r t s lead into the foregut and then the m i d g u t of which there are three regions, anterior (Mgl), middle (Mg2) and posterior (Mg3). The foregut and hindgut are of ectodermal origin lined by a thick impermeable cuticle and the m i d g u t is endodermal with a soft inner epithelial cell layer. The
TABLE 11.5 Transmission of viruses by thrips Thrip species
Virus
Virus family (group)
Virus-vector relationship
Frankiniella occidentalis
GRSV" INSV PDV PFBV PNRSV TCSV TSWV TSWV GRSV TCSV TSWV GRSV TCSV TSWV PNRSV TSV PNRSV PNRSV GBNV WSMV TSWV PNRSV SoMV TSV TSWV
Tospovirus Tospovirus Ilarvirus Carmovirus Ilarvirus Tospovz ru s Tospovz rus Tospovz rus Tospovzrus Tospovz rus Tospovz rus Tospovz rus Tospovz rus Tospovz rus Ilarvirus Ilarvirus Ilarvirus Ilarvirus Tospovirus Tospovirus Tospovirus Ilarvirus Sobemovirus Ilarvirus Tospovirus
PP PP Pollen Pollen Pollen PP PP PP PP PP PP PP PP PP Pollen Pollen Pollen Pollen PP PP PP Pollen Pollen Pollen PP
F. flgsca F. intensa F. schulzei
Microcephalothrips abdominalis Thrips australis T. imaginis T. pahni T. setosus T. tabaci
"See Appendix 1 for virus acronym. PP, persistent propagative. Adapted from Ullman et al. (1997), with permission.
5 15
516
I1 T R A N S M I S S I O N 1" BY INVERTEBRATES, NEMATODES AND FUNGI
Salivary gland
Tubular salivary gland
Ligament Foregut
Midgut 3
Midgut 1 Muscle
Midgut 2 Hindgut Malpighian tube
Fig. 11.16 Composite drawing of the alimentary tract and associated organs of the thrips, Hercinothrips femoralis. From Nagata (1999), with permission.
salivary gland complex comprises two lobular and two tubular glands, which may correspond to the accessory salivary glands of other insects. It has been suggested that the tubular glands connect to Mgl (Ullman et al., 1989) but no direct connection could be found (Del Bene et aI., 1991). Mgl is also connected to the salivary glands, this time the lobular glands by thin thread-like structures, termed ligaments (Ullman et al., 1989). B, T o s p o v i r u s t r a n s m i s s i o n (reviewed by Nagata, 1999) Transmission of tospoviruses by thrips has several unusual features (Fig. 11.17). Only the larvae, and not adults, can acquire the virus, and the competence to acquire decreases with age of the larvae. Effectively, the virus has to be acquired by a first instar nymph for realistic
transmission (van de Wetering et al., 1996). The virus could be acquired or transmitted by first instar nymphs of Franklinielta occidentalis in feeding periods of as short as 5 minutes but the median acquisition access period on infected Impatiens plants was 106 minutes and the median inoculation access period to petunia leaf discs was 58 minutes. The median latent period varied with temperature, being 84 hours at 27~ and 171 hours at 20~ Individuals may retain infectivity for life, but their ability to transmit may be erratic. The virus is not passed through the egg. Although the titer of virus was higher in females, males were more efficient transmitters, probably because of their feeding behavior (Sakurai et al., 1998; van der Wetering et al., 1999). There are distinct levels of species and biotype specificity (Wijkamp et al., 1995). Franktiniella occidentalis was the most efficient vector for four tospoviruses, TSWV, INSV, TCSV and GRSV. The dark form of F. schultzei transmitted TSWV, TCSV and GRSV, whereas the light form of this species transmitted TSWV and TCSV poorly. Only one of four populations of T. tabaci from different geographical regions transmitted only TSWV of the four viruses tested, and that with low efficiency. If TSWV is cultured by successive transfers only in plants, the isolate loses the ability to be transmitted by thrips. C. V i r u s - v e c t o r r e l a t i o n s h i p The accumulation of the nucleocapsid and a non-structural protein of TSWV was studied in developing nymphs and adults of F. occidentalis (Wijkamp et al., 1993) and both proteins were shown to increase in the vector. The proteins and virus particles accumulated in the salivary glands and other tissues. These observations taken with the times for virus acquisition and the latent period indicated that TSWV multiplied in its vector. As with other internally-borne persistently transmitted viruses, tospoviruses have to pass several barriers in the vector, which suggests that there is a receptor-mediated mechanism(s). TSWV is enveloped with spikes, made up from two virus-coded glycoproteins, extending from
VI.
the envelope (see Chapter 5, Section VIII.B). Feeding F. occidentalis on plants infected with wild type and an envelope-deficient isolate showed that the thrips became infected only when they acquired intact virus particles (Nagata, 1999). Similarly the envelope-deficient isolate did not infect primary E occidentalis cell cultures, nor did the nucleocapsids of wild type which had had the envelope removed. These observations suggested that the viral glycoprotein(s) contain the binding site for a receptor in the vector's midgut. Two proteins from F. occidentalis have been shown to bind to TSWV glycoproteins. Gel overlay assays and immunolabeling identified a 50-kDa thrips protein and anti-idiotype antibodies against each of the two TSWV glycoproteins labeled a 50-kDa thrip protein which was localized to the larval thrip midgut (Bandla et al., 1998). A 94-kDa protein that bound to TSWV was identified in E occidentalis and T. tabaci but not in the non-vector aphid M. persicae. This protein bound to the TSWV G2 glycoprotein and was present throughout the thrips body (Kikkert et al., 1998). The binding properties of these two thrip proteins would suggest that they may be
TttRIPS
(THYSANOI'TERA)
5 17
associated with receptor sites but the detailed sites have not yet been identified.
D. Route through the thrips A detailed study of the route that TSWV takes through F. occidentalis has been made by Nagata (1999), who used immunofluorescent staining of nymphs at various times after virus acquisition. The first infections were found in the Mgl region (see Fig. 11.16) about 24 hours postacquisition (hpa). These infections increased in intensity but remained restricted to the Mgl epithelium for some time. In late larval stage, it spread to the circular and longitudinal midgut muscle tissues. By the adult stage the visceral muscle tissues of the midgut and foregut were infected. Infection of the salivary glands was first observed 72 hpa and at the same time the ligaments connecting the midgut with the salivary glands became infected. There was no evidence for TSWV in either the h e m o c o e l or the midgut basal lamina. It appeared that the virus reached the salivary glands through the ligaments connecting Mgl to the salivary glands. This is a different route to
\'-,,
9 Symptom expression, larval eclosion & feeding
' h ~ 1st instar ~.....___....~~ larva ,ir ~ ~
'
Only 1st instar larvae can acquirevirus
~
ar
Dispersal, inoculation & oviposition by adults Prepupa
Adult
are quiescent & nonfeeding
Fig. 11.17 Line drawing depicting the relationship between thrips and tospoviruses. Transmission by adults occurs only if larvae acquire the virus. *Second instar larvae occasionally transmit tospoviruses, but are wingless and seldom disperse widely. The adult is the primary dispersal stage of the insect and thought to be of greatest epidemiological significance. From Ullman et al. (1997), with permission.
518
11 TRANSMISSION 1: BY INVERTEBRATES, NEMATODES AND FUNGI
that conventionally proposed for persistent viruses, which is m o v e m e n t t h r o u g h the hemocoel from the gut cells to the salivary glands. Thrips that ingested TSWV at the second instar stage had less infection of the midgut and rarely infection of the salivary glands.
VII. OTHER SUCKING A N D PIERCING VECTOR GROUPS
A. Mealybugs (Coccoidea and
Pseudococcoidea ) Mealybugs are much less mobile on the plant than other groups of vectors such as aphids and leafhoppers, a feature that makes them relatively inefficient as virus vectors. They spread from one plant to another in contact with it, and the crawling n y m p h s move more readily than adults. Ants that tend the mealybugs may move them from one plant to another (Sether et al., 1998) and occasional long-distance dispersal by wind may occur. Mealybugs feed on the phloem. They have been established as the vectors of many of the badnaviruses and several closteroviruses (GLRaV-3, LCV and PMWaV) and trichoviruses (GVA and GVB). Six vector species of differing abundance for CSSV were recorded by Bigger (1981) but there are transmitting and non-transmitting races of certain species (Posnette, 1950). CSSV could be acquired by Pseudococcus njaletlsis within 20 minutes but the mealybug took at least 16 minutes to penetrate the leaf tissue. The virus persisted in the mealybug for less that 3 hours (Posnette and Robertson, 1950). The relationship between the CSSV and mealybugs has some similarities to the non-persistent aphidtransmitted viruses. Presumably, the virus is carried on or near the stylets of the mealybug. The closterovirus PMWaV is transmitted by the pink pineapple mealybug Dysmicoccus brevipes and the grey pineapple mealybug D. neobrevipes, the second and third instars being more effective at acquiring the virus than first instars and gravid females (Sether et al., 1998).
GLRaV-3 is retained in its vector, Planococcus citri, for approximately 24 hours (Cabaleiro and Segura, 1997), which suggests a semi-persistent relationship.
B. Bugs (Miridae and Piesmatidae) The mirid bugs feed by means of stylets but their biology and taxonomy are not well understood. Cyrtopettis nicotianae has been shown to be a vector of VTMoV, SBMV and several other, but not all, sobemoviruses (Gibb and Randles, 1988). Minimum acquisition time was I minute, a characteristic of non-persistent viruses. However, rate of transmission increased with increasing acquisition feeding time, a property characteristic of semi-persistent or circulative transmission. Other characteristics were like those of semi-persistent or circulative transmission. There was no evidence for virus replication in the vector. The viral antigen of VTMoV was detected in the gut, h e m o l y m p h and feces of its vector, Cyrtopeltis nicotianae, but not in the salivary glands, and infective virus was found in feces 6 days after acquisition (Gibb and Randles, 1990). Non-infective myrids were able to inoculate plants from infectious sap deposits on the upper epidermis and an ingestiondefecation model, not involving salivary glands, is suggested for this virus-vector association. Thus, the transmission was like that of beetles in several respects. Mirids are important crop pests, so it is possible that other virus vectors will be found. BLCV is transmitted in a persistent, propagative manner by the piesmatid bug Piesma quadratuJJ~. There is no evidence for transmission through the egg (Proesler, 1980).
VIII. INSECTS WITH BITING MOUTHPARTS A. Vector groups and feeding habits A few vectors have been reported from the orders Orthoptera and Dermaptera. Important vectors are found in the Coleoptera (beetles). Almost all the vectors belong in a few families. Interest centers on the family Chrysomelidae,
VII1. INSE(;TS WITH BITIN(; MOUTHPARTS
which consists of 55 000 species of plant-eating beetles. About 30 of these are known to transmit plant viruses, and each species feeds on a limited range of host plants. Twenty vector species are found in the subfamilies Galerucinae and Halticinae (flea-beetles), two are in the Crysomelinae and two in Criocerinae. As pointed out by Selman (1973), this distribution almost certainly reflects the interests of investigators rather than the actual situation in nature. Many beetle vectors probably remain to be discovered. In the Curculionidae and Apionidae a few species of weevils have been shown to be virus vectors. Leaf-feeding beetles do not have salivary glands. The chrysomelid beetles tend to eat the parenchyma tissues between vascular bundles, thus leaving holes in the leaf but, with heavy infestation, damage may be more severe. They regurgitate during feeding, which bathes the mouthparts with plant sap, as well as with viruses, if the plant fed upon is infected. It was once thought that transmission by beetles involved simply a mechanical process of wounding in the presence of virus. This is not so because: (1) some very stable saptransmissible viruses such as TMV are not easily transmitted by beetles (but see Orlob, 1963); (2) some transmitted viruses may be retained by beetle vectors for long periods; and (3) there is a substantial degree of specificity between viruses and vector beetles. Beetle transmission has been reviewed by Fulton et al. (1987). B. Viruses t r a n s m i t t e d by beetles The viruses transmitted by beetles belong to the Tymovirus, Comovirus, Bromovirus and Sobemovirus groups. Most viruses in these groups are not transmitted by members of other arthropod groups (but see transmission of sobemoviruses by myrids; Section VII.B above) and are usually quite stable, reaching high concentrations in infected tissues. They have small isometric particles ( 2 5 - 3 0 n m diameter) and are readily transmitted by mechanical inoculation. The viruses tend to have relatively narrow host ranges, as do their beetle vectors.
519
C. B e e t l e - v i r u s r e l a t i o n s h i p s Beetles can acquire virus very quickly--even after a single bite--but efficiency of transmission increases with longer feeding, as does retention time (Fulton et al., 1987). Some viruses appear quickly in the hemolymph after certain beetle species have fed on an infected plant; others do not (Wang et al., 1992, 1994). Insects become viruliferous after injection of virus into the hemocoel. Retention time varies between about 1 and 10 days with different beetles. However, under dormant, overwintering conditions, beetles may stay viruliferous for periods of months. Beetles can transmit the virus with their first bite on a susceptible plant. There is no good evidence for the existence of a latent period following virus acquisition, and no evidence for virus replication in beetle vectors. These observations suggest that beetletransmitted viruses are externally borne. It has been established that the regurgitant fluid is a key factor in determining whether a virus will or will not have beetle vectors (Gergerich et al., 1983). This discovery was made possible by a gross wounding technique, which involved cutting discs from a leaf with a glass cylinder contaminated with virusregurgitant mixture, thus mimicking the kind of wounds made by feeding beetles. When virus was mixed with regurgitant, only viruses normally transmitted by beetles were transmitted by the gross wounding technique. In ordinary mechanical inoculation using abrasives, all mixtures were non-infectious (Gergerich et al., 1983; Monis et al., 1986). Regurgitant does not irreversibly inactivate viruses not transmitted by beetles, because infectious virus could be recovered by dilution of the regurgitant-virus mixture or by isolation of the virus. Regurgitant from several leaf-feeding beetle species was found to contain an RNase activity equivalent to 0.1-1.0 m g / m l of pancreatic RNase. This enzyme, used in this concentration range, inactivated beetle-non-transmitted viruses such as TMV when inoculated by the gross wounding technique (Gergerich et al., 1986), but the transmission of viruses that are normally transmitted by beetles was not affected. Thus, pancreatic RNase could mimic
520
11 T R A N S M I S S I O N 1: BY INVERTEBRATES, N E M A T O D E S ANt) FUNGI
the effect of beetle regurgitant. In further work, three kinds of RNase differing in the way they cleave RNA were found to act with the same discrimination as pancreatic RNase. Other basic proteins did not inhibit transmission of viruses not transmissible by beetles. Thus, it appears to be the enzymatic activity of the proteins that affects transmissibility (Gergerich and Scott, 1988a). Why this should be so is not clear. Perhaps the RNase activity affects establishment of the beetle-non-transmitted viruses in the initially inoculated cells as suggested by Gergerich and Scott. Alternatively, RNases may bind more firmly to viruses such as TMV, preventing uncoating in the cell. Gergerich and Scott (1988b) showed that several beetle-transmitted viruses could m o v e through cut stems. Furthermore, such viruses, inoculated below a steam-killed section of stem in an intact plant, could move and infect the upper parts of the plant, whereas beetle-nontransmitted viruses could not. These results suggest that the ability to be translocated in the xylem and to infect n o n - w o u n d e d tissue is a feature of beetle-transmitted viruses. However, TYMV, which is beetle transmitted, can move into a leaf from a cut petiole but cannot infect the leaf (Matthews, 1970). When sodium azide was included in the inoculum mixture, cells in a zone around the gross w o u n d i n g site were rapidly killed but infection by SBMV still occurred. Transmission of a beetle-non-transmitted virus was severely affected (Fulton et al., 1987). It was suggested that the ability of beetle-transmitted viruses to move in the xylem might take them to ceils unaffected by the sodium azide treatment. The apparently simple transmission of viruses by beetles is now seen to be quite a complex process. For example, the experiments summarized above shed little light on the problem of specificity among beetle speciesmwhy some species are highly efficient vectors of a particular virus and others are not.
IX. MITES (ARACHNIDA) Members of the mite families Eriophyidae and Tetranychidae feed by piercing plant cells and
sucking the contents, but they differ in many other respects. Several members of the first of these families are vectors for rymoviruses. One unassigned virus (Peach mosaic virus) is transmitted by the Tetranychidae.
A. Eriophyidae The eriophyid mites are not closely related to other groups of mites. Members are k n o w n to transmit at least six plant viruses, and they feed by puncturing plant cells with stylets and sucking in the cell contents. The stylets are held inside the groove of the rostrum, which has two pads that act as ducts for the saliva (Fig. 11.18). Eriophyid mites are very small arthropods (about 0.2 m m in length) and have limited powers of independent movement. They frequently infest buds and young leaves, where they often cause little damage and may quite easily be overlooked. They are readily killed by desiccation. In spite of this, their main method of spread from plant to plant is by wind (Slykhuis, 1955). Most species are quite specific for the host plant on which they feed, usually being confined to one plant genus or, at most, the members of a single family. These mites cannot survive for long periods away from a host plant and, thus, most of the plant species on which they feed are perennials. They have a relatively simple developmental history that may be completed in 6-14 days. There are two n y m p h a l instars followed by a resting 'pseudopupa'. Males are not often seen. Some species have two kinds of female, one being specialized for hibernation. One of the best-studied mite vectors is Aceria *ulipae (Keifer), which can transmit two viruses simultaneously: WSMV and wheat spot mosaic virus (Slykhuis, 1962). Figure 11.19 illustrates the main internal structures of an adult of this species. The relationship of this vector with WSMV has been studied in considerable detail. Like non-persistent aphid-transmitted viruses, it has a rod-shaped particle and is readily sap transmitted. Aceria tulipae can acquire the virus in a 15-minute period on an infected leaf. A similar m i n i m u m period is required to transmit the
IX. MITES (ARACHNIDA)
521
Fig. 11.18 Head of an adult mite Aceria tulipae. S, shield; CO, coxa; R, rostrum; CH, stylets. (A) Electron micrograph of stylets. A hair lies along the upper side of stylets. Stylets are about 5 ~tm long, and only about one-third of this length penetrates during feeding. (B) Electron micrograph of feather claw. From Orlob (1966), with kind permission of the copyright holder, 9 Blackwell Science Ltd..
i i,
\,,\ .s
~ ~SG
.... AS
Fig. 11.19 Diagram of female adult Aceria tulipae, fc, Feather claws; r, rostrum; sd, salivary duct; c, chelicera; SG, salivary gland; NS, neurosynganglion; Fg, foregut; GF, genital flap; ME, mature egg; yp, yolk platelets; MT, microtubercles; Mg, midgut; Mv, microvilli; Dev. Oocytes, developing oocytes; NC, nurse cells; o, ovariole; Ov, oviduct region; Hg, hindgut; RS, rectal sac; T, rectal tube; AS, anal sucker. From Whitmoyer et aI. (1972), with permission. virus. The mite r e t a i n s infectivity t h r o u g h m o u l t s a n d m a y r e m a i n infective, at least in the greenhouse, for 6-9 d a y s after r e m o v a l f r o m an infected plant. O n a ' v i r u s - i m m u n e ' host held at 3~ they r e m a i n e d infective for over 2
m o n t h s . The m i t e s b e c o m e i n f e c t i v e as n y m p h s b u t not as adults w h e n fed on an infected plant. A b o u t 30% of mites f r o m a dise a s e d w h e a t p l a n t tested singly will t r a n s m i t the virus (Slykhuis, 1965; Orlob, 1966). The
522
11 T R A N S M I S S I O N 1: BY INVERTEBRATES, NEMATODES AND FUNGI
uncharacterized wheat spot mosaic virus has a similar relationship with its A. tulipae vector (Nault and Styer, 1970). Virus-vector relationships are difficult to study because of the small size of the mite. There is no good evidence for replication of viruses in mite vectors. Paliwal (1980) found particles of WSMV in the midgut, body cavity and salivary glands of the mite Eriophyes tulipae, suggesting that the virus is circulative in this vector.
B. Tetranychidae The Tetranychidae consists of medium-sized mites (approximately 0.8 mm) that are all plant feeders, usually with a wide host range. The spider mite Tetranychus urticae (Koch) was claimed to be a vector of PVY, but this was not confirmed by Orlob (1968). Orlob showed that T. urticae was unable to transmit nine viruses, but could acquire several viruses as revealed by presence of virus within the insect. Several nontransmitted viruses reached high concentrations in the mite (Orlob and Takahashi, 1971). BaYSMV is transmitted by the brown wheat mite (Petrobia latens) (Robertson and Carroll, 1988). Pre-adult mites readily acquired BaYSMV and both they and adult mites efficiently transmitted the virus to barley plants (Smidansky and Carroll, 1996). There is indirect evidence for transovarial passage of BaYSMV (Robertson and Carroll, 1988; Smidansky and Carroll, 1996). However, the virus was lost on keeping an originally viruliferous mite colony on a cucurbit (a non-host for BaYSMV), which is not in accord with a propagative relationship between the virus and vector.
X. POLLINATING INSECTS RBDV (Murant et al., 1974) and other viruses transmitted through infected pollen, and having insect-pollinated host plants, probably have the infecting pollen distributed by pollinating insects. Field experiments showed that BLMV is transmitted via pollen carried by foraging honeybees. About half the honeybees trapped in a field containing infected blueberry bushes
had the virus in pollen from their pollen baskets, as determined by ELISA (Childress and Ramsdell, 1987). Caging experiments demonstrated that bees and infected pollen were both essential for new infections to occur. As noted in Section VI, plants can be infected with some carmoviruses, ilarviruses and sobemoviruses through pollen carried by thrips.
XI. NEMATODES (NEMATODA) (reviewed by Brown et al., 1995, 1996b) Since Hewitt et al. (1958) demonstrated that the fanleaf virus of grapes is transmitted by a dagger nematode, several widespread and important viruses have been shown to be transmitted through the soil by nematodes. Vector nematodes belong to the order Dorylaimida, family Longidoridae, in which three closely-related genera, Xiphinema, Longidorus and Paralongidorus, transmit viruses, and to the order Triplonchida, family Trichodoridae, in which the genera Paratrichodorus and Trichodorus transmit viruses. The three genera are all ectoparasitic and feed on epidermal cells of the root (Fig. 11.20) with feeding punctures occurring frequently near the root cap. Of the approximately 375 species in the Longidoridae only eight Longidorus, one Paralongidorus and seven Xiphenema species are virus vectors. Similarly only seven Paratrichodorus and four Trichodorus species are vectors out of the 80 or so species in the
TrichodorMae. Two genera of plant viruses are transmitted by nematodes. Nepoviruses are transmitted by species in the genera Xiphinema and Longidorus, and tobraviruses are transmitted by species of Trichodorus and Paratrichodorus. All three tobraviruses are nematode transmitted but only about one-third of the nepoviruses are transmitted by these vectors. With the exception of TRSV, none of the viruses in these two genera is known to have invertebrate vectors other than nematodes; some nepoviruses are pollen transmitted (see Chapter 12, Section III.A.4).
XI. NEMAT~.~I)ES (NEMA'rC:~>A)
Fig. 11.20 Nematodes of the species Trichodorus christei feeding on a root of blueberry (Vaccinium corymbosum). (Courtesy of B.M. Zuckermann.) A. C r i t e r i a for d e m o n s t r a t i n g n e m a t o d e transmission Nematodes are difficult vectors to deal with experimentally because of their small size and their s o m e w h a t critical requirements with respect to soil moisture content, type of soil and, to a lesser extent, temperature. To overcome these problems five criteria have been proposed for establishing the nematode vectoring of viruses (Trudgill et al., 1983; Brown et al., 1989a). 9 Infection of a bait plant must be demonstrated. 9 Experiments should be done with handpicked nematodes. 9 Appropriate controls should be included to show unequivocally that the nematode is the vector. 9 The nematode should be fully identified. 9 The virus should be fully characterized.
523
A common method for detecting nematode transmission has been to set out suitable 'bait' plants (such as cucumber) in a sample of the test soil. These plants are grown for a time to allow any viruliferous nematodes to feed on the roots and transmit the virus, and for any transmitted virus to replicate. Extracts from the roots and leaves of the bait plants are then inoculated mechanically to a suitable range of indicator species. Various modifications of the procedure can be used. For example, Valdez (1972) described a small-scale procedure for testing individual handpicked nematodes, while van Hoof (1976) showed that detached tobacco leaves buried in soil could become infected with TRV by nematodes within a few hours. The proportion of nematodes ingesting virus can be determined by crushing whole nematodes and examining the extract by immunosorbent electron microscopy (Roberts and Brown, 1980) or by inoculating the suspension to suitable test plants Yassin, 1968). TRV can be detected in trichodorids by an RT-PCR method (van der Wilk et al., 1994). However, it m u s t be r e m e m b e r e d that, a l t h o u g h these methods show that the nematode ingested the virus, they do not show that it transmits it. Estimation of the extent of transmission of a particular virus in the field by examination of the nematodes present may be complicated by various factors. For example, the distribution of individuals carrying the virus may be different from that of the p o p u l a t i o n as a whole. Subpopulations within a given area, for example those in the surface layer and those in subsoil, may be of differing importance (Gugerli, 1977). Populations from different geographical areas may also differ in efficiency of transmission (Brown, 1986). B. N e m a t o d e f e e d i n g Longidorids are large nematodes ( 2 - 1 2 m m long) with long hollow feeding stylets (60-250 ~tm) which enable them to penetrate deep into root tips. The stylet comprises the anterior odontostyle, which penetrates the root cells, and the posterior odontophore in which there is nerve tissue adjacent to the food canal. The e s o p h a g u s connects the stylet to a
524
11 T R A N S M I S S I O N 1: BY INVERTEBRATES, NEMATODES ANI) FUNGI
muscular p u m p which withdraws the plant cell contents and forces them through a one-way valve into the gut. Most longidorids induce the formation of galls when feeding at root tips, this being the first of two feeding phases. The second phase is exploitive involving repeated bouts of salivation causing liquefaction of the cytoplasm followed by ingestion. Longidorus spp. have long periods of ingestion during which a volume approximately equivalent to 40 normal root-tip cells is removed each hour. Trichodorids are much smaller, with adults about 1 m m long. They typically feed on root hairs and epidermal and subepidermal cells of the root elongation zone. Their feeding has five phases: exploration, perforation of the cell wall and penetration to a depth of 2-3 ~tm, salivation, ingestion and withdrawal from the cell. C. V i r u s - n e m a t o d e r e l a t i o n s h i p s Brown and Weischer (1998) divided the nematode transmission of a virus into seven discrete
but inter-related processes: ingestion, acquisition, adsorption, retention, release, transfer and establishment (reviewed by Visser, 2000). Ingestion is the intake of virus particles from the infected plant and, although it does not require a specific interaction between nematode and virus, it needs a specific interaction between the nematode and plant. In the acquisition phase the ingested viral particles are retained in an intact state and specific features on the surface of the particle are recognized by receptor sites in the nematode feeding apparatus leading to adsorption (Fig. 11.21). Once adsorbed, infectious particles can be retained in the nematode for months or even years, but not after moulting. Release of the viral particles is t h o u g h t to occur by a change in pH caused by saliva flow when the nematode commences feeding on a new plant. In the transfer and establishment phases, the viral particles are placed in the plant cell, and start replicating and causing infection.
Fig. 11.21 Localization of viruses within nematode vectors. (A) Diagram of anterior portion of vectors. Broken lines indicate portions of the alimentary tract where virus particles are retained. (B) Longitudinal section of the buccal region of Longidorus elongatus carrying the nepovirus, RRSV. Note numerous virus-like particles (V) lining the guide sheath (G) and that none is associated with the stoma cuticle (C). Bar marker 1 ~tm. From Harrison er al. (1974), with permission.
XI. NEMATODES (NEMATODA)
There is specificity in the relationships between nematodes and the viruses they transmit with often an apparent unique association between the virus isolate and the vector species. There are some cases of different virus isolates sharing the same vector species or, conversely, one particular virus isolate being transmitted by several nematode species (Brown et al., 1995; Vassilakos et al., 1997; Brown and Weischer, 1998). There are 13 trichodorid species k n o w n to be tobravirus vectors, but only one or two of these transmits each tobravirus. By undertaking virus transmission bait tests with single trichodorid nematodes from England, the Netherlands, Scotland and Sweden, Ploeg et al. (1992) showed that there was a substantial degree of specificity between the n e m a t o d e vector and the tobravirus serotype. This specificity was more marked with Paratrichodorus species than with Trichodorus species (Table 11.6). Several nepoviruses are transmitted by more than one vector species but there can be differences in the observations under laboratory and field conditions. The Scottish and English isolates of RRSV are each transmitted in the laboratory by both Longidorus elongatus and L. macrosoma. However, under field conditions the Scottish isolate is only associated with L. elongatus and the English isolate with L. macrosoma (Brown et al., 1995). TABLE 11.6 Complementarity of tobravirus transmission by trichodorid nematodes Virus
Serotype
Sequenced isolates
Vectors
PEBV
English
TpA56
P. anemones T. cylindricus T. primitivus T. virutiferus P. pachydermus P. teres P. anemones P. pachydermus P. nanus P. pachydermus T. cylindricus T. primitivus T. virutiferus
PEBV
Dutch
-
TRV
PaY4
PaY4
TRV
PRN
PpK20
TRV
RQ
TpO1
a
Generic names: P., Paratrichodorus; T., Trichodorus. From Visser (2000), with permission.
525
Once acquired, viruses may persist in transmissible form in starved Longidorus for up to 12 weeks, in Xiphinema for about a year, and much more than a year in Trichodorus (van Hoof, 1970). Transmission does not appear to involve replication of the virus in the vector. Plant virus particles have never been observed within nematode cells. Consistent with this is the fact that no evidence has been obtained for virus transmission through eggs of nematode vectors. Specificity of transmission does not appear to involve the ability to ingest active virus since both transmitted and non-transmitted viruses have been detected within individuals of the same nematode species (Harrison et al., 1974). Sites of retention of virus particles within nematodes have been identified by electron microscopy of thin sections (reviewed by Brown et al., 1995). Nepovirus particles are associated with the inner surface of the odontostyle of various Longidorus species and with the cuticular lining of the odontophore and e s o p h a g u s of Xiphenema species. Tobravirus particles have been observed absorbed to the cuticular lining of the esophageal lumen.
D. Virus-vector molecular interactions i Nepoviruses
The genetic determinants for the transmissibility of RRSV and TBRV are encoded by RNA2 which expresses, a m o n g other proteins, the viral coat protein (reviewed by Brown et al., 1996b). ArMV particles were found associated with carbohydrate-like material on the food canal walls of X i p h i n e m a d i v e r s i c a u d a t u m (Robertson and Henry, 1986). ii Tobraviruses
By m a k i n g reciprocal p s e u d o - r e c o m b i n a n t s between a nematode-transmissible and a nontransmissible isolate of TRV, Ploeg et al. (1993b) showed that transmissibility segregated with RNA2. As noted in Chapter 6 (Section VIII.H.2.a), tobravirus RNA2 is variable in size and, as well as encoding the viral coat protein, encodes one to three non-structural proteins. A recombinant virus, in which the coat protein gene of a nematode non-transmissible isolate of
526
11 T R A N S M I S S I O N 1: BY INVERTEBRATES, N E M A T O D E S A N D F U N G I
PEBV was replaced with that of a highly nematode transmissible isolate of TRV, was not transmitted by nematodes indicating that more than one of the RNA2 genes was involved (MacFarlane et al., 1995). Mutations in both the 29-kDa and 23-kDa non-structural genes of PEBV abolished nematode transmission without affecting particle formation, as did removal of the C-terminal mobile region of the coat protein (MacFarlane et al., 1996). However, only mutation in the 40-kDa (formerly designated 29.5 kDa) non-structural gene of TRV isolate PpK20 abolished transmission by Paratrichodorus pachydermis, whereas that of the 32.8-kDa gene did not (Hernandez et al., 1997); it was suggested that the 32.8-kDa protein might be involved in transmission by other vector nematode species. Using the yeast twohybrid system an interaction was detected between the TRV-PpK20 coat protein (CP) and both the 32.8- and 40-kDa proteins (Visser and Bol, 1999). Deletion of the C-terminal 19 amino acids interfered with the CP-40K interaction but not with the CP-32.8K interaction, whereas deletion of the C-terminal 79 amino acids affected both interactions. It is suggested that the non-structural proteins may be transmission helper components analogous to those in some aphid and leafhopper virus transmission systems (reviewed by Visser, 2000).
XII. F U N G I (reviewed by Campbell, 1996) Several viruses have been shown to be transmitted by soil-inhabiting fungi. The known vectors are members of the class Plasmodiophoromycetes in the division Myxomycota, or in the class Chytridiomycetes in the division Eumycota. Both classes include endoparasites of higher plants. Species in the chytrid genus Olpidiunl transmit viruses with isometric particles, while species in two plasmodiophorus genera (Polymyxa and Spongospora) transmit rod-shaped or filamentous viruses (Table 11.7). The two chytrid vectors, Olpidium brassicae and O. bornavirus, are characterized by having posteriorly uniflagellate zoospores, whereas those of the three plasmodiophoral vectors,
Polymyxa graminis, P. betae and Spongospora subterranean, are biflagellate. All five species are obligate parasites of plant roots and have similar development stages (Fig. 11.22). They survive between crops by resting spores that produce zoospores, which infect the host. The zoospores form thalli in the host cytoplasm. In the early stages of infection the cytoplasm of thalli is separated from the host cytoplasm by a membrane, but later the thalli form a cell wall. The entire thallus is converted into vegetative sporangia or resting spores. A detailed study of the infection of sugar beet roots by P. betae is described by Barr and Asher (1996) (Fig. 11.23). The invasion of a root cell by Olpidium is illustrated in Fig. 11.24. Various degrees of host specificity exist in both the chytrid and plasmodiophoral vectors, with some isolates having a wide host range and others a narrow host range. The wide host range isolates tend to be better vectors than do the narrow ones. Two types of virus-fungal vector relationships have been recognized, termed in vitro and in vivo (Campbell, 1996).
A. In vitro fungal transmission The in vitro virus-vector relationship is found between the isometric viruses of the Tombusviridae and two Olpidium species (Table 11.7). Virions from the soil water adsorb on to the surface of the zoospore membrane and are thought to enter the zoospore cytoplasm when the flagellum is 'reeled in'. It is unknown how the virus passes from the zoospore cytoplasm to the host cytoplasm, but it is thought that this occurs early in fungal infection of the root. Reciprocal exchange of the coat proteins of TBSV (not transmitted by O. bornavarus) and CNV (transmitted by O. bornavarus) showed that the coat protein is involved in the uptake of the virus by the zoospore (McLean et al., 1994). One amino acid in the coat protein of CNV was identified as being important for transmissibility, and binding studies showed that this was associated with recognition of the virus by O. bornavarus zoospores (Robbins et al., 1997).
XIII. I'~ISCUSSI(3NAND SUMMARY 527 PL 4 $ MODIOPHORA
.''t
~"x ~
@
~
}
"'"
"~
11 t i
% //
\
//
'I I
I
.
'
? MeiOsis t
In r o o t cells
I
,
I I I
! ! t
1
I
~ ~
In soil
i I
I
..- ....PZ--~'ql
,~,s~,oo, u,, -- : ""'.'N
9
I DIPLOID
HAPt.O~D
/
In root hairs
'
I
l1III
B. In vivo f u n g a l t r a n s m i s s i o n The in vivo virus-vector relationship occurs b e t w e e n the r o d - s h a p e d viruses of the Bymovirus, Furovirus and Varicosavirus genera and O. brassicae and the three p l a s m o d i o p h o r a l species. The model for this relationship is based on observations on O. brassica and LBVV, P. graminis and SBWMV, and P. betae and BNYVV (Campbell, 1996). The virus is w i t h i n the zoospores w h e n they are released from the vegetative sporangia or resting spores and infects the new host w h e n these zoospores establish their own infection of the root. The processes of virus acquisition and release by the zoospores are u n k n o w n . Studies on various b y m o v i r u s e s and benyviruses suggest that a r e a d - t h r o u g h domain from the coat protein (for g e n o m e structure of bymoviruses and benyviruses see
Fig. 11.22 Life cycle of a plasmodiophoral fungus. On the left-hand side is the diploid stage in root cells; on the right-hand side is the haploid stage in root hairs. Between are the phases in the soil where plant-to-plant transmission of viruses can occur. Chapter 6, Sections VIII.C.2 and H.13 respectively) is implicated in the fungal transmission of these viruses (reviewed by Campbell, 1996). BNYVV RNAs 3 and 4 also have an indirect effect on the transmission, most likely t h r o u g h controlling factors such as spread and accumulation of the virus in the root system.
XIII.
DISCUSSION
AND
SUMMARY
Study of the invertebrate vectors that transmit plant viruses is i m p o r t a n t for two reasons. First, these vectors play a major role in disseminating virus diseases of economic importance in all countries. Second, virus-vector relationships are of considerable biological interest, especially those where the virus replicates in the animal vector as well as in its plant host.
528
11 T R A N S M I S S I O N 1" BY INVERTEBRATES, NEMATODES AND FUNGI
TABLE 11.7 Viruses and virus-like agents for which fungal vectors have been proven or suggested Virus genus or group Virus Polyhedral virions, in vitro acquisition Tombusvirus
Obr
Obo
CNV
+u
Carmovirus
MNSV CLSV CSBV SqNV
+u +u +u +u
Necrovirus
TNV ChNV LNV
+u + +
RCNMV STNV
-flu
D ian thovi ru s Satellite virus Polyhedral virions, acquisition unknown
Fungal vector" Pbe Pgr
Sss
Ssn
WYSV Virion not characterized, acquisition unknown Virion rod-shaped, in vivo acquisition Furovirus
WCLA SBWMV OGSV RSNV
Pectuvirus
PCV IPCV
Benyvirus
BNYVV BSBV
Pomovirus
PMTV
Bymovirus
BaMMV BaYMV OMV RNMV WSSMV
+u +
-flu + -fl + +u
LBVV TSV FLNV
+u + +
Other rod-shaped, not characterized, in vivo LRNA acquisition PYVA
+u
Varicosavirus
+u
" Vectors: Obr, Olpidium brassicae; Obo, O. bornavanus; Pgr, Polymyxa graminis; Pbe, P. betae; Sss, Spongospora subterranea f.sp. subterranea; Ssn, S. subterranea f.sp. nasturtii. +, Specific fungus associated with virus transmission; + u, unifungal or equivalent culture of fungus demonstrated to transmit virus, not necessarily the association of vector with virus; WCLA, watercress chlorotic leafspot agent; LRNA, lettuce ring necrosis agent; PYVA, pepper yellow vein agent. From Campbell (1996), with kind permission of the copyright holder, 9 Annual Reivews. www.AnnualReviews.org Only two invertebrate phyla have members that feed on living green land plants and both of t h e s e m t h e Nematoda a n d A r t h r o p o d a - c o n t a i n v e c t o r s of p l a n t viruses. T h r e e g e n e r a of n e m a t o d e s are v e c t o r s . Two, Xiphinema a n d Longidorus, c o n t a i n vec-
t o r s for t h e p o l y h e d r a l v i r u s e s of t h e Nepovirus g r o u p . R o d - s h a p e d t o b r a v i r u s e s are t r a n s m i t t e d b y s p e c i e s of Trichodorus. T h e r e is a s u b s t a n t i a l d e g r e e of specificity in the v i r u s - n e m a t o d e r e l a t i o n s h i p , a l m o s t cert a i n l y i n v o l v i n g a t t a c h m e n t of viral p a r t i c l e s
XIII.
I)IS(:USSI(~N A N D S U M M A R Y
529
Fig. 11.23 {see Plate 11.2) Invasion of sugar beet roots by Polymyxa betae. (A) Young plasmodium in epithelial cells with blue-stained lipid and nuclei, and pink dots of unidentified material. (B) Young lobed zoosporangium in cortical cell. Note pink ground color, blue zoospores and dark blue nuclei. (C) Mature zoosporangium containing blue-stained zoospores. {D) Distribution of P. betae 8 days after inoculation. Plasmodia in various stages of development are restricted to epidermal cells. (E) Distribution of P. betae 18 days after inoculation. Plasmodia and zoosporangia are present throughout the root cortex but not in the endodermis or stele. Bar markers: (A) and (C), 5 ~tm; (B), 10 ~tm; (D) and (E), 50 lam. From Barr and Asher (1996), with permission. to specific zones in the lining of the n e m a tode gut. A m o n g the classes of the Arthropoda p h y l u m , t w o have m e m b e r s that feed on living green land p l a n t s m t h e Arachnida a n d Insecta--and both of these contain viral vectors. The m o s t i m p o r t a n t g r o u p of vectors n u m e r i c a l l y is to be f o u n d in the insect order Homoptera, w h i c h
includes the aphids, leafhoppers, p l a n t h o p p e r s , whiteflies a n d m e a l y b u g s . As these g r o u p s h a v e s u c k i n g m o u t h p a r t s that p e n e t r a t e leaf cells a n d tissues, they are ideally suited to t r a n s m i t v i r u s e s f r o m d i s e a s e d to h e a l t h y plants. A b o u t 66% of the a r t h r o p o d - b o r n e v i r u s e s are t r a n s m i t t e d by aphids. There are v a r i o u s
530
11 TRANSMISSION 1: BY INVERTEBRATES, NEMATODES AND FUNGI
Fig. 11.24 Fungal transmission. Infection of a root cell by Olpidium. Electron micrograph showing contents of an encysted zoospore entering the host cell. CC, cyst cytoplasm; HC, host cytoplasm; HW, host wall; CW, cyst wall; CT, cyst tonoplast; CE, cyst ectoplast; V, vacuole. From Temminck and Campbell (1969), with permission. types of v i r u s - a p h i d relationship, as s u m m a rized in Table 11.2. None of these involves a 'flying pin' kind of transmission. Several members of the Rhabdoviridae replicate in their aphid vectors, and vectors may remain infective for their lifetime. Even w h e n the aphid remains infective for a relatively brief period, there is specificity in the virus-vector relationship. This specificity involves virus-coded proteins. (1) Potyviruses and caulimoviruses each produce a virus-coded protein, a helper fiTctor, that is essential for aphid transmission. They probably act by facilitating the binding of virus to some internal surface of the aphid. (2) Some semipersistent viruses cannot be transmitted by an aphid vector w h e n the RNA is coated with its own virus-coded coat protein. They can, however, be transmitted w h e n the RNA is encapsulated in the coat protein of some unrelated
virus, a helper virus, in a plant infected with both viruses. (3) Viral coat proteins are also involved in the specific retention of particular viruses by particular aphids. (4) Specificity of aphid vector transmission m a y involve different strains of the same virus, as occurs with luteoviruses. Virus strains will be transmitted by a particular vector only if the RNA is in the protein coat appropriate for that vector. Details of the viral side of the interactions of some of the the helper c o m p o n e n t and coat protein systems are now understood at the molecular level but little is k n o w n about the vector side of the interaction. The leafhoppers and planthoppers constitute a second important group of vectors. No viruses are transmitted in a non-persistent m a n n e r by hoppers. Two important viruses, M C D V and RTSV, are transmitted in a semipersistent m a n n e r but most are circulative, either being non-propagative or propagative. Viruses transmitted by hoppers are not as n u m e r o u s as those transmitted by aphid vectors, but they include a n u m b e r of economically i m p o r t a n t viruses, especially those infecting food crops belonging to the Poaceae. There is a substantial degree of vector specificity between virus and hopper, and frequently both have s o m e w h a t narrow host ranges. genera--Reoviridae, Four families and Rhabdoviridae, and the Tenuivirus and Marafivirus g e n e r a - - c o n t a i n m e m b e r s that replicate in their leafhopper vectors. Such replication usually has little effect on the hoppers. However, from the point of view of the virus, replication in the vector has two important consequences: (1) once they acquire virus the vectors normally remain infective for the rest of their lives, and (2) replication in the vector is often associated with transovarial passage of the virus, thus giving it a means of survival over winter that is quite i n d e p e n d e n t of the host plant. In the plant reoviruses, particular g e n o m e s e g m e n t s code for gene p r o d u c t s required for replication in the insect, but not in the plant. With viruses that replicate in their vectors, there may be a high degree of specificity between vector and virus, or even strains of a virus.
XIII.
Leaf-feeding beetles have chewing mouthparts and do not possess salivary glands. They regurgitate during feeding, which bathes the mouthparts in sap. This regurgitant will contain virus if the beetle has fed on an infected plant. Beetles can acquire virus after a single bite and can infect a healthy plant with one bite. However, beetle transmission is not a purely mechanical process. There is a high degree of specificity between beetle vector and virus, and some very stable viruses such as TMV are not transmitted by beetles. The viruses that are transmitted belong to the Tymovirus, Comovirus, Bromovirus and Sobemovirus genera. Sometimes one beetle species will transmit a particular virus with high efficiency while a related species does so inefficiently. The reasons for this sort of specificity are not understood. However, we are beginning to understand w h y some stable viruses are not beetle transmitted. The regurgitant fluid of the beetles contains an inhibitor that prevents the transmission of non-beetle-transmitted viruses but does not affect those that are transmitted. There is good evidence that the inhibitor is an RNase. Other insect vectors are found among the mealybugs, whiteflies, mirid bugs and thrips. The viruses transmitted by these groups are not
I)IS(~USSION AN1) SUMMARY
531
numerous, but the first two are vectors for some viruses causing important diseases in tropical crops. In the Arachnida, eriophyid mites are vectors for several viruses. For viruses transmitted through the pollen, pollinating insects can transfer infected pollen to healthy plants, thus transmitting the virus in an indirect manner. Other biological vectors include plasmodiophoral fungi with which viruses have specific interactions. Little is k n o w n about the detailed molecular interactions involved in this form of transmission. Thus, for biological vectors, analyzes of the viral d e t e r m i n a n t s of the specificity of v i r u s - v e c t o r interactions have given m u c h detailed information. This is relatively easy because of the relatively small size of the viral g e n o m e and the ability to m a n i p u l a t e it. However, the genomes and molecular biology of the vectors are much more complex and there has been little research on them. The application of m o d e r n technologies should enable these aspects to be studied in the future. Findings from such research could lead to new approaches in generic control of important viruses of crops by interfering with key aspects of the virus-vector interaction.
This Page Intentionally Left Blank
C H A P T E R
12
Transmission 2:
Mechanical, Seed, Pollen and Epidemiology Many factors have to be taken into account in performing efficient experimental mechanical inoculation of viruses; these are described in detail in Walkey (1991) and Dijkstra and de Jager (1998), and are outlined below.
In the previous chapter, I described how viruses are transmitted by 'biological' vectors: arthropods, nematodes and fungi. This chapter covers three further aspects of virus m o v e m e n t between plants. As viruses need mechanical damage to enter a host, mechanical inoculation is a widely used technique for infecting plants experimentally; some viruses are also transmitted naturally by mechanical means. In this chapter, I discuss various considerations in experimental mechanical inoculation of viruses and also some other means (seed and pollen) that do not involve a biological vector by which viruses can move from the initially infected plant. Then I will describe how viruses move in the field and the various factors involved in the occurrence of epidemics.
A.
S o u r c e a n d p r e p a r a t i o n of i n o c u l u m
The most c o m m o n source of virus for mechanical inoculation is infected leaf tissue. Since the virus has to be released from cells in infected plants, many of the considerations for obtaining the best quality inoculum are similar to those in the extraction of viruses from plants for purification (Chapter 4). It is usual that plant material likely to contain the highest a m o u n t of virus is used as a source but one also has to consider the possible presence of inhibitors of virus infection. In many cases, infected y o u n g leaves showing strong symptoms are used but for some viruses, other tissues may be better. For instance, roots of plants infected with TNV contain greater amounts of virus than do leaves (Smith, 1937). CMV is transmitted more efficiently from cucumber flower petals than from leaves which contain more inhibitors (Sill and Walker, 1952). The infected plant material is sometimes ground in tap water (Dijkstra and de Jager, 1998) but, more frequently, an extraction buffer is used. It has been recognized for some years that phosphate buffers enhance infectivity of many viruses (Yarwood, 1952; Fulton, 1964). Breaking the plant cells exposes the virus to secondary metabolites that can affect infectivity and that are more prevalent in some hosts than in others. As well as plant products such as
I. M E C H A N I C A L T R A N S M I S S I O N Mechanical inoculation involves the introduction of infective virus or viral RNA into a w o u n d made through the plant surface. When virus establishes itself successfully in the cell, infection occurs. This method of transmission is of great importance for many aspects of experimental plant virology, particularly the assay of viruses often by local lesion production (see Chapter 15) in the propagation of viruses for purification (see Chapter 4) and in the study of the early events in the interaction between a virus and susceptible cells (see Chapter 10). When intact virus is used as inoculum, the viral nucleic acid must be partly or entirely uncoated at an early stage. This process is discussed in Chapter 7 (Section II). Virus acronyms are given in Appendix 1.
533
534
12 TRANSMISSION 2: MECHANICAL, SEED, POLLEN AND EPIDEMIOLOGY
nucleases and the products from the oxidation of polyphenols which may affect virus structure and stability, there are many other substances in crude sap that could affect the number of successful infections produced by a virus. The former can be countered by additives similar to those used in virus purification (see Chapter 4), the latter often by dilution of the sap. Buffer mixtures with additives m a y be specially designed for successful inoculation from particular hosts (e.g. Martin and Converse, 1982). An alternative method for preparing inoculum which overcomes inhibitor problems is to grind infected tissue frozen in liquid nitrogen to a fine powder and inoculate that directly, for instance with a fine brush (Lawson and Taconis, 1965: Ragetli et al., 1973).
B. Applying the inocul'am The efficiency of mechanical inoculation is greatly increased when some abrasive material is added to the inoculum or sprinkled over the leaves before inoculation. The most commonly used abrasives are c a r b o r u n d u m (400-500 mesh) or diatomaceous earths such as Celite. The increase in n u m b e r of local lesions obtained by the use of abrasives varies with different hosts and viruses, but may be 100-fold or more. The time of addition of these materials may be important. Celite added after grinding and dilution was much more effective than when added before grinding (Yarwood, 1968). The early method of placing drops of the inoculum on the leaf and scratching or pricking the leaf surface with a needle to cause wounding was very inefficient and has been largely superseded by gently rubbing the leaf surface with some suitable object wetted with the inoculum to give more efficient transmission. A wide variety of objects has been used, depending on the preference of the operator and the volume of inoculum available (see Walkey, 1991; Dijkstra and de Jager, 1998). The objective in mechanical inoculation is to make numerous small w o u n d s in the leaf surface without causing death of the cells. The pressure required to do this depends on many factors, such as plant species, age and condition of leaf, and additives present in the inoculum.
Macroscopic areas of dead tissue appearing on the inoculated leaf within a day or so indicate that the w o u n d i n g was excessive. With a few viruses and hosts severe abrasion is more effective (Louie and Lorbeer, 1966). For some pathogens, for example CTV, which is probably confined to the phloem, TBRV and CEVd, cutting or slashing the plant stem with a contaminated blade is the most effective method of mechanical transmission (Garnsey and Whidden, 1973; Garnsey et al., 1977; Bitterlin et al., 1987). The vascular puncture technique in which the virus is inoculated to kernels of maize and barley is successful for MWLMV and MRFV (Louie, 1995; Madriz-Orde~ana et al., 2000). Particle bombardment has been used for infection with viruses (Franz et al., 1999) and with viral RNA (Klein et al., 1987), cDNA clones of RNA viruses (see Gal-On et al., 1995, 1997; Fakhfakh et al., 1996) and cloned DNA viruses (see Gilbertson et al; 1991; Garzon-Tiznado et al., 1993; Hagen et al., 1994). The cDNA constructs of RNA viruses included a promoter, usually the CaMV 35S promoter (see Chapter 7, Section IV.C.1), to express an infectious transcript; those of DNA viruses use the viral promoter but have to be such that the introduced genome can effectively replicate (see below). Particle bombardment inoculation has resulted in infection with several viruses that had proved to be difficult, if not impossible, to transmit mechanically but some, for instance RTBV, are not transmissible by this procedure (Dasgupta et al., 1991). Holmes (1929) considered that washing inoculated leaves with water immediately after inoculation increases the n u m b e r of local lesions formed, and this has become a fairly widespread practice. However, washing leaves after inoculation, spraying with water or dipping leaves in water may substantially reduce the number of local lesions produced by several viruses or have variable effects depending on other conditions (Yarwood, 1973). The effect of washing or dipping leaves in water on the n u m b e r of lesions probably depends on many factors, and particularly on whether inhibitors of infection are present in the inoculum. If such inhibitors are present, washing may minimize their effect.
I1.
FA(.'TORS INFLUEN(.'IN(~ THE (.'OURSE OF INFE(.'T1ON A N t ) DISEASE
If the leaves are dried rapidly after inoculation, either by blotting or with an air jet, there may be a marked increase in the number of local lesions, but again the effect is variable (Yarw ood, 1973). Polson and von Wechmar (1980) used electroendosmosis to introduce MSV into leaves through cut petioles in a way that gave rise to infection even though this virus is normally transmitted only by leafhoppers. Konate and Fritig (1984) described an efficient microinoculation procedure that allowed early events following mechanical inoculation to be studied at predetermined individual infection sites on the leaves. Injection of virus into petioles or stems with a hypodermic syringe has been used occasionally but, apart from the use of a high-pressure medical serum injector to transmit BCTV (Mumford, 1972), is generally a very inefficient method, although micro-injection of trichome cells has proved useful in studying cell-to-cell spread of viruses (Derrick et al., 1992). However, the injection of infective constructs of viruses cloned into Agrobacterium tumefaciens has proved to be very effective, especially with viruses that are difficult to transmit mechanically. In this process, termed agro-inoculation, the viral construct, cloned into the T-DNA of A. tumefaciens Ti plasmid is transformed into A. tumefaciens, which is then injected into the host. The system was initially developed for CaMV, which infects dicotyledonous species (Grimsley at al., 1986). A. tumefaciens is thought not to infect monocotyledons, but Grimsley et al. (1987) showed that, when cultures of the bacterium containing a plasmid with tandemly repeated copies of MSV DNA were inoculated to whole maize plants, the plants developed symptoms caused by MSV. Agro-inoculation has been used for several other geminiviruses (e.g. Donson et al., 1988; Kheyr-Pour et al., 1994) and for RTBV (Dasgupta et al., 1991), which are not mechanically transmissible. As with particle bombardment, agro-inoculation constructs of these DNA viruses has to be such that transcription and replication can be initiated on their introduction into the plant. For geminiviruses, dimers of the viral genome are usually used; for reverse transcribing viruses (e.g. CaMV, RTBV) a 'one and a
535
bitmer' construct that can transcribe the 35S RNA (see Chapter 7, Section IV.C.1) is produced. There appears to be some specificity on the strains of A. tumefaciens that give efficient transmission of viruses to monocotyledons (Boulton et al., 1989; Dasgupta et al., 1991). cDNAs of RNA viruses have also been agroinoculated effectively (see Leisner et al., 1992a; Turpen et al., 1993); as with particle b o m b a r d m e n t inoculation, these constructs have to include a suitable promoter. Inoculating large numbers of plants can be a time-consuming process, and various procedures have been adopted to reduce the time involved. For example, dipping and moving the leaves of seedling plants in the inoculum may provide a rapid method for inoculating large numbers of seedlings at the time of transplantation. Where large numbers of leaves are to be inoculated with the same inoculum, airbrushes of the type used by artists may prove useful (Whitham et al., 1999). Alternatively, the inoculum may be applied in a solid stream (Louie et al., 1983). A sensitive and rapid method using a specially designed airgun has been described by Laidlaw (1987). To avoid washing pestles and mortars or other glassware where large numbers of individual tests have to be made, it may be possible to rub a piece of diseased leaf directly on a leaf of the test plant (Murakishi, 1963).
II. FACTORS I N F L U E N C I N G THE COURSE OF I N F E C T I O N A N D DISEASE For a given species of plant and a given virus, there are many factors that can influence the course of infection and the disease that develops. The way in which a virus is inoculated into the plant may be important. Sometimes many different strains of a virus occur and these may cause quite different kinds of disease in the same host plant under the same conditions (see Chapter 17). In this section, I shall discuss inherent variables in the host plant itself, and environmental factors. Interactions between unrelated viruses and between viruses and
536
12 TRANSMISSION 2: MECHANICAL, SEED, POLLEN AND EPIDEMIOLOGY
some other agents of disease are described in Chapter 10, Section V. A. T h e plant being inoculated Sometimes a virus can be transmitted mechanically by inoculating the cotyledon leaves, but not the first true leaves, for example with a virus of sweet potato (Alconero, 1973). Mechanical inoculation of roots is inconvenient and is often less successful than with leaves. However, transmission by this means has been achieved for several viruses (e.g. Moline and Ford, 1974). 1. Susceptibility to infection Most commonly, small young leaves and old leaves are less susceptible than well-expanded younger leaves. There may be a marked gradient of susceptibility with age (Fig. 12.1). The curves in Fig. 12.1 can be taken to indicate the changes in susceptibility that individual leaves undergo with time. One has to consider the recent
40 N
E
u o
I/I C
.9
30
20
B
ID O U 0 .J
10-
o
I
l
1
2
1
4
1
1
6
1
1
8
1
1
1o
PI
Fig. 12.1 Effect of leaf age on number of local lesions produced by TMV in leaves of Samsun N N tobacco. Since y o u n g e r leaves are expanding, the n u m b e r is expressed per unit leaf area at the time of inoculation. A plastochron is the time interval between corresponding developmental stages of successive leaves. The leaf plastochron index (PI) is an arbitrary measure that allows plants at slightly different developmental stages to be compared. Each curve represents successive leaves from
a single plant. PI O, 15.10; I , 15.12; A, 15.30; 9 15.52; E] 15.53. From Takahashi (1972b), with kind permission of the copyright holder, 9 Blackwell Science Ltd.
understanding of host response to virus infection (see Chapter 10). The gradient may not always be of the form shown in Fig. 12.1. Thus, on Nicotiana glutinosa plants with 8 to 10 well-developed leaves, TMV will produce more local lesions on the middle and lower leaves than on the younger ones. In contrast, TBSV may produce no lesions on the oldest leaves and most on the y o u n g e s t (Bawden, 1964). Abscisic acid accelerates senescence processes in leaves. Exogenously applied abscisic acid increased susceptibility of Samsun tobacco leaves to TMV (Balazs et al., 1973), indicating a possible role for this hormone in age-dependent changes in leaf susceptibility. 2. Environmental factors The environmental conditions under which plants are grown before inoculation, at the time of inoculation, and during the development of disease can have profound effects on the course of infection. A plant that is highly susceptible to a given virus under one set of conditions may be completely resistant under another. If infection occurs, the plant may support a high or low concentration of virus and develop severe disease or remain almost symptomless, depending on the conditions. 3. Facu~rs affecting susceptibility to infection Any factor that alters the ease with which the surface of the leaf is w o u n d e d will alter the probability of successful entry of virus introduced by mechanical inoculation, while physiological changes in the leaf may make the cell more or less suitable for virus establishment. As a broad generalization, greenhousegrown plants will have greatest susceptibility when they are grown and used under the following conditions: (1) mineral nutrition and water supply that do not limit growth; (2) moderate to low light intensities; (3) a temperature in the range of 18-30~ depending on virus and host; and (4) inoculation carried out in the afternoon. These plants are termed horticulturally 'soft' and may be different to plants grown for normal horticultural purposes.
II.
F A C T O R S I N F I . U E N C I N G T H E C O U R S E OF INFE(.'TION A N D DISEASE
a. Light There are two general situations in which light affects susceptibility: short-term changes in light intensity or deviation over a period of no more than a day or two, and long-term effects that may markedly influence the growth of the plant. Bawden and Roberts (1947) found that in summer reducing light intensity or giving plants complete darkness for periods of days increased the susceptibility of several hosts to certain viruses. This effect of shading or darkening plants for about a day before inoculation is often used as a practical measure to increase the susceptibility of test plants with viruses that are difficult to transmit. Various experiments suggest that light may have two opposing kinds of effects. Excessive illumination may reduce susceptibility. On the other hand, after a period of darkness, even a short burst of light may increase the number of local lesions. For example, when bean plants that had been in the dark for 18 hours were given a 1-minute exposure to 800 foot-candles before inoculation, the number of local lesions produced by a TNV inoculum was double that of plants inoculated under minimal illumination required to see the leaves (Matthews, 1953b). In almost all studies on the effects of light, the experimental plants have been raised under ordinary greenhouse conditions. Thus, any experimentally imposed change will have been superimposed on the natural daily cycle of variation in susceptibility, discussed in a subsequent section. The time of day at which an experiment is performed can have a marked effect on results (Matthews, 1953b). As far as the long-term effects of light on susceptibility are concerned, it is generally found that high light intensities over the period of growth of the test plant give rise to 'hard' plants that may have very low susceptibility compared with 'soft' plants raised under lower light intensities. A further consideration is the quality of light. Plant grown in the spring and autumn in temperate climates are often more susceptible and produce more virus than those grown in the summer or winter. Even with supplementary lighting, it appears that the quality of natural
537
light and possibly the changing of natural day length might have an influence on virus yield. b. Temperature
In general, pre-incubation of plants at slightly higher temperatures than normal before inoculation increases susceptibility. The effect of treating plants at a higher temperature after inoculation may vary with the virus or the strain of virus tested (e.g. Kassanis, 1952). A brief shock treatment, for example at 50~ after inoculation may affect the kind of lesion that subsequently develops (e.g. Foster and Ross, 1975). Where necrotic ring spot local lesions are produced, light and temperature conditions interact in the production of rings (Harrison and Jones, 1971). Temperature may also affect the speed and efficiency of graft transmission (Fridlund, 1967a). c. Water supply and humidity
The immediate effect of washing inoculated leaves with water was discussed in Section I.B. Generally speaking, if plants are grown with a minimal supply of water they become more or less stunted, leaf texture is 'hard', and they will give greatly reduced numbers of local lesions compared with plants raised under conditions where water supply is not limiting growth. For example, Tinsley (1953) found that as many as 10 times more local lesions were produced by several viruses on well-watered N. glutinosa and tobacco plants as on poorly watered plants. Most reports suggest that moderate wilting near the time of inoculation increases susceptibility. For example, when detached bean leaves were wilted before inoculation with TMV, water deficits of 0-15% of the green weight of the leaf increased infection. More severe deficits in the range of 15-29% decreased infection. When leaves were wilted after inoculation, numbers of local lesions increased with increasing water deficit over the range of 0-353/o (Yarwood, 1955). Maximum increase was 4-8-fold over that of unwilted leaves. d. Nutrition
The nutrition of the host plant may have a marked effect on the numbers of local lesions
538
12 TRANSMISSION 2" MECHANICAL, SEED, POLLEN AN[) EPIDEMIOLOGY
produced by viruses. As would be expected, interactions between different nutrients are quite complex. For example, Bawden and Kassanis (1950) investigated the effects of various levels of nitrogen, phosphorus and potassium on susceptibility of tobacco and N. glutinosa plants to strains of TMV. The range of fertilizer treatments used p r o d u c e d large effects on plant growth and many significant differences in susceptibility. In general, nutritional conditions that were most favorable for plant growth were also those giving greatest susceptibility. There was no evidence that one particular element increased susceptibility on its own. Trace elements may also affect transmissibility. For example, increased m a n g a n e s e supply caused a marked increase in the number of local lesions produced by PSTVd in Scopolia leaves (Singh et al., 1974).
e. Time of day Since many basic processes in leaves are influenced by the diurnal cycle of night and day, it is not surprising that the susceptibility of leaves to mechanical inoculation varies systematically with time of day. The number of local lesions produced rises to a maximum in the afternoon and falls to a m i n i m u m during the night, usually just before dawn (Fig. 12.2).
A diurnal variation has been found for various host-virus combinations. This diurnal change is not dependent on immediate changes in environmental conditions, but appears to be 'built in', as are some other physiological processes in leaves. Thus, bean plants placed in darkness at constant temperature at 4 p.m. and maintained in darkness through the following day showed a very similar cycle of susceptibility to plants that were exposed to the normal daylight cycle (Matthews, 1953b).
f. Time of year Where there are large climatic differences between the seasons, plants may vary widely in their susceptibility to a virus at different times of the year. For example, certain varieties of bean (Phaseolus aulgaris) grown at Rothamsted in England during the s u m m e r appeared to be almost i m m u n e to CMV. In the winter, they produced substantial numbers of necrotic local lesions (Bhargava, 1951). Similar, but less extreme, differences have been noted with other viruses producing local lesions. The major environmental factor concerned in greenhousegrown plants is probably light, as Bawden and Roberts (1947) showed that typical winter reactions of plants can be reproduced in summer by appropriate shading. g. Chemical and mechanical injury
Nematicides and fungicides may increase susceptibility to virus infection. For example, atrazine increased susceptibility of maize hybrids to MDMV (MacKenzie et al., 1970). The mechanical injury involved in transplanting wheat plants led to increased transmission of WSSMV from infected soil (Slykhuis, 1976).
0.4 0.2 oc~ .._J
0.0
-O2 -0.4 J_
I
l
1
l
1
1
1
1
1
I
1
1
1
l
6 8 1012 2 4 6 8 1012 2 4 6 8 I0 AM PM AM Time of ~noculation
Fig. 12.2 Effect of time of day of inoculation on the numbers of local lesions produced by TNV in beans. Log R relates the number produced to the number produced at a base time (8 a.m.). Bar A represents the difference between any two points required for significance at P = 1%. From Matthews (1953a), with permission.
B. Development of disease 1. Virus c~mcentration in the inoculum The amount of infecting virus may influence the extent to which growth is subsequently depressed (compare (A) and (B) in Fig. 12.3). This may in fact be an age-of-plant effect because it is probable that virus moves systemically through the plant sooner after a heavy
II.
F A ( . ' T O R S INFI_UENCING T i l E (.'OURSE OF I N F E ( ' T I O N A N t ) DISEASE
539
2. Environmental factors a. Light Light intensity and duration affect virus production and disease expression in different ways with different viruses, but, generally speaking, high light intensities and long days favor replication. For instance, local lesions formed in 17 hours under continuous illumination at 25~ on detached P. vulgaris leaves inoculated with AMV maintained in moist Petri dishes, whereas they took nearly 30 h under low lighting on the laboratory bench (R. Hull, unpublished observation). b. Temperature
Fig. 12.3 Interaction between time of infection and number of infecting aphids on subsequent depression in size of barley plants by BYDV. (A} About eight aphids per plant. (B) About 150 aphids per plant. The inoculative feed was 2 days, so that feeding damage was insignificant. Stages of growth when infected: (left to right) control; flowering; boot; tillering; three-leaf; oneleaf. From Smith (1967), with permission.
inoculation, either mechanically or by insects. However, other workers have not found a decrease in plant yield with increasing numbers of aphids (e.g. Skaria et at., 1984). The apparent discrepancy may be in the n u m b e r of aphids used. To obtain effects such as that shown in Fig. 12.3, 100 or more aphids were used per plant for the heavy inoculation.
Over the range of temperatures at which plants are normally grown, increasing temperature usually increases the rate at which viruses replicate and move through the plant. Like other biological phenomena, however, increase in temperature above a certain point leads to a reduction in the rate of replication. The species of host plant, strain of virus, and age of the leaf in which the virus is replicating may have a major effect on the way the virus behaves with changes in temperature. Figure 12.4 illustrates some effects of temperature on TMV replication. The temperature at which plants are grown frequently affects the kind of disease that develops. For example, the type of pigmentation d e v e l o p i n g in s u b t e r r a n e a n clover infected with red-leaf virus was dependent on the temperature at which the plants were grown (Helms et al., 1985). A severe stem tip necrosis developed in certain soybean cultivats held at 24~ following infection with SMV. At 28~ most plants developed typical mosaic disease (Tu and Buzzell, 1987). Plantains and bananas infected with BSV exhibited marked symptoms when grown at 22~ but the s y m p t o m s were reduced or even disappeared when the plants were transferred to 28-35~ (Dahal et al., 1998); s y m p t o m s reappeared w h e n the plants were returned to the lower temperature. The viral titer was significantly higher in the s y m p t o m a t i c plants. However, this may not necessarily be a temperature effect: it could be due to changes in light quantity or quality.
540
12 TRANSMISSION 2" MECHANICAL, SEED, POLLEN AND EPIDEMIOLOGY 40
A
35
~o
3O
, 9 25
Ot~9 o o z
.--
1500
-6
1200
,~ c~
20 15
B
I
4~ 32 ~
9oo 600
"6
10
3oo
12
24
36
48
60
72
1 96
o
Hours after infections
High growth temperatures may annul the effect of some hypersensitivity genes. For example, in Nicotiana cultivars with the N gene, TMV causes systemic disease at 36~ whereas at 28~ or lower it gave a hypersensitive response (Kassanis, 1952). On the other hand, the activity of genes N x and N b in potato, causing hypersensitivity to PVX, were not affected by higher temperatures (Adams et al., 1986b). Increasing temperature, up to a certain point, increases the rate of systemic movement and decreases the time before the first appearance of systemic symptoms (e.g. Jensen, 1973). The shortest incubation period for systemic invasion does not necessarily occur at the temperature that leads to maximum production of virus. In plants kept at different temperatures there may or may not be an approximate correlation between severity of disease symptoms and virus concentration reached. For example, such a correlation is observed with TMV and PVX in tobacco in the range 16-28~ (Bancroft and Pound, 1954). Changes in the temperature at which plants are grown may lead to a selective multiplication of certain strains adapted to the particular conditions (see Chapter 17, Sections III and IV.A.3) . Practical applications of heat therapy are discussed in Chapter 16 (Section II.C.2.b). c. Water supply
The effects of water supply on virus replication have not been studied systematically. A chronic deficiency of water giving stunted 'hard' plants
2
4
6
8
10
12
]4
]6
Time (days)
]8
20
22
Fig. 12.4 Effect of temperature on TMV replication. (A) A linear plot of the time course of TMV replication in tobacco mesophyll protoplasts. From Dawson et al. (1975), with permission. (B) Effect of temperature on the replication of TMV (common strain) in tobacco leaf discs. Concentration of virus measured serologically. From Lebeurier and Hirth (1966), with permission.
will usually give rise to less obvious symptoms. A liberal supply of soil water increases the incidence of internal browning disease in tomatoes due to ToMV infection (Boyle and Bergman, 1967). d. Nutrition
A number of investigations have been made into the effects of host plant nutrition on virus replication, particularly of TMV in tobacco plants. Many of the results are conflicting, probably due to variations in methods of growing the plants, the effective concentration of nutrients supplied, strain of virus used, method of estimating virus, and the basis for expressing the results. Supply of the major elements that give best plant growth usually allows for the greatest replication of virus. For example, Bawden and Kassanis (1950) concluded that the effects of nitrogen and phosphorus supply on TMV replication were fairly closely correlated with effects on plant growth. Treatments giving better plant growth led to a greater increase in virus production, both per unit fresh weight and per plant. Pea plants provided with low calcium nutrition developed severe stunting when infected with pea leafroll virus (Thompson and Ferguson, 1976). Minor element nutrition (zinc, molybdenum, manganese, iron, boron and copper) has a variable effect on the capacity of plants to support virus replication. Effects on virus accumulation often parallel effects on plant growth, but exceptions have been observed. For example,
II.
FA(.'TORS I N F L U E N C I N G T H E C O U R S E OF I N F E C T I O N A N D D I S E A S E
manganese deficiency led to an increase in TMV concentration in the leaves. By contrast, toxic concentrations of this element in cowpea led to greatly enhanced replication of CCMV (Dawson, 1972). e. Time of year The complex factors including day length, light intensity and quality, air and soil temperatures, and water supply that change during the seasonal cycle will affect plant growth, and thus the disease produced by a given virus, and the extent to which a virus replicates. For example, there was a seasonal variation in the concentration of AMV in alfalfa, the highest concentrations being recorded in spring and the lowest in a u t u m n (Matisovfi, 1971). Some viruses, such as PLRV in potato and BYV in sugar beet, cause much more distinct disease symptoms in summer than in winter. Others, such as PVX in potato, may cause more severe disease under winter conditions. Chloroplasts in the leaves of Abutilon infected with AbMV showed a marked seasonal variation in the severity of ultrastructural changes, being most severe in the summer (Schuchalter-Eicke and Jeske, 1983).
3. Systemic spread In a fairly mature tobacco or tomato plant inoculated with TMV in the younger leaves, the lower leaves may never become infected with virus unless directly inoculated. With increasing age there is a greater tendency for infection to remain localized. Physiological age rather than actual age is the significant factor as discussed under virus m o v e m e n t (see Chapter 9). When potato plants are infected with PVX, either naturally in the field or by inoculation, some tubers may be free of the virus at the end of the first season. Beemster (1966) found that the proportion of tubers and eyes infected with PVY is closely related to the time at which the plants are inoculated. For example, the percentage of infected tubers was 100% for plants 8 weeks old on inoculation and only 25% for plants that were 13 weeks old when inoculated. Time of infection is often an important factor in d e t e r m i n i n g loss of yield in economically important crops. For example, loss of yield is generally more severe when cereals are infected
541
at an early stage of growth with BYDV (Smith, 1967) (Fig. 12.3). In field experiments with Capsicum annuum, plant growth and fruit yield improved almost in direct proportion to the lateness of inoculating the plants with CMV (Agrios et al., 1985). The term 'mature plant resistance' has been used for these age-of-plant effects, but the mechanism for them is unknown. Many metabolic changes occur as leaves age. For example, ribosome content decreases (Venekamp and Beemster, 1980), but no causal relationship has been established between such changes and increased resistance to the effects of inoculation with a virus. Barker and Woodford (1987) describe a longer-term effect of late infection. Potato plants grown from tubers of plants infected with PLRV late in the previous season developed unusually mild symptoms, although they contained as m u c h virus as plants with severe symptoms. Perhaps a strain selection process was at work or there is a change in the natural plant infection response (see Chapter 17). 4. The kinds of host response Genetic and molecular aspects of resistance to viruses are discussed in Chapters 16 and 10 respectively. C. Viral n u c l e i c acid as i n o c u l u m For any virus that has (+)-sense ssRNA, and many with ssDNA or dsDNA, as its genetic material it should be possible, in principle, to prepare an extract of total nucleic acids from infected tissue and use this to inoculate healthy plants. This procedure may allow mechanical transmission w h e n whole-leaf extracts are ineffective. Success may be due to removal of virus inhibitors into the phenol phase or interface (or by another method), or it may be due to the existence of unstable or incomplete virus that is inactivated by nucleases unless these are removed by the phenol. With some tissues and viruses, grinding or blending in the presence of phenol may release virus from sites where it remains bound in normal sap extracts. Any infectivity in phenol extracts or RNA preparations will be fully susceptible to nucleases on
542
12 TRANSMISSION 2" MECHANICAL, SEED, POLLEN AND EPIDEMIOLOGY
the cuticle and the cell wall are probably effective in allowing virus to enter, and broken hairs are probably a common route of entry (Fig. 12.6). There have been some suggestions of direct contacts between plant cells and the environment through ectodesmata and hydathodes. Ectodesmata, retermed ectoteicoides, are not analogous to plasmodesmata (Franke, 1971) and it is considered that, in most cases, they are not sites of virus entry. Experiments with several viruses indicate that, following mechanical inoculation, some underlying mesophyll cells may be infected directly with the inoculum (Salinas Calvete and Wieringa-Brants, 1984; Matthews and Witz, 1985). Hydathodes are structures containing water pores located at leaf margins (Cook et al., 1952) that connect to the intracellular spaces and to the xylem vascular system. Under conditions of water uptake and limited transpiration, such as w a r m soils and high humidity in the dark, liquid is expelled through the hydathodes in a process termed guttation. Particles of TMV have been found in the guttant of tomato (Johnson, 1937), of ToMV in tomato and Go,ll#lmna %lobosa, of PPMV from Capsicum annut~Jll (French et al., 1993), of 10 genera of viruses in the guttant of cucumber (French and Elder, 1999) and of BMV in barley (Ding et al., 2000). BMV was found in the intercellular spaces (Ding et al., 2000) and ZYMV in the xylem (French and Elder, 1999) but how they got there has not yet been determined.
contaminated glassware once the phenol is removed, as it must be before inoculation using the extract.
D. Nature and number of infectible sites 1. Nature of the leaf surface On the upper surface of leaves commonly used for mechanical inoculation, there will be about 1-5 • 106 cells of various types. The structure of the leaf surface is shown in Fig. 12.5 and details can be found in Esau (1953). The outermost layer is the epicuticular wax, which in different species may be very different in amount, fine structure and chemical constitution. The epicuticular wax layer will have a strong influence on the wettability of the leaf surface during mechanical inoculation. 2. Nature of the infectible sites Virus placed on an intact leaf surface cannot infect directly and w o u n d s must be made that break through the inert leaf surface. It is possible that some types of cell on the leaf surface are more susceptible to w o u n d i n g than others. It has been shown directly, by microinfection methods that virus can be introduced into leaf hairs, but the infection rate was low (Zech, 1952). It seems probable that all the cell types making up the epidermis are potentially capable of being infected by mechanical inoculation, the cuticle being probably the major barrier to infection. Wounds that penetrate right through
Isotropic
~
Opt negative Isotropic Opt positive
~ .~ ~
~
~'
~
~
~ ~
~
~ ~
~ ~
Epicuticular wax Cuticle Cutinized wall With e m b e d e d wax Pectin layer Cellulose cell wall Epidermal cells Palisade layer
Fig. 12.5 The barriers to virus infection. Diagrammatic representation of the epicuticle of the plant seen in cross-section. The lines dividing the layers above the epidermal cells indicate regions of major change in the construction of components rather than sharp boundaries. Individual plant species may depart greatly from this general arrangement. From Eglinton and Hamilton (1967), with kind permission of the copyright holder, 9 The American Association for the Advancement of Science.
II.
F A ( ; T O R S I N F L U E N ( . ' I N G T I l E C O U R S E ('~F INFE(~TION A N D D I S E A S E
543
Fig. 12.6 Scanning electron micrographs of the surface of Nicotiana glutinosa leaves before and after mechanical inoculation. (Top) An untreated leaf showing intact leaf hairs and epidermis. (Bottom) the broken hairs following mechanical inoculation. The particles of the carborundum abrasive may be clearly seen. Bar marker 0.1 nm. (Courtesy of M.J.W. Webb.) 3. Evidence from the infectivity dilution curve The relation b e t w e e n n u m b e r s of local lesions p r o d u c e d a n d the dilution of the i n o c u l u m is discussed in relation to the assay of viruses in Chapter 15. Various theoretical m o d e l s have been d e v e l o p e d in a t t e m p t s to explain the nature of the dilution curve (Fulton, 1962; F u r u m o t o and Mickey, 1967; Gokhale and Bald, 1987). Dilution curves for AMV (requiring t h r e e particles) a n d T N V ( a s s u m e d to
require only one particle) are s h o w n in Fig. 15.2 for d a t a o b t a i n e d in the s a m e h o s t species. A l t h o u g h the difference b e t w e e n the t w o curves in Fig. 15.2 is quite clear, a dilution curve on its o w n c a n n o t be u s e d to decide w h e t h e r a viral g e n o m e is h o u s e d in one, two or three particles. There are m a n y other factors that cause the slope of the curve to v a r y from e x p e r i m e n t to e x p e r i m e n t (see Section II.A).
544
12 TRANSMISSION 2: MECHANICAL, SEED, POLLEN AND EPIDEMIOLOGY
4. Lifetime of infectible sites following wounding The time for which wounds remain infectible has usually been studied by dipping leaves into inoculum at various times after abrasion and counting the number of lesions that subsequently develop. Generally speaking, the number of infectible sites falls off very rapidly after abrasion. For example, Furumoto and Wil dman (1963a) found that about 70 % of the sites on N. glutinosa leaves that were susceptible to infection with TMV by dipping 2 seconds after wounding had lost their susceptibility by 90 seconds. The remaining sites lost their susceptibility much more slowly over a period of about 1 hour. However, not all viruses and hosts follow this pattern. Jedlinski (1964) found that during the first 10 minutes after abrasion the number of infectible sites could either increase, decrease or remain constant, depending on the host-virus system tested. There are many treatments that, applied to the leaf before or after wounding, can alter the course of events (see Section II.A). E. N u m b e r of particles r e q u i r e d to give an i n f e c t i o n There are three aspects to consider concerning the number of particles required for infection. 1. Number of particles to give one successful infection Mechanical inoculation of leaves has been widely regarded as a very inefficient process. Various estimates suggested that between about 1 0 4 and 107 virus particles have to be applied to the leaf for each local lesion that subsequently develops (e.g. Walker and Pirone, 1972a). With multi-particle viruses the requirement for more than one particle makes the inoculation process even less efficient than for single-particle viruses. For CPMV (two component) 10~' to 10s particles were applied mechanically for each lesion produced (van Kammen, 1968) and for AMV (three component) this number was 10s to 10 l~ (van Vloten-Doting, 1968). In most estimates of the efficiency of mechanical inoculation, relatively large volumes of inoculum were applied to the test leaves and the calculation involved determining the number of particles applied per lesion produced.
However, if a very small volume of inoculum is applied using a series of dilutions of the virus, the limiting dilution at which an infection is obtained gives a much lower estimate. Thus, Walker and Pirone (1972b) found that about 450 TMV particles in 2.5 gl of inoculum were sufficient to infect a tobacco plant. Using even smaller volumes of inoculum (0.1-1.0 gl) as few as 10 to 30 particles of TYMV were required to produce a single local lesion in Chinese cabbage (Fraser and Matthews, 1979a). These very substantial increases in efficiency of the mechanical inoculation process may be due to three effects: (1) for a given number of virus particles applied, the smaller the volume, the higher the virus concentration; (2) a virus particle in a very small volume will have a greater probability of finding an infectible site in a short time than one in a large volume; and (3) the fact that rapid drying increases the number of local lesions (see Section I.B) may also be a factor. Two other methods have been used to determine efficiencies of infection: inoculation of protoplast suspensions and micro-injection of cells. The number of particles adsorbed to each viable cell for each successful infection can be measured for inoculated protoplasts, and is usually about 0.1-1.0% of the applied inoculum. Efficiency of infection of protoplasts can be expressed as the average number of virus particles adsorbed per protoplast to give infection in one-half of the protoplasts (IDs0). Values of the IDs0 calculated from published data for several viruses fell in the range 50-500 (Fraser and Matthews, 1979a). The numbers of particles actually supplied in the inoculum were of the order of 100 to 1000 times these numbers. Thus, the efficiency of inoculation of protoplasts may not be intrinsically any higher than mechanical inoculation of leaves. Furthermore, the amount of virus binding to protoplasts is influenced by factors such as temperature, pH, ionic strength and the presence of added compounds such as polyethylene glycol (Roenhorst et al., 1988). The number of particles actually entering a cell for each successful infection can be measured only with micro-injection. For example, Halliwell and Gazaway (1975) obtained an IDs0 of 310 TMV particles injected in I pL into single tobacco cells.
II.
FA(~TORS I N F L U E N C I N G T H E C O U R S E OF INFE(~'TION A N [ ) DISEASE
2. Proportion of particles containing an infectious genome In the absence of an efficient method for inoculation, it is not possible to obtain unequivocal estimates of the proportion of infectious particles in a virus preparation, but Furumoto and Wildman (1963b) concluded that at least one in 10 of the TMV particles in a purified preparation was infectious. The use of infectious constructs or infectious transcripts should provide a higher proportion of infectious molecules, but these unencapsidated nucleic acids are prone to degradation by nucleases before they can reach the site of initial replication. 3. Can one infectious genome give an infection? It is generally agreed that for many viruses only a single infectious particle or infection unit is needed actually to infect a cell and give rise to the visible lesion (e.g. Boxall and MacNeil, 1974). Theoretical consideration of the dilution curve is consistent with these data. Experiments with protoplasts indicate that about one TMV particle is sufficient to infect one protoplast (Takebe, 1977). Reddy and Black (1973) found that two to three WTV particles were sufficient to infect a cell in insect vector cell monolayers. However, the results from protoplasts may not be directly applicable to inoculation of whole leaves. In practice, a large number of viral particles is needed for each successful infection, and a major factor is almost certainly the inefficiency of the process as it is usually performed. There are probably several reasons for this: (1) it is very likely that only a very small proportion of epidermal cells have potentially infectible w o u n d s made in the leaf surface above them; (2) the lifetime of the infectible w o u n d s is short, and many of the viral particles in the fluid above a w o u n d never make contact with the site; (3) the distribution of the virus applied over the leaf surface is probably very uneven in terms of surface areas of the size of single cells; (4) much of the virus may be adsorbed to inactive sites on the leaf surface and remain trapped there; and (5) some viral particles may enter potentially infectible cells, but not become successfully established, unless 'rescued' in some way. The existence of such cen-
545
ters has been shown for some host-virus combinations (Rappaport and Wu, 1963).
E Mechanical transmission in the field Compared with transmission by invertebrate vectors or vegetative propagation, field spread by mechanical means is usually of minor importance. However, with some viruses it is of considerable practical significance. TMV can readily contaminate hands, clothing and implements, and be spread by workers and, for instance, birds in tobacco and tomato crops (Broadbent, 1963, 1965a; Broadbent and Fletcher, 1963). This is particularly important during the early growth of the crop, for example during the setting out of plants. Plants infected early act as sources of infection for further spread either during cultural operations as disbudding or by rubbing together of healthy and infected leaves by wind. TMV may be spread mechanically by tobacco smokers, because the virus is commonly present in processed tobacco leaf. For example, all 37 brands of cigarette sold in West Germany contained TMV (Wetter, 1975) and the virus was detected in 60 of 64 smoking tobaccos (Broadbent, 1962). PVX can also be readily transmitted by c o n t a m i n a t e d i m p l e m e n t s or machines and by workers or animals that have been in contact with diseased plants (Todd, 1958). On some materials, the virus persists for several weeks. PVX can spread either by direct leaf contact between n e i g h b o ring plants (Clinch et al., 1938) or between n e i g h b o r s when leaves are not in contact. This has been assumed to be due to mechanical inoculation by contact between roots, but a soil-inhabiting vector has not been excluded. Some viruses of fruit trees have been shown to be spread in orchards by cutting tools (e.g. Wutscher and Shull, 1975). Field trials have suggested that RCMV can be readily transmitted by mowing machines (Ryd6n and Gerhardson, 1978). On the other hand, Heard and Chapman (1986) concluded that in m o w n ryegrass swards most of the local spread of RGMV was due to transmission by mites rather than the mowing operation.
546
12 TRANSMISSION 2: MECHANICAL, SEED, POLLEN AND EPIDEMIOLOGY
G. Abiotic transmission in soil 1. Above ground Allen (1981) showed that, in glasshouses, TMV could be transmitted from soil containing the virus coming into contact with leaves. Some other stable viruses may well be transmitted in this way. 2. Below ground Several viruses may infect roots from viruscontaminated soil apparently without fungi or arthropods being involved, for example TMV (Broadbent, 1965b), TBSV (Kleinhempel and Kegler, 1982), CRSV (Brown and Trudgill, 1984) and SBMV (Teakle, 1986). Levels of infection of glasshouse tomatoes from ToMV-containing plant debris was affected by inoculum concentration (Pares et al., 1996). As the inoculum concentration decreased, infection levels decreased and a greater proportion of infections were symptomless or were restricted to the roots.
The excretion of viruses in the guttant from a range of plants points to another possible route of virus infection. McKinney (1953) suggested that h u m a n activity in pastures could transmit BMV between plants. The recent finding (Ding et al., 2000) that BMV is present in guttation fluid from infected barley plants suggests a source of virus for mechanical transmission by the activities of humans and other animals. Plant pathogenic bacteria are transmitted through guttation fluid that is w i t h d r a w n into the plant through hydathodes (see Carlton et al., 1998; Hugouvieux et al., 1998). However, it is not known whether virus infection could result from entry by this route as the virus particles would have to find a means of movement from the apoplast or xylem vessels to the cytoplasm of susceptible cells.
III. DIRECT PASSAGE IN LIVING HIGHER PLANT MATERIAL A. Through the seed
H. Summary and discussion Mechanical inoculation is used for many purposes in plant virology such as host range studies, assessing the infectivity of a virus preparation and diagnosis of a virus. The outcome of the inoculation is, to a great extent, dependent upon the condition of the plants being inoculated. As a generality, the inoculated plants should be horticulturally 'soft'. This does not necessarily apply to natural conditions, but it is only a few viruses that are transmitted mechanically in the field. These viruses, such as TMV and PVX, occur at high concentrations in infected plants and can be spread to adjacent plants through broken hairs. We should also remember that h u m a n s are very effective at transmitting viruses (and viroids) through cutting tools and through various agronomic practices. As described in Section II.A.3 there is diurnal variation in the susceptibility of plants to viral infection. Plants have circadian expression of some genes (see Beator and Kloppstech, 1996) and it would be interesting to know whether there was any link between this and their diurnal susceptibility.
About one-seventh of the known plant viruses are transmitted through the seed of at least one of their infected host plants. Seed transmission provides a very effective means of introducing virus into a crop at an early stage, giving randomized foci of primary infection throughout the planting. Thus, when some other method of transmission can operate to spread the virus within the growing crop, seed transmission may be of considerable economic importance. Viruses may persist in seed for long periods so that commercial distribution of a seed-borne virus over long distances may occur. Table 12.1 lists the approximate frequency with which seed transmission has been found among some viruses of various groups and among viroids. Two general types of seed transmission can be distinguished. With TMV in tomato, seed transmission is largely due to contamination of the seedling by mechanical means. This type of transmission may occur with other tobamoviruses. The external virus can be readily inactivated by certain treatments eliminating all, or almost all, seed-borne infection. A small but
III. I)IRECT PASSA(}E IN LIVING tilGHER PLANT MATERIAL
547
TABLE 12.1 Relative importance of seed transmission for viruses of various virus groups Virus group
No. of members In group
Alfamovirus Bromovirus Capillovirus Carlavirus Carmovirus Caulimoviridae Closterovirus Comovirus Cryptovirus Cucumovirus Dianthovirus Enamovirus Fabavirus Geminivirus Hordeivirus Ilarvirus Luteovirus Marafivirus Nepovirus
Plant reovirus Potexvirus Potyvirus Rhabdovirus Sobemovirus Tenuivirus Tobamovirus Tobravirus Tospovirus Tombusvirus Tymovirus
Viroids
Seed-borne
1 6 4
1 1 1
60 18 34 28 15 31
2 2 ic 1 6 31
3 5 1 4
3 0 1 0
102
1
4
1
17
8
7 3
0 0
40 14 36 179 15 14 11 17
17 0 4 16 1 4 0 7
3
3
13 13 23 15
1 1 3 5
Type of potential injury ~ A
B
C
+ +
+ +
+ +
+
+
4.
+
+
+
D
Seed transmission E
F
(ro)b 1-23 4. 1-60
2-90 10-40 Up to 100 4. 1-90 100 S-labeled sulfate is about 3.7 • 10s Bq (10 mCi) per plant. Using these amounts, a purified virus is obtained containing roughly 1000cpm/~tg counting with an efficiency of about 5% for both labels. Most methods of viral isolation from infected tissue lead to losses of some virus. The extent of such losses can be estimated by adding a very small known amount of radioactive virus to the starting material and estimating the loss of radioactivity as the isolation proceeds (e.g. Fraser and Gerwitz, 1985).
1. Antibodies The term 'immunoglobulin' is often used interchangeably with 'antibody'. However, strictly an antibody is a molecule that binds to a k n o w n antigen, whereas immunoglobulin refers to this group of proteins irrespective or not of whether their binding target is known. Antibodies are secreted by B lymphocytes. They are a large family of glycoproteins that share key structural and functional features. Structurally they are composed of one or more copies of a characteristic unit that can be visualized as forming a Y shape (Fig. 15.6). Each Y contains four polypeptides: two identical copies of the heavy chain and two identical copies of the light chain joined by disulfide bonds. Antibodies are divided into five classes, IgG, IgM, IgA, IgE and IgD, based on the number of Y-like units and the type of heavy-chain polypeptide they contain (Table 15.1) IgG molecules have three protein domains. Two of the domains, forming the arms of the Y, are identical and are termed the Fab domain. They each contain an antigen-binding site at the end, making the IgG molecule bivalent. The third domain, the Fc domain, forms the stem of the Y. The three domains may be separated from one another by cleavage with the protease papain. The Fc region binds protein A, a protein obtained from the cell wall of Staphylococcus aureus, with very high affinity. This property is used in several serological procedures.
IV. M E T H O D S D E P E N D I N G ON PROPERTIES OF VIRAL PROTEINS Some of the most important and widely used methods for assay, detection and diagnosis depend on the surface properties of viral proteins. For most plant viruses, this means the protein or proteins that make up the viral coat. Different procedures may use the protein in the intact virus, the protein subunits from disrupted virus or proteins expressed from cloned cDNA or DNA in a system such as Escherichia coli or insect cells. More recently non-structural proteins coded for by a virus have been used in diagnosis. A.
Serological procedures
Serological procedures are based on the interaction between a protein or proteins (termed the antigen) in the pathogen with antibodies raised TABLE 15.1 The five classes of immunoglobulins Characteristics Heavy chain Light chain Molecular formula Valency Serum concentration (mg/mL) Function 'n=1,2or3.
!gG
IgM
IgA
Y
D
0~
~: or ;~ 2
~: or X s: or ;~ (!%K2)sor (1%;(2)5 (% K2),, or (0t 2 )re),' 10 2,4or6
K or Z 82K2or c2Z2 2
8-16 Secondary response
0.5-2 Primary response
0.01-0.4 Protects against parasites?
Y2K2 or
72Z2
641
1-4 Protects mucous membranes
IgE 8
IgD 8 ~ork _
_
2 0-0.4
642
~5 M E T H O D S FOlK ASSAY, I)ETECTION A N D D I A G N O S I S
The N-terminal regions of both the light and heavy chains in the arms of the Y-shaped IgG molecule comprise very heterogeneous sequences. This is k n o w n as the variable (V) region. The C-terminal region of the light chains and the rest of the heavy chains form the constant (C) region. The V regions of one heavy and one light chain combine to form one antigen-binding site. The heterogeneity of the V regions provides the structural basis for the large repertoire of binding sites used in an effective immune response. For more details of the variation and how it arises see Harlow and Lane (1988). 2. Antigens Antigens are usually fairly large molecules or particles consisting of, or containing, protein or polysaccharides that are foreign to the vertebrate species into which they are introduced. Most have a molecular weight greater than 10 kDa, although smaller peptides can elicit
antibody production. There are two aspects to the activity of an antigen. First, the antigen can stimulate the animal to produce antibody proteins that will react specifically with the antigen. This aspect is known as the immunogenicity of the antigen. Second, the antigen must be able to combine with the specific antibody produced. This is generally referred to as the antigenicity of the molecule. Some small molecules with a specific structure such as amino acids may not be immunogenic by themselves but may be able to combine with antibodies produced in response to a larger antigen containing the small molecule as part of its structure. Such small molecules are k n o w n as haptens. Large molecules are usually more effective i m m u n o g e n s than small ones. Thus, plant viruses containing protein macromolecules are often very effective in stimulating specific antibody production. The subunits of a viral protein coat are much less efficient.
IV.
M E T t t O D S I)EPENDIN(~ ON P R O P E R T I E S OF V I R A L P R O T E I N S
It is specific regions on antigens, termed epitopes, that induce and interact with specific antibodies. Epitopes can be composed of amino acids, carbohydrates, lipids, nucleic acids and a wide range of synthetic organic chemicals. About 7-15 amino acids at the surface of a protein may be involved in an antigenic site. However, there are difficulties in defining such sites precisely. These are discussed by van Regenmortel (1989b).
a. Type of epitope There are several different grouping of epitopes. In one, the continuous epitope is a linear stretch of amino acids that is distinguished from the discontinuous epitope, which is formed from a group of spatially adjacent surface residues brought together by the folding of the polypeptide chain or from the juxtaposition of residues from two or more separate peptide chains. In another grouping system there are cryptotopes, which are hidden epitopes revealed only on dissociation or denaturation of the antigen, neotopes formed by the juxtaposition of adjacent polypeptides (e.g. adjacent viral coat protein subunits), metatopes, which are epitopes present on both the dissociated and polymerized forms of the antigen, and neutralization epitopes, which are specifically recognized by antibody molecules able to neutralize the infectivity of a virus. The conformational specificity of viral epitopes is discussed by van Regenmortel (1992). The epitope type giving rise to a monoclonal antibody (MAb) and the relative proportions of different epitopes recognized by a polyclonal antiserum can affect the outcome of certain serological tests. For instance, an antibody to a cryptotope is unlikely to recognize an antigen in DAS-ELISA but is likely to in a Western blot. It has been suggested that if the surface area of a protein that is recognized by an antibody molecule is about 2.0 • 2.5 nm, then all protein antigenic determinants are likely to be discontinuous (Barlow et al., 1986). Antigenic determinants specific for quaternary structure in the virus particle have been demonstrated for all the groups of viruses (van Regenmortel, 1982). The existence of such determinants in the TMV
643
rod was confirmed by the finding that, of 18 MAbs raised against the virus, eight did not react with viral subunits in ELISA tests (Altschuh et al., 1985). The molecular dissection of antigens by monoclonal antibodies has been discussed by van Regenmortel (1984b). Three methods have been used to localize antigenic determinants: (1) peptides obtained from the protein by chemical or enzymatic cleavage are screened for their reactivity with antibodies; (2) short synthetic peptides representing k n o w n amino acid sequences in the protein can be similarly screened; and (3) immunological crossreactivity is assayed between closely related proteins having one or a few amino acid substitutions at known sites. There are substantial limitations for all these procedures. Peptides derived by cleavage or synthesis may not maintain in solution the conformation that the sequence possessed in the intact molecule. Furthermore, such peptides will only rarely represent all the amino acids in the original antigenic determinant. For these reasons reactivity of peptides with antibodies to the intact protein is usually very low. The limitations of the peptide approach can be further illustrated from work with TMV. Tests with a set of synthetic and cleavage peptides indicated that the dissociated TMV protein possesses seven antigenic determinants (Altschuh et al., 1983). Work with a set of synthetic peptides representing almost the full length of the protein showed that almost the entire sequence possesses antigenic activity (A1-Moudallal et al., 1985). 3. Interaction between antibodies and antigens The interaction between an epitope and antibody is affected by both affinity and avidity.
a. Affinity Affinity is a measure of the strength of the binding of an epitope to an antibody. As this binding is a reversible bimolecular interaction, affinity describes the a m o u n t of a n t i b o d y antigen complex that will be found at equilibrium. High-affinity antibodies perform better in all immunochemical techniques, not only
644
15 METHODS FOR ASSAY, DETECTION AND DIAGNOSIS
because of their higher capacity but also because of the stability of the complex. b. Avidity
Avidity is a measure of the overall stability of the complex between antibodies and antigens, and is governed by three factors: the intrinsic affinity of the antibody for the epitope, the valency of the antibody and antigen, and the geometric arrangement of the interacting components. c. Titer
Titer is a relative measure of the concentration of a specific antibody in an antiserum. It is often used to define the dilution endpoint of the antiserum for detection of an antigen. Thus, as the sensitivities of various serological tests differ, the apparent titer is applicable only to the test under discussion. d. Serological differentiation index The serological differentiation index (SDI) is a measure of the serological cross-reactivity of two antigens. It is the number of 2-fold dilution steps separating the homologous and heterologous titers. The homologous titer is that of the antiserum with respect to the antigen used for immunizing the animal, while the heterologous titer is that with respect to another related antigen. Because of the variation in the response of animals to antigens, the SDI is reliable only if it is the average of several measurements. It is measured by a variety of techniques such as ELISA and by single radial immunodiffusion (Krajacic et al., 1992). 4. Types of antisera There are two basic types of antisera: polyclonal, which contain antibodies to all the available epitopes on the antigen, and monoclonal, which contain antibodies to one epitope. There is much discussion as to which is the best for diagnosis but this will depend on what question the diagnostician is addressing. Monoclonal antisera are much more specific than polyclonal antisera and can be used to differentiate strains of many pathogens. There are some examples of a monoclonal antiserum that is virus group
specific (e.g. potyviruses). On the other hand specificity can be a disadvantage and a variant of the pathogen may not be detected. 5. Production of antisera Antisera have been produced against plant viruses in a variety of animals. Rabbits have been used most often because they respond well to plant virus antigens, are easy to handle and produce useful volumes of serum. Individual animals may vary quite widely in their response to a particular antigen. The amount and specificity of antibody produced in response to a given plant virus antigen may be determined genetically. Unstable plant viruses may have their immunogenicity increased by chemical stabilization (e.g. Francki and Habili, 1972). Protocols for the production of antisera and for the removal of any antibodies reacting with host antigens are given by van Regenmortel (1982), Hampton et al. (1990) and van Regenmortel and Dubs (1993). A recent new approach is to inject a cDNA to the antigen protein sequence cloned in a mammalian expressing vector. Using this approach, an antiserum was made to BYDV (Pal et aI., 2000). Although mice provide relatively small volumes of antiserum, their use may be an advantage when very small amounts of viral antigen are available. Highly inbred strains of mice also minimize variation in response between animals. Mice are, of course, used in the production of monoclonal antisera, as are rats. Chickens are convenient animals to use for the production of polyclonal antisera. When laying hens are immunized, large quantities of purified IgG can be obtained from the egg yolks in a relatively short time (van Regenmortel, 1982; Polson et al., 1985; Bollen et al., 1996; Schade et al., 1997). Purified IgG from rabbit antisera can be obtained by isolating a specific precipitate of virus and IgG, dissociating virus and IgG at low pH, and removing the virus by high-speed centrifugation. Alternatively, to avoid lengthy acid treatment, antibodies can be bound to virus made insoluble by polymerization with glutaraldehyde. This procedure requires only a short low-speed centrifugation after acid treatment to remove the virus (Maeda and Inouye, 1985).
IV. M E T H O D S DEPENDING ON PROPERTIES OF VIRAL P R O T E I N S
It was assumed for a long time that plants could not produce antibodies. However, as described in Chapter 16 (Section VII.C.3), antibodies can be produced in transgenic plants. 6. Monoclonal antibodies B lymphocytes cannot be cultured in vitro. To overcome this problem, KOhler and Milstein (1975) took B lymphocytes from an immunized mouse and fused these in vitro with an 'immortal' mouse myeloma cell line. Selection of appropriate single fused cells gave 'hybridomas' producing only a single kind of a n t i b o d y - a monoclonal antibody. Since monoclonal antibodies against TMV were described (Briand et al., 1982; Dietzgen and Sander, 1982), there has been an explosive growth of interest in this type of antibody for many aspects of plant virus research, and particularly for detection and diagnosis. The term monoclonal antibody has been variously abbreviated as MAb, McAb, MA and McA. I will use the first of these. By 1986, MAbs had been prepared against more than 30 different plant viruses (van Regenmortel, 1986), and many additions to the list have since been reported. a. The nature of antibody specificity in relation to MAbs
The binding site of a MAb, or any individual antibody in a polyclonal serum, is able to bind to different antigenic determinants with varying degrees of affinity, that is, an individual binding site on an antibody is polyfunctional. Furthermore, antibodies against a single antigenic determinant (or epitope) may vary in affinities from one that is scarcely measurable with standard techniques to one that is 100 000fold higher. Thus, it is quite possible for an antibody to bind more strongly to an antigenic determinant differing from the one that stimulated its production. Such a phenomenon may well be obscured by the many different antibodies in a polyclonal antiserum, but with MAbs it will show up clearly. For example, when MAbs were produced following immunization of a mouse with a particular strain of TMV it was found that some of the MAbs
645
reacted more strongly with other strains of the virus (A1-Moudallal et al., 1982; van Regenmortel, 1982). This has important implications for the use of MAbs in the delineation of virus strains (Chapter 17, Section II.B.3). As individual MAbs react with a specific epitope, a panel of MAbs will have ones that react against the different types of epitope. For instance, Dore et al. (1988) distinguished MAbs that reacted against dissociated TMV rods (probably cryptotopes) and those that reacted against assembled virions (probably neotopes). Electron microscopy showed that the neotopes reacted with the entire surface of the virion, whereas the cryptotopes reacted only with the extremities of the viral rod and with disc aggregates of the protein. Another aspect of the specificity of MAbs must be borne in mind. A particular protein antigen may have only one site for binding a particular MAb. If this single site is shared with another protein, a significant cross-reaction could occur even in the absence of general structural similarity between the two proteins. This may be an u n c o m m o n phenomenon. Nevertheless, cross-reactions detected with MAbs cannot be taken to indicate significant structural or functional similarity between two proteins without other supporting evidence (Carter and ter Meulen, 1984). b. Production of MAbs
Technical details for the production of MAbs directed against plant viruses are given by Sander and Dietzgen (1984), Harlow and Lane (1988) and Hampton et al. (1990). The use of rats is described by Torrance et al. (1986a,b). An outline of the procedure using mice is shown in Fig. 15.7. The kind of screening test used to test for MAb production is of critical importance (van Regenmortel, 1986). Most workers use some form of ELISA. Multi-layered sandwich procedures, especially using chicken antibody in one of the layers, appeared to be the most sensitive (A1-Moudallal et al., 1984). The results obtained depend very much on the quality of the reagents used and the exact conditions of pH and other factors in the medium.
646 15 METHODSFOR ASSAY, DETECTION AND DIAGNOSIS I.
Inject mouse with viral antigen.
x Spleen cell suspension
II.
Cultured myeloma cells-will not survive in medium containing hypoxanthine, thymidine and aminopterin (HAT).
Fuse spleen cells with myeloma cells and transfer to 96-well plates.
R)-o-o-o-o-crW~q
In HAT medium: Myeloma cells die. Spleen cells alone do not survive in culture. Only hybridoma cells grow.
Ill.
Observe hybridoma cells growing. Test culture supernatants for antibody of required specificity.
IV.
Clone and amplify the cell cultures that are selected.
Reclone the hybridoma to ensure single clones. Freeze a sample of cells to ensure the cells are not lost. Store culture supernatants as antibody sample of low titre.
Grow cells as ascitic tumour.
The ascitic fluid contains monoclonal antibody of very high titer.
V.
7. Single-chain antibodies and phage displays The bringing together of two technologies has resulted in the ability to produce a large range of molecules with the properties of single-type antibodies similar to MAbs w i t h o u t involving injection of animals. These technologies are the production of single-chain antibodies and the display of recombinant proteins on the surface of bacteriophage (phage display). The antibody repertoire of a vertebrate is estimated to consist of more than 10 s different antibodies, the range being given by variation in the sequence of the variable regions of the
Fig. 15.7 Steps in the production of monoclonal antibodies. From Matthews (1991).
h e a v y (V~) and light (Vl) i m m u n o g l o b u l i n chain (Fig. 15.6). Forced cloning of cDNA produced by RT-PCR of the sequences encoding the VH and VJ chain d o m a i n s allowed amplification of the a n t i b o d y repertoire e n c o d i n g sequences (Orlandi et al., 1989; Sastry et al., 1989). The two variable chains could be expressed in E. coli as a single-chain variable f r a g m e n t (scFv) fusion protein by joining t h e m with a flexible linker protein (Huston et al., 1988). Cloning of r a n d o m l y combined VH and VL sequences gave a pool of scFv-encoding genes, allowing the generation of large combi-
IV.
METttr
natorial a n t i b o d y libraries with a diversity c o m p a r a b l e to the n a t u r a l i m m u n e repertoire (Huse et al., 1989; M a r k s et al., 1991). P h a g e display d e p e n d s u p o n the insertion of c o d i n g sequences into a p h a g e structural gene which, if it does not d i s r u p t an essential function of the gene product, will be d i s p l a y e d on the surface of the viral particle. If the insert is a set of r a n d o m sequences, the resulting particles will present a library of peptides, each disp l a y e d on a viral scaffold (reviewed in Rodi a n d M a k o w s k i , 1999). The m o s t c o m m o n l y u s e d p h a g e for display are the f i l a m e n t o u s M13 a n d fd. These p h a g e h a v e m i n o r structural proteins, usually at the e n d s of the particles as well as the major coat proteins (see Gao et al., 1999). Both these type of protein h a v e b e e n u s e d to d i s p l a y r e c o m b i n a n t proteins. In b r i n g i n g these t w o technologies together, the repertoire of scFv molecules is d i s p l a y e d on
I~EPENI)IN(~ ()N I'R()PEI~,TIES OF V I R A l . I ' R ( ) T E I N S
647
the surface of the p h a g e (see McCafferty et al., 1990). Those p h a g e d i s p l a y i n g a n t i b o d i e s to the a n t i g e n of interest, such as a viral coat protein, are selected by b i n d i n g to the a n t i g e n i m m o b i lized in, say, a microtiter plate ( t e r m e d panning). The b o u n d p h a g e are then released by elution a n d u s e d to infect E. coli (Fig. 15.8). This cycle of p a n n i n g , e l u t i o n a n d i n f e c t i o n is r e p e a t e d several times until antibodies w i t h the desired affinity are obtained. The Fv c D N A library does not necessarily h a v e to be f r o m a n i m a l s that h a v e an i m m u n e r e s p o n s e to the a n t i g e n of interest; if they are f r o m u n i m m u nized a n i m a l s they are t e r m e d naive.
B. Methods for detecting antibody-virus combination A w i d e v a r i e t y of m e t h o d s has been d e v e l o p e d for d e m o n s t r a t i n g a n d e s t i m a t i n g c o m b i n a t i o n
Fig. 15.8 Outline of antibody engineering and selection of antigen-specific antibodies by phage display technology. The minimum requirements of an antibody for antigen recognition are located in the antigen-binding site. The V~ and VL domains, joined by a linker peptide-encoding DNA fragment, are inserted into a phagemid vector as a single-chain Fv segment. In the scFv-encoding phagemid an ampicillin resistance gene, an origin of replication for E. coli and a packaging signal which is required for phage assembly, are present. Depending on the E. coli strain used, the scFvs can be expressed as soluble proteins in HB2151 bacteria, wherein translation halts at an amber stop codon (located between scFv and g3p). Fusion proteins of scFv with the minor coat protein of filamentous phage Fd are obtained from TG1 bacteria in which translation proceeds at the amber codon. Selection of BNYVV-specific scFV-antibodies from a combinatorial antibody library can be achieved using phage display. Helper phage, which contain the entire phage genome but lack an efficient packaging signal, are used to 'rescue' phagemids from a combinatorial antibody library. When both helper phage and phagemid are present within the same bacterium, phage antibodies are assembled which carry scFv antibodies on their surface and contain the scFv-encoding phagemid vector. Consequently, within PhAbs, the genotype and phenotype are linked. To select for antigen specificity, PhAbs, rescued from a combinatorial antibody library, are allowed to bind to immobilized BNYVV (panning). Washing removes the PhAbs that lack affinity for BNYVV. Bound PhAbs are eluted and the selected PhAbs are applied to sequential rounds of panning until the desired affinity is obtained. From Griep et al. (1999), with kind permission of the copyright holders, 9 Kluwer Academic Publishers.
648
~5 M E T H O D S FOR ASSAY, D E T E C T I O N A N D D I A G N O S I S
between antibodies and antigens. Some of these, such as complement fixation and anaphylaxis, have rarely been used with plant viruses and will not be further discussed. Most traditional methods for using antisera with plant viruses involved direct observation of specific precipitates of virus and antibody, either in liquid media or in agar gels. Over about the past 20 years or so, these methOds have been progressively superseded by the use of the enzyme-linked immunosorbent assay (ELISA), immunosorbent electron microscopy (ISEM) and 'dot blots' employing either polyclonal or monoclonal antibodies. These are described in more detail in subsequent sections. 1. Advantages and limitations Serological tests have provided rapid and convenient methods for the identification and estimation of plant viruses, the main advantages being: (1) the specificity of the reaction allows virus to be measured in the presence of host material or other impurities; (2) results are obtained in a few hours or overnight compared with days, or even weeks, for infectivity assays; (3) the methods give an answer that is directly proportional to viral concentration over a wide range of concentrations; (4) some serological detection and assay procedures are more sensitive than infectivity measurements; (5) serological tests are particularly useful with viruses that have no good local lesion host or that are not sap transmissible; and (6) antisera can be stored and comparable tests made over periods of years and in different laboratories. It must be borne in mind that serological tests detect and measure the virus protein antigen, not the amount of infective virus. This fact may, of course, be used to advantage in some situations. The possibility of chance cross-reactions must be considered. Just a few amino acids in a protein sequence are able to elicit an antibody combining with that sequence, provided it is suitably displayed at the surface of the protein. Thus, it is theoretically possible that two quite unrelated protein antigens might elicit crossreacting antibodies purely because of a chance short amino acid sequence similarity. Such similarity has been established between the coat
protein of the U1 strain of TMV and the large subunit of the host protein ribulose bisphosphate carboxylase (Dietzgen and Zaitlin, 1986). There was no cross-reaction with distantly related TMV strains or other viruses. Aside from forming the basis for a range of assay and diagnostic methods, serological reactions can be used in a variety of other ways, such as investigating virus structure (see Chapter 5), virus relationships (see Chapters 2 and 17) and virus activities in cells (see Chapter 10). Technical details of methods are given by van Regenmortel (1982), and van Regenmortel and Dubs (1993). 2. Precipitation methods a. Precipitation in liquid media The formation of a visible specific precipitate between the antigen and antibody is one of the most direct ways of observing the combination between antibody and virus, but relatively high concentrations of reagents are needed. Plant viruses are polyvalent, that is, each particle can combine with many antibody protein molecules. The actual number that can combine depends on the size of the virus antigen. Polyvalent antigen combines with divalent antibody to form a lattice-structured precipitate. The precipitation reaction of a virus with a particular antiserum is clearly delineated if a series of tests are made with 2-fold dilutions of both reagents in all combinations, under standard conditions of temperature and mixing (Matthews, 1957). In a modification of the precipitation reaction in tubes, k n o w n as a ring test, a small volume of undiluted or slightly diluted serum is placed in a small glass tube and overlaid carefully with the virus antigen solution. With time, antibody diffuses into the virus solution and virus diffuses into the antiserum. Somewhere near the boundary, a zone of specific precipitate will form, provided both reagents are sufficiently concentrated. It is a useful quick test, but somewhat insensitive. When a drop of freshly expressed leaf sap from plants infected with viruses occurring in high concentration is mixed on a microscope slide with a drop of antiserum, clumping of
IV.
M E T H O D S D E P E N D I N G ON PROPERTIES OF VIRAL P R O T E I N S
small particles of host material occurs. This may be seen with the naked eye but is viewed more readily with a hand lens or low-powered microscope. Chloroplasts and chloroplast fragments are prominent in the clumped aggregates. However, some viruses such as WSMV may undergo rapid antigenic modification in leaf extracts unless glutaraldehyde fixation is used (Langenberg, 1989). b. Precipitation methods using virus bound to larger objects
The sensitivity of the precipitation reaction depends on the smallest amount of antigen that will form a visible precipitate. The smaller the antigen, the greater the weight of antigen required. Various methods have been developed in which the virus is adsorbed to particles substantially larger than viruses, such as latex or red blood cells, before reaction with antiserum. Compared with simple precipitation tests, these procedures allow the detection of 100- to 1000-fold smaller quantities of virus (van Regenmortel, 1982). Latex agglutination allowed the detection of between 0.1 and 0.5 btg/mL of six rod-shaped viruses infecting legumes (Demski et al., 1986). A simple field kit using latex agglutination for three viruses infecting potatoes allowed results to be obtained within 3-10 minutes (Talley et al., 1980). Coating of latex particles with protein A has been used to increase the sensitivity of the procedure (Torrance, 1980). Reverse passive hemagglutination (RPH) is a procedure in which antibody is linked non-specifically to red blood cells by some appropriate treatment. A solution containing virus is added to a suspension of the antibody-linked red blood cells in round-bottom microtiter plates. In a negative test, the cells settle as a compact button. Agglutination occurs in a positive test, giving a shield of cells over the bottom of the tube. Sander et al. (1989), using TMV as a model system, have shown that RPH is equal in sensitivity and specificity to double-antibody sandwich ELISA. The method involves only one step; readings can be made by eye within 90 minutes without the need of equipment; and
649
appropriately stabilized red blood cells can be stored for long periods. The protocol for a chloroplast agglutination test is described by Dijkstra and de Jager (1998). c. Immunodiffusion reactions in gels
The great advantages of immunodiffusion reactions carried out in gels are: (1) mixtures of antigenic molecules and their corresponding antibodies may be physically separated, either because of differing rates of diffusion in the gel or because of differing rates of migration in an electric field (in immunoelectrophoresis), or by a combination of these factors; and (2) direct comparisons can be made of two antigens by placing them in neighboring wells on the same plate. The method may not be as sensitive as the tube precipitation method in terms of a detectable concentration of virus, but, of course, much smaller volumes of fluid are required. In early work, immunodiffusion tests were carried out in small tubes (i.e. in one dimension). However, this type of test has been superseded by double diffusion tests in two dimensions on glass slides. Ouchterlony (1962) gives a general account of these methods. Wells are punched in the agar or agarose gel in a defined geometrical arrangement. It is usual in double diffusion tests to have the antiserum in a central well and the antigen solutions being tested in a series of wells surrounding the central well. However, if a single antigen sample is being tested against several antisera, say in virus identification, the sample can be placed in the central well. Antigen and antibody diffuse toward each other in the gel, and after a time a zone will form where the two reagents are in suitable proportions to form a precipitating complex. Both reactants leave the solution at this point and more diffuses in to build up a visible line of precipitation that traps related antigen and antibody. Unrelated antigens or antibodies can pass through the band of precipitation. Bands can be recorded by direct visual observation with appropriate lighting, by the use of protein stains or by photography. When radioactive virus is used, radioautography can be used to detect the bands.
650
15 M E T H O D S FOR ASSAY, D E T E C T I O N A N D D I A G N O S I S
W h e n comparing bands formed by two antigens in neighboring wells, several types of patterns have been distinguished. M o v e m e n t of antigen in the agar gel is strongly d e p e n d e n t on the size and shape of the virus. For small isometric plant viruses, the m e t h o d s are satisfactory. Rods m a y diffuse slowly or not at all. For routine tests, a suitable detergent in the i m m u n o d i f f u s i o n s y s t e m m a y allow r a p i d migration in the gel of virus degradation products. Elongated virus particles can also be m a d e to diffuse by sonication. This has the advantage that the original antigenic surface of the virus survives in the sonicated f r a g m e n t s (e.g. Moghal and Francki, 1976). Serological relationships between viruses can be determined by the interactions of bands from adjacent wells. The bands from serologically identical or very closely related viruses fuse, whereas those from more distantly related viruses can form spurs (Fig 15.9). A modification of gel diffusion uses an agarlike polysaccharide from Pst'udoll~oi~as elodea (Gelrite) instead of agar. A n t i g e n s were detected more rapidly in Gelrite, and with about a 100-fold increase in sensitivity. This w o u l d make the test almost comparable with ELISA in this respect (Ohki and Inouye, 1987).
A
(A-B-C)
d. Immunoelectrophoresis in gels In i m m u n o e l e c t r o p h o r e s i s , a mixture of antigens is first separated by migration in an electric field in an agar gel containing an appropriate buffer. Antiserum is then placed in a trough parallel to the path of electrophoretic migration and an immunodiffusion test is carried out. This is a powerful method for resolving mixtures of antigens as two i n d e p e n d e n t criteria are involved: electrophoretic mobility and antigenic specificity. In i m m u n o - o s m o p h o r e s i s two sets of wells are cut in agar buffered near pH 7. Antiserum is placed in one well of each pair and a virus dilution in the other, and a voltage gradient is applied. Antibody protein moves because of endosmotic flow; the virus moves because of a net negative charge. Thus, the reagents are brought together in the gel much more rapidly than by simple gel diffusion (e.g. John, 1965).
Fig. 15.9 Possible precipitin line patterns in gel doublediffusion tests indicating the degree of serological relationship between two viruses, I and II. Arrangement of reactants is indicated in the wells. Epitopes of viral coat proteins are A, B, C, D, P, Q or R. Their linkage in coat protein molecules and virus particles is indicated by hyphen and parentheses. The antiserum is against virus I (epitopes A-B-C) and contains antibodies a, b and c, which occur separately. From Dijkstra and de Jager (1998), with kind permission of the copyright holders, 9 Springer-Verlag GmbH and Co. KG.
IV.
METtt(31)S 13EPENI)IN(} ON PROPElS, TIES OF V I R A l . P R O T E I N S
651
In rocket immunoelectrophoresis, antigen migrates in an electric field in a layer of agarose containing the a p p r o p r i a t e antibody. The migration of the antigen toward the anode gives rise to rocket-shaped patterns of precipitation. The area u n d e r the rocket is proportional to antigen concentration. This procedure has been adapted to the assay of 25-ng to 10-btg amounts of CaMV in leaf extracts from about 25 mg of tissue (Hagen et al., 1982).
various morphological types in both purified preparations and crude extracts. It is particularly convenient w h e n large n u m b e r s of tests are needed. It is very sensitive, detecting concentrations as low as 1 - 1 0 n g / m L . Detailed protocols are given in m a n y manuals and journals, including Clark and Bar-Joseph (1984), Hamilton et al. (1990), van Regenmortel and Dubs (1993) and Dijkstra and de Jager (1998). Two general procedures have been developed.
3. Radioimmune assay
a. Direct double-antibody sandwich method
R a d i o i m m u n e assay, u s i n g radioactively labeled viral antigens, provides a sensitive and specific m e t h o d for measuring plant viral antigens (Ball, 1973). In this method, viral antibody is irreversibly attached to the walls of disposable tubes. The a m o u n t s of labeled virus that bind to the surface of such tubes can then be measured. In view of the sensitivity n o w available with ELISA procedures, this m e t h o d of assay is unlikely to be widely used.
The principle of the direct double-antibody s a n d w i c h procedure is s u m m a r i z e d in Fig. 15.10. The kinds of data obtained in ELISA tests are illustrated in Fig. 15.11. This is the m e t h o d described by Clark and A d a m s (1977). It has been widely used, but suffers two limitations. First, it may be very strain specific. For discrimination between virus strains, this can be a useful feature; however, for routine diagnostic tests it means that different viral serotypes may escape detection. This high specificity is almost certainly due to the fact that the coupling of the enzyme to the antibody interferes with weaker combining reactions with strains that are not closely related. Second, this procedure requires a different antivirus e n z y m e - a n t i b o d y complex to be prepared for each virus to be tested.
4. ELISA procedures Clark and A d a m s (1977) s h o w e d that the microplate m e t h o d of ELISA could be very effectively applied to the detection and assay of plant viruses. Since that time the method has come to be used more and more widely. M a n y variations of the basic procedure have been described, with the objective of optimizing the tests for particular purposes. The method is very economical in the use of reactants and readily adapted to quantitative measurement. It can be applied to viruses of 1
2
v
v
b. Indirect double-antibody sandwich methods
In the indirect procedure, the enzyme used in the final detection and assay step is conjugated to an anti-globulin antibody. For example, if the 3
;l
4
E.~O4E
o d o of_ 9oEE 9_~.O0/
v
Fig. 15.10 Principle of the ELISA technique for plant viruses (direct double-antibody sandwich method). (1) The ~, globulin fraction from an antiserum is allowed to coat the surface of wells in a polystyrene microtiter plate. The plates are then washed. (2) The test sample containing virus is added and combination with the fixed antibody is allowed to occur. (3) After washing again, enzyme-labeled specific antibody is allowed to combine with any virus attached to the fixed antibody. (Alkaline phosphatase is linked to the antibody with glutaraldehyde.) (4) The plate is again washed and enzyme substrate is added. The colorless substrate p-nitrophenyl phosphate (~/~) gives rise to a yellow product (0), which can be observed visually in field applications or measured at 405 nm using an automated spectrophotometer. An example of the kind of data obtained is given in Fig. 15.11. Modified from Clark and Adams (1977), with permission.
652
1.2
15
METHODS
FOR ASSAY, DETECTION
AND
DIAGNOSIS
concentrations cause no problems. The broadest range of serologically related viruses is detected by indirect ELISA using unpre-coated plates, but this procedure is open to interference by crude plant sap. Pre-coating of plates with antibodies or their F(ab')2 fragments (Barbara and Clark, 1982) eliminated the interference problem but narrowed the specificity. Other forms of interference may occur. For example, roots of herbaceous plants may contain a factor that makes the use of protein A-horseradish peroxidase unsatisfactory as an enzyme conjugate (Jones and Mitchell, 1987). Other virus-coated proteins such as the cylindrical inclusion body protein produced by potyviruses can be used for diagnosis with ELISA methods (Yeh and Gonsalves, 1984).
o~ 9
!.0
E o 0.8
2 0.6-
0.4
0.2-
c. D o t E L I S A !
!
i
i
i
i
i
-
1
1
r--
o,,
i
Reciprocal of extract dilution
Fig. 15.11 Example of data obtained in an ELISA test. Detection of lettuce LNYV in Nicotiana glutinosa plants systemically infected for various periods. O, Uninfected plants; plants infected for 7 days (no systemic symptoms, A), 15 days (prominent systemic symptoms, II), 19 days (prominent systemic symptoms, : ',) and 30 days (severe chlorosis and stunting, G). From Chu and Francki (1982), with permission. virus antibodies were raised in a rabbit, a chicken anti-rabbit globulin might be used. Thus, one conjugated globulin preparation can be used to assay bound rabbit antibody for a range of viruses. Furthermore, indirect methods detect a broader range of related viruses with a single antiserum (Koenig, 1981). Many variations of these procedures are possible. Koenig and Paul (1982) studied nine of them with the aim of optimizing the tests for different objectives (Fig. 15.12). Their results emphasized the versatility of the assay for different applications. They concluded that the direct double-antibody sandwich method is the most convenient for the routine detection of plant viruses in situations where strain specificity and very low virus
Several procedures have been developed that use nitrocellulose or nylon membranes as the solid substrate for ELISA tests. In one procedure, the virus in a plant extract is electroblotted on to the membrane as the first step (Rybicki and von Wechmar, 1982). As an alternative first step, the membrane is coated with antiviral IgG by soaking in an appropriate solution (Banttari and Goodwin, 1985). For the final color development, a substrate is added that the enzyme linked to the IgG converts to an insoluble colored material. The intensity of the colored spot can be assessed by eye or by using a reflectance densitometer. This kind of assay was twice as sensitive as the test carried out in microliter plates for PLRV (Smith and Banttari, 1987). An example of a dot blot assay is given in Fig. 15.13. The dot i m m u n o b i n d i n g procedure for TMV used by Hibi and Saito (1985) was only about one-tenth as sensitive as a similar ELISA test in microliter plates. Nevertheless, 1.0 ng of virus could be detected, and its simplicity and speed (a few hours) make the test useful as a practical diagnostic technique. Graddon and Randles (1986) described a single antibody dot immunoassay for the rapid detection of SCMoV. The method was found to be 12 times as sensitive as ELISA in terms of total antigen detectable. The main advantages
IV. M E T H O D S DEPENDING ON PROPERTIES OF VIRAL PROTEINS
E
= Virus
= intact virus specitic antibody
= Enzyme
= Fab of F(ab')2fragment of virus specific antibody
653
= antibodies specific for IgG or Fc fragments (i) Black and white symbols indicate antibodies derived from two different animal species. (ii) Shaded areas indicate a reaction between the Fc portion of a virus-specific antibody and an Fc-specific antibody. (iii) Reagents added in order from bottom of diagram except where vertical bars indicate two reagents are preincubated together before addition to the plate. Direct procedures
Indirect procedures .
.
.
.
.
,t
.
.
Jk.
Short procedures with preincubation of two ingredients.
[ ]
[ Fig. 15.12 Diagrammatic representation of some variants of ELISA. Modified from Koenig and Paul (1982), with kind permission of the copyright holders, 9 Elsevier Science. of the method were speed (3 hours to complete a test), low cost and the small a m o u n t of reagents required. Dot blot immunoassays may be particularly useful for routine detection of virus in seeds or seed samples, especially for
laboratories where an inexpensive and simple test is needed (Lange and Heide, 1986). However, e n d o g e n o u s insect enzymes may interfere with tests using insect extracts (Berger et al., 1985). Dot blot i m m u n o b i n d i n g using
654
15 METHODS FOR ASSAY, DETECTION AND DIAGNOSIS
Fig. 15.13 The sensitivity of a dot immunobinding assay. V, purified SPMYEV; I, crude sap from infected leaves; H, crude sap from healthy leaves. The conjugated enzyme was alkaline phosphatase and the substrate for color development was Fast Red TR salt. From Yoshikawa et al. (1986), with permission. plain paper has been developed as a diagnostic method for five potato viruses using direct extraction from green leaves (Heide and Lange, 1988). The technique of dotting the antigen on to nitrocellulose or other papers can be used with methods other than ELISA for detection of the antigen. For example, Hsu (1984) used goldlabeled goat antirabbit IgG to detect rabbit antibodies bound to TMV on nitrocellulose paper. The gold-labeled antibody is directly visible because of its pink color. The method could detect 1-5 pg of TMV protein and is also simple and economical. The use of electrophoretic separation of proteins followed by immunoblotting (western blotting) is discussed in Section IV.I.2. Koenig and Burgermeister (1986) discuss the application of various immunoblotting techniques and consider the problem of unexplained or false-positive reactions that sometimes occur. These reactions may be due to non-specific binding of immunoglobulins to coat proteins of certain plant viruses (Dietzgen and Francki, 1987). d. Immuno-tissue printing
As an extension to dot ELISA, immuno-tissue printing involves applying the cut surface of a leaf or stem (or any other plant organ) to a nitrocellulose m e m b r a n e and revealing the presence of an antigen (say viral coat protein) by immunoprobing. Standard protocols for
tissue printing have been published (Holt, 1992) and its use has been described in several papers (see Knapp et al., 1995). Tissue printing has several advantages. (1) It gives detailed information on the tissue distribution of a virus. (2) Extraction of sap from leaves in which a virus has limited tissue distribution leads to the dilution of the virus with sap from uninfected cells. Since this technique samples the contents of each cell on the cut surface individually, there is increased sensitivity. (3) The technique is easily applicable to field sampling: tissue blots can be made in the field and there is no need to collect leaf samples for sap extraction in the laboratory. e. Other modifications
The following examples illustrate the versatility of the basic ELISA procedure. i. Biotin-avidin
This detection system is based on the very high affinity of the protein avidin for biotin. The biotin is chemically coupled to the globulin fraction of the antiserum, while the enzyme to be used for detection is coupled to the avidin. Biotin coupling does not interfere with the binding capacity of the antiviral antibody so that a broad range of viral serotypes can be detected using the direct double-antibody s a n d w i c h procedure, and the assays have increased sensitivity (Zrein e, al., 1986). Further
IV.
M E T t t O I ) S I ) E P E N I ) I N G ON P R O P E R T I E S OF VIRAL P R O T E I N S
advantages are speed of development and versatility (Hewish et al., 1986). ii. Use of a fluorogenic substrate
Torrance and Jones (1982) showed that a fluorogenic substrate of alkaline phosphatase (4methyl-umbelliferyl phosphate) gave a more sensitive ELISA test than the standard chromogenic substrate, but the technique requires an expensive plate-reading machine. Reichenbacher et al. (1984) developed an ultramicro ELISA test using the same fluorogenic substrate and plates with a 10-0uL test volume. Time-resolved fluoroimmunoassay has been developed to increase sensitivity by reducing background (Siitari and Kurppa, 1987). This is done by using a europium chelate that produces an intense fluorescence with a very long decay time, during which background fluorescence has decayed. Using this procedure with MAbs, Sinijarv et al. (1988) were able to detect PVX at 5 p g / m L in potato tuber extracts diluted 7 • 104-fold, and in leaf extracts diluted 2 • 107-fold. iii. Enzyme amplification
Torrance (1987) described a procedure for increasing the sensitivity of ELISA assays in which the enzyme bound to the antibody catalyses the conversion of NADP to NAD, which then takes part in a second enzyme-mediated cyclic reaction to produce a red-colored end product. The method could be used for rapid diagnosis of a luteovirus occurring in very low concentrations in plants and also for detecting the virus in individual aphid vectors. iv. Use of F(ab') : and a protein A-enzyme conjugate
Barbara and Clark (1982) described a simple indirect ELISA test in which the virus was trapped by F(ab')2 fragments of homologous antibody bound to the plate. Trapped virus was then allowed to react with intact antivirus immunoglobulin. Thus, only one viral antiserum was needed. Bound immunoglobulin was then assayed by a conjugate of protein A and enzyme. Protein A will bind only to the Fc portion of the intact immunoglobulin. The procedure has been used effectively with several viruses in small fruits (Converse and Martin, 1982).
655
v. Radioimmune ELISA
Ghabrial and Shepherd (1980) developed a simple and highly sensitive radioimmunosorbent assay using the principle of the direct doubleantibody sandwich technique in microliter plates, except that 125I-labeled IgG is substituted for the enzyme-linked IgG in an ELISA assay. The bound and labeled IgG is dissociated from the sandwich before being assayed. vi. Measurement of specific activity of viruses in crude extracts
Konate and Frieig (1983) combined two procedures to enable the specific activity of virus radioactively labeled in vieo to be determined. Indirect double-antibody sandwich ELISA was used to estimate the amount of virus in a well. Then, as a second step, the radioactivity of the virus trapped in the well was assayed. vii. ELISA on polystyrene beads
Polystyrene beads 6.5 mm in diameter have been used as the solid phase (one bead per test) instead of microtiter plates (Chen et al., 1982). This system was more discriminating for detecting differences among isolates of SMV than the standard plate procedure.
C. Collection, preparation and storage of samples As with other diagnostic tests, it is necessary to define and optimize time of sampling and tissue to be sampled to achieve a reliable routine detection procedure (e.g. Torrance and Dolby, 1984; Rowhani et al., 1985; P6rez de San Roman et al., 1988). The main factor limiting the number of tests that can be done with ELISA procedures is the preparation of tissue extracts. Various procedures have been tested to minimize this problem (e.g. Mathon et al., 1987). On the other hand, with some viruses at least, it may be possible to avoid the extraction step by placing several small discs of leaf tissue, cut from the leaves to be tested, directly in the well of a microtiter plate (Romaine et al., 1981). When many hundreds of field samples have to be processed, it is often necessary to store them for a time before ELISA tests are carried out. Storage conditions may be critical for reliable results. For this reason conditions need to
656
15 METHODS FOR ASSAY, DETECTION ANt) DIAGNOSIS
be optimized for each virus and host (e.g. Ward et al., 1987).
D. Monoclonal antibodies 1. Tests using MAbs Many MAbs do not provide a precipitation reaction, especially with protein antigen monomers, for which there may be only one antibody combining site per molecule. However, the use of a MAb and an anti-MAb is a convenient way of extracting or precipitating a peptide or protein antigen with a single determinant. For detection and diagnosis, MAbs have been used most commonly in conjunction with ELISA tests or sometimes with radioimmune assays. Again, exact conditions with respect to pH and other factors may be vitally important. For many applications it may be best to use the same ELISA protocol that was used to screen for MAbs during the isolation procedure because quite different MAbs may be selected depending on the ELISA procedure used (van Regenmortel, 1986). If highly concentrated preparations of MAbs are used in ELISA tests, spurious cross-reactions between viruses in different groups may be detected. Such effects can be avoided by the use of milk proteins instead of bovine albumin as a blocking agent and as a diluent (Zimmermann and van Regenmortel, 1989). The reactivity of a MAb with a given antigen may differ greatly depending on the kind of ELISA procedure used (Dekker et al., 1989). Diaco et al. (1985) used MAbs in a biotin-avidin ELISA for the detection of SMV in soybean seed, while Omura et al. (1986) used MAbs with latex flocculation to detect RSV in plants and insects. Sherwood et al. (1989) obtained a MAb that reacted in ELISA and dot blot tests with the inner core of TSWV, a virus with a lipoprotein envelope. For diagnostic purposes it may be useful to select for MAbs that react with a common antigenic determinant on several strains, for example PVY (Gugerli and Fries, 1983), PVX (Torrance et al., 1986a) and CTV (Vela et al., 1986). Alternatively, several strain-specific or virus-specific MAbs may be pooled to give the required reagent (e.g. Dore et al., 1987a; P6rez de San Roman et al.,
1988). Alternatively, MAbs can be used to distinguish between strains or even isolates with single amino acids differences in their coat proteins (e.g. Chen et al., 1997). Epitope profiling using a panel of MAbs to distinguish virus strains is described in Chapter 17 (Section II.B.4). MAbs can also be used to establish relationships between viruses (e.g. Seddas et al., 2000). 2. Advantages of MAbs i. Requirements for immunization
Mice and rats can be immunized with small amounts of antigen (approximately 100 jug or less). If the virus preparation used is contaminated with host material or other viruses it is still possible to select for MAbs that react only with the virus of interest. ii. Standardization
MAbs provide a uniform reagent that can be distributed to different laboratories, eliminating much of the confusion that has arisen in the past from the use of variable polyclonal antisera. Furthermore, MAbs can be obtained in almost unlimited quantities in suitable circumstances. iii. Specificity
MAbs combine with only one antigenic site on the antigen. Thus, they may have very high specificity and can provide a refined tool for distinguishing between virus strains (see Chapter 17). They can also be used to investigate aspects of virus architecture (see Chapter 5) and transmission by vectors (see Chapter 11). iv. High affinity The screening procedure for detecting MAbs allows for the possibility of obtaining antibodies with very high affinity for the antigen. High-affinity antibodies m a y be used at very high dilutions, minimizing problems of background in the assays. They can also be used for virus purification by affinity chromatography. v. Storage of cells
Hybridomas can be stored in liquid nitrogen to provide a source of MAb-producing cells over a long period.
IV.
M E T H O D S D E P E N D I N G C)N I ' R O P E R T I E S OF V I R A L P R O T E I N S
3. Disadvantages of MAbs a.
Preparation
Polyclonal antisera are relatively easy to prepare. The isolation of new MAbs is labor intensive, time consuming and relatively expensive. For any particular project, these realities must be weighed against the substantial advantages discussed above.
it is much less expensive; and (4) it overcomes the problem of immunological epitopes being lost by proteolysis during the production of antibodies. For instance, the important C-terminal epitope of the coat protein of BNYVV is readily lost by proteolysis during the preparation of MAbs in hybridomas. Phage display obviates this problem (Uhde et al., 2000).
Serologically specific electron microscopy
E
b. Specificity
MAbs may be too specific for some applications, especially in diagnosis. This is an important consideration. c. Sensitivity to conformational
657
changes
Because of their high specificity, MAbs may be very sensitive to conformational changes in the antigen brought about by binding to the solid phase or by other conditions in the assay. 4. Summary Although MAbs provide an excellent tool for many aspects of plant virus research, it is unlikely that they will replace the use of polyclonal antisera in all applications. As far as detection and diagnosis are concerned, appropriate MAbs are available commercially.
E. Phage-displayed single-chain antibodies The theory and use of phage display for the production of antibodies comprising scFv regions of immunoglobulins is described in Section IV.A.7. This technology has been used to provide epitope-specific antibodies for a range of viruses, e.g. BNYVV (Griep et al., 1999; Uhde et al., 2000), PLRV (Harper et al., 1999c; Toth et al., 1999) and TSWV (Griep et al., 2000). A phage library displaying random nonapeptides bound to CMV coat protein (Gough et al., 1999). This shows that proteins other than antibodies that bind to viruses can be discovered. This technique has several advantages over the production of MAbs by hybridomas including: (1) it is more rapid, taking weeks rather than months to produce the panel of antibodies; (2) it does not require the use of animals; (3)
A combination of electron microscopy and serology was first used by Larson et al. (1950) and also illustrated by Matthews (1957), but the technique of negative staining was needed before the procedure could be developed as an effective and widely applicable diagnostic tool. Derrick (1973) described a procedure in which the support film on an electron microscope grid was first coated with specific antibody for the virus being studied. Grids were then floated on appropriate dilutions of the virus solution for 1 hour. They were then washed, dried, shadowed and examined in the electron microscope. Under appropriate conditions, many more viral particles are trapped on a grid coated with antiserum than on one coated with normal serum. Thus, the method offers a diagnostic procedure based on two properties of the virus: serological reactivity with the antiserum used and particle morphology. The method was not popular for some years but is now widely used as a diagnostic tool in appropriate circumstances. Various terms have been used to describe the process: serologically specific electron microscopy (SSEM), immunosorbent electron microscopy (ISEM), solid-phase immune electron microscopy and electron microscope serology. The subject has been reviewed by Milne (1991, 1993). 1. Decoration technique Milne and Luisoni (1977) introduced to plant virology a modification of the general procedure in which, after being adsorbed on to the EM grid, viral particles are coated with virusspecific antibody. This produces a halo of IgG molecules around the virus particles that can be readily visualized in negatively stained
658
15 M E T H O D S FOR ASSAY, D E T E C T I O N AN[) I_)IAGNOSIS
preparations. The p h e n o m e n o n was termed 'decoration'. This procedure probably offers the most convincing d e m o n s t r a t i o n by electron microscopy of specific combination between virus and IgG. For a critical test, a second serologically unrelated virus of similar m o r p h o l o g y should be a d d e d to the preparation (Fig. 15.14). Technical details of protocols are given and difficulties with the procedure are discussed by Milne (1984), Milne and Lesemann (1984) and Katz and Kohn (1984). Particles of some viruses m a y disintegrate following decoration (Langenberg, 1986b). Nevertheless, the unstable rod-shaped virus-like particles associated with lettuce big vein disease have been recognized using the technique and s h o w n to be related to TStV (Vetten et al., 1987). The main advantages of the method are: (1) the result is usually clear, in the form of virus particles of a particular morphology, and thus false-positive results are rare; (2) sensitivity
may be of the same order as with ELISA procedures and m a y be 1000 times more sensitive for the detection of some viruses than conventional EM (Roberts and Harrison, 1979); (3) w h e n the support film is coated with antibody, m u c h less host b a c k g r o u n d material is b o u n d to the grid; (4) antisera can be used without fractionation or conjugation, low-titer sera can be satisfactory, and only small volumes are required; (5) very small volumes (about1 ~tL) of virus extract may be sufficient; (6) antibodies against host components are not a problem inasmuch as they do not bind to virus; (7) one antiserum may detect a range of serological variants (on the other hand, the use of monoclonal antibodies may greatly increase the specificity of the test) (e.g. Diaco et al., 1986); (8) results may be obtained within 30 minutes; (9) w h e n decoration is used, u n r e l a t e d u n d e c o r a t e d virus particles on the grid are readily detected; (10) different proteins on a particle can be decorated
Fig. 15.14 Immune specific electron microscopy (ISEM). Bottom: Mixture of purified TNV and TBSV adsorbed to the grid and then treated with saturating levels of antibodies to TBSV. TBSV particles are 'decorated'. TNV particles remain clean, with sharp outlines. Antibody molecules appear in the background. Bar marker 100 nm. Top: A natural mixture of two potyviruses from a perennial cucurbit, Bryonia cretica. The antiserum used has decorated particles of only one of the viruses in the mixture. One particle near the center is longer than normal and decorated for only part of its length. This particle probably arose by end-to-end aggregation between a particle of each of the two viruses. Bar marker 100 nm. From Milne (1990), with kind permission of the copyright holders, 9 Springer-Verlag GmbH and Co. KG.
IV.
M E T t t O I ) S 13EI'ENI)INC; C'~N PRC'}PERTIES OF VIRAl. P R O T E I N S
659
(e.g. the two coat proteins on closteroviruses and criniviruses; see Fig. 5.12); and (11) prepared grids may be sent to a distant laboratory for application of virus extracts and returned to a base laboratory for further steps and EM examination. Some disadvantages of the procedure are: (1) it will not detect virus structures too small to be resolved in the EM (e.g. coat protein monomers); (2) sometimes the m e t h o d works inconsistently or not at all for reasons that are not well understood (Milne and Lesemann, 1984); (3) it involves the use of expensive EM equipment, which requires skilled technical work and is labor i n t e n s i v e - f o r these reasons it cannot compete with, say, ELISA for largescale routine testing; and (4) w h e n quantitative results are required, particle counting is laborious, variability of particle numbers per grid square m a y be high and control grids are required. In summary, SSEM and ISEM cannot replace ELISA tests w h e n large numbers of samples have to be tested. Their main uses are (1) in the identification of an u n k n o w n virus, (2) in situations where only a few diagnostic tests are needed; and (3) w h e n ELISA tests are equivocal, or not sufficiently sensitive and need direct confirmation (e.g. detection of BSV; Lockhart, 1986; Thottappilly et al., 1998). 2. Immuno-gold labeling Gold particles are highly electron dense and thus show the position on an EM grid of any particle to which they are attached. Protein A forms a reasonably stable complex with gold particles. This complex can then be used to locate any IgG molecules b o u n d to virus coat protein on the EM grid. Van Lent and Verduin (1985, 1986) prepared gold particles of two average diameters (7 and 16 nm). The size of these particles in relation to virus particles and the specificity of the reaction are illustrated in Fig. 15.15A. I m m u n o - g o l d labeling is particularly valuable in locating viral antigens in thin sections of infected cells (see Fig. 8.24). Van Lent and Verduin found that the larger gold particles were more readily seen in stained thin sections. The technique has been adopted for localiza-
Fig. 15.15 Gold particle labeling of virus particles. (A) Protein A labeled with gold particles combining with virus-specific IgG near the virus surface. Virus particles were labeled in suspension with protein A-gold complexes. A mixture of purified CCMV labeled with 7nm protein A-gold and TMV labeled with 16-nm protein A-gold. Bar marker 100 nm. From van Lent and Verduin (1986). (B) RTBV particle treated first with antiRTBV protease rabbit antiserum and then with goldlabeled goat anti-rabbit serum. Bar marker 200nm. From Hay et al. (1994), with permission. tion of viral antigens by light microscopy (van Lent and Verduin, 1987) and has been reviewed by Patterson and Verduin (1987). G. Fluorescent antibody The fluorescent antibody m e t h o d has been applied to the study of the intracellular location and distribution of plant viruses within tissues
660
~5 M E T H O D S FOR ASSAY, D E T E C T I O N AN[) D I A G N O S I S
of the host plant and in insect vectors. Nagaraj (1965) gives the detailed procedures that must be carried out in the preparation and storage of TMV antibody conjugated with fluorescein to avoid non-specific staining. Appropriate controls are essential. Fluorescent antibody staining is particularly suited to detecting viral antigens in isolated infected protoplasts (Otsuki and Takebe, 1969). Chiu and Black (1969) and others have used standard fluorescent antibody methods very successfully for detecting virus antigens in insect vectors.
H. Neutralization of infectivity When viruses are mixed with a specific antiserum, infectivity is reduced to a greater extent than when mixed with a non-immune serum. At high dilutions of serum, inactivation may occur only with the specific antiserum (Rappaport and Siegel, 1955). This phenomenon has been little used in plant virology mainly because of the lack of precision in assays for infectivity. Neutralization tests have been used to demonstrate serological relationships between viruses that are not transmissible mechanically (e.g. Rochow and Duffus, 1978). With a sufficiently stable virus such as TMV, infectivity of the virus can be restored by removal of the antibody at low pH. The mechanism of neutralization has not been established. The binding of IgG to the virus protein may prevent effective release of the nucleic acid from the protein coat. Alternatively, it may block virus from attaching to some site within the cell at an early stage of infection. Dietzgen (1986) selected a monoclonal antibody specific for a C-terminal antigenic determinant of the TMV coat protein using a chemically synthesized tetrapeptide. This MAb neutralized the infectivity of TMV. A systematic study using these procedures may illuminate the mechanism of neutralization.
I. Electrophoretic procedures Electrophoresis in a suitable substrate separates proteins according to size and net electric charge at the pH used. Polyacrylamide gel electrophoresis in a medium containing sodium dodecyl sulfate (SDS-PAGE) is commonly used.
The position and amount of the proteins can then be visualized by a non-specific procedure such as staining or by a specific procedure such as immunoassay (termed western blotting). 1. Gel electrophoresis followed by staining Provided a virus occurs in sufficient concentration, and provided it can be freed sufficiently from any interfering host proteins by some simple preliminary procedure, gel electrophoresis is a rapid method for detecting viral coat proteins. Provided they occur in high enough concentration, other virus-coded proteins may be detected. 2. Electrophoresis followed by electroblot immunoassay O'Donnell ef al. (1982) and Rybicki and yon Wechmar (1982) were the first to use the proteinfractionating power of electrophoresis together with the sensitivity and specificity of solidphase immunoassay to identify and assay viral proteins. The main stages in the technique are: (1) fractionation of the proteins in infected plant sap by SDS-PAGE; (2) electrophoretic transfer of the protein bands from the gel to a membrane (usually nitro-cellulose); (3) blocking of remaining free binding sites on the membrane with protein, usually serum albumin or fat-free dried milk powder; (4) probing for the viral coat protein band with specific antiserum; and (5) detection of the antibody-antigen complex with 125I-labeled protein A or by some ELISA procedure. A great advantage of this technique is that it identifies the virus by two independent properties of its coat protein: molecular weight and serological specificity. Shukla et al. (1983) found that aminophenylthioether (APF) paper was more efficient than nitrocellulose in binding coat proteins. As little as 0.5 ng of viral protein could be detected, compared with 500 ng for Coomassie blue staining. Furthermore, APT paper binds the proteins very stably so that a sheet can be cleaned of antibody and re-probed sequentially with a number of antisera. Electroblot immunoassay can be used following electrophoresis of intact virus particles in agarose gels and transfer of the virus by diffusion blotting to the paper (Koenig and Burgermeister, 1986).
V. M E T H O D S I N V O L V I N G PROPERTIES OF THE VIRAL N U C L E I C A C I D
As noted in Section III.A.2, virus particles can be separated by gel electrophoresis. In a technique termed native electrophoresis and western blot analysis (NEWeB), particles of PPV were electrophoresed directly from plant extracts in agarose or mixed acrylamideagarose gels and blotted on to nitrocellulose membranes (Manoussopoulos et al., 2000). The position of the particles was then identified by immunoprobing. Two different strains of PPV could be distinguished in double infections.
V. METHODS INVOLVING PROPERTIES OF THE VIRAL NUCLEIC ACID General properties of a viral nucleic acid, such as whether it is DNA or RNA, double stranded (ds) or single stranded (ss), or consists of one or more pieces, are fundamental for allocating an unknown virus to a particular family or group. However, with the exception of dsRNA, these properties are usually of little use for routine diagnosis, detection or assay. The ability to make DNA copies (cDNA) of parts or all of a plant viral RNA genome has opened up many new possibilities. The nucleotide sequence of the DNA copy can be determined, but this is far too time consuming to be considered as a diagnostic procedure, except in special circumstances. Basically, there are four approaches to the use of nucleic acids for detection and diagnosis of viruses: 1. the type and molecular sizes of the virionassociated nucleic acids 2. the cleavage pattern of viral DNA or cDNA 3. hybridization between nucleic acids 4. the polymerase chain reaction. In this section, I discuss the principles of these approaches, and then their applications.
A. Type and size of nucleic acid As noted above, the properties of dsRNAs associated with RNA viral infections have been used for diagnosis (reviewed by Dodds, 1993).
661
Double-stranded RNAs are associated with plant RNA viruses in two ways: (1) the plant reoviruses (Chapter 6, Section VI) and cryptoviruses have genomes consisting of dsRNA pieces; and (2) in tissues infected with ssRNA viruses, a double-stranded form of the genome RNA accumulates that is twice the size of the genomic RNA. This is known as the replicative form (see Fig. 8.6). These dsRNAs have been used for diagnosis either following characterization of the double-stranded form by PAGE or by the use of antibodies reacting with dsRNA. a. Electrophoresis of isolated dsRNAs
Dodds et al. (1984) and Dodds (1993) summarize procedures for the isolation of dsRNAs from infected tissue, and for their separation and characterization by PAGE or an agarose gel, using appropriate molecular weight markers. Bands of dsRNA are normally revealed by staining with ethidium bromide or by hybridization (Bar-Joseph et al., 1983). The dsRNA bands can also be eluted from the gel and used to make radioactive probes for use in dot blot analyzes. In principle, each RNA virus should give rise to a distinctive band of dsRNA. This has been found for some virus groups, for example a series of rod-shaped viruses with monopartite genomes (Valverde et al., 1986). However, a series of dsRNA molecules smaller than the full-length dsRNA are almost always present. The pattern of these smaller virus-specific RNAs is characteristic for the virus. Sometimes strains of the same virus can be distinguished by the pattern of smaller bands. The use of dsRNAs in diagnosis is complicated by the fact that some uninoculated and apparently healthy plants contain a series of dsRNA species. For example, N a m e t h and Dodds (1985) found that 40 of 50 glasshousegrown uninoculated cucurbit cultivars contained readily detectable dsRNA species in the molecular weight range 0.5-11.0 • 1 0 6 Da. Each cultivar had a characteristic pattern, indicating seed transmission. The nature of these dsRNAs has not been established, but in view of the high molecular weight of the largest species, they may be associated with
662
15 METHODS FOR ASSAY, DETECTION AND DIAGNOSIS
undescribed cryptic viruses. On the other hand, Wakarchuk and Hamilton (1985) described high-molecular-weight dsRNAs from one variety of Phaseolus vulgaris. These dsRNAs had sequence similarity to the genome DNA of the bean variety in which it was found, as well as that of other bean varieties that contained no detectable dsRNA. Other large dsRNA species from several plants are noted in Chapter 17. Another source of dsRNA could be fungi adventitiously associated with the plant being tested. Many of the viruses of fungi have dsRNA genomes and there are dsRNA molecules associated with plant pathogenic fungi (see Nuss and Koltin, 1990). In spite of these difficulties, the dsRNA method has a role to play in virus diagnosis for some host species. Thus TRSV, RBDV and Raspberry leaf spot virus could be readily detected in infected Rubus spp. Field samples of diseased plants often contained two or more viruses that could be readily identified from the dsRNA patterns (Martin, 1986). The procedure provided an alternative to diagnosis by grafting or aphid transmission to indicator hosts. Similarly, the method was valuable for diagnosis of CTV and strains of the virus, in various citrus species, provided that the tissue sampled was in optimal condition (Dodds et al., 1987). A low-molecular-weight dsRNA associated with groundnut rosette disease has been used as a diagnostic tool (Breyel et al., 1988).
b. Antibodies against dsRNA Antibodies reacting non-specifically against dsRNAs were found in antisera prepared against plant reoviruses (Ikegami and Francki, 1973; van der Lubbe et al., 1979). Such nonspecific antisera can be prepared using synthetic poly(I):poly(C) antigen (Stollar, 1975), but the procedure has not received wide application for plant viruses. MAbs have been produced that recognize dsRNA and have been used in immuno-blots to detect this form of RNA in extracts from plants infected with 18 viruses (Luk~ics, 1994). Immuno-blot analyzes in combination with temperature gradient electrophoresis could distinguish between the
dsRNA genomes of BCV-1 and BCV-2 from one another. As MAbs to dsRNAs do not react with double-stranded sequences of less than 11 bp, they can be used as probes of RNA structure in crude nucleic acid extracts (Schonborn et al., 1991). c. Polymorphisms in heteroduplex RNAs
The electrophoretic migration of single- and double-stranded nucleic acids in gels differs and can be used as a basis for diagnostic techniques. The technique of single-stranded conformation polymorphism (SSCP) was originally based on the migration rates of ssDNA molecules conferred by their particular primary and tertiary structures (Orita et al., 1989a). As described in Chapter 17 (Section II.A.l.b), SSCP was applied to the ssRNA of BNYVV in an attempt to differentiate strains (Koenig et al., 1995) but proved to be variable in detecting mutations. By adapting the method using heterologous duplexes of RNA transcripts polymorphisms were detected between strains of PNRSV (Rosner et al., 1998; 1999). B. C l e a v a g e p a t t e r n s of D N A Cleaving cDNAs of RNA genomes and the genomes of DNA viruses at specific sites with restriction enzymes and determining the sizes of the fragments by PAGE is a possible procedure for distinguishing viruses in a particular group. For instance, Hull (1980) showed that isolates of CaMV could be distinguished on the restriction endonuclease patterns (subsequently termed restriction length polymorphism mapping; RFLP) of the virion DNA. However, there is considerable variation in viral populations (see Chapter 17, Section I.A) as shown by RFLPs between individual clones of RTBV DNA from a single virus preparation (Villegas et al., 1997). For the Geminivirus group, Haber et al. (1987) used labeled DNA from one virus of the group to identify a range of group members by probing restriction enzyme fragments that gave a pattern of bands after PAGE characteristic for each member. An example of the application of RFLP analysis to cDNA of an RNA virus is given by Kruse et al. (1994), who
663
V. M E T H O D S I N V O L V I N G PROPERTIES OF THE V I R A L N U C L E I C A C I D
used the technique to distinguish two strains of BNYVV. This technique may be of use in specific situations but it has not yet achieved any wide acceptance in plant virology. C. H y b r i d i z a t i o n p r o c e d u r e s These procedures depend on the fact that single-stranded nucleic acid molecules of opposite polarity and with sufficient similarity in their nucleotide sequence will hybridize to form a double-stranded molecule. The motivation to develop hybridization methods came first with the viroids, which have no associated proteins that can be used for diagnosis (Palukatis et al., 1981; Allen and Dale, 1981). 1. Basis for hybridization procedures The theory concerning nucleic acid hybridization is complex. It has been discussed by Britten et al. (1974) and Hull (1993, 2000). The Watson and Crick model for the structure of dsDNA showed that the two strands were held together by hydrogen bonds between specific (complementary) bases, namely adenine and thymidine, cytosine and guanine. This interaction of base pairing is the basis of all molecular hybridization. The early studies revealed several important features of this process. The ability to use various physical and chemical procedures to disrupt base pairing and hence separate the strands (termed melting or denaturing the nucleic acid) and then to reinstate base pairing and thus renature the doublestranded nucleic acid (termed hybridization) enabled the various factors controlling the stability of the duplex to be examined. Most of these basic studies were performed using both the target and probe DNAs in solution (liquid-liquid hybridization); there are also some data available for RNA:RNA and DNA:RNA interactions in solution. Although most of plant pathogen diagnosis now involves mixed-phase hybridization with the target immobilized on a solid matrix, the theory developed for liquid-liquid hybridization is still very relevant. Below is a general account of some of the major factors that have to be considered in
hybridization experiments. More details on the theory of hybridization can be found in Ausubel et al. (1998), Britten et al. (1974), Britten and Davidson (1985) and Young and Anderson (1985). a.
Denaturation
There are various denaturation.
factors
that
affect
i. Temperature
Double-stranded nucleic acid will denature at high temperatures. Strand separation takes place over a relatively narrow range of temperatures, the temperature at which 50% of the sequences are denatured being called the melting temperature (Tm). The major factors affecting the T are the composition of the nucleic acid, the concentration of salt in, and the pH of, the solution, and the presence of materials that can disrupt hydrogen bonding such as formamide. ii. Nucleic acid composition Guanosine + cytosine (G+C) base pairs, which have three hydrogen bonds, are more stable than adenosine + thymidine (uridine) (A+T) base pairs, which have two. For perfectly basepaired DNA in 1• SSC (SSC = 0.15 M sodium chloride, 0.015 M sodium citrate), T m is related to the G + C content by:
Tm = 0.41 (%G+C) + 69.3 iii. Salt
The salt concentration has a marked effect on the T of a duplex. The T increases by almost 16~ for each 10-fold increase in the concentration of monovalent cation over the range of 0.01-0.1 M; the effect is less at higher concentrations. Over the lower concentration range the T m i s given by: Tin= (ToGC/2.44) + 81.5 + 16.6 logM
where M is the molarity of the monovalent cation. Divalent cations have an even greater effect and should be removed by the use of chelators such as EDTA.
664
15 M E T H O D S FOR ASSAY, D E T E C T I O N A N D D I A G N O S I S
iv. pH
The T is insensitive to pH in the range 5-9. Below p H 5, depurination will start and will become more rapid as the pH is lowered. This can be of use as a method for introducing nicks into DNA as apurinic acid is alkali labile. Above pH 9 denaturation of doublestranded nucleic acid sets in, first in A+T-rich regions. Most duplex DNAs are fully denatured at p H 12. The phosphodiester bonds of RNA are degraded above pH 8, the higher the pH and temperature, the more rapid the degradation. v. Organic solvents
Some organic solvents such as formamide, dimethyl formamide and dimethylsulfoxide, and also urea, lower the T . The T is reduced by 0.7~ for each percent of formamide. vi. Base-pair mismatch
Mismatched sequences are less stable than perfectly base-paired duplexes. In nucleic acids of more than 100 bp, a mismatch of 1% reduces the Tm by about 1~ Thus, mismatching can be assessed by varying the hybridization conditions (see stringency below). vii. R N A : R N A and R N A : D N A duplexes
The Tm of dsRNA is significantly higher than that of dsDNA. A general value for the Tm of DNA in 1• SSC is about 85~ and of RNA is close to 100~ The T m o f RNA:DNA hybrids is about 4-5~ higher than that for DNA:DNA duplexes under the same conditions. b. Re.association
The two main considerations relevant to reassociation are the rate at which it occurs and the stability of the products. The stability of the products is affected by the same factors that control denaturation. The main factors affecting rate of re-association are described below. i. Temperature
For a typical D N A ' D N A re-association reaction, the graph relating the rate of formation of duplexes to temperature is a broad curve with the m a x i m u m rate at about 25~ below the Tm . This is taken to be the o p t i m u m temperature
for re-association. At temperatures well below the optimum, re-association m a y effectively cease. Thus, it is possible to maintain the two strands of a duplex separated for considerable periods of time by melting at high temperature and then lowering the temperature rapidly (by plunging the tube into ice) to well below the T (termed quenching). If the T is lowered, for example by the use of formamide, the o p t i m u m rate of re-association is at the new T-25~ However, because the t e m p e r a t u r e is lower, the overall rate of h y b r i d i z a t i o n will be slower than in the absence of formamide. ii. Salt concentration
The concentration of monovalent cations affects the rate of re-association markedly. Below 0.1 M s o d i u m chloride, a 2-fold increase in salt concentration increases the rate by 5-10-fold or more. The rate continues to rise with increasing salt concentration, but becomes constant above 1.2 M sodium chloride. Divalent cations have very pronounced effects on the rate and should be removed by the use of chelators such as EDTA. iii. Base mismatch
The also 10% tion
precision with which base-pairing occurs affects the rate of re-association. For each mismatch the rate is halved when the reacis under optimal conditions (e.g. T-25~
iv. Fragment length
If the cDNA strands are the same length, the rate of re-association rises with the square root of the length. When the strands are of different length, the rate depends upon which fragment is in excess and interpretation can be very complicated. v. Concentration of nucleic acid
The time required for the formation of duplexes is directly proportional to the initial concentration of the interacting single-stranded molecules. Reactions are normally measured as the product of the initial concentration (C o) of the nucleic acid in moles of nucleotide per liter and time (t) of reaction in seconds (C~ expressed as moles per second per liter). C~ values are of great use in liquid-liquid hybridization in d e t e r m i n i n g features such as a m o u n t s of
V.
METtlODS
unique sequence (complexity) and repeated sequences in nucleic acids. vi. Polymers The anionic polymer, dextran sulfate, accelerates the rate of re-association both in solution and on filters. A 10% solution of dextran sulfate (MW 500 000 Da) increases the rate of hybridization in solution by about 10-fold and of hybridization to immobilized nucleic acids by up to 100-fold. This increase is thought to be due to the exclusion of the nucleic acid from the volume occupied by the polymer, thus effectively increasing its concentration. vii. R N A : D N A hybridization The factors affecting hybridization of RNA to cDNA are s o m e w h a t different from those affecting D N A : D N A interactions, probably because of the greater a m o u n t of strong secondary structure in the RNA. If RNA is in excess u n d e r m o d e r a t e salt conditions (0.18 M sodium chloride), the rate is almost the same as with DNA:DNA hybridization. However, the rate does not rise as rapidly as does DNA:DNA hybridization with increasing salt concentration. If DNA is in excess, the rate is four to five times slower than that expected for DNA:DNA re-association. c. M i x e d . p h a s e
hybridization
The kinetics of mixed-phase hybridization have been less well studied than those of liquidliquid hybridization. It is generally assumed that the effects of the reaction conditions are qualitatively, and in most cases quantitatively, the same for filter hybridization as they are for liquid-liquid hybridization (for details see Anderson and Young, 1985). However, with the c o m m o n practice in filter hybridization of using double-stranded probes, the strandedness of the probe can have effects that may be important if kinetics are being studied. i. Probe strandedness When double-stranded probes are used in filter hybridization, two sets of hybridization take place, that between the probe and the immobilized target sequence and the self-annealing of the probe. The latter can have two effects:
I N V O L V I N ( ~ P R O P E R T I E S OF T H E VIRAl_ NUCI.EIC" A(.'ID
665
removal of the effective probe, which reduces sensitivity, and the formation of concatenates, which may then hybridize to the target nucleic acid and thus increase sensitivity but decrease specificity. To minimize these effects, the sequences in the probe complementary to the target nucleic acid should be relatively short, the probe should be at low concentration in a small reaction volume, and the reaction should be at as high a temperature as possible. Single-stranded probes overcome many of these problems. However if 32p-labeled singlestranded probes are used at concentrations above 100 n g / m L , non-specific binding to the membrane may occur. ii. Length of probe For short oligonucleotide probes (less than 30 base pairs) the T m c a n be estimated from: V m
--
2(A + T) + 4(G + C)
where A, T, G and C are the numbers of adenosine, thymidine, guanosine and cytosine bases respectively. For filter hybridization the dissociation temperature (Td) is calculated: Td - T - k where the constant k has been determined experimentally to be 7.6~ The T of DNA longer than 50 bp can be calculated from: Tm - 81.5 + 16.6 • log[Na] + 0.41 • ( % G + C ) - 6 7 5 / l e n g t h - 0 . 6 5 x (%formamide) iii. Stringency (see Anderson and Young, 1985; Meinkoth and Wahl, 1984) The term "stringency" is often used imprecisely. It relates to the effect of hybridization a n d / o r wash conditions on the interaction between complementary nucleic acids, which may be incompletely matched. The use of different stringencies is one of the more powerful tools of the hybridization technology. Filter hybridization can be used to determine the degree of relatedness between sequences. To achieve this, one has to be able to estimate
666
~5 M E T H O D S FOR ASSAY, D E T E C T I O N A N D D I A G N O S I S
approximately the proportion of mismatches in the hybrid. This is done by adjusting the reaction conditions so that the desired interaction can be examined. If close relationships are to be distinguished from distant ones, more stringent conditions are used; if distant relationships are to be detected, the conditions should be less stringent. The stringency can be varied at two stages in the procedure' at hybridization and at the post-hybridization wash. As a general rule, for distantly related sequences the hybridization conditions should be relaxed and the washes carried out under increasing stringency; for closely related sequences hybridization and washing should be under stringent conditions. Stringency can be varied by changing the temperature a n d / o r salt concentration a n d / o r formamide concentration. Also to be considered in experiments to determine the relationship between nucleic acids are the relative concentrations of probe and target nucleic acid and the size of the probe. The probe nucleic acid should never be in excess, to ensure that it does not saturate the target nucleic acid. For probes larger than 100 bases, the Tm of a DNA duplex is decreased by approximately 1~ for every 1% mismatch; for hybrids shorter than 20 bp the Tm decreases by about 5~ for each mismatched base pair. As a general guideline for nucleic acids of more than 200 bp and 40-50% G + C content, the following are conditions for various stringencies" Lc~w stringency
50~
Stringent
65~
t tigh stringency 65~
5 • SSC
allowing apprc~x. 25% mismatch 2 • SSC allowing appr~x. 10% mismatch 0.1• SSC allowing 0.05 ppm 0.06 g/L
Ultraviolet light
Dependent on flow rate
Walsh (1992) Tomlinson and Faithfull (1979) Gharbi and Verhoyen (1993)
Soil amendments and fungicides PMTV S. subterranea TNV O. brassicae BNYVV P. betae
Zinc sulfate Captan Fluazinam
1320 kg/ha 1 g/kg soil 10-50 ppm
Cooper et al. (1976) Thomas (1973) Uchino et al. (1993)
Soil treatments LBVV LBVV WSSMV
Mercuric chloride Jet 5 Methyl bromide
0.05 To 2% 1.5 ml/L
Grogan et al., (1958) Walsh (1998) Slykhuis (1970)
Nutrient or aquatic culture WYSV S. subterranea LBVV 0. brassicae LBVV and lettuce ring necrosis
0. brassicae
0. brassicae 0. brassicae P. graminis
1V. PROTECTING THE PLANT FROM SYSTEMIC DISEASE
and genetic protection (conventional resistance and transgenic resistance). The first two of these approaches are discussed in this section and genetic protection in the next two sections. A . M i l d s t r a i n p r o t e c t i o n (crossp r o t e c t i o n ) (reviewed by Urban et al., 1990; Lecoq, 1998) Infection of a plant with a strain of virus causing only mild disease symptoms (the protecting strain; also k n o w n as the mild, attenuated, hypovirulent or avirulent strain) may protect it from infection with severe strains (the challenging strain) (Chapter 17, Section II.C.4). Thus, plants might be purposely infected with a mild strain as a protective measure against severe disease. This was first reported by McKinney (1929), who observed that tobacco plants systemically infected by a 'green' strain of TMV were protected from infection by another strain that induced yellow mosaic symptoms. While such a procedure could be worthwhile as an expedient in very difficult situations, it is not to be recommended as a general practice, for the following reasons: 1. So-called mild strains often reduce yield by about 5-10%. 2. The infected crop may act as a reservoir of virus from which other more sensitive species or varieties can become infected. 3. The dominant strain of virus may change to a more severe type in some plants. 4. Serious disease may result from mixed infection when an unrelated virus is introduced into the crop. 5. For annual crops, introduction of a mild strain is a labor-intensive procedure. 6. The genome of the mild strain may recombine with that of another virus, leading to the production of a new virus. In spite of these difficulties, the procedure has been used successfully, at least for a time, with some crops. A suitable mild isolate should have the following properties: 9 It should induce milder symptoms in all the cultivated hosts than isolates commonly encountered and should not alter
699
the marketable properties of the crop products. 9 It should give fully systemic infections and invade most, if not all, tissues. 9 It should be genetically stable and not give rise to severe forms. 9 It should not be easily disseminated by vectors to limit unintentional spread to other crops. 9 It should provide protection against the widest possible range of strains of the challenging virus. 9 The protective inoculum should be easy and inexpensive to produce, check for purity, provide to farmers and apply to the target crops. Mild protecting strains are produced from naturally occurring variants, from r a n d o m mutagenesis or from directed mutagenesis of severe strains. Broadbent (1964) suggested that, because late infection of greenhouse tomatoes with ToMV often causes a severe reduction in quality of the fruit compared with early infection, growers who regularly suffer those losses should inoculate their plants at an early stage with a mild strain of the virus. Rast obtained a nitrous acid mutant of ToMV that was symptomless in tomato (Rast, 1975). Seedling inoculation with this strain was widely practiced for a time in some w e s t e r n E u r o p e a n countries and in Canada, N e w Zealand and Japan. However, cultivars with resistance genes or with multiple genes for tolerance to ToMV have largely replaced seedling inoculation with the attenuated strain in commercial practice. CTV p r o v i d e s the most successful example for the use of cross-protection. Worldwide, this is the most important virus in citrus orchards. In the 1920s, after its introduction to South America from South Africa, the virus virtually destroyed the citrus industry in many parts of Argentina, Brazil and Uruguay. The successful application of cross-protection by inoculation with mild CTV isolates in Brazil has been detailed by Costa and Muller (1980). The method has been particularly successful with Pera oranges, with more than eight million trees being planted in Brazil by 1980. Protection
700
~6 C O N T R O L A N t ) USES OF PLANT VIRUSES
continues in most individual plants through successive clonal generations. However, in an 8year assessment of the ability of four mild isolates to suppress severe CTV isolates in Valencia sweet orange on sour orange rootstock in Florida, about 75 % of the mild-strain protected trees had severe s y m p t o m s compared with about 85% of the unprotected trees (Powell et al., 1992). The use of the same isolates gave better protection of Ruby Red grapefruit on sour orange rootstock (Powell et al., 1999). Thus, there are differences in the responses of the scion-rootstock combination, but it is also important to have a compatible mild strain. The search for i m p r o v e d attenuated strains of the virus continues (e.g. Muller and Costa, 1987; Roistacher et al., 1987), and the technique is being adopted in other countries. Other viruses and crops for which attempts are being made to develop effective crossprotection include PRSV in papaya (Yeh et al., 1988; Gonsalves, 1998), TMV in pepper (Tanzi et al., 1986) and ZYMV in courgette (Walkey et al., 1992). In principle, the cacao swollen shoot disease of cacao in Ghana should be controllable to some degree by cross-protection with mild strains of the virus (Posnette and Todd, 1955; Hughes and Ollennu, 1994), but various difficulties have prevented its effective application (Fulton, 1986). In particular, the use of the technique was incompatible with the objective of treating all known outbreaks by removal of infected trees. This is no longer feasible, so that there is now scope for using mild strain protection in the worst affected areas of Ghana, where other control measures have been abandoned (Matthews, 1991). In Chapter 10 (Section IV), I describe the recent developments in understanding host response to virus infection by posttranscriptional gene silencing. It is most likely that mild strain protection operates by the mild strain 'priming' this d e f e n s e system so that it operates against the superinfecting severe strain. With this in mind, it is important that the sequence of the protecting mild strain is similar to that of the superinfecting severe strain. An analysis of sequence of CTV isolates from different sites and collected at different times showed that the mild strains used in Florida and Spain were very similar to a
diverse range of isolates (Albiach-Marti et al., 2000a). This was unexpected as the first two CTV isolates to be sequenced had up to 60% sequence divergence. However, it is possible that, in some instance of mild strain protection, other mechanisms, such as competition for replication sites, operate.
B. S a t e l l i t e - m e d i a t e d p r o t e c t i o n (reviewed by Tien and Wu, 1991; Jacquemond and Tepfer, 1998) Satellite viruses and RNAs are described in Chapter 14 (Section II) and, as far as potential biocontrol agents, fall into three categories: those that enhance the helper virus symptoms, those that have no effect and those that reduce the helper virus symptoms.. It is the latter that have potential as control agents. Most of the work has focused on the satellites of CMV. Tien et al. (1987) obtained mild strains of CMV by adding selected satellite RNAs to a CMV isolate. This procedure has been tested, with increased yields, in many areas of China for the control of CMV in peppers. The extent of the protective effect depended on such factors as inoculation time, percentage of plants inoculated, the nature of virulent CMV strains already in the field, and variety of pepper. Wu et al. (1989) compared the degree of protection obtained by pre-inoculating tobacco and pepper plants with a mild strain of CMV with or without satellite RNA. The presence of satellite RNA increased the protection obtained. Similar results were obtained in greenhouse experiments with CMV in tomato. In field trials, protection was maintained when the virus was introduced by aphids (Montasser et at., 1991). This work has been taken a stage further by Gallitelli et al. (1991), who inoculated several hundred young tomato seedlings using several varieties, with CMV strain S carrying a nonnecrogenic satellite called S-CARNA5. These were planted out at the seedling stage in spring on a farm in southern Italy, where a severe epidemic of the tomato necrosis disease was expected. The epidemic occurred, with 100% of plants being destroyed in some fields. In the field containing the inoculated plants, protection
IV. PROTECTING TIlE PLANT FROM SYSTEMIC DISEASE
C. Antiviral chemicals
against necrosis was almost 100%, while 40% of uninoculated plants developed lethal necrosis. Fruit yields were about doubled in the protected plants. Satellite protection against CMV infection of tomato has also been used in Japan (Sayama et at., 1993) and China (Tien and Wu, 1991), and against CMV in pepper and melon plants in the USA (Montasser et al., 1998). A model suggested that inoculation with CMV containing a mild satellite RNA prior to challenge with a severe satellite isolate with or without CMV interfered with the replication and symptom production of the severe strain (Smith et al., 1992). However, there has been concern over the durability of using satellites as biocontrol agents. There is a wide range of necrogenic and other virulent strains of satCMV (Jacquemond and Tepfer, 1998). Passage of a benign satellite of CMV through Nicotiana tabacum led to the satellite rapidly mutating to a pathogenic form (Palukaitis and Roossinck, 1996) and mutations of a single or a few bases can change a nonnecrogenic variant to a necrogenic one (Fig. 16.12). Necrogenic variants of the CMV satellite have a greater virulence than non-necrogenic variants (Escriu et al., 2000), but, as they depress the accumulation of the helper virus more than do non-necrogenic variants, the necrogenic variants are not so efficiently aphid transmitted. y S a t - R N A (a)
1. Chemical control of viruses (reviewed by Hansen, 1989) Considerable effort has gone into a search for inhibitors of virus infection and multiplication that could be used to give direct protection to a crop against virus infection in the way that fungicides protect against fungi. There has been no successful control on a commercial scale by the application of antiviral chemicals. The major difficulties are: 1. An effective compound must inhibit virus infection and multiplication without damaging the plant. This is the first, and major, problem. Virus replication is so intimately bound up with cell processes that any c o m p o u n d blocking virus replication is likely to have damaging effects on the host. 2. An effective antiviral compound would need to move systemically through the plant if it were to prevent virus infection by invertebrate vectors. 3. A compound acting systemically would need to retain its activity for a reasonable period. Frequent protective treatments would be impracticable. Many compounds that have some antiviral activity are inactivated in the plant after a time.
CUAAGGCUUAUGCUAUGCUGAUCUCCGUGAAUGUCUAUACAUUCCUCUACAGGACC
C
U
u Y S a t - R N A (b)
A WII
S a t - R N A (c)
G R S a t - R N A (d)
UU
CGUGAAUGUCUAUACAUUCCUCUACA~C
CC
GU
U
necrogenic ameliorative
CUUAGACUUAGGUUAUGCUGAUCUC
U
necrogenic ameliorative
CUAA~AUGCUAUGCUGAUCUC
U
701
U~
CGUGAAUGUCUACACAUUC
CUCUACAGGACCC
ameliorative
necrogenic
UC
CUAA~AUGCUACGCUGAUCUCCGUGAAUGUCUA
g
u
.U C A U U C C U C .A C A G G A C C C
ameliorative necrogenic
Fig. 16.12 Alignment of the 55 3'-terminal residues of the satellite RNA variants mutated from a necrogenic form towards a non-necrogenic one, or vice versa. Arrows indicate the positions found to be determinant for necrogenicity. From Jaquemond and Tepfer (1998), with kind permission of the copyright holder, 9 The American Phytopathological Society.
702
~6 C O N T R O L
A N D USES OF P L A N T V I R U S E S
4. For most crops and viruses, the compound would need to be able to be produced on a large scale at an economic price. This might not apply to certain relatively small-scale, high-value crops, such as greenhouse orchids. 5. For use with many crops, the compound would have to pass food and drug regulations. Many of the compounds that have been used experimentally would not be approved under such regulations. Many substances isolated from plants and other organisms, as well as synthetic organic chemicals, have been tested for activity against plant viruses. Almost all the substances showing some inhibition of virus infectivity do so only if applied to the leaves before inoculation or very shortly afterwards. An example is a glucan preparation obtained from Phytophthora megasperma fs.p. glycinea, which appears to inhibit infection by several viruses by a novel, but unknown mechanism (Kopp et al., 1989). Work in this field has been reviewed by Matthews (1981), Tomlinson (1981), White and Antoniw (1983) and Verma et al. (1998). Synthetic analogs of the purine and pyrimidine bases found in nucleic acids have been widely studied, and the search for inhibitory compounds of this type continues (Dawson and Boyd, 1987a). The substituted triazole, 4(5)amin o-1H-1,2,3- tria z o 1e-5 (4)-c a r b o xy amid e, can be regarded as an aza analog of the substituted imidazole compound which is known to be a precursor of the purine ring in some systems. This compound showed some plant virus inhibitory activity but was less effective than 8azaguanine. The riboside of this compound was suggested as a possible antiviral agent by Matthews (1953c). It was later found to have broad-spectrum activity in experimental animal virus systems and given the name virazole (Sidwell et al., 1972). Virazole, also known as ribavirin, has been studied in a variety of plant-virus systems. For example, pretreatment of tobacco plants with the compound delayed or prevented systemic infection with TSWV (De Fazio et al., 1980). Virazole reduced the concentration of CMV and AMV in cultured plant tissues. However, virus-free plants were
obtained from meristem tip cultures whether or not virazole had been in the medium (Simpkins et al., 1981). Other virus-inhibitory compounds have been included in culture media in attempts to improve the efficiency of meristem tip culture for obtaining virus-free plants. None has found widespread use. 2. Suppression of disease symptoms by chemicals For a time, there was considerable interest in the use of certain systemic fungicides to suppress the symptoms of virus infection without necessarily having any effect on the amount of virus produced in the leaves. The fungicides concerned (e.g. benlate and bavistin) break down in aqueous solution to give methylbenzimidazole-2-yl-carbamate (MBC or carbendazim). It has been reported that these systemic fungicides possess cytokinin-like activity, although at a low level compared with kinetin. Application of MBC as a soil drench caused a substantial reduction in leaf symptoms caused by TMV in tobacco (Tomlinson et al., 1976). However, this compound or others with similar effects have not found commercial application. The severity of infection of tomato plants by ToMoV was reduced by treatment with plant growth-promoting rhizobacteria (PGPR) (Murphy et al., 2000). It was suggested that the use of PGPR could become a component of an integrated program for management of this virus in tomato. It should be noted that, in some cases, high applications of nitrogen fertilizers can suppress virus disease symptoms. However, virus concentrations are not diminished.
V. C O N V E N T I O N A L RESISTANCE TO P L A N T VIRUSES A. Kinds of host response The genetic makeup of the host plant has a profound influence on the outcome following inoculation with a particular virus. The kinds of host response are defined in Table 10.1 and molecular aspects are discussed in Chapter 10. Figure 16.13 illustrates the dramatic effect that a single mendelian gene can have on disease
V. C O N V E N T I O N A L R E S I S T A N C E TO P L A N T VIRUSES
703
Fig. 16.13 Reactions of two varieties of tobacco to a strain of TMV from tomato. Left: Var. White Burley showing typical systemic mosaic. Center: Vat. Warnes, necrotic local lesions with no systemic spread of virus. Right: An F1 hybrid (Warnes 9 • White Burley ~) showing necrotic local lesions and a severe systemic necrosis and stunting. From Matthews (1991). response. Defining the response of a particular host species or cultivar to a particular virus must always be regarded as provisional. A new m u t a n t of the virus m a y develop in the stock culture, or a new strain of the virus m a y be found in nature that causes a different response in the plant. This applies particularly to nonhost immunity. For example, for m a n y decades the potato seedling USDA 41956 was considered i m m u n e to PVX. However, in due course a strain of this virus was discovered that can infect this genotype (Moreira et al., 1978). Nevertheless, i m m u n i t y w i t h l o n g - t e r m durability must occur frequently in nature. It has been suggested with respect to cellular parasites that long-lived plants such as trees, and species naturally d o m i n a n t over large areas, such as prairie grasses and aggressive weeds, m u s t have d e v e l o p e d greater resistance to lethal parasites than other species. Similar considerations m a y apply as far as virus infection is concerned, but little relevant experimental evidence is available. It is probable that m a n y of the plants described in the past as i m m u n e to a particular virus were in fact infectible, but resistant and showing extreme hypersensitivity as defined in Table 10.1. The viruses concerned m a y have had cell-to-cell m o v e m e n t proteins that were defec-
tive in the particular plant, resulting in multiplication only in the initially infected cells. This p h e n o m e n o n may be detectable only by special procedures (for example, see Fig. 9.4). The following points concerning the effects of host genes on the plant's response to infection emerge from m a n y different studies: 1. Both d o m i n a n t a n d recessive m e n d e l i a n genes m a y have effects. However, while most genes k n o w n to affect host responses are inherited in a mendelian manner, cytoplasmically transferred factors m a y sometimes be involved (Nagaich et al., 1968). 2. There m a y or may not be a gene dose effect. 3. Genes at different loci m a y have similar effects. 4. The genetic background of the host m a y affect the activity of a resistance gene. 5. Genes m a y have their effect with all strains of a virus, or with only some. 6. Some genes influence the response to more than one virus. 7. Plant age and environmental conditions m a y interact strongly with host genotype to produce the final response. 8. Route of infection m a y affect the host response. Systemic necrosis m a y develop following introduction of a virus by grafting
704
~6 C O N T R O L A N D USES OF P L A N T V I R U S E S
into a high-resistant host that does not allow systemic spread of the same virus following mechanical inoculation (e.g. ToMV in resistant tomato lines; Stobbs and McNeill, 1980). 9. Resistance originally thought to be to the virus may be really to the vector (Dahal et al., 1990a, b). As pointed out by Fraser (1988, 1992), there are three main types of resistance and immunity to a particular virus, considered from the point of view of the complexity of the host population involved: (1) i m m u n i t y involves every individual of the species; little is k n o w n about the basis for immunity, but it is related to the question of the host range of viruses discussed in Chapter 3 (Section V); (2) cultivar resistance describes the situation where one or more cultivars or breeding lines within a species show resistance while others do not; and (3) acquired or induced resistance is present where resistance is conferred on otherwise susceptible individual plants following inoculation with a virus. This last p h e n o m e n o n was discussed in Chapter 10 (Section III.K). Some authors have considered that immunity and cultivar resistance are based on quite different underlying mechanisms. However, studies with a bacterial pathogen in which only one pathogen gene was used show that, for this class of pathogen at least, the two phenomena have the same basis (Whalen et al., 1988). B. G e n e t i c s of r e s i s t a n c e to v i r u s e s This section briefly outlines the kinds of genetics involved in cultivar resistance to virus infection that can be used for virus control. The possible mechanisms for such resistance are discussed in Chapter 10. Plant resistance to viruses has been reviewed by Ponz and Bruening (1986),
Fraser (1987a, 1988, 1992), Evered and Hamett (1987), Khetarpal et al. (1998) and H a m m o n d (1998). The s u m m a r y in Table 16.4 shows that resistance to viruses in most crop virus combinations is controlled by a single d o m i n a n t gene. However, this may merely reflect the fact that most resistant cultivars were developed in breeding programs aimed at the introduction of a single resistance gene. Furthermore, incomplete dominance may be a reflection of gene dosage or be due to environmental factors (Fischer and Rufty, 1993). There have not been m a n y studies of the inheritance of resistance in wild species. Some specific examples of dominant, incompletely dominant, and apparently recessive genes for resistance are given in Table 16.5. Sometimes response to virus infection is associated with resistance to some other kind of disease agent. Thus, the necrotic response of tobacco to infection by a strain of PVY may be a pleiotropic effect of the gene controlling resistance to a root-knot nematode (Rufty et al., 1983). 1. The gene-for-gene hypothesis Gene-for-gene relationships are well known between host plant and fungal or bacterial pathogens. They have been established primarily based on genetic analyzes of both plants and pathogens. With these parasites, each allele in the host that confers resistance may be reflected in a complementary virulence locus in the parasite that can overcome the resistance. Virulence and avirulence genes and plant resistance are discussed in Chapter 10. Viruses and virus strains may be described in relation to the various host responses defined in Table 10.1. Thus, if a particular plant species or cultivar shows immunity or resistance to a
TABLE 16.4 Summary of number of virus resistance genes reported Resistance gene Dominant Recessive Incompletely dominant Nature unknown Total number of resistance genes
Monogenic
Oligogenic or polygenic
81 43 15 -
10 20 6 4
139
40
From Khertarpal et al. (1998), with kind permission of the copyright holder, 9 The American Phytopathological Society.
V. C O N V E N T I O N A L R E S I S T A N C E TO P L A N T V1RUSES
705
TABLE 16.5. Examples of host genes for resistance to plant viruses Gene
Host species
Controlled by dominant genes N Nicotiana glutinosa N' N. sylvestris Zym" Cucurbita moschata Tm-2 Lycopersicon esculentum Tm-22 L. esculentum Nx, Nb Solanum tuberosum By, By-2 Phaseolus vulgaris RSVl, R s v 2
Glycine max
Controlled by incompletely dominant gene Tm-1 L. esculentum L 1, L, L3 Capsicum spp. Two genes Multiple genes
Hordeum vulgare Vigna sinensis
Apparently recessive genes By-3 P. vulgaris Sw 2, SW3, Sw 4 L. esculentum
Virus
Virulent virus isolates known
Selected references ~
TMV TMV ZYMV TMV
No Yes
Holmes (1929) Melchers et aI. (1966) Paris et al. (1988) Pilowski et al. (1981)
TMV PVX BYMV
Yes Yes Yes
SbMV
Yes
TMV TMV
Yes Yes
BSMV SCPMV
Yes
BYMV TSWV
No Yes
No B
Yes
Cirulli and Alexander (1969) Jones (1982) Schroeder and Provvidenti (1968) Buzzell and Tu (1984) Fraser et al. (1980) Berzal-Herranz et al. (1995), Dardick and Culver (1997), Dardick et al. (1999) Sisler and Timian (1956) Hobbs et al. (1987) Provvidenti and Schroeder (1973) Finlay (1953)
Resistance considered to be conferred by three genes, zym-1, zym-2 and zym-3, of which zym-1 is essential and zym-2 and zym-3 lower the degree of susceptibility (Paris and Cohen, 2000), with permission.
virus, that virus is said to be n o n - p a t h o g e n i c for that species or cultivar. A virus is pathogenic if it usually causes systemic disease in a species or cultivar. A gene for i m m u n i t y or resistance i n t r o d u c e d into such a species or cultivar m a y m a k e the virus avirulent. H o w e v e r , a m u t a n t strain m a y o v e r c o m e the host gene resistance and it w o u l d then be virulent. Gene-for-gene relationships have been proposed by some a u t h o r s for v i r u s - h o s t interactions (e.g. Fraser, 1987a). A well-studied example of the gene-for-gene h y p o t h e s i s applied to a plant virus a n d its host is the resistance of t o m a t o to ToMV (Table 10.3). There are three resistance genes: Tm-1, Tm-2 a n d Tm-22. The virus has evolved variants that can o v e r c o m e all three host resistance genes. The virulent virus strains are n u m b e r e d according to the host genes they can overcome. For example, strain 0 is avirulent. Strain 2 overcomes gene Tm-2; strain 1.22 o v e r c o m e s genes 1 a n d 22. N o virus strains are yet k n o w n that o v e r c o m e host genes 2 plus 22 or 1, 2 plus 2.22. The host genotypes differ in their 'durability' in the field.
Thus, strain I isolates a p p e a r e d w i t h i n a year in commercial crops containing only the Tm-1 gene. By contrast, m o s t c o m m e r c i a l ToMVresistant t o m a t o cultivars n o w c o n t a i n the genes Tin-l~ T m - l : T m - 2 / Tm-22 or T m - 1 / + :Tm2 / T m - 2 2 , a n d these a p p e a r to be highly durable in their resistance to TMV (Fraser, 1985). The relationships s u m m a r i z e d in Table 10.3 bear a superficial similarity to the gene-forgene relationships seen b e t w e e n bacterial a n d f u n g a l p a t h o g e n s a n d their hosts. Such p a t h o g e n s contain large n u m b e r s of genes a n d could easily m a i n t a i n or d e v e l o p a suite of genes that either allow or o v e r c o m e host resistance. By contrast, ToMV contains four genes, all w i t h functions i n v o l v e d in virus replication or m o v e m e n t . The change to virulence by the virus m u s t involve m u t a t i o n a l events in one or m o r e of the four genes or in the controlling elem e n t s of the viral genome. A n e x a m p l e is prov i d e d by the w o r k of Meshi et al. (1988). They e x a m i n e d the nucleotide sequence of a resistance breaking strain of ToMV, Ltal, w h i c h is able to replicate in t o m a t o e s w i t h the Tm-1 gene
706
~6 C O N T R O L AND USES OF PLANT VIRUSES
(i.e. a strain of genotype 1 in Table 10.3). They found two base substitutions resulting in amino acid changes Gln979-~ Glu and His984-* Tyr in the 130- and 180-kDa viral proteins. They demonstrated that these were indeed the changes responsible for resistance breaking by introducing these mutations singly or together into the parent strain. All three constructed mutants replicated in tomato protoplasts with the Tm-1 ~gene, but the change His984 ~ Tyr did so to a greatly reduced extent. Another resistance breaking strain of ToMV, Ltbl, is able to multiply in tomatoes with the Tm-2 gene. Nucleotide sequence analysis revealed two changes in the 30-kDa protein compared with the parent ToMV (Cys66 -* Phe and Glu133 ~ Lys) (Meshi et al., 1989). Thus, the Tm-2 resistance in tomato may involve some aspect of cell-to-cell movement. Mutations in virus genotypes 2 and 2 2 of Table 10.3 involve nucleotide changes in the TMV gene coding for the movement protein, as has been found for the ts mutant LsI (see Chapter 9, Section II.D) (Nishiguchi and Motoyoshi, 1987). Thus, for viruses, virulence or avirulence appears to be controlled by single point mutations in genes essential for virus replication or movement, rather than some change in genes dedicated to maintaining or overcoming host resistance. 2. Mechanisms of host immunity and resistance (reviewed in Fraser, 1998) Details of what is known of some molecular mechanisms of host immunity and resistance are described in Chapter 10. Here I discuss some aspects related to the control of plant viruses. a. Resistance of cowpea to C P M V
Seedlings of the cowpea cultivar Arlington are resistant to CPMV, and isolated protoplasts are also resistant. This cultivar is therefore immune, as defined in Table 10.1. This immunity is governed by a single dominant gene as determined by crosses with the susceptible Black Eye variety. Ponz et at. (1988a) found three inhibitory activities in extracts of Arlington cowpea protoplasts that were at
higher levels than those in Black Eye extracts. These were: (1) inhibitor(s) of the translation of CPMV RNAs; (2) proteinase(s) that degrades CPMV proteins; and (3) an inhibitor of proteolytic processing of a CPMV polyprotein. The proteinases were not specific for CPMV proteins and were not coinherited with the immunity to CPMV. The inhibitor of polyprotein processing was specific for CPMV and had the co-inheritance expected for an agent mediating the immunity to CPMV. It can be reasonably concluded that the proteinase inhibitor is the host-coded gene product responsible for immunity to CPMV. This is the first such gene product identified for plant viruses and the first example of immunity for which a clear molecular mechanism has been established. The inheritance of the translation inhibitor activities was complex but one or more of these may play an accessory role in immunity. In another study, Ponz et al. (1988b) found that the Arlington cowpea line also had a single dominant gene for resistance to TRSV, but that this was quite distinct from the gene for immunity to CPMV.
b. Impaired systemic movement of the virus In some cultivars that have a genotype conferring resistance to a particular virus, movement through the vascular tissue is affected in a manner that appears not to involve a virus-coded cell-to-cell movement protein in the way discussed in a subsequent section. For example, the spread of PLRV infection within the phloem of resistant potato cultivars appears to be impaired, leading to much less efficient acquisition of the virus by the aphid vector M y z u s persicae (Barker and Harrison, 1986). Similarly, in corn (Zea mays) resistance to MDMV, the pattern of virus spread in inoculated leaves suggested that the plant inhibited the virus from moving through the vascular system (Lei and Agrios, 1986). The molecular mechanism for such host effects is unknown. Expression of the TMV 30-kDa movement protein is strongly and selectively enhanced in tobacco protoplasts by actinomycin (Blum et al., 1989). It was suggested that the drug may act by selectively inhibiting a host factor that nor-
V. CONVENTIONAL RESISTANCE TO PLANT VIRUSES
mally suppresses the expression of the 30-kDa viral protein. c. Non.specific virus inhibitors
Extracts of many plants contain substances that inhibit infection by viruses when mixed with the inoculum. Some may inhibit virus replication in experimental systems, and many are known to be proteins or glycoproteins. The relevance of any of these substances to plant resistance to viruses has not been established. One of the most studied is a 29-kDa protein isolated from Phytolacca americana that is known to inhibit a wide range of eukaryotic ribosomes. Ready et al. (1986) showed that the Phytolacca protein is located mainly in the cell wall. They suggested that when injury to a cell occurs during the process of virus infection, the inhibitor enters the cytoplasm and shuts off virus synthesis.
d. Ineffective viral genes As discussed in Chapter 9 (Section II.D), many viruses code for a gene product that is necessary for cell-to-cell movement. If a movement protein is ineffective in a particular host plant then virus replication is confined to the initially infected cells. Consequently, the plant shows extreme hypersensitivity (see Table 10.1) (although not necessarily a necrotic response) and is effectively resistant to the virus. A virus that does not normally move systemically in a particular plant species may do so in the presence of an unrelated virus that does infect systemically and complements the movement of the defective virus (see Chapter 9, Section II.D.6). This indicates that a mismatch between viral movement protein and plant species may be a quite common mechanism for plant resistance. 3. Clustering of resistance genes In several plant species, the resistance virus resistance genes are clustered to specific loci on the chromosomes; in this, virus resistance resembles that for fungi (Boiler and Meins, 1992; Pryor and Ellis, 1993). For instance, in Pisum sativum, resistances to the lentil strain of PSbMV, BYMV, WMV-2, CYVV and BCMV
707
NL-8 strain are controlled by tightly linked recessive genes on chromosome 2 (Provvidenti and Alconero; 1988a; Provvidenti, 1990). Other examples include the close linkages between resistance to WMV and ZYMV in melon (Cucumis melo) (Anagnostou et al., 2000) and between resistance to PRSV-W, ZYMV, WMV and MWMV in cucumber (C. sativa) (Grumet et at., 2000). C. T o l e r a n c e The classical example of genetically controlled tolerance is the Ambalema tobacco variety. TMV infects and multiplies through the plant, but in the field, infected plants remain almost normal in appearance. This tolerance is due to a pair of independently segregating recessive genes rml and rm2 , and perhaps to others as well, with minor effects. Other examples are known where either a single gene or many genes control tolerance. For example, tolerance of a set of barley genotypes to BYDV was controlled by a single major gene probably with different alleles giving differing degrees of tolerance (Catherall et al., 1970). From a study of the relative abundance of D N A forms and viral RNAs in plants infected with CaMV, Covey et al. (1990) concluded that the expression of the CaMV minichromosome is a key phase in the replication cycle that is regulated by the host. Kohlrabi is a host that is tolerant of infection compared with turnip. In this host, high levels of supercoiled DNA accumulated, with very little generation of RNA transcripts, viral products or virus. D . U s e of c o n v e n t i o n a l r e s i s t a n c e for control A review of the consideration in a breeding p r o g r a m for resistance to an important virus, that causing rhizomania of sugar beet, is given in Scholten and Lange (2000). Many of the aspects that they discuss are applicable to breeding programs for resistance to other viruses. In this section, I describe some examples of the application of conventional resistance to the control of viruses.
708
16 C O N T R O L A N D USES OF P L A N T VIRUSES
1. Immunity The apparent i m m u n i t y of seven raspberry cultivars to RVCV was confirmed by the fact that they could not be infected by graft inoculation (Jennings and Jones, 1986). Certain cultivars of swede (Brassica napus) appear to be i m m u n e to TuMV (Tomlinson and Ward, 1982), and certain cultivars of barley appear to be i m m u n e to BaYMV. Plants remained uninfected with the virus following root inoculation with virusbearing cultures of the fungal vector Polymyxa graminis. Transmission by zoospores from the fungus growing on roots of an i m m u n e cultivar was rare or absent. However, although m a n y searches have been made, true i m m u n i t y against viruses a n d viroids, which can be incorporated into useful crop cultivars, is a rather u n c o m m o n p h e n o m enon. For example, Singh and Slack (1984) found no i m m u n i t y to PSTVd a m o n g 555 introductions belonging to 81 tuber-bearing Solanum species. 2. Resistance Where suitable genes can be introduced into agriculturally satisfactory cultivars, breeding for resistance to a virus provides one of the best solutions to the p r o b l e m of virus disease. Attempts to achieve this objective have been made with m a n y of the virus diseases of major importance. Genes for resistance or hypersensitivity have often been found, but it has frequently p r o v e d difficult to incorporate such factors into useful cultivars.
A hypersensitive reaction to a virus involving necrotic local lesions with no systemic spread, as discussed in Chapter 10 (Section III), can give effective resistance in the field. For example, the necrotic reaction of N. glutinosa to TMV was bred into some commercial lines of N. tabacum (Valleau, 1946). It has since been used widely with Virginia-type tobacco cultivars, but not with flue-cured types. Occasionally, very high resistance to a virus has been discovered even in a plant where most k n o w n cultivars were highly susceptible. However, the designation of a variety or genotype as very highly resistant to a particular virus must always be provisional, because it is always possible that a m u t a n t of the virus will arise that can overcome the plant resistance. The major problem with resistance of any sort as a control measure is its durability. H o w long can it be d e p l o y e d successfully before a resistance breaking (virulent) strain of the virus emerges? Fraser (1992) listed 87 host-virus combinations for which resistance genes have been found. Of those tested, virulent virus isolates able to overcome these resistances are k n o w n for more than 75% of these (Table 16.6). Fewer than 10% of the resistance genes listed have remained effective w h e n tested against a wide range of virus isolates over a long period. However, some of the virulent isolates were found only in laboratory tests rather than field outbreaks. The costs of a breeding p r o g r a m must be w e i g h e d against the possible gains in crop
TABLE 16.6. Summary of occurrence of resistance-breaking isolates Immunity or subliminal infection
Local lesion
Partial localization
Resistance phenotype Dominant alleles Incompletely dominant Recessive dominant
5 0 6
22 0 0
Virulent isolates reported Dominant alleles Incompletely dominant Recessive
Yes 20 9 9
No 4 3 4
Not tested 16 3 8
Total
38
11
27
1 4 3
Systemically effective
3 11 9
From Fraser (1992), with kind permission of the copyright holder, 9 Kluwer Academic Publishers.
Not known
V. CONVENTIONAL RESISTANCE TO PLANT VIRUSES
yield (see Buddenhagen, 1981). M a n y factors are involved, such as: (1) the severity of the viral disease in relation to other yield-limiting factors; (2) the 'quality' of the available resistance genes (for example, resistance genes against CMV are usually 'weak' and short lived, which m a y be due, at least in part, to the m a n y strains of CMV that exist in the field); (3) the i m p o r t a n c e of the crop (compare, for instance, a minor ornamental species with a staple food crop such as rice); and (4) crop quality. Good virus resistance that gives increased yields may be accompanied by poorer quality in the product, as h a p p e n e d with some TMVresistant tobacco cultivars (Johnson and Main, 1983). a. Sources of resistant genotypes
Quite frequently, no resistance could be found for particular crops and viruses. For example, no resistance to BWYV was found in 70 cultivars and 500 breeding lines of lettuce (Watts, 1975) but some was found later (Maisonneuve et al., 1991). Russell (1960) found no resistance to BYV in 100 000 beet seedlings. In such circumstances, it m a y be necessary to search for resistance a m o n g wild species. In general, there is certainly a need to broaden the search for resistance genes. For example, in the UK, 22 horticultural crops have no k n o w n source of
resistance to a total of 25 viruses k n o w n to affect them (Fraser, 1988). Nevertheless, m a n y effective sources of resistance have been found and are currently in use (for list see Khetarpal et al., 1998). A m o n g the main sources of genetic resistance to pest and pathogens are the centers of origin and regions of diversification of the crop species. In these places plants have been exposed to selective pressure from the pest or pathogen over long periods of time and thus have developed resistance ((Leppik, 1970). This is well illustrated by the S o l a n u m species that contain sources of resistance to PLRV (Table 16.7). Sometimes certain existing lines or varieties have been found to have a useful degree of resistance, as with the resistance of Corbett refugee b e a n to BCMV ( Z a u m e y e r a n d Meiners, 1975). Resistance m a y not always be uniform, however. For example, even within inbred lines of corn there was variation in resistance to M D M V (Louie et al., 1976). In cabbage, variation in degree of resistance to TuMV was found b e t w e e n different lines of the same variety (Polak, 1983). Occasionally, useful sources of resistance can be identified by m a k i n g initial selections from plants s h o w i n g good g r o w t h in o t h e r w i s e severely infected fields (e.g. Cope et al., 1978). With i m p o r t a n t crops, searches for sources of resistance have been m a d e on a w o r l d w i d e
TABLE 16.7 Sotanum species containing sources of resistance to PLRV Species
Series
Ploidy
Country of origin
Cultivated or wild
Tuber-bearing
S. S. S. S.
Etuberosa Etuberosa Megistacroloba Yungasensa
2• 2• 2• 2•
Wild Wild Wild
No No Yes
Wild
Yes
S. acaule
Acaulia
4•
S. demissum S. phureja
Demissa Tuberosa
6• 2•
Wild Wild
Yes Yes
Cultivated
Yes
S. tuberosum s s p . andigena
Tuberosa
4•
S. tuberosum s s p . tuberosum
Tuberosa
4•
Chile Argentina, Chile Peru Argentina, Bolivia, Paraguay, Uruguay Argentina, Bolivia, Peru Mexico Bolivia, Colombia, Ecuador, Peru, Venezuela Argentina, Bolivia, Colombia, Ecuador, Peru, Venezuela Chile
Cultivated Cultivated
Yes Yes
etuberosum brevidens raphanifolium chacoense
From Barker and Waterhouse (1999), with permission.
709
710
]6 C O N T R O L A N D USES OF P L A N T VIRUSES
basis. Timian (1975) tested 4889 entries in the Barley World collection and found 44 that showed no symptoms of infection with BSMV in the field. As another example, sources of resistance to CSSV in Ghana have been found in the u p p e r A m a z o n region (Legg and Lockwood, 1977), and attempts have been m a d e to combine resistance from various sources (Kenten and Lockwood, 1977). Useful resistance has sometimes been found among a collection of mutants induced by physical or chemical means. For example, Ukai and Yamashita (1984) identified such a barley mutant that was highly resistant to BaYMV. In principle, culture of plant cells as protoplasts offers several possibilities for obtaining new sources of resistance to viruses. However, no commercially successful virus-resistant cultivars have yet been derived by this means. The topic has been reviewed by Shepard (1981). When plant cells are grown in culture for a time and then plants regenerated from single cells or small clusters of cells, considerable genetic variation or somaclonal variation m a y be observed in various properties, including resistance to disease. Attempts have been made to increase the frequency and range of variations by treatment of the tissue in culture with mutagens. Somaclonal variants of tomatoes p r o d u c e d by adventitious shoot formation from leaf discs showed some resistance to ToMV (Smith and Murakishi, 1993). Protoplast fusion can take place in vitro even between protoplasts belonging to different genera. In principle, this offers the possibility of introducing virus resistance genes into a crop cultivar from a quite distantly related species. The donor cells are often irradiated before fusion to f r a g m e n t the chromosomes. Following fusion, chromosomes and parts of chromosomes are eliminated in a fairly random manner, until a more or less stable set of chromosomes is achieved. The hope is that a cell line will arise that contains a functional set of chromosomes from the crop cultivar, together with a minimal a m o u n t of alien genetic material containing the desired resistance gene or genes. In model experiments, transfer of methotrex-
ate and 5-methyltryptophan resistances from carrot to tobacco was achieved by fusion between leaf mesophyll protoplasts of tobacco and irradiated cell culture protoplasts of carrot (Dudits et al., 1987). Some of the regenerated plants had tobacco morphology and independently segregating genes for the two resistance markers from carrot. Intergeneric hybridization in a search for resistance to BYDV in wheat has been reviewed by Comeau and Plourde (1987). Substantial resistance has been found in about 20 species in the tribe Triticeae, which could be, or have been, h y b r i d i z e d to cultivated w h e a t or close relatives.
b. Low seed transmission The discussion in the previous section was concerned with resistance of the growing plant to virus infection. For those viruses affecting annual crops that are transmitted through the seed, resistance to seed transmission may be an important method for limiting infection in the field. For example, a single recessive gene conditions resistance to seed transmission of BSMV in barley (Carroll et al., 1979). The resistance gene, found in an Ethiopian barley called Modjo, was introduced into a new variety, Mobet, with good agronomic qualities and high resistance to seed transmission of Montana isolates of BSMV (Carroll et al., 1983). Resistance to seed transmission has also been found for LMV in lettuce and for some legume viruses.
c. Adequate testing of resistant material Varieties or lines showing resistance in preliminary trials must then be tested under a range of conditions. Important factors to be considered are strains of the virus, climatic conditions (e.g. Thompson and Hebert, 1970) and inoculum pressure (e.g. Kenten and Lockwood, 1977).
d. The need for resistance to multiple pathogens The difficulties in finding suitable breeding material are c o m p o u n d e d w h e n there are strains of not one but several viruses to consider. Cowpeas in tropical Africa are infected to
V. ( : O N V E N T I O N A L RESISTANCE TO PLANT VIRUSES
a significant extent by at least seven different viruses. In such circumstances, a breeding p r o g r a m may utilize any form of genetic protection that can be found. Sources of resistance, hypersensitivity or tolerance have been found for five of the viruses (Matthews, 1991). However, several of these viruses have different strains or isolates that may break resistance to other isolates (van Boxtel et at., 2000). There is of course the further problem of combining these factors with multiple resistance to fungal and bacterial diseases. For example, genetic resistance to TMV, cyst nematodes, root-knot nematodes, and wildfire from Nicotiana repanda has been incorporated into N. tabacum (Gwynn et al., 1986).
e. Durability of resistance and the emergence of resistance-breaking strains For some crop plants and viruses, resistance has proved to be remarkably durable. Thus, the resistance to BCMV found in Corbett Refugee bean has been bred into most varieties of dry and snap beans in the United States, and the resistance had not broken d o w n after 45 years (Zaumeyer and Meiners, 1975). One of the most noteworthy examples of durability has been the resistance in sugar beet to the BCTV (Duffus, 1987), the original selections for which were made in the 1920s. The resistance is multigenic and appears to involve a lower concentration of virus in resistant plants that do become infected and a m u c h longer incubation period in resistant varieties. However, over a period of years a series of more aggressive strains of the virus has emerged. Resistance to TuMV in lettuce described more than 40 years ago has been tested with strains of the virus from m a n y countries, but the resistance has not been broken (Duffus, 1987; Robbins et al., 1994). A hypersensitive type of resistance to CCMV in cowpeas was overcome by certain strains of the virus (Paguio et al., 1988). Resistance breaking strains of PVX have been described in the UK for resistant potato varieties, but these have not yet become a serious practical problem (Jones, 1985). RBDV has been controlled in
711
Scotland by the use of cultivars that are i m m u n e to the prevalent strain of the virus. This situation has been threatened by the discovery of a resistance-breaking strain (Murant et al., 1986). Once a substantial population of resistant plants is exposed in the field there is a good probability that a new strain of the virus will evolve, or be introduced, that can overcome the resistance. The problem of virus strains in the development of a breeding p r o g r a m is well illustrated by the tests carried out by Rast (1967a,b) with 64 different isolates of ToMV on 30 clones of Lycopersicum peruvianum. Different isolates (even from the same strain of ToMV as judged by s y m p t o m s on tomato or tobacco) differed in the range of L. peruvianum clones they would infect. Every clone could be infected by at least one isolate. Even w h e n an apparently successful resistant variety has been developed, it may be important to maintain other measures to minimize contact of the resistant variety with the virus concerned. For example, in tomatoes resistant to ToMV, the virus did not move systemically for several weeks and reached only a low concentration (Dawson, 1967). Extracts from infected resistant plants were more infective for healthy resistant plants than virus from susceptible lines. Resistant plants infected with such virus showed more obvious symptoms. The Tm22 gene in tomato was very useful for protection against ToMV for more than 10 years. However, ToMV strains have now been found that can, in the laboratory, overcome the resistance due to the Tm22 gene. It is probably only a matter of time before these become prevalent in commercial glasshouses. We must conclude that, on present knowledge for most crops and most viruses, the search for new sources of resistance and their incorporation into useful cultivars will be a continuing and very long-term process, as it is with m a n y fungal and other parasitic agents. The interrelations between host genetic factors and other epidemiological aspects of virus diseases are discussed by B u d d e n h a g e n (1983).
712
16 C O N T R O L A N D USES OF PLANT VIRUSES
3. Tolerance Where no source of genetic resistance can be found in the host plant, a search for tolerant varieties or races is sometimes made. However, tolerance is not nearly as satisfactory a solution as genetic resistance for several reasons. 1. The infected tolerant plants may act as a reservoir of infection for other hosts. Thus, it is bad practice to grow tolerant and sensitive varieties together under conditions where spread of virus may be rapid. 2. Large numbers of virus-infected plants may come into cultivation. The genetic constitution of host or virus may change to give a breakdown in the tolerant reaction. 3. The d e p l o y m e n t of tolerant varieties removes the incentive to find immunity to the virus until the tolerance breaks d o w n in an 'out of sight, out of mind' attitude. 4. Virus infection may increase susceptibility to a fungal disease (see Chapter 10, Section V.G). However, tolerant varieties may yield very much better than standard varieties where virus infection causes severe crop losses and where large reservoirs of virus exist under conditions where they cannot be eradicated. Thus, tolerance has, in fact, been widely used (see Posnette, 1969). Cultivars of wheat and oats commonly grown in the US Midwest have probably been selected for tolerance to BYDV in an incidental manner, because of the prevalence of the virus (Clement et al., 1986). Tolerance to MSV has been found in maize and rapidly incorporated into high-yielding maize populations for use in tropical Africa (Soto et al., 1982). Walkey and Antill (1989) obtained an unusual result with a variety of garlic called Fructidor. The yield of selected stocks was significantly less than that of unselected infected stocks. Salomon (1999) argues that breeding for tolerance has an advantage over breeding for resistance in that the selection pressures for the development of resistance-breaking strains are less. This may be so for fungal and bacterial pathogens but it is open to question as to whether this consideration pertains to viruses as well.
VI. T R A N S G E N I C P R O T E C T I O N A G A I N S T P L A N T VIRUSES A. Introduction It is n o w possible to introduce almost any foreign sequence into a plant and obtain expression of that sequence. In principle, this should make it possible to transfer genes for resistance or immunity to a particular virus, across species, genus and family boundaries. Furthermore, genes can be designed to interfere with directly, or induce the host to interfere with, the virus replication cycle. Several approaches to producing transgenic plants resistant to virus infection are being actively explored. There are essentially three sources of transgenes for protecting plants against viruses: (1) natural resistance genes; (2) genes derived from viral sequences, giving what is termed pathogen-derived resistance (PDR); and (3) genes from various other sources that interfere with the target virus. These are discussed in the following sections. There have been numerous reviews on the subject, including Beachy (1993, 1997), Fitchen and Beachy (1993), Wilson (1993), Scholthof et al. (1993), Baulcombe (1994), Hull (1994b), Wilson and Davies (1994), Lomonossoff (1995), Prins and Goldbach (1996), Palukaitis and Zaitlin (1997), Kaniewski and Lawson (1998), Bendahmane and Beachy (1999), H a m m o n d et al. (1999), Waterhouse and Upadhyaya (1999). This selection of reviews, over 7 years, illustrates the rapid development of understanding of this subject.
B. Natural resistance genes Molecular aspects of genes found in plant species that confer resistance to various viruses are discussed in Chapter 10. When a resistance gene has been identified, it can be isolated and transferred to another plant species. The R x l gene, which gives extreme resistance to PVX, has been isolated from potato and transformed into Nicotiana benlhamiana and N. tabacum (Bendahmane et al., 1999) where it gives resistance to the virus. Similarly, the N gene, found naturally in N. glulinosa, and which confers hypersensitive resistance to
VII. PATHOGEN-DERIVED RESISTANCE
TMV, gives resistance to TMV when transferred to tomato (Whitham et at., 1996).
VII. PATHOGEN.DERIVED RESISTANCE The ideas leading up to the concept of pathogen-derived resistance for plant viruses were first postulated by Hamilton (1980) and are encapsulated as a general concept in a paper by Sanford and Johnson (1985). They suggested that the transgenic expression of pathogen sequences might interfere with the pathogen itself terming this concept 'parasitederived resistance'. Since then, several names have been used for this approach including 'non-conventional protection', 'transgenic resistance' and 'engineered virus resistance', but the generally accepted term is now pathogen-derived resistance (PDR). Since the mid-1980s, this approach has attracted major interest and is the main one by which transgenic protection is being produced against viruses in plants. The first demonstration of PDR against plant viruses was by Powell-Abel et at. (1986), who showed that the expression of TMV coat protein in tobacco plants protected those plants against TMV. This opened the floodgates for extensive research both on protecting crop species against viruses and on the mechanisms involved.
TABLE
713
The basic idea arising out of Sanford and Johnson's (1985) concept is that, if one understands the molecular interactions involved in the functioning of a pathogen, mechanisms can be devised for interfering with them. Although this concept applies to all pathogens and invertebrate pests, it has mainly been used against viruses because of their relatively simple genomes. In developing the concept it was recognized that the interactions of interest occur at all stages of the virus infection cycle and that they can potentially be interfered with in various ways, for example by blocking the interaction or by decoying one or more of the molecules involved in the interaction. This then led to the idea that the overall strategy as being one of attacking specific viral 'targets' with specific molecular 'bullets'. Some examples of targets and attacking mechanisms, or bullets, are given in Table 16.8. However, in practice, much of the development of this approach was done without detailed knowledge of the precise molecular mechanisms involved, and analysis of these results has thrown light on several new mechanisms. Perhaps the most important is the gene-silencing p h e n o m e n o n described in Chapter 10 (Section IV) and which is further discussed in Section VII.B.3 below. In this rapidly expanding subject, there are various terminological problems. The main one, whether to term this phenomenon resistance or protection, is discussed in Section
16.8 Examples of 'targets' and 'bullets' for pathogen-derived resistance
Targets
Bullets
Viral gene products Coat protein Replicase Cell-to-cell spread function Protease
Molecular blockers Viral gene products Mutated viral gene products Antisense nucleic acid + ribozymes Sense nucleic acid Antibodies
Control sequences Replication control sequences Origins of replication Primer binding sites Expression control sequences Subgenomic RNA promoters Translational leader sequences Splice sequences
Decoys Nucleic acid control sequences Satellites Protease sites Other Non-host resistance
From Hull (1994b), with kind permission of the copyright holder, 9 Kluwer Academic Publishers.
714
~4 C O N T R O L A N D USES OF P L A N T V I R U S E S
VII.B.7. I will use the term protection wherever possible, but in situations where it has been used widely (such as pathogen-derived resistance) I will retain the term resistance. Currently, there are two basic molecular mechanisms by with PDR is thought to operate. In some systems the expression of an unmodified or a modified viral gene product interferes with the viral infection c y c l e - this I will term protein-based protection. The second mechanism does not involve the expression of a protein product, and I will call this nucleic acid-based protection.
A. P r o t e i n - b a s e d p r o t e c t i o n As noted above, the first demonstration of PDR involved the expression of TMV coat protein (Powell-Abel et al., 1986). Since then, there have been many examples of the use of this coat protein-mediated protection. The expression of other viral gene products also gives protection to a greater or lesser extent against the target virus. 1. Transgenic plants expressing a viral coat protein The sequences encoding viral coat proteins are the most widely used for conferring protection in plants (Fitchen and Beachy, 1993); coat protein genes from at least 35 viruses, representing 15 viral taxonomic groups, have been transformed into many different plant species (Palukaitis and Zaitlin, 1997). This is because this gene was used in the first example of this approach and because coat protein genes are relatively easy to identify and clone. The phenomenon is often referred to as 'coat proteinmediated resistance' (CP-MR). a. Tobacco mosaic virus coat protein (reviewed by Beachy, 1999) Beachy et al. (1986) and Bevan et al. (1985) first reported the expression of the coat protein of TMV in tobacco plants into which a cDNA containing the coat protein gene had been incorporated. This was quickly followed by reports on the expression of the coat protein gene of AMV by several groups (e.g. Loesch-Fries et al., 1987),
that of TRV by van Dun et al. (1987) and that of PVX by Hemenway et al. (1988). Some laboratories used the CaMV 35S promoter and others the 19S promoter to obtain expression. The 35S promoter was considerably more effective. Powell-Abel et aI. (1986) showed that transgenic plants expressing TMV coat protein either escaped infection following inoculation or developed systemic disease symptoms significantly later than plants not expressing the gene. Plants that showed no systemic disease did not accumulate TMV in uninoculated leaves (Nelson et at., 1987). Transgenic plants produced only 10-20% as many local lesions as controls when inoculated with a strain of TMV causing local lesions. The idea that transgenic plants resist initial infection rather than subsequent replication was suggested by results obtained using transgenic Xanthi nc tobacco plants, in which fewer local lesions were produced than on control plants. However, the lesions that did develop were just as big as those on control leaves, indicating that once infection was initiated there was no further block in the infection cycle. In transgenic plants it would be possible, in principle, for the resistance to infection to be due to the coat protein mRNAs transcribed from the cDNA, to the coat protein itself, or to both of these molecules. To test these possibilities, Register et al. (1988) constructed a series of cDNAs generating mRNA sequences that would produce no coat protein, or mRNA that lacked the replicase recognition site but that would produce coat protein. These experiments conclusively implicated the coat protein rather than the mRNA in causing resistance to superinfection. Further experiments with TMV confirmed the earlier results and showed that the 3' tRNA-like sequence was not necessary to generate resistance (Powell et al., 1990). Register et al. (1988) and Register and Beachy (1988) showed that protoplasts made from transgenic plants expressing coat protein were specifically protected against infection with TMV. When tobacco protoplasts took up coat protein, they were transiently protected from infection with TMV (Register and Beachy, 1989). Thus, coat protein outside the cell is probably not involved in coat protein-mediated
VIl. PAT~tOGEN-I)ERIVED RESISTANCE
protection. Pathogenesis-related proteins (see Chapter 10, Section III.K.1) also do not appear to be involved in this resistance (Carr et al., 1989). b. Dose and sequence dependency of protection by T M V coat protein
The greater the amount of virus inoculum, the lower is the protection afforded by TMV coat protein (Nejidat and Beachy, 1990; Bendahmane et al., 1997b). There is a positive correlation between the level of protections and the sequence similarity between the transgene coat protein and that of the challenge virus (Nejidat and Beachy, 1990). For instance, the coat proteins of ToMV and TMGMV have 82% and 72% sequence identity, respectively, with that of TMV, whereas that of RMV is only 45% identical; TMV coat protein gave much better protection against ToMV and TMGMV than against RMV. It appears likely that it is the structure of the coat protein and possibly differences in the sites of carboxylcarboxylate interactions (see Chapter 5, Section III.B.5) that influence the protection given against different tobamoviruses (Bendahmane and Beachy, 1999). There was little or no protection against viruses from other families or genera (e.g. AMV, CMV, PVX or PVY) (Anderson et al., 1989). Transgenic expression of TMV coat protein does not protect against inoculation with viral RNA (Nelson et al., 1987). c. Mechanism of T M V coat protein protection (reviewed by Reimann-Philipp and Beachy, 1993; Bendahmane and Beachy, 1999)
A careful analysis of virus spread in single lesions in transgenic and control tobacco plants showed that, when local infection did take place in tissues expressing the coat protein, there was no inhibition in subsequent cellto-cell movement (Register et al., 1988). On the other hand, when a leaf-bearing stem segment from a transgenic plant was grafted between lower and upper sections of a non-transgenic plant, systemic movement of TMV into the leaves above the graft was inhibited. Transgenic tobacco plants expressing the 30kDa movement protein of TMV were not pro-
715
tected against infection or disease development (Deom et al., 1987). When TMV is treated at pH 8.0, translatability in vitro is greatly enhanced (see Chapter 7, Section II.B.4). Register et al. (1988) and Register and Beachy (1988) found that TMV treated in this manner was able to overcome the resistance of transgenic plants in the same way as RNA. This result certainly supports an early event as being important in the resistance of transgenic plants. Osbourn et al. (1989b) showed that TMV-like pseudo-particles containing the GUS reporter gene expressed this gene 100 times less efficiently in protoplasts from coat protein-transgenic tobacco plants than in control protoplasts. The data suggested that about 97% of the GUS pseudo-particles remained uncoated. However, other experiments indicated that inhibition of virus disassembly is insufficient to account entirely for coat protein-mediated resistance, and that some later event or events in virus replication must be involved. The TMV mutant DT-IG produces no coat protein and does not normally move systemically. When this mutant is inoculated into transgenic tobaccos expressing the coat protein gene, rod-shaped particles were found in the systemically invaded leaves (Matthews, 1991). However, the viral rods isolated from the systemic infection were unable to infect fresh transgenic plants, supporting the view that some early uncoating event is involved in resistance. Osbourn ef al. (1989a) tested the possibility that coat protein expressed in transgenic plants inhibits virus replication by recoating uncoated viral RNA from a challenge inoculum. They produced double-transformed tobacco plants that were expressing TMV coat protein and a reporter gene (CAT) whose transcripts contained a copy of the TMV origin of assembly sequence. No rods could be detected in cell extracts of these plants by electron microscopy, and there was no significant reduction in CAT activity. However, transformed plants retained their ability to resist infection by TMV. Thus, it seems unlikely that re-encapsidation of uncoated RNA of challenge virus by endogenous coat protein is involved in the resistance of transgenic plants expressing coat proteins.
716
16 C O N T R O L
A N D USES OF P L A N T V I R U S E S
In a comparison of the co-translational disassembly (see Chapter 7, Section II.B.4, for a description of co-translational disassembly) of TMV in CP(+) and CP(-) protoplasts, Wu et al. (1990) showed that TMV recruited polyribosomes within 5-10 minutes and that the input virus was largely undetectable within 60 minutes. In contrast, in the CP(+) protoplasts input virus was not recruited to polyribosomes and was largely intact at 60 minutes. Thus, it appeared most likely that coat proteinmediated resistance is an early event after virus entry into the cell and it was suggested that the transgenically expressed coat protein blocks disassembly (Wu et al., 1990). There are several hypotheses as to how the coat protein can block disassembly. The first suggests an inhibition of viral uncoating (Register and Nelson, 1992; Bendahmane et al., 1997b) with the transgenic coat protein driving the disassembly reaction towards assembly. The major point in this hypothesis is that the balance between uncoating of the virus particle and particle assembly is controlled by the amount of free coat protein in the relevant cellular compartment. Thus, in the nontransgenic situation, there is little or no free coat protein in the cell during disassembly, and assembly occurs when the amount of 'local' free coat protein becomes significant. In the transgenic plant, the expression of the transgene produces significant amounts of free coat protein, thus switching the balance to particle assembly. The second hypothesis proposes that a cellular site for TMV disassembly is blocked by the transgenic coat protein. To examine the second hypothesis, Clark et al. (1995) postulated that, if the recognition binding transgenic coat protein to a putative receptor site was on the radial surface of the coat protein, the level of resistance against a challenging virus containing SHMV sequences on its surface would be similar to the resistance against SHMV itself. However, plants expressing TMV coat protein were just as resistant to the chimeric virus as they were to TMV. Plants challenged with a TMV mutant in which the coat protein was replaced by that of SHMV showed the same low level of resistance as that to SHMV. This experiment gave a
strong indication that the binding site hypothesis was not tenable. The structure of TMV coat protein and factors involved with the assembly of TMV particles are discussed in Chapter 5 (Section IV). A series of experiments on the effects of mutations of TMV coat protein designed to affect virus assembly are reviewed in Bendahmane and Beachy (1999). Although the experiments showed differences in protection, they did not fully resolve the mechanism of the protection. d. Other viral coat proteins
Coat protein-mediated resistance (protection) has been demonstrated for many other virus families and genera (Waterhouse and Upadhyaya, 1999). Plants that were expressing AMV coat protein were also resistant to infection with AMV (e.g. Loesch-Fries et al., 1987; Turner et al., 1987a; van Dun et al., 1987), even though this protein is required for virus replication (see Chapter 8, Section IV.G). Similar results were obtained with a mutated AMV coat protein (van Dun et al., 1988b). Transgenic tobacco plants expressing the PVX coat protein gene were significantly protected against PVX infection, as shown by a reduced number of local lesions on inoculated leaves, delayed or no systemic symptom development, and a reduction in virus accumulation in both inoculated and systemically infected leaves. The higher the level of coat protein expression, the higher was the level of protection (Hemenway et al., 1988). Plants expressing an antisense coat protein transcript were resistant to infection with PVX, but only with low concentrations of virus in the inoculum. The extent of protection provided by such transgenic plants is correlated with the degree of expression of the coat protein gene. Figure 16.14 shows the striking protection obtained with the transgenic expression of CMV coat protein. The extent of protection is reduced as inoculum concentration is increased. Like TMV, the coat proteins of AMV and TRV do not protect against RNA inoculum (LoeschFries et al., 1987; van Dun et al., 1987; Angenent et al., 1990) but, unlike TMV, plants expressing
VII.
PATHO(}EN-DERIVED RESISTANCE
717
Fig. 16.14 Protection of transgenic tobacco plants expressing the CMV coat protein gene against mechanical inoculation with CMV at 25 btg/mL. CP§ plants (back row) and CPplants (front row) were photographed 1 month after inoculation. From Cuozzo et al. (1988), with permission.
high levels of PVX coat protein were resistant to infection with PVX RNA (Hemenway et al., 1988). The expressed coat protein does not necessarily have to be that of the target virus, but it has to be sufficiently closely related. For instance, field resistance in tobacco to PVY was conferred by by expression of LMV coat protein (Dinant et al., 1998). Two coat proteins can be expressed from one construct. Marcos and Beachy (1997) designed a construct comprising the coat proteins of the tobamovirus TMV and the potyvirus SMV together with the highly specific TEV NIa proteinase. In plants transformed with this construct, the proteinase processed the multifunctional polypeptide to give accumulation of the two viral coat proteins. These plants were protected against both TMV and PVY. Similarly, Nicotiana benthamiana plants expressing sequences of the N gene of TSWV and the coat protein of TuMV were protected against both viruses (Jan et al., 2000a). Squash lines expressing the coat protein genes from CMV, WMV 2 and ZYMV were resistant to all three viruses (Fuchs et al., 1998), indicating that these genes can be pyramided. Of interest is 'coat' protein-mediated protection against viruses either that have several
coat protein species or that do not have conventional capsids. Rice plants transformed with one of three structural proteins ($5, $8 or $9) of the oryzavirus RRSV showed resistance to the virus (Waterhouse and Upadhyaya, 1999) but it was not noted whether any of these genes were expressed as proteins. The particles of tenuiviruses are ribonucleoproteins, the main protein being the nucleocapsid (N) protein (see Chapter 2, Section III.I). Rice plants transformed with, and expressing, the N protein of RSV were protected against the virus (Haykawa et al., 1992). 2. Viral movement proteins The proteins encoded by viruses that facilitate their cell-to-cell movement are described in Chapter 9 (Section II.D.2). The relationship between both structure and function of movement proteins indicated that, if one could block the function by, for instance, using a defective mutant protein, a broader resistance might result (Cooper et al., 1996). Expression of the BMV movement protein gave partial resistance to TMV (Malyshenko et al., 1993) and that of the potexvirus triple gene block will confer protection against other viruses with a similar genome organization (Beck et al., 1994; Seppanen et al., 1997). However, a functional
718
16 C O N T R O L A N D USES OF P L A N T VIRUSES
movement protein can complement defective proteins; when the 30-kDa movement protein gene was expressed in transgenic tobacco plants, movement of the transport-defective LSI mutant of TMV was possible (Deom et al., 1987). As noted in Chapter 9, Section II.D, TMV and RCNMV movement proteins will complement the movement of the insect virus, flock house virus, in transgenic plants (Dasgupta et al., 2001). The transgenic expression of the movement proteins of BDMV had a deleterious effect on the plant development (Hou et al., 2000). 3. Viral replicase proteins (reviewed by Palukaitis and Zaitlin, 1997) Transformed tobacco plants containing cDNA copies of AMV RNAs 1 or 2 showed no resistance to AMV (van Dun et al., 1988a). It is not certain whether this result was due to low levels of expression or whether both genes in the same plant might confer some resistance. No resistance was detectable in tobacco plants transformed with the non-structural 13- and 16-kDa genes of strain PLB of TRV, or the 29-kDa gene unique to strain TCM (Angenent et al., 1990). Replicase sequences expressing proteins were shown to provide protection against three plant viruses, AMV (Brederode et al., 1995), CMV (Palukaitis and Zaitlin, 1997) and TMV (Golemboski et al., 1990). Transformation with the Pl or P2 replicase genes of AMV did not give protection and neither did mutants of the P2 gene, N-truncated to resemble TMV 54kDa gene or with the GDD motif (the RNAdependent RNA polymerase catalytic site; see Chapter 8, Section IV.B.1) changed to VDD (Brederode et al., 1995). However, mutation of the GDD motif to GGD, GVD or DDD gave high levels of protection against AMV. As described in Section V.B below, in most cases use of replicase sequences give RNAmediated protection. However, the AMV, CMV and TMV systems show relatively high steady-state levels of accumulation of transgene mRNA. In each case, the construct comprised only part of the replicase protein, the 54-kDa moiety of the TMV replicase (see Chapter 7, Section V.E.1), truncated CMV 2a protein and the AMV 2a protein. The truncated 2a protein encoded by the transgene was
detected for CMV (Carr et al., 1994) but not for the other two viruses, and transgene translatability increased the effectiveness of the protection (Wintermantel and Zaitlin, 2000). Mutations in the sequence encoding the TMV and AMV proteins interfered with protection, suggesting that the protein itself was involved (Carr and Zaitlin, 1992; Brederode et al., 1995). Expression of PVY NIb gene is possibly protein mediated (Audy et al., 1994). The transgenic plants expressing the TMV 54-kDa protein were resistant to infection with TMVU1 or its RNA at high concentrations. They were also resistant to a U1 mutant, but not to two other tobamoviruses or to CMV. Experiments in protoplasts derived from plants transgenic for the 54-kDa protein have shown that a very early event is involved. Carr et aI. (1993) demonstrated that expression of the 54kDa protein was required for protection, but Marano and Baulcombe (1998) suggested that the situation is more complicated, with at least part of the protection being nucleic acid based. Various observations discussed by Patukaitis and Zaitlin (1997) indicate that this form of protection is not a dominant negative mutant effect but point to complex and subtle mechanisms. It is suggested that the interactions between the transgenic replicase proteins and other virus-encoded proteins may affect the processes of replication and cell-to-cell movement at the wrong time in the infection cycle, leading to the arrest of some stage of the replication process. When a construct of ACMV replicationassociated gene (AC1) mutated to alter the putative NTP-binding site was expressed in Nicotiana ben thamiana, the plants showed a delay in symptom appearance a n d / o r mild symptoms (Sangare et al., 1999). A high level of the mutated gene was necessary for protection.
B. Nucleic acid-based protection Four potential forms of protection based on the expression of viral RNA sequences have been recognized: (1) that induced by the viral RNA sequence expressed in (+) sense; (2) that induced by the viral RNA expressed as antisense molecules; (3) that induced by the expres-
VIl. PATHOGEN-DERIVEr) RESISTANCE
sion of satellite RNAs; (4) that in which ribozymes are targeted to viral genomes. Early in the development of coat proteinmediated protection, there were some unexpected observations. For instance, there was no correlation between resistance and the expression of potyvirus, PVY or luteovirus, PLRV coat proteins in potato (e.g. Kawchuk et al., 1990; Lawson et al., 1990). Lindbo and Dougherty (1992a, b) found that the untranslatable coat protein gene of TEV gave higher levels of protection than either full-length or truncated translatable constructs. Similar observations were made for protection given by untranslatable sequences of TSWV (de Haan et al., 1992) and PVY (van der Vlugt et al., 1992). These and other observations suggested that the protection, at least in these cases, was mediated by nucleic acid rather than by protein. 1. RNA-mediated protection (reviewed by PrinT and Goldbach, 1996) As no promoterless transgenes have been shown to confer protection (Lomonossoff, 1995), it must be assumed that either the RNA transcript or a protein that is encoded give the protection. In a plant transformed with a construct that does not give a protein, any protection is obviously due to the RNA. However, not all the plant lines expressing a non-coding RNA show protection. When a plant is transformed with a construct designed to produce a viral protein, it can often be difficult to distinguish between protection due to expression of the protein itself or that due to the RNA transcript. However, there are Various features of the protection that tend to be characteristic for RNAmediated protection (Smith et al., 1994): 9 There is no direct correlation between RNA expression levels and the level of protection (see Pang et al., 1993). 9 Usually, no transgene-encoded protein can be detected and the steady state of the transcript in inoculated plants is often in low amount. 9 The protection is usually narrow and against strains of the virus that have very similar sequences to that of the transgene.
719
9 Unlike coat protein-mediated protection, the protection is not overcome by inoculating RNA. 9 Also, unlike coat protein-mediated protection, RNA-mediated protection is not dose dependent and operates at high levels of inoculum. 9 The insert in the host genome comprises multiple copies of the transgene, particularly with direct repeats of coding regions (Sijen et al., 1996). 9 Copies of the transgene may be truncated and/or in an antisense orientation (Waterhouse et al., 1998; Kohli et al., 1999b). 9 Transgene sequences and sometimes their promoter(s) may be methylated (Jones et al., 1999; Kohli et al., 1999b; Sonoda et al., 1999). When transcript levels have been examined, three general classes of resistance phenotype have been recognized: 1. Plants that are fully susceptible. These plants have low to moderate levels of transgene transcription and steady-state RNA. 2. Plants that become infected and then recover. These have moderate to high levels of transgene transcription and steady-state RNA in uninfected plants, but low-level steady-state RNA in recovered tissues. 3. Plants that are highly resistant. These plants have high levels of transgene expression but low steady-state levels. Since the recognition of the differences between protein- and RNA-mediated protection, there have been many examples of plants transformed with viral sequences that show the properties of RNA-mediated protection. 2. Sequences for RNA-mediated protection RNA-mediated protection has been given by a range of sequences from viral genomes. In many cases, it has resulted from attempts to transform plants with the viral genes described above. Cowpea plants transformed with the CPMV movement protein gene are protected against CPMV by an RNA-based mechanism (Sijen et al., 1995). The resistance also operates against PVX RNA containing the CPMV movement
720
16 C O N T R O L A N D USES OF P L A N T V I R U S E S
protein gene. This protection against PVX is initiated using sequences as small as 60 nucleotides, particularly situated in the 3' part of the transcribed region of the movement protein transgene (Sijen et al., 1996). Transformation with a direct repeat of the movement protein sequences increased the frequency of resistant lines by 20-60%. Transgenic expression in rice of four separate constructs derived from the BMV genome, an artificial DI RNA from RNA2, a sense tRNAlike structure corresponding to the 3' end of RNA2, an antisense sequence to the intergenic region of RNA3, and RNA encoding the virus coat protein, all gave protection against BMV (Huntley and Hall, 1996). Protection is given by the expression of the NIa, NIb and coat protein regions of potyviruses. To determine whether combinations of these genes would give additively greater protection, Maiti et al. (1999) compared transgenic lines expressing TVMV coat protein with those expressing NIa, NIb and coat protein. The plants with the combination of three genes were invariably less resistant to TVMV than those expressing the coat protein gene alone. Furthermore, plants with the three-gene combination had virtually no resistance to TEV in contrast to the coat protein lines. Immunity to BYDV-PAV was conferred on barley plants transformed to express hairpin (hp)RNA (see Chapter 10, Section IV.D) containing the polymerase gene of BYDV (Wang et al., 2001a). Previous attempts to produce transgenic protection against this luteovirus have had only limited success (McGrath et al., 1997; Koev et al., 1998; Wang et al., 2001b). This indicates that the hpRNA construct is important in obtaining high levels of protection. 3. Molecular basis of RNA-mediated protection The recovery phenomenon associated with low steady states of transgene RNA in recovered tissues and the low steady-state RNA levels in highly resistant plants, coupled with the narrow range of protection against viruses with homologous sequences to the transgene, are all suggestive of homology-dependent or posttranscriptional gene silencing. This is described in detail in Chapter 10 (Section IV).
4. Protein- and RNA-mediated protection A single type of construct may confer protection by both protein- and RNA-mediated mechanisms. For instance, transformation with the TSWV N gene sequence conferred resistance to heterologous tospoviruses in plant lines with the highest levels of expressed protein but the most effective protection to the homologous virus in plant lines that had the lowest steadystate levels of RNA and little protein (Pang et al., 1993). Similarly, when barley plants were transformed with constructs to express the coat protein of BYDV-PAV, some resistant lines had detectable levels of coat protein whereas others did not (McGrath et al. 1997). From an analysis of protection by TMV replicase sequences, Goregaoker et aI. (2000) concluded that both RNA and protein sequences were involved in conjunction with the speed of the infecting challenging virus. Nicotiana b e n t h a m i a n a transformed with PMTV coat protein gene showed strong resistance to the virus irrespective of the amount of transcript RNA or coat protein detected (Barker et al., 1998). Lines transformed with a nontranslatable form of PMTV coat protein gene were either not resistant or showed low levels of resistance. Barker et al. (1998) suggested that the form of protection given by PMTV coat protein is unique because, although it depends on coat protein translation to be effective, it mediates very strong resistance. Transformation of potato with constructs expressing a mutant form of PLRV ORF 4, the gene for the movement protein gave transgenic lines that showed broad-range protection against PLRV and also against PVY and PVX (Tacke et al., 1996). One of the transgenic lines had strong protection against PLRV but was susceptible to PVY and PVX. From an analysis of the RNA levels associated with these two phenotypes, it was concluded that the protection against PLRV was RNA mediated and that against PVY and PVX was mediated by the mutant movement protein binding to important sites in the plasmodesmata (Tacke et al., 1996). Palukaitis and Zaitlin (1997) draw attention to the need to test large numbers of independent transformants, not only to obtain lines with
VII.
the best protection characteristics, but also to rule out the possibility that protection is not given by a particular construct. 5. Transgenic plants expressing antisense RNAs (reviewed by Tabler et al., 1998) One method of gene regulation in organisms is by complementary RNA molecules that are able to bind to the RNA transcripts of specific genes and thus prevent their translation. Such RNA has been called antisense or micRNA (messenger-RNA-interfering complementary RNA). Appropriate antisense sequences incorporated into a plant genome have been shown to block the activity of specific genes (e.g. Delauney et at., 1988; van der Krol et al., 1988). The possibility of using this strategy for the control of plant viruses has been explored. Various laboratories have carried out in vitro studies with oligonucleotides complementary to some plant virus RNA sequences. Oligodeoxynucleotides complementary to genomic PVX RNA caused translation arrest in a Krebs2 cell-free system. This was thought to be due to endogenous RNase H activity in the cellfree system (Miroshnichenko et al., 1988). Antisense sequences complementary to sequences near the 5' end of TMV RNA inhibited in vitro translation of this RNA in a rabbit reticulocyte lysate. The inhibition was probably due to direct interference with ribosome attachment (Crum et al., 1988). Morch et al. (1987) found that the 'sense' nucleotide sequences corresponding to the replicase recognition site near the 3' end of genomic TYMV RNA specifically inhibited in vitro the activity of the TYMV replicase isolated from virus-infected plants. The relevance of these various experiments to possible virus inhibition in vivo remains to be determined. Among transgenic tobacco plants containing genes for the production of antisense RNAs for three regions of the CMV genome, only one showed some resistance to the virus (Rezaian et al., 1988). Cuozzo et al. (1988) compared the extent of protection provided in transgenic tobacco plants by the coat protein gene of CMV or its antisense transcript. Symptom development and virus accumulation were reduced or absent in plants transgenic for the sense gene,
PATHOGEN-DERIVED RESISTANCE
721
and this was unaffected by inoculum concentration over the range used. By contrast, antisense plants were protected only at low inoculum concentrations. Transgenic tobacco plants expressing RNA sequences complementary to the coat protein gene of TMV were not protected as strongly from TMV infection as were plants expressing the coat protein gene itself (Powell et al., 1989). In a few cases expression of antisense RNA has given significant protection, for example to BYMV (Hammond and Kamo, 1995). Similarly, both sense and antisense constructs of TRSV coat protein gave protection against the virus, apparently by an RNA-mediated mechanism (Yepes et al., 1996). Antisense RNAs also give protection against geminiviruses (Frischmuth and Stanley, 1993). The targets have ranged from the rare mRNA of the Rep protein of TGMV and TYLCV (Day et al., 1991; Bendahmane and Gronenborn, 1997) to the coat protein of ToMoV (Sinisterra et al., 1999). Antisense RNA has given some protection against viroids (reviewed by Tabler et al., 1998). It is likely that antisense protection operates by mechanisms similar to those of RNAmediated protection described above. 6. Ribozymes (reviewed by Tabler et al., 1998) As described in Chapter 14 (Sections I.C.3.b and II.B.4.b), ribozymes are catalytic RNAs that can cleave at specific sites in complementary target RNAs. Since the ribozyme has to be complementary to the target viral sequence, it can be considered to be an antisense RNA. Incorporation of a ribozyme into an antisense RNA to TMV gave no significant advantage over the antisense RNA itself (de Feyter et al., 1996), but constructs directing PPV that contained a hammerhead ribozyme gave stronger protection than did the ordinary antisense RNA construct (Liu et al., 2000). 7. Levels of protection The reactions of various forms of transgenic protection give a great range of responses. These vary from delay in symptom production for just a few days to complete immunity. This raises the question of definitions of resistance and protection. Many transgenic
722
~6 C O N T R O L
A N [ ) USES OF P L A N T V I R U S E S
plant responses do not fit the definitions outlined in Table 10.1. For instance, can the delay of symptom expression by a few days really be called resistance? Hull and Davies (1992) discuss this point, and suggest the following definitions:
Resistance to a virus is a property of the plant that reduces virus multiplication and reduces or prevents virus spread within the plant and~or symptom expression. Protection is a property conferred to a plant that interferes with the virus infection cycle (the virus infection cycle includes transmission ,from an infected to a healthy plant and the full systemic infection of the healthy plant). Thus, most of the phenotypes described above should be classed as protection. Hull and Davies (1992) also go on to suggest that there should be categorization of field protection, with the seven levels describing the behavior of the transgenic plant (Table 16.9). This would enable the reader or listener to understand the level of protection afforded by the 'gene' being discussed. However, there are some problems in adopting such a system. First, the level of protection may vary between siblings in a transgenic line. In many reports on protection, results are quoted as percent or number of plants showing (or not showing) symptoms. Second, the level of protection may vary according to conditions such as temperature (Nejidat and Beachy, 1989). Third, the level of protection may vary with the generation of progeny from the transformant. Thus, there are certain difficulties in categorizing levels of protection. For instance, the level of protection may vary according to the plant developmental stage (Table 16.10) (Jan et al., 2000b) or possibly due to environmental factors. 8. Relationship between natural cross-protection and protection in transgenic plants The mechanism for transgenic protection against a viral infection, especially coat protein-
mediated protection, has been compared with natural cross-protection or mild strain protection (see Section IV.A). There are several similarities that have been used to support the idea. (1) In both situations the degree of resistance depends on the inoculum concentration, with high concentrations reducing the observed resistance. (2) Both are effective against closely related strains of a virus, less against distantly related strains, and not against unrelated viruses. (3) In some circumstances cross-protection can be substantially overcome when RNA is used as inoculum rather than whole virus (Sherwood and Fulton, 1982; Dodds et al., 1985). Similarly, the resistance of transgenic plants expressing the coat protein of several viruses is substantially but not completely overcome when RNA is used as inoculum (see Section VII.A.1). (4) In classic cross-protection experiments, no crossprotection was observed between two rather similar viruses, AMV and TSV. In experiments with transgenic plants expressing their viral coat proteins, high resistance to infection was observed against the homologous virus and none against the heterologous virus (van Dun et al., 1988b). On the other hand, there appear to be some differences between natural cross-protection and coat protein-induced resistance. When cross-protection between related strains of a virus is incomplete, the local lesions produced may be much smaller than in control leaves (illustrated for PVX in Fig. 17.6). This indicates reduced movement a n d / o r replication of the superinfecting strain. Local lesions that formed in transgenic tobacco plants expressing the PVX coat were smaller than those of the controls (Hemenway et al., 1988), in line with the result for PVX shown in Fig. 17.6. However, the reduced number of local lesions that do form on transgenic plants infected with TMV became as large as those of controls, indicating no block in replication or local movement once infection was successful. It is quite possible that there are several mechanisms that give cross-protection. One of them is likely to involve the post-transcriptional gene silencing host defense system and, thus, to resemble RNA-mediated protection.
VII.
PATHO(]EN-DERIVED RESISTANCE
723
TABLE 16.9 Proposed levels of protection Level
Finding Full immunity to a range of viruses Full immunity to a range of strains of a virus Full immunity to a few closely related strains of a virus Subliminal infection with the virus unable to spread from initially infected cells Delay in systemic infection with a description of duration Reduction in severity of systemic symptoms associated with a reduction in virus titer Reduction of systemic symptoms with no reduction in virus titer
From Hull and Davies (1992), with kind permission of the copyright holder, 9 CRC Press LLC. TABLE 16.10 Reactions of transgenic R1 squash plants to inoculation with SqMV at different developmental stages under field conditions Genotype Control SqMV-22 SqMV-3 SqMV-127
Percentage of plants showing resistance after inoculation at:" 17 DAG 31 DAG 45 DAG 0
0
0
0 0 (21) 79 (7)
7 21 93
7 17 (8) 92
Three upper leaves of each transgenic plant were inoculated at 1, 3 or 5 weeks after transplanting (17, 31 or 45 DAG) (DAG, days after germination). With 1:15 diluted extract of infected squash leaves; 12 to 14 plants were inoculated in each treatment. Values in parentheses indicate plants that displayed the recovery phenotype. From Jan et aI. (2000b), with permission. a
9. Satellite-mediated protection The general n a t u r e of satellite R N A s is described in C h a p t e r 14 (Section II.B), including the ability of some satellite RNAs to attenuate the s y m p t o m s of the helper virus. It has been s h o w n for t w o satellite RNAs that transgenic plants expressing the satellite R N A are less severely diseased w h e n inoculated w i t h the helper virus. H a r r i s o n et al. (1987) s h o w e d that, w h e n transgenic tobacco plants containing D N A copies of a CMV satellite R N A were inoculated w i t h a satellite-free CMV isolate, satellite replication occurred. At the same time, CMV replication w a s r e d u c e d a n d disease s y m p t o m s were greatly attenuated. In untransf o r m e d plants the CMV isolate caused mosaic disease and stunting. In the t r a n s f o r m e d plants, no mosaic a p p e a r e d and plants g r e w almost as well as healthy ones. These differences persisted for 14 weeks, the longest p e r i o d tested. F u r t h e r m o r e , the same result w a s obtained in plants raised from seed of the transf o r m e d plants. W h e n t r a n s f o r m e d plants were inoculated w i t h TAV, there w a s a similar a t t e n u a t i o n of disease s y m p t o m s but w i t h o u t a m a r k e d decrease in TAV g e n o m e
synthesis. J a c q u e m o n d et al. (1988) s h o w e d that tobacco p l a n t s t r a n s g e n i c for a CMV satRNA w e r e tolerant to infection by aphids, the m a i n m e t h o d of field t r a n s m i s s i o n for CMV. Results similar to those of H a r r i s o n et al. (1987) w e r e r e p o r t e d at the same time for another satellite-virus combination, satTRSV a n d TRSV (Gerlach et al., 1987). Tobacco plants that expressed full-length satTRSV or its comp l e m e n t a r y s e q u e n c e as R N A t r a n s c r i p t s increased their synthesis of satTRSV R N A foll o w i n g inoculation w i t h TRSV, but virus replication w a s r e d u c e d a n d disease s y m p t o m s were greatly ameliorated. This protection w a s m a i n t a i n e d for the life of the plants. Two distinct m e c h a n i s m s of resistance were f o u n d in N. benthamiana w i t h full-length of sequences of a mild variant of satGRV inoculated w i t h GRV plus severe s a t R N A (Taliansky et al., 1998). In one set of t r a n s f o r m e d lines, there w e r e high levels of transcript R N A a n d the replication of both severe s a t R N A a n d GRV genomic R N A was inhibited. In the second set of plants there were 10w levels of transcript RNA, a n d replication of severe satRNA, but not that of GRV genomic RNA, w a s inhibited. It
724
16 C O N T R O L AND USES OF PLANT VIRUSES
was concluded that in the first set of plants both GRV genomic and severe satRNA replication was down-regulated by the mild satRNA and in the second there was homology-dependent gene silencing of the severe satRNA. Resistant plants were also produced using only the 5' terminal one-third of the mild satRNA. The use of satellite RNAs in transgenic plants to protect against the effect of virus infection has both advantages and disadvantages. The protection afforded is not affected by the inoculure concentration, as it is with viral coat protein transformants. The losses that do occur in transgenic plants because of slight stunting will affect only the plants that become naturally infected in the field, whereas if all plants are deliberately infected with a mild CMV-satellite combination they will all suffer some loss (Section IV.B). Furthermore, the resistance may be stronger in transgenic plants than in plants inoculated with the satellite. Inoculation is not needed each season, and the mutation frequency is lower. Nevertheless, there are distinct risks and limitations with the satellite control strategy. The satellite RNA could cause virulent disease in another crop species or could mutate to a form that enhances disease rather than causing attenuation (see Chapter 14, Section II.B.5). Another risk is the reservoir of virus available to vectors in the protected plants. Lastly, the satellite approach will be limited to those viruses for which satellite RNAs are known. 10. DI nucleic acid-mediated pr~tection Defective interfering (DI) nucleic acids are described in Chapter 8 (Section IX.C). They are mutants of viral genomes that are incapable of autonomous replication but contain sequences that enable them to be replicated in the presence of the parent helper virus. In many cases, they are amplified at the expense of the parent virus, ameliorating the symptoms induced by that virus. When such nucleic acids are transgenically expressed, infection with the parent virus mobilizes and amplifies them. Transgenic expression of DI RNAs was shown to protect N. benthamiana plants against the apical necrosis and death usually caused by CymRSV (Kollar et al., 1993). N. benthamiana plants transformed to express the TBSV DI
RNA were protected against that virus and closely related tombusviruses (CNV and CIRV) but were susceptible to a distantly related tombus-like virus (CymRSV) and unrelated viruses (BDMV, PVX, TMV) (Rubio et al., 1999). Transformation of N. benthamiana to express the DI DNA of ACMV interferes with the replication of both genomic components of that geminivirus (Frischmuth and Stanley, 1991, 1993). Serial transmission from the transgenic plants led to increasing numbers of asymptomatic plants with undetectable levels of viral DNA. The protection afforded by the DI DNA is confined to closely related strains of the virus. Similarly, the accumulation of the Logan strain of BCTV is reduced in N. benthamiana plants transgenically expressing the DI DNA of that virus strain (Stenger, 1994); however, there was no effect on other BCTV strains. C. O t h e r f o r m s of t r a n s g e n i c p r o t e c t i o n Various forms of protection against viruses have been shown for a variety of transgenes that are not derived from viruses themselves. Some of these are described in this section. 1. Transgenic plants expressing PR proteins The PR host proteins induced following infection with viruses causing necrotic local lesions were discussed in Chapter 10 (Section III.K.1). These proteins, which are part of a non-specific host defense reaction, are involved in the phenomenon of local acquired resistance. Treatment of leaves with salicylic acid induces certain PR proteins and inhibits AMV replication in such leaves. Hooft van Huijsduijnen e/ al. (1.986) have isolated and cloned the mRNAs for some of these proteins. In principle, it might be possible to provide protection against certain viruses by using 'transgenic' plants in which PR protein genes are expressed constitutively under the control of a suitable promoter, but this has not yet been proven. The evidence is against this, with no protection arising from transgenic expression of PR proteins (see Lusso and Kuc, 1996). 2. Antisense to [3-1,3-glucanase The [3-1,3-glucanases are proteins believed to be part of the constitutive and induced defense
VII.
system of plants against fungal infection. Unexpectedly plants deficient in these enzymes as a result of the expression of an antisense RNA show markedly reduced lesion size and number in the local lesion response of N. tabacum Havana 425 to TMV and in N. sylvestris to TNV (Beffa et al., 1996). The mutant plants also showed reduced severity and delay of symptoms of TMV in N. sylvestris. 3. Transgenic plants expressing virus-specific antibodies Plants do not have an immune system like that of animals in which specific antibody proteins are formed in response to an infection, and it has long been assumed that plants could not produce such proteins. However, the work of Hiatt et at. (1989) demonstrates that this is possible. They obtained cDNAs derived from mouse hybridoma mRNA, transformed tobacco leaf segments, and regenerated plants. Plants expressing single 7 (heavy) or ~c (light) chains were crossed to produce plants in which both chains were expressed simultaneously. A functional antibody made up over 1% of leaf proteins. Production of antibodies in plants (plantibodies) has been reviewed by Smith (1996) and by Zhang and Wu (1998). The expression in plants of a single-chain Fv antibody, derived from a panel of monoclonal antibodies against AMCV coat protein, reduced the incidence of infection and delayed s y m p t o m development (Tavladoraki et al., 1993). A further suggestion is to express antiidiotypic antibodies to important binding sites (Martin, 1998). Anti-idiotypic antibodies are antibodies made against the antigen-interacting portion of an antibody and thus have the same surface conformation as the antigen. Such an antibody should bind to a virus-specific domain in, say, a coat protein, movement protein or replicase. The transgenic expression of plantibodies has potential for the control of plant viruses and also for determining functions of plant proteins (Fig. 16.15) (de Jaeger et al., 2000), including those involved in disease determination. Antibodies directed against other antigens, e.g. lymphoma-associated protein, have been
PATHOGEN-DERIVED RESISTANCE
725
produced in plants (McCormick et al., 1999). 4. Transgenic plants expressing 2',5'oligoadenylate synthetase In mammalian systems, interferons are effective antiviral molecules. When one of the components of the virus-inhibiting pathway, 2',5'-oligoadenylate synthetase, was expressed in potato plants, it gave protection against PVX (Truve et al., 1993). The virus concentration in transgenic plants was lower than that in plants expressing PVX coat protein. 5. Transgenic plants expressing ribosomeinactivating proteins (reviewed by Wang and Tumer, 2000) Ribosome-inactivating proteins (RIPs) deglycosylate a specific base in the 28S rRNA and prevent binding of elongation factor 2. RIPs have been isolated from several plant species. Transgenic plants expressing the pokeweed (Phytolacca americana) RIP (termed pokeweed antiviral protein; PAP) showed protection against PVX, PVY and PLRV (Lodge et al., 1993). Plants expressing a low level of a Cterminal deletion mutant of PAP were resistant to PVX (Tumer et al., 1997). As the intact C-terminus is required for toxicity and depurination of tobacco ribosomes in vivo, it was concluded that the antiviral activity of PAP can be dissociated from its toxicity. Another RIP from pokeweed, PAP-II, is less toxic than PAP and plants expressing it are protected against TMV and PVX (Wang et al., 1998c); a similar protein, PIP-2 from P. insularis, exhibits antiviral activity against TMV (Song et al., 2000). The pokeweed antiviral protein and its applications are reviewed in Tumer et al. (1999). Other RIPs that show antiviral activity include BAP from Bougainvillea spectabilis (Balasaraswathi et al., 1998) which protected against TSWV, trichosanthin (Lam et al., 1996) which gave protection against TuMV, and dianthin (Hong et al., 1996) giving protection against ACMV. 6. Transgenic plants expressing ribonuclease gene pac-1 RNA viruses replicate via a complementary strand and, thus, are thought to have a dsRNA
726
~6 C O N T R O L
A N D USES OF P L A N T V I R U S E S
Fig. 16.15 Potential mechanisms of antibody-mediated in vivo modulation of protein or signal molecule activity. From de Jaeger et aI. (2000), with kind permission of the copyright holder, 9 Kluwer Academic Publishers.
stage in their infection cycle. To attack these replication intermediates the yeast-derived dsRNA-specific RNase gene, pac-1, was transformed into tobacco plants (Watanabe et al., 1995). Transformed plants showed a decrease in lesion numbers when inoculated with TMV and a delay in symptom appearance when inoculated with CMV or PVY. 7. Human cystatin C As described in Chapter 7 (Section V.B.1), potyviruses express their genetic information as a polyproeein that is cleaved by virusspecific proteases to functional proteins. One of the viral enzymes, HCPro, is a papain-like cysteine proeease. H u m a n cystatin C, an inhibitor of cyseeine proteases, interfered with the autoprocessing of PPV polyprotein by HCPro and, unexpectedly had an inhibitory effect on the NIa protease, a serine protease
(Garcia et al., 1993). It was suggested that the transgenic expression of such a protease inhibitor might protect plants against viruses that had polyprotein processing in their infection cycle. 8. Transgenic plants expressing an insect toxin The bacterium Bacillus thurin,giensis produces a polypeptide that is toxic to insects. Different strains of this bacterium produce toxic polypeptides with specificity for different insect groups. Vaeck et al. (1987) produced transgenic plants expressing the toxin gene that were protected against insect attack. Among the virus vector groups of insects, only the Coleoptera appear to be affected by the toxin from some strains of the bacterium. In principle, it might be possible to produce transgenic plants protected against beetle vectors of some plant viruses.
VII. PATHOGEN-DERIVED RESISTANCE
D. Field releases of t r a n s g e n i c plants (reviewed by Kaniewski and Thomas, 1993) Concerns have been expressed about the release and use of plants modified by genetic manipulation. This has led to plants being produced by this means being treated in a different manner to those produced by conventional breeding techniques and being subject to specific regulatory structures. In most countries that have regulatory structures, there are two stages in the field releases of transgenic plants. In the first stage, the plants are released under contained conditions directed by the provisions of the country's biosafety regulatory structure. This stage of release essentially has two purposes: (1) to address any potential problems that a risk assessment might identify; and (2) to assess the field performance of the transgenic lines being tested. Plant lines that satisfy both the regulators and the releasers then go to the second stage of more general release, termed commercial release or farmer release. In this section, I will discuss field performance of transgenic plants and in the next section the possible risks of transgenes that protect plants against viruses. Testing the field performance of transgenic plants is essentially no different to testing plant lines that have been obtained by traditional breeding (Delannay et al., 1989). The testing objectives include evaluating the plant appearance, typeness, growth vigor, yield and quality. Of especial importance is to assess the stability and durability of the protecting transgene under these conditions. Two main factors can affect stability and durability: (1) possible climatic effects on the expression of the transgene; and (2) the presence of protection-breaking strains or isolates of the virus that are present in the viral ecosystem but were not recognized in the initial glasshouse tests. There have been numerous contained field trials of virus-protected transgenic plants, and lines of some crops such as potatoes and papaya are in more general release (see Perlak et al., 1995; Thomas et al., 1997; Gonsalves, 1998). Here I will give some examples of the results obtained. Field experiments with tomatoes suggest
727
that transgenic expression of TMV coat protein may be commercially useful in this host (Fig. 16.16). The transgenic plants were partially resistant to TMV and to strains L, 2, and 22 of ToMV. In the field, no more than 5% of transgenic plants showed systemic disease symptoms, compared with 99% for the control plants. Lack of visual symptoms was correlated
U3
o
80
m 60 Z
~ ~0 0
m 20 Z
< ..._i
0
EL
i i i //-~~ 10 20 30 40 62 76 DAYS POST INOCULATION
~100 (Jr)
oI--"
80
n
>-
m 60 Z
=E 40
O "r
20
._J
13_
-// : i ~ i //-tO 10 20 30 40 61 DAYS POST INOCULATION
Fig. 16.16 D e v e l o p m e n t of systemic s y m p t o m s of TMV infection u n d e r field conditions in t o m a t o e s nontransgenic for coat protein. C), N o n - t r a n s f o r m e d plants; r-! and ~r, plants of two lines of transgenic tomatoes expressing TMV coat protein. Plants were inoculated on terminal leaflets of three successive leaves with TMV strain U1 at 10 btg/mL, 8 days after planting out in the field. Observations were m a d e on 48 plants from each line. From Nelson et al. (1988), with permission.
728
16 C O N T R O L
A N D USES OF P L A N T V I R U S E S
with an absence of ToMV. In inoculated control plants, fruit yields were depressed by 26-35%. There was no evidence that expression of the coat protein gene reduced plant growth or fruit yield compared with uninoculated nontransgenic plants (Nelson et al., 1988). In Australia, most lines of potato cvs Kennebec and Atlantic containing the PLRV coat protein gene showed no measurable differences in agronomic performance when compared with non-transgenic lines in the absence of PLRV (Graham et al., 1995). However, under conditions where PLRV was prevalent, one transgenic Kennebec line gave approximately 30% increased tuber yield over its nontransgenic counterpart (Barker and Waterhouse, 1999). In the USA, some Russet Burbank potato lines containing the PLRV coat protein gene showed low levels of protection against primary spread of the virus but marked reductions in the secondary spread (Thomas et al., 1997). Tomato lines expressing CMV coat protein showed field resistance to the target virus but at a lower level to that found in growth chamber experiments (Tomassolit et al., 1999). The expression of CMV mild satRNA in tomato confers field tolerance to CMV (Stommel et al., 1998). Yie et al. (1995) described the rapid production of homozygous tobacco lines protected against CMV by mild satRNA. Three of these lines were highly protected against CMV under field conditions. Transgenic lines that are protected against mechanically inoculated target virus may not be protected when the virus is inoculated by natural vectors. For example, one line of potatoes expressing the PVY coat protein gave promising results on mechanical inoculation but was not protected against aphid inoculation (Lawson et al., 1990). Similarly, transgenic plants expressing TRV coat protein were not protected against nematode transmission (Ploeg et al., 1993a). Of particular concern is the stability of the protection conferred by the transgene. There are several ways in which the protection can be overcome or broken down, including: 1. There are many factors that may affect the stability of a transgene including rearrange-
ment during meiosis, and methylation. These usually show up in early generations of a transgenic line and would be unlikely to occur in later generations that were released to the field. 2. As noted in Section VII.B.7, environmental conditions may affect the expression of a transgene. 3. Protection may not always be effective with different strains of a virus. For example, tobacco plants transformed to express the coat protein of strain PLB or TCM of TRV were resistant only to the strain whose coat protein was being expressed (Angenent et al., 1990). On the other hand, Nejidat and Beachy (1990) found that transgenic tobacco plants expressing the coat protein gene of TMV were resistant to ToMV and TMGMV but not to RMV. The protective effect against TMV in transgenic plants expressing coat protein was greatly reduced when tobacco plants were held at 35~ instead of 25~ However, plants held in a 25~176 night-day cycle retained resistance (Nejidat and Beachy, 1989). 4. As described in Chapter 10 (Section IV.H), some viruses encode gene products that suppress PTGS. There is concern that the suppression of PTGS by one virus may overcome the transgenic protection against another. However, Wang et al. (2001a) reported that coinfection by CYDV did not compromise the transgenic protection against BYDV-PAV. 5. The CaMV 35S promoter is widely used in constructs to express virus-protecting transgenes. However, in oilseed rape plants in which this promoter is being used to express transgenes, it can be silenced on infection with CaMV (Covey et al., 1997; A1-Kaff et al., 1998, 2000).
E. Potential risks associated with field release of virus transgenic plants (reviewed by Hammered et al. 1999) A major consideration for the use of plant lines transgenically protected against viruses involves the possible risks that might arise from their release to the general environment. This has been discussed in many places (e.g. Hull, 1990a,b; de Zoeten, 1991; Hull and Davies, 1992; Tepfer, 1993; Miller et al.,
VII.
1997b). Essentially, there are three areas of potential risk: risk to humans and other animals, risks to the environment, and commercial risks. Risks to humans are basically any potential deleterious effects that the transgenic product may have to food and whether it has any allergenic properties. For instance, the potential risks of long-term consumption of food expressing relatively high levels of RIPs would need special consideration. However, it must be remembered that humans have been eating virus-infected plants for millenia without any obvious deleterious effects. The potential commercial risks involve the agronomic properties of the transgenic line, the durability of the protection and any possible effects of spread of the transgene to other crops. These considerations are not limited to viral transgenes but apply to most, if not all transgenes. The area of virus-protecting transgenes that has attracted specific interest is the use of virus sequences. The basic question that is asked is: what is the risk of any interactions that might arise between a virus or virus-related sequence integrated in the host genome and another virus superinfecting that plant? Three scenarios are considered: heteroencapsidation, recombination and synergism. 1. Heteroencapsidation Heteroencapsidation involves the superinfection of a plant expressing the coat protein of virus A by the unrelated virus B, the expression of the virus A coat protein not protecting the plant against virus B. The risk is that the coat protein of virus A might encapsidate the genome of virus B, thereby conferring on it other properties such as different transmission characteristics. Heteroencapsidation by transgenically expressed coat protein has been reported for closely related viruses, such as CMV and AMV (Candelier-Harvey and Hull, 1993), and between potyviruses (Lecoq et al., 1993). In a broader study using unrelated viruses, the particles of viruses which have different forms of stabilization, P. CandelierHarvey and R. Hull (unpublished results) showed that heteroencapsidation occurs only between closely related viruses, the particles of which have similar forms of stabilization.
PAWHOGEN-I3ERIVED RESISWANCE
729
2. Recombination The concern here is that recombination between the transgene and superinfecting virus might lead to a new virus. Recombination involving integrated sequences is discussed in Chapter 8 (Section IX.B.5). A hybrid virus, made in vitro by replacing the 2b gene of CMV by its homolog from TAV, is significantly more virulent than either of its parents (Ding et at., 1996). This potential risk is particularly pertinent to luteoviruses (Miller et al., 1997b). 3. Synergism As described in Chapter 10 (Section V.F), synergistic interactions between two unrelated viruses are potentiated by distinct virus sequences. Thus, there is a possibility that the effect of a superinfecting virus could be exacerbated by a transgene expressing a synergisminducing sequence. 4. Avoiding risk Understanding the molecular interactions involved in the potential risk situations can lead to methods for the 'sanitizing' of the transgene to avoid that risk. For example, as described in Chapter 11 (Section III.E.7), aphid transmission of potyviruses involves an amino acid triplet (Asp, Ala, Gly; DAG) in the coat protein. Mutation of this in a PPV coat protein transgene abolished the aphid transmissibility but did not affect the protection offered by the transgene (Jacquet et al., 1998). Similarly, mutation in the PPV coat protein gene suppressed particle assembly, heterologous encapsidation and complementation in transgenic N. benthamiana but retained protection against ChiVMV and PVY (Varrelmann and Maiss, 2000). The understanding of the factors involved in recombination (Chapter 8, Section IX.B) will lead to transgene constructs that lessen the possibility of new molecules being formed between the transgene and a superinfecting virus. Similarly, sequences that potentiate synergism, such as the potyvirus HCPro or the cucumovirus 2b genes (see Chapter 10, Section IV.H), can be avoided. In all these risk assessments, it is important to compare the transgenic situation with the non-transgenic situation. Thus, there is the
730
~6 C O N T R O L A N D USES OF P L A N T VIRUSES
possibility of the above potential risks occurring in mixed infections between viruses. This is discussed in H a m m o n d et al. (1999).
VIII. D I S C U S S I O N A N D CONCLUSIONS As discussed in Chapter 3, it is impossible to give precise measures of crop losses due to viruses. For a given virus, losses may vary widely with season, crop, country and locality. Nevertheless, there are sufficient data to show that continuing effort is needed to prevent losses from becoming more and more extensive. Essentially, once a plant becomes infected with a virus, in most cases it is not feasible to try to cure it. The approaches of heat treatment and meristem culturing (Sections II.C.2.b and II.C.2.c) are in practicality applicable only to high-value crops. Three kinds of situation are of particular importance: (1) annual crops of staple foods such as grains and sugar beet that are either grown on a large scale or are subsistence crops and that, under certain seasonal conditions, may be subject to epidemics of viral disease; (2) perennial crops, mainly fruit trees with a big investment in time and land, where spread of a virus disease, such as citrus tristeza or plum pox, may be particularly damaging; and (3) high-value cash crops such as tobacco, tomato, cucurbits, peppers and a number of ornamental plants that are subject to widespread virus infections. Possible control measures have been classified under three headings: (1) removal or avoidance of sources of infection; (2) control or avoidance of vectors; and (3) protecting the plant from systemic disease. In principle, by far the best method for control would be the development of cultivars that resist a particular virus on a permanent basis. Experience has shown that viruses continually mutate in the field with respect to both virulence and the range of crops or cultivars they can infect. Thus, it has been generally considered that breeding for resistance or the development of transgenic plants is unlikely to give a permanent solution for any particular virus and crop. It has been suggested for conventional resistance to pests and dis-
eases that one should not put too strong a selection pressure on a viral population or else a resistance-breaking strain will arise. However, Hull (1994b) has argued that, as variation arises from the replication of a virus, the more that one can inhibit the replication the less chance there is that variants will arise that can overcome the protection. Furthermore, if transgenic protection is targeted at highly conserved sequences or motifs that are essential for replication, it is highly unlikely that any variant would be able to compensate for the defect. With almost all crops affected by viruses, an integrated and continuing program of control measures is necessary to reduce crop losses to acceptable levels. Such programs will usually need to include elements of all three kinds of control measure identified at the beginning of this discussion. As noted above, the deployment of resistant or protected cultivars or lines is one of the most promising approaches. Genetic technology is, and will increasingly, contribute to this approach. Breeding programs are being speeded up by the use of molecular marker techniques, and the co-linearity of the organization of many plant genomes is enabling the prediction of loci of potential resistance genes. Transgenic technology is providing a range of protecting genes, some of which should be relatively durable. Thus, future breeding programs for virus resistance should bring together the conventional and transgenic approaches with stacking, pyramiding a n d / o r deployment of protecting genes to tackle the problem being addressed. In developing strategies for the integrated approach it is essential to have a full understanding of the disease, its epidemiology and ecology, and of the pathogen, its genetic make-up and functioning, and its potential for variation. However, as well as the science involved in improving agriculture, the use of transgene technology has raised many public issues such as the impact of agricultural practices on the environment and on the food that this industry provides, and the ethics of overcoming the species barrier. Thus, the application of this technology to improving the world's food security will involve many more issues than just the science and its application.
IX.
IX. P O S S I B L E U S E S OF V I R U S E S FOR GENE TECHNOLOGY A. Viruses as gene vectors (reviewed in Porta and Lomonossoff, 1996; Scholthof et al., 1996) In the early 1980s there was considerable interest in the possibility of d e v e l o p i n g plant viruses as vectors for introducing foreign genes into plants. At first, interest centered on the caulimoviruses, the only plant viruses with dsDNA genomes, because cloned DNA of the viruses was shown to be infectious (Howell et al., 1980). Interest later extended to the ssDNA geminiviruses, and then to RNA viruses when it became possible to reverse-transcribe these into dsDNA, which could produce infectious RNA transcripts. The main potential advantages of a plant virus as a gene vector were seen to be: (1) the virus or infectious nucleic acid could be applied directly to leaves, thus avoiding the need to use protoplasts and the consequent difficulties in plant regeneration; (2) it could replicate to high copy number; (3) there would be no 'position effects' of a site in the plant chromosomal DNA; and (4) the virus could move throughout the plant, thus offering the potential to introduce a gene into an existing perennial crop such as orchard trees. Such a virus vector would have to be able to carry a non-viral gene (or genes) in a way that did not interfere with replication or m o v e m e n t of the genomic viral nucleic acid. Ideally, it would also have the following properties: (1) inability to spread from plant to plant in the field, providing a natural containment system; (2) induction of very mild or no disease; (3) a broad host range, which would allow one vector to be used for many species, but would be a potential disadvantage in terms of safety; and (4) maintenance of continuous infection for the lifetime of the host plant. The major general limitations in the use of plant viruses as gene vectors are: (1) they are not inherited in the DNA of the host plant, and therefore genes introduced by viruses cannot be used in conventional breeding programs; (2) plants of annual crops would have to be
POSSIBLE USES OF V I R U S E S FOR GENE T E C ' H N O L O G Y
731
inoculated every season, unless there was a very high rate of seed transmission; (3) by recombination or other means the foreign gene introduced with the viral genome may be lost quite rapidly with the virus reverting to wild type; and (4) it would be necessary to use a virus that caused minimal disease in the crop cultivar. The virus used as vector might mutate to produce significant disease, or be transmitted to other crops that were susceptible. Infection in the field with an unrelated second virus might cause very severe disease. 1. Caulimoviruses Howell et al. (1981) inserted an eight-base-pair EcoRI linker molecule into the large intergenic region of cloned CaMV DNA (see Chapter 6, Section IV.A.1, for CaMV genome organization) and showed there was no impairment of infectivity. Gronenborn et aI. (1981) reported the successful propagation of foreign DNA in turnip plants using CaMV as a vector. However, they found that the size of the DNA insert that could be successfully p r o p a g a t e d was limited to about 250 base pairs or less. Brisson et al. (1984) reported the first successful expression of a foreign gene in plants using CaMV as a vector. They used the 234-bp dihydrofolate reductase gene from Escherichia coli, which confers methotrexate resistance in the bacterium. They inserted the gene into a derived strain :of CaMV from which most of gene II had been deleted. The chimeric DNA was stably propagated in turnip plants and the bacterial gene was expressed, as shown by assays for methotrexate resistance. De Zoeten et al. (1989) replaced the ORF II of CaMV DNA with the h u m a n interferon (IFNaD) gene. They obtained a stable strain of CaMV that replicated in B rassica rapa and led to the production of IFNaD. The interferon was located in the CaMV viroplasms and it had antiviral activity in an animal cell assay. Paszkowski et al. (1986) constructed a hybrid CaMV g e n o m e containing the selectable marker gene n e o m y c i n p h o s p h o t r a n s f e r a s e type II, which replaced the gene VI coding region. This construct was not viable in plants and could not be complemented in trans by w i l d - t y p e CaMV. However, inoculation of
732
~6 ( ; O N T R O L A N D USES OF P L A N T V I R U S E S
B rassica campestris protoplasts under DNAuptake conditions gave rise to stable cell lines genetically transformed for the marker gene. This occurred only w h e n the hybrid was coinoculated with wild-type CaMV. The mechanism for this effect is not understood. CaMV as a probe for studying gene expression in plants is discussed by Pfeiffer and Hohn (1989). Thus, there appear to be several constraints to the use of CaMV as a gene vector. These include the packing capacity of the CaMV particle, the amount of viral DNA that can be removed without affecting the functioning of the genome, and the interactions between different parts of the genome in expression and replication. Removal of non-essential regions of the genome should enable about 1000 bp of sequence to be inserted (Ftitterer et al., 1990), but it is not certain whether all this sequence is really non-essential. The use of pairs of vectors with overlapping deletions has been suggested (Hirochika and Hayashi, 1991).
2. Geminiviruses Much attention has been focused on the geminiviruses as potential gene vectors because of their DNA genomes and the fact that the small size of the genomes makes them convenient for in vitro manipulations (reviewed by Mullineaux et al., 1992; Stanley, 1993; Timmermans et al., 1994; Palmer and Rybicki, 1997). Nevertheless, this small size may restrict the amount of viral DNA that can be deleted (see Davies et al., 1987). However, this is counterbalanced by the fact that for some geminiviruses a viable coat protein and encapsidation are not necessary for successful inoculation by mechanical means, or for systemic movement through the plant (Gardiner et al., 1988). There are other potential difficulties (Davies et al., 1987). Recombination can occur to give parent-type molecules. Most geminiviruses are restricted mainly to the phloem and associated cells. However, the wide host range of the geminiviruses (compared with the caulimoviruses) makes them of considerable interest. The ability of some members to infect cereal crops would be particularly useful, except for the fact that they are not seed transmitted and are mechanically transmitted only with difficulty. In any
event, inoculations on the scale needed for cereal crops would be impractical. Nevertheless, model experiments have shown that a geminivirus can be used to introduce and express foreign genes in plants. Various workers have shown that an intact coat protein gene is not essential for TGMV replication and movement. Hayes ei al. (1988d) constructed a chimeric TGMV DNA in which most of the coat protein gene had been deleted and replaced by the bacterial neomycin phosphotransferase gene (neo gene). They used Agrobaclerium inoculation (agro-infection) to introduce this construct into tobacco plants. The TGMV DNA was replicated in the transgenic plants and the neo gene expressed. Ward et al. (1988) isolated a coat protein mutant of ACMV DNA1 in which 727 nucleotides in the coat protein gene were deleted. This mutant was non-infectious. However, when the coat protein ORF was replaced by the coding region of the bacterial chloramphenicol acetyltransferase (CAT) gene under control of the coat protein promoter, infectivity was restored, virus spread through the plant, and the CAT gene was expressed. Nevertheless, systemic movement of infectious constructs with larger than normal DNAs in the coat protein position exerted a strong selection pressure favoring derivatives of normal size (Elmer and Rogers, 1990). Transient expression systems allow for the rapid screening of DNA constructs designed to study the activities of promoter sequences, RNA-processing signals, and so on, in cells, and as a preliminary screen in the construction of transgenic plants. In principle, viral DNA might be altered to provide plasmid-type vectors for high copy number and rapid expression of modified or foreign genes. The work of Elmer et al. (1988a) with a geminivirus progressed some distance toward the construction of such a plant plasmid. Deletion derivatives of TGMV DNA A defined the minimal DNA fragment capable of self-replication as about 1640 bp or 60% of the A sequence. Hanley-Bowdoin et al. (1988) showed that following agro-inoculation of petunia leaf discs with TGMV DNA A, the viral coat protein gene was transiently expressed, beginning about 2
IX.
days after agroinfection. Constructs were then made in which the coat protein ORF was replaced by the bacterial C A T or [~glucuronidase (GUS) genes, under the control of the coat protein promoter. Following agroinoculation, these foreign genes were also transiently expressed in petunia leaf discs. GrOning et al. (1987) found that the DNA of a geminivirus infecting Abutilon setlovianum was located in the chloroplasts, raising the possibility of constructing a chloroplast-specific transformation vector. As noted above, some geminiviruses such as mastreviruses have the potential for expressing genes in monocots. However, although constructs of MSV with the V2 gene substituted by a C A T gene or with the GUS gene inserted into the short intergenic region expressed the inserted gene, they did not move systemically (Lazarowitz et al., 1989; Shen and Hohn, 1994). A 3.7-kbp plant cell-E, coli shuttle vector based on WDV has been developed (Ugaki et al., 1991). Although this shuttle vector has been useful in studying viral replication maize cells (Timmermans et al., 1992), its use in whole plants has yet to be demonstrated. 3. RNA viruses The ability to manipulate RNA virus genomes by means of a cloned cDNA intermediate has opened up the possibility of using RNA as well as DNA viruses as gene vectors. In principle the k n o w n high error rate in RNA replication (see Chapter 17, Section IV.A) might place a limitation on the use of RNA viruses (Siegel, 1985; van Vloten-Doting et al., 1985). The experimental evidence to date suggests that mutation may not be a major limiting factor, at least in the short term. Viruses with isometric and rodshaped particles have been studied as potential vectors but those with rod-shaped particles have better potential as there are fewer constraints on the amount of nucleic acid that can be inserted. a. TMV Takamatsu et al. (1987) prepared a TMV cDNA construct in which most of the coat protein gene was removed and the C A T gene was in its place.
POSSIBLE USES OF VIRUSES FOR GENE T E C H N O L O O Y
733
When in vitro transcripts from this construct were inoculated on to tobacco leaves, the local lesions that formed were smaller than normal, but biologically active CAT was produced, and this activity increased in the inoculated leaves for 2 weeks. RNA was extracted from these leaves and re-inoculated on to tobacco after being encapsulated in TMV protein in vitro. CAT activity was again detected, indicating that this replicating RNA had some degree of stability in vivo. Yamaya et al. (1988), using a disarmed Ti plasmid vector, introduced a cDNA copy of the TMV genome under the control of the 35S CaMV promoter into the genomic DNA of tobacco plants. This experiment demonstrated that a non-seed-transmitted RNA virus can be made seed transmissible. Dawson et al. (1989) constructed a hybrid TMV in which the C A T gene was inserted between the coat protein and the 30-kDa genes (Fig. 16.17A). This construct replicated efficiently, produced an additional subgenomic RNA and CAT activity, and assembled into 350n m virus rods. However, during systemic infection the insert was precisely deleted, giving rise to wild-type virus; it was thought that this deletion resulted from homologous recombination between the two copies of the coat protein gene subgenomic promoter. Takamatsu et al. (1990b) constructed, via cDNA, a TMV RNA in which an additional sequence encoding Leuenkephalin, a pentapeptide with opiate-like activity, was incorporated just 5' to the termination codon for coat protein. The pentapeptide was expressed in protoplasts as a fusion product with coat protein. To overcome the problem of loss of the insert through recombination between homologous sequences, Donson et al. (1991) developed a hybrid vector containing sequences from two t o b a m o v i r u s e s (TMV-U1 and ORSV (Fig. 16.17B)). In this vector, the gene of interest is expressed from the TMV coat protein subgenomic promoter and the coat protein from the ORSV promoter. There was a problem with partial deletion of one of the inserts, NPTII, but this was considered to be insert specific. This vector system has been used to express cztrichosanthin (Kumagai et al., 1993), and a derivative of this vector expressing the coat
734
16 C O N T R O L
A N D USES OF P L A N T V I R U S E S
CAT-CP
species and retained the inserted gene especially in N. clevelandii plants. SPC
c. BMV
A
CP-CAT SPC
hybrid TMV vector
!i
SPC
TMV/DHFR
~i:: i:iilili!ii ~:I30KII.III:I!::I~ O:CP ~ -
B
SPC
TMV/NPTIi SPC
Fig. 16.17 Schematic diagrams of TMV expression vectors. (A) TMV mutants in which the gene coding for CAT is added into the complete TMV-U1 genome. The CAT gene is fused behind the subgenomic promoter for the coat protein in order to produce an additional sgRNA. The resulting sequence duplications are indicated by arrows beneath each construct. (B) Hybrid TMV expression vector in which all ORFs are TMV-U1 sequences, except the coat protein sequence, which originates from ORSV and is transcribed into mRNA from its own ORSV-derived, subgenomic promoter (shown as a small open box). Foreign genes coding for DHFR and NPTII are expressed via subgenomic mRNAs transcribed from the heterologous TMV-U1 promoter located in the 30-kDa ORF. SPC, subgenomic promoter for the TMV-U1 coat protein. From Porta and Lomonossoff (1998), with permission.
French et al. (1986) constructed variants of BMV RNA3 in which the coat protein gene was replaced with the bacterial C A T gene, or in which the C A T gene was inserted near the 5' end of the coat gene. When inoculated on to barley protoplasts together with normal BMV RNAs 1 and 2, these RNA3 constructs replicated and p r o d u c e d subgenomic RNAs equivalent to the normal coat protein subgenomic RNA (see Chapter 7, Section V.E.3.a). When the C A T gene was inserted in-frame with the upstream coat protein initiation codon, CAT expression exceeded that in plant cells transformed by Ti plasmid-based vectors. Pacha et al. (1990) s h o w e d that all the cisacting elements required for the replication of the 2.1-kb CCMV RNA3 can be contained in a 454-base replicon made of 5'- and 3'-terminal sequences. Thus, it may be possible to express a foreign gene of significant size using this replicon. d. T R V
protein of ToMV instead of ORSV has been used to manipulate a biosynthetic p a t h w a y in planta (Kumagai et aI., 1995).
Angenent et al. (1989c) found that sequences of 340 nucleotides at the 5' end and of 405 at the 3' end of the RNA2 of TRV (strain PLB) were sufficient for replication. Constructs in which the deleted viral nucleotides were replaced with a 1401-nucleotide sequence from plasmid D N A were replicated in protoplasts co-inoculated with the complete genome of strain PLB, indicating that this virus may have potential as a gene vector. Tobravirus vectors express foreign proteins efficiently in roots (MacFarlane and Popovich, 2000).
b. P V X
e. T B S V
A similar construct to that of TMV s h o w n in Fig. 16.17A was m a d e with PVX with a duplication of the coat protein promoter ( C h a p m a n et al., 1992). The PVX construct with the GUS gene being expressed from one of the promoters gave systemic infection of various Nicotiana
The coat protein of TBSV is dispensable for systemic infection of certain Nicotiana species (Scholthof e/ al., 1993) a n d hence can be replaced by foreign sequences such as the GUS gene (see Scholthof et al., 1993; Scholthof, 1999). By placing a c D N A copy of the TBSV genome
IX.
between the CaMV 35S promoter and the hepatitis A ribozyme, followed by the nopaline synthase gene polyadenylation signal, Sholthof (1999) was able to gain infections by directly rub-inoculating the construct. The 35S promoter and the NOS terminator controlled transcription of the construct and the cleavage of the transcript by the ribozyme delimited the 3' end of the infectious genome. This vector enables expression of inserts into the coat protein gene region.
B. Viruses as sources of control elements for transgenic plants As explained in Chapter 7 (Section III.D.2.d), constructs for transforming plants contain various control elements. Certain plant viral nucleic acid sequences have been found to have useful activity in these chimeric gene constructs as promoters of DNA and RNA transcription and as enhancers of mRNA translation.
POSSIBLE USES OF VIRUSES FOR GENE T E C H N O L O G Y
735
of 250 bp of upstream sequences. This modification gave about a 10-fold increase in transcriptional activity. The '34S' promoter from FMV had about the same activity as the 35S CaMV promoter (Sanger et al., 1990). The 35S promoter is nominally a constitutive one. However, Benfey and Chua (1989) showed that there was marked histological localization of expression of GUS activity in petunia under control of this promoter. By contrast, the 19S promoter directed expression of the CAT gene in a range of tissues in tobacco plants (Morris et al., 1988b). b. Other DNA virus promoters
Properties of promoters from other DNA viruses, such as badnaviruses and geminiviruses, are discussed in Chapter 7 (Sections IV.C.2). Several of these promoters have been shown to have activity in transformation constructs but are not used as widely as the CaMV 35S promoter.
1. Promoters a. C a u l i m o v i r u s
promoters
Transcription of CaMV DNA gives rise to a 19S and a 35S mRNA (see Chapter 7, Section IV.C.1). Shewmaker et al. (1985) introduced a full-length copy of CaMV DNA into the T DNA of the Agrobacterium Ti plasmid, and this was integrated into various plant genomes. They showed that the 19S and 35S promoters were functional in this form in several plant species. They are both strong constitutive promoters and have found wide application for the expression of a range of heterologous genes (e.g. Balazs et al., 1985; Bevan et al., 1985; Odell et al., 1985, Lloyd et al., 1986; Nagy et al., 1986). The 35S promoter has been found to be much more effective than the 19S promoter in several systems. For example, expression of the ~subunit of [3-conglycinin in petunia plants under control of the 35S promoter was 10 to 50 times greater than that from the 19S promoter (Lawton et al., 1987). The 35S promoter was also found to be 10 to 30 times more effective than the nopaline synthase promoter from Agrobacterium tumefaciens (Garcia et al., 1987). Kay et al. (1987) constructed a variant 35S promoter that contained a tandem duplication
2. Untranslated leader sequences as enhancers of translation Untranslated leader sequences of several viruses have been shown to act as very efficient enhancers of mRNA translational efficiency both in vitro and in vivo, and in prokaryotic and eukaryotic systems (see Chapter 7, Section V.C.5). AMV RNA4 is known to be a welltranslated message for AMV coat protein. Jobling and Gehrke (1987) replaced the natural leader sequences of a barley and a human gene with AMV RNA4 leader sequence. These constructs showed up to a 35-fold increase in mRNA translational efficiency in the rabbit reticulocyte and wheatgerm systems. Sleat et al. (1987) made similar constructs involving the uncapped mRNAs for two vertebrate genes and the bacterial GUS gene with or without a 5'-terminal 67-nucleotide sequence derived from the untranslated region of TMV RNA (the f2 sequence). These were tested in vitro in the rabbit reticulocyte, wheatgerm and E. coli systems. The TMV leader sequence enhanced translation of almost every mRNA in every system. Gallie et al. (1987a,b) extended these results to show that the 67-nucleotide sequence
736
16 C O N T R O L
A N D USES OF P L A N T V I R U S E S
was also a potentially useful enhancer in vivo in mesophyll protoplasts and Xenopus oocytes. A deletion derivative of f2 appeared to be functionally equivalent to a Shine-Dalgarno sequence in several bacterial systems (Gallie and Kado, 1989). The translational enhancement brought about by the TMV f2 sequence is mediated by the ribosomal fraction of the in vitro system used (Gallie et al., 1988). The experiments of Sleat et at. (1988c) support the view that the untranslated viral leader sequences reduce RNA secondary structure, making the 5' terminus more accessible to scanning by ribosomal subunits or by interaction with initiation factors. 3. In vitro studies on transcription In vitro translation systems which faithfully reproduce in vivo gene expression have proved
very useful in animal and other systems for developing an understanding of nucleotide sequences and protein factors involved in the control of transcription. Cooke and Penon (1990) have taken a step in this direction for plant systems. They obtained a partially purified extract from tobacco cell suspensions that contained all the factors necessary for transcription from the 19S promoter. 4. Use of the TMV origin of assembly (OAS) fi~r the introduction of fi)reign RNAs Sleat et al. (1986) showed that a correctly oriented OAS sequence located 3' to a foreign RNA sequence could initiate the efficient encapsulation of the foreign RNA in vitro. In an extension of these experiments, Gallie et al. (1987c) prepared transcripts that encoded the OAS together with the RNA sequence for the CAT enzyme. When these were encapsulated in TMV coat protein in vitro and inoculated to a wide range of cell types, the CAT mRNA was transiently expressed. Immuno-gold labeling located the site of disassembly and transient gene expression in epidermal cells of inoculated tobacco leaves (Plaskitt et al., 1987). Sleat et al. (1988c) constructed a plasmid derivative containing the 5' leader sequence of TMV followed by the CAT sequence and the OAS. This was introduced into the DNA of tobacco plants using an Agrobacterium vector. Transcripts from this nuclear DNA were encap-
sulated into TMV-Iike rods when the transgenic plants were infected with TMV. These experiments demonstrated an efficient complementation between functions encoded in the host genome and those of the infecting virus. 5. Other possible RNA virus sequences The subgenomic promoters for several RNA viruses have now been defined (see Chapter 7, Section V.B.2). These are used in the RNA virus vectors described above and may prove useful for gene amplification, as might 5'- or 3'terminal sequences that have not yet been tested. The potential of simple and specific RNA enzymes (ribozymes) from satellite RNA sequences is discussed in Section VII.B.6. C. V i r u s e s for p r e s e n t i n g h e t e r o l o g o u s peptides (reviewed by Lomonossoff and Johnson, 1996; Porta and Lomonossoff, 1996, 1998; Johnson et al., 1997) The production of small peptides is required for several reasons, including acting as epitopes for vaccines and biologically active peptides. As described in Chapter 15 (Section IV.A.2), epitopes are patches of amino acids that adopt specific conformations. Free peptides can act as epitopes but the immunogenicity is enhanced by presentation on the surface of a macromolecular assembly. One approach to presenting the peptide sequence in the correct conformation is to incorporate them into a viral coat protein sequence in such a way that they are exposed on the surface of the virus particle. The virus particle can then be used as a vaccine. There are several advantages to doing this with plant viruses, including: (1) the virus can be produced in large amounts and in less developed countries where the technology for animal virus vaccine production may be limited; (2) such vaccines may be given orally as part of the normal food supply; (3) the virus will not infect the human or other animal and thus is completely inactive; and (4) the system is not subject to contamination by other virulent animal pathogens. A potential disadvantage is that the high rate of mutation of RNA viruses could result in the deletion or loss of inserted sequences, especially as they would not be
IX. POSSIBLE USES OF VIRUSES FOR GENE TECHNOLOGY
737
Fig. 16.18 Generation of chimeric CPMV particles. Foreign sequences are inserted into the gene for the S coat protein borne by RNA2. Both RNA1 and RNA2 are translated into polyproteins and undergo a cascade of cleavages whose sites and final products are shown. RNA2 (bearing the heterologous sequence) needs to be co-inoculated with RNA1 (unmodified) to initiate an infection in cowpea plants. S protein harboring a foreign epitope in its ]3B-[3C loop and native L protein assemble at 60 copies each into icosahedral virus particles on which the foreign insert is expressed around the 5-fold axes of symmetry. From Porta and Lomonossoff (1998), with kind permission of the copyright holder, 9 John Wiley and Sons Ltd.
738
16 C O N T R O L AND USES OF PLANT VIRUSES
under selection pressure (van Vloten Doting et al., 1985); however, experience is indicating that this problem may not be as significant as at first feared (Johnson et al., 1997). Several plant viruses have been used for the presentation of foreign peptides. 1. CPMV (reviewed by Lomonossoff and Hamilton, 1999) The structure of CPMV has been solved to atomic resolution (see Chapter 5, Section VI.B.6.a) (Lomonossoff and Johnson, 1991). The capsid comprises two types of protein: the L protein, which has two [3-barrel domain domains, and the S protein, which has one [3-barrel domain. Analysis of the three-dimensional structure suggested that loops between the 13 strands would be suitable for the insertion of sequences to be expressed as epitopes, as these loops are not involved in contacts between protein subunits. The 13B-13C loop of the S protein is highly exposed (Lomonossoff and Johnson, 1995) and was used for most of the insertions (Fig. 16.18); some insertions have been made in other loops (Lomonossoff and Hamilton, 1999). Early studies on inserting sequences at the 13B-13C loop site gave guidelines for construction of viable, genetically stable, chimeras (Porta et al., 1994). These included: (1) foreign sequences should be inserted as additions and not replacements of the CPMV sequence; (2) sequence duplication should be avoided as this led to loss of insert by recombination; and (3) the precise site of insertion was important for maximizing growth of chimeras. Understanding of these guidelines gave a standard procedure for inserting foreign DNA into the [3B-[3C loop of the S protein (Spall et al., 1997). Chimeras with inserts for the sequence for up to 38 amino acids have been successfully made in which the presence of foreign DNA did not significantly affect the ability of the modified virus to replicate. Various epitopes have been inserted, including ones from Human rhinovirus 14, Human immunodejiciency virus type 1 and Canine parvovirus (Lomonossoff and Hamilton, 1999).
2. TMV The initial attempts to express foreign peptides on TMV coat protein used an E. coli expression vector to produce a self-assembling chimera of the viral coat protein and a poliovirus epitope (Haynes et al., 1986). With the development of infectious cDNA clones to TMV, it was possible to use a self-replicating system in plants. Fusion of a foreign sequence to the C-terminus of the coat protein prevented particle assembly (Takamatsu et al., 1990b). To overcome this problem, Hamamoto et al. (1993) placed the insert after an amber stop codon at the C-terminus of the coat protein gene (Fig. 16.19A) so that it could be expressed as a readthrough protein. Particles were assembled with about 5% of the coat protein subunits expressing the inserted sequence (Sugiyama et al., 1995). Replacement of two amino acids on a surface loop near the C-terminus of the coat protein gave particles with 100% of the subunits containing the insert (Fig. 16.19B) (Turpen et al., 1995), as did replacement into another part of the C-terminal region not involved in particle assembly (Fig. 16.19C) (Fitchen et al., 1995). 3. Other viruses Chimeric fusions to the coat protein of the potyvirus JGMV have been expressed in E. coli (Jagedish et al., 1993, 1996). The extracted protein could be polymerized into potyvirus-like particles. These particles could be formed with inserts at the N-terminus or replacing some of the C-terminal sequence. A construct that gave both fused and unfused versions of the green fluorescence protein at the N-terminus of PVX coat protein was produced by Santa Cruz et al. (1996). In this construct the GFP gene was separated from the coat protein by the footand-mouth disease virus 2A sequence which mediates processing at its C-terminal end. The presence of some unfused coat protein molecules was essential for the assembly of virus particles. In subsequent experiments, proteins ranging in size from 8.5 to 31kDa were expressed as 'overcoats' on the surface of PVX particles (Santa Cruz et al., 1996).
IX.
POSSIBLE USES OF VIRUSES FOR GENE T E C H N O L O G Y
739
Fig. 16.19 Production of chimeric TMV particles in planta. Foreign oligonucleotide sequences are introduced at one of three positions, labeled A, B and C, in the gene of the TMV coat protein, which is expressed from the most 3' of the three viral subgenomic mRNAs. In vitro transcripts of the altered full-length genomic cDNA are inoculated on to tobacco. The resulting recombinant TMV coat proteins are represented as ribbon drawings with the numbers indicating each insertion site. Upon assembly of these coat proteins, chimeric rodshaped virions are formed on which the foreign peptides are differentially displayed and distributed; a maximum of 5% of the coat proteins present an insert in position A. All of the coat proteins are modified in positions B and C. From Porta and Lomonossoff (1998), with kind permission of the copyright holder, 9 John Wiley and Sons Ltd. TBSV particles w e r e a s s e m b l e d f r o m the cons t r u c t s h o w n in Fig. 16.20, in w h i c h a n i n s e r t w a s m a d e at t h e C - t e r m i n u s of the coat p r o t e i n g e n e (Joelson et al., 1997). Similarly, the H I V p24 ORF w a s e x p r e s s e d as a n i n - f r a m e f u s i o n w i t h a 5 ' - t e r m i n a l p o r t i o n of the TBSV coat prot e i n ORF ( Z h a n g et al., 2000). T h u s , it w o u l d a p p e a r that TBSV c a n t o l e r a t e i n s e r t i o n s in b o t h
the N- a n d C - t e r m i n a l r e g i o n s of the c o a t p r o t e i n gene.
D. Viruses in functional genomics of
plants (reviewed in Lindbo et al.,
2001)
The g e n e s i l e n c i n g i n d u c e d b y v i r u s i n f e c t i o n of p l a n t s is d e s c r i b e d in C h a p t e r 10 (Section
740
16 C O N T R O L
UAG
5' ---~
33K
A N D U S E S OF P L A N T V I R U S E S
read through I 92K polymerase
t
41K I I119K ~'--~ J 41K coat protein I I 22K I ~ l 19K I
/
movement proteins
insertion point
S
N~ x 180
I
IV.G). In virus-induced gene silencing (VIGS) (see Fig. 10.18) a gene incorporated into a virus vector, for instance a TMV-based vector (Kumagai et al., 1995), a PVX-based vector (Ruiz et al., 1998) or a TGMV-based vector (Kjemtrup et al., 1998), will silence a h o m o l o g o u s gene in a plant. As described in Chapter 10 (Section IV.H), some viruses can suppress PTGS; these w o u l d not be effective for VIGS. The VIGS phenomen o n was initially s h o w n for transgenes and has n o w been d e m o n s t r a t e d for various plant genes such as p h y t o e n e desaturase (Ruiz et al., 1998) and a m a g n e s i u m chelatase gene (Kjemtrup et al., 1998). Baulcombe (1999b) p r o p o s e d that a VIGS approach could be an approach for determining the function(s) of genes identified from genome sequencing. The current approach of insertional mutagenesis has various drawbacks, such as the inability to identify genes, w h o s e d i s r u p t i o n is lethal before the plant has developed. With the VIGS approach, the lethality w o u l d be apparent from the death of the mature plant that had been inoculated. Thus, virus vectors could be a powerful tool in functional genomics.
E. Summary and discussion
Fig. 16.20 Generation of chimeric TBSV particles. Fusions of heterologous sequences are made at the 3' end of the coat protein gene in a full-length cDNA clone of the viral genome. The coat protein is translated from the larger of the two viral subgenomic mRNAs and comprises three domains, designated R, S and P, with P bearing the foreign amino acids (shown in black). When 180 copies of the viral coat protein, labeled A, assemble to form icosahedral particles, the fused peptides, each represented by a black half-circle, are well exposed at the surface of the virions. From Porta and Lomonossoff (1998), with kind permission of the copyright holder, 9 John Wiley and Sons Ltd.
The discussion in this section demonstrates that viruses do not always cause problems and some have their uses. As well as the uses in gene technology, some viruses have a positive horticultural importance, with cultivars being i d e n t i f i e d a n d n a m e d a c c o r d i n g to their virus symptoms. Thus, Abutiton striatum cv. Tompsonii is infected with AbMV and derives its attractiveness from the variegation induced by that virus (similar to Plate 3.7). Similarly, Lonicera japonica cv. Aureo-reticulata shows s y m p t o m s of HYVMV. A wide range of gene technology has been suggested and proven for plant viruses, and several of these basic approaches have been described above. The ones that are making the m o s t impact are the use of viral control sequences in transgenic constructs and the use of viral vectors for expressing foreign sequences. The use of VIGS in functional genomics is not yet widespread, but has great potential.
IX.
Representative
Virus
Gene Replacement
--.[
.....
i ( llJ. . . .
.
.H
POSSIBLE USES OF VIRUSES FOR (;ERE T E C H N O L O G Y
" ',H
II I
1
741
- liP-
~
Gene Insertion
Comparison of strategies used to express foreign genes (black box) from different viruses. White boxes indicate viral genes. The epitope presentation method (a) involves translational fusion of a small sequence inside the coat protein gene or (b) translational read through of an amber stop codon (*) at the 3' end. From Scholthof et al. (1996), with kind permission of the copyright holder, 9 Annual Reviews. www.AnnualReviews.org Fig. 16.21
Epitope Presentation
--t,,
Cornplernentation
--[
7~~~ ~
a
.....................~ ~
,k----
_J
-
-
I
Gene Replacement
Gene Insertion
t~pitooe Presentation
Bromovirus Caulimovirus Furovirus Geminivirus Hordeivirus Potexvirus Tobamovirus Tobravirus Tombusvirus
Caulimovirus Geminivirus
Comovirus Tobamovirus Tombusvirus
Potexvirus Potyvirus
Tobamovirus
Approaches to using viruses as expression vectors have been g r o u p e d u n d e r four headings by Scholthof et al. (1996) (Fig. 16.21). Most of these uses have been discussed above and can be found in more detail in the review of Scholthof et al. (1996). The D N A plant viruses did not realize their initial expectations as gene vectors. There are several reasons for this including their high rates of recombination, the close integration of viral gene functions and packaging constraints. However, these viruses provide i m p o r t a n t control sequences for transformation constructs and will probably continue to do so, especially with those that give tissue-specific expression. A l t h o u g h it was initially suggested that RNA viruses w o u l d have problems with accumulation of errors due to replication u n d e r nonselecting conditions, this has not p r o v e d to be so. One m a j o r a d v a n t a g e that r o d - s h a p e d viruses have is that they do not have the con-
, b ........1 -
Complernentation
Alfamovirus Caulimovirus
Dianthovirus Geminivirus Potexvirus Tobamovirus Tombusvirus
straints on p a c k a g i n g large increases in the size of the viral g e n o m e that isometric viruses have. Theoretically, there are no limits on the size of nucleic acid that could be p a c k a g e d in a rods h a p e d particle, but it is likely that limits will be f o u n d in the future. Plant viruses have great potential in the production of biologically active peptides and vaccines. H o w e v e r , there are v a r i o u s constraints that will have to be borne in m i n d if they are to be used on a commercial scale. These include the need to reduce the s y m p toms caused by the virus so that the host is not overly d a m a g e d , prevention of spread of the modified virus to other crops w h e r e it could cause severe infections, control of spread of any food products from the crop bearing the m o d i f i e d virus into the food s u p p l i e s of h u m a n s and other animals, and, of course, quality control of the modified virus to detect any non-functional mutants.
This Page Intentionally Left Blank
C
H
A
P
T
E
R
17
Variation, Evolu tion and Origins of Plant Viruses I. S T R A I N S OF V I R U S E S
Like other living entities, viruses substantially resemble the parent during their replication, but can change to give rise to new types or 'strains'. This inherent variation enables viruses to adapt to new and changing situations. Over longer periods of time new viruses arise and there must have been a time at which the archetypical virus arose. For many groups of organisms an u n d e r s t a n d i n g of the evolutionary pathways can be obtained from the fossil record over geological time, but viruses do not leave the conventional form of fossils. However, the increasing molecular information that is accruing on plant (and other) viruses is revealing what can be termed 'molecular fossil' information. Even so, our knowledge of the pathways of virus evolution is quite fragmentary, although there is no doubt that viruses have undergone, and continue to undergo, evolutionary change that is sometimes rapid. New strains provide the raw material for such change. From the plant pathologist's point of view, the existence of viral strains in the field that cause different kinds of disease is often a matter of considerable practical importance. Reliable criteria are needed for distinguishing and identifying strains. For these reasons the study of virus strains is an interesting and important aspect of plant virology. In this chapter, I discuss the variation of plant viruses and h o w this information is giving an u n d e r s t a n d i n g of viral origins and evolution.
A. Quasi-species As pointed out in Chapter 2 (Section I.C.2), a virus species is not a uniform population but is a quasi-species. The concept of quasi-species is fundamental to the u n d e r s t a n d i n g of virus variation and evolution; it is discussed in detail by Eigen (1993), Domingo et al. (1995), Smith et al. (1997), Domingo (1999) and Domingo et al. (1999a). The term quasi-species describes a type of population structure in which collections of closely related genomes are subjected to a continuous process of genetic variation, competition and selection. Usually, the distribution of mutants or variants is centerd on one or several master sequences. The selection equilibrium is meta-stable and may collapse or change when an advantageous mutant appears in the distribution. In this case, the previous quasi-species will be substituted by a new one characterized by a new master sequence and a new mutant spectrum. The stability of a quasi-species depends upon the complexity of the genetic information in the viral genome, the copy fidelity on replication of the genome and the superiority of the master sequence. A quasi-species has a physical, chemical and biological definition. In the physical definition, a quasi-species can be regarded as a cloud in sequence space which is the theoretical representation of all the possible variants of a genomic sequence. For a ssRNA virus of 10 kb the sequence space is 4 ~~176176176 Thus, the quasispecies cloud represents only a very small proportion of the sequence space and is constrained by the requirements of gene and nucleic
Virus acronyms are given in Appendix 1.
743
744
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
acid functions. Chemically, the quasi-species is a rated d i s t r i b u t i o n of related non-identical genomes. Biologically, a quasi-species is the phenotypic expression of the population, most likely d o m i n a t e d by that of the master sequence.
B. Virus strains The definition of a virus species is given in Chapter 2 (Section I.C) and, as discussed in the p r e v i o u s section, is essentially a cloud in sequence space. This makes a strict definition of a strain difficult, if not impossible. However, one has to describe variants within a species and, in reality, take a pragmatic approach. Characters have to be weighed up as to h o w they w o u l d contribute to making subdivisions and to communication, not only between virologists, but also to plant pathologists, extension workers, farmers and m a n y other groups. This can be illustrated by two examples. BWYV is a luteovirus with a wide host range including sugar beet in the USA and for m a n y years BMYV, which infected sugar beet in Europe, was regarded as a strain of BWYV (see Waterhouse et al., 1988). Confusion arose w h e n it was realized that the European luteovirus that was most closely related to BWYV did not infect sugar beet but was c o m m o n in the oilseed rape crop. This caused m a n y problems in explaining to farmers that the BWYV (Beet western yellows virus) in their overwintering oilseed rape crop would not infect their beet crop the next year. In an analysis of nucleotide sequences of 38 isolates of BYMV and BMYV from Europe, Iran and the USA, de Miranda et al. (1995b) identified three distinct sequence groups. The first contained isolates that could infect both oilseed rape and beet, the second infected only oilseed rape and the third only sugar beet. It was suggested that groups 1 and 2 be n a m e d BWYV (possibly BWYV-1 and BWYV-2) and the third group, BMYV. The second example is RTBV that has two major strains or isolate groups, one in southeast Asia and the other in the Indian subcontinent (Druka and Hull, 1998). These two strains differ in nucleotide sequence by about 25 To and at the a m i n o acid level by a b o u t 12-40% d e p e n d i n g on ORF (Hull, 1996). Thus, using
the criteria used for potyviruses (see below), these should be considered as two distinct viruses. However, the strains cause the same s y m p t o m s and cross-interact with their transmission helper virus, RTSV, in a similar manner. Thus, it w o u l d be confusing to farmers and extension w o r k e r s to give t h e m different names, although useful to virologists to indicate that they are different strains.
II. CRITERIA FOR THE R E C O G N I T I O N OF S T R A I N S A virus species might be defined simply as a collection of strains w i t h similar properties. Sometimes we wish to ask whether two similar virus isolates are identical or not; on other occasions we have to decide whether two isolates are different virus species or strains of the same species. Two kinds of properties are available for the recognition and delineation of virus strains: (1) structural criteria based on the properties of the virus particle itself and its components; and (2) biological criteria based on various interactions between the virus and its host plant and its insect or other vectors. These criteria were discussed in Chapter 2 in relation to the general problem of virus classification. Serological properties are based on the structure of the viral protein or proteins, but, because of their practical importance, serological criteria are considered here in a separate section.
A. Structural criteria 1. Nucleic acids
a. Heterogeneity in viral R N A genomes As described above, all the RNA genomes that have been examined have been found to exist not as a single nucleotide sequence but as a distribution of sequence variants around a consensus sequence. Most of the variants in a culture of a particular virus strain will normally consist of base substitutions at various sites, perhaps with some deletions or additions of nucleotides. However, more substantial variation can occur w h e n one or more segments of a multi-partite genome are not under selection
If.
pressure. Variants may arise quite rapidly, and these could lead to confusion in strain identification. For example, genomic RNAs 3 and 4 of BNYVV code for functions required for fungal transmission in the soil (Chapter 11, Section XII). When virus was isolated from leaves of infected sugar beets, where there is presumably no selection pressure to maintain the integrity of these genes, a wide range of deletion mutants was recovered (Burgermeister et al., 1986). This potential for rapid change must be borne in mind w h e n considering nucleic acid differences as criteria for recognizing virus strains.
b. Methods for assessing nucleic acid relationships i. Nucleotide sequences Nucleoeide sequences give valuable information about the extent of relationships between viruses. Determination of even relatively short sequences can provide useful information. However, to study the relationships between strains of a virus effectively it is necessary to determine the full nucleotide sequence of at least one, or preferably several, isolates. In earlier studies on nucleotide sequences, where sequence was d e t e r m i n e d directly on the labeled viral RNA, the microheterogeneity noted above was not detected. The sequence determined would normally be the consensus sequence for that particular culture (e.g. Swinkels and Bol, 1980). When cDNA clones of RNA viruses are used for sequencing, the resulting sequence may, by chance, contain sequences that are not typical of the consensus sequence. This possibility is checked by the sequencing of several clones covering the same section of the genome. Care should also be taken on possible sequence errors contributed by the reverse transcriptase in making cDNA from virion RNA (Bracho et at., 1998). To ensure that a functional genome has been sequenced, it is necessary to use a cloned DNA that has been shown to be infectious (e.g. Lazarowitz, 1988) or a cloned DNA that can be transcribed into infectious RNA (e.g. Dawson et al., 1986). In considering nucleotide sequences as a measure of relatedness between virus strains, it
C R I T E R I A FOR T H E R E C O G N I T I O N
OF S T R A I N S
745
must be remembered that the extent and distribution of such differences may vary quite widely depending on the virus and the part of the genome examined. To take some examples: (1) the coat protein gene and 3' untranslated region of three strains of PVY and two strains of WMV2 had 83% and 92% identity respectively (Frenkel et al., 1989). There were a few clusters of non-identical nucleotides but most were distributed more or less r a n d o m l y throughout the sequence studied. (2) By contrast, the coat protein and 3' untranslated regions of four strains of PPV and two strains of TEV had 97-99% identity (Frenkel et al., 1989). Similarly, the 39nucleotide 5' leader sequences of the RNA4 of seven strains of AMV were identical except at position 26, where A, G or U residues were found (Swinkels and Bol, 1980). Sequences with one or more recognition function(s) may often be highly conserved. For example, based on sequence homology in the 3' non-translated region, it has been proposed that PepMoV should be regarded as a strain of PVY (van der Vlugt et al., 1989). However, in considering these examples it must be remembered that certain sequences, especially the 5'- and 3'terminal regions are important for the control of replication and expression of the genome, and hence may be less variable than other regions. ii. Hybridization Nucleic acid hybridization experiments can give valuable information concerning the degree of base sequence homology between the nucleic acids of different virus isolates, but interpretation of the data may not be straightforward. The basis for nucleic acid hybridization was discussed in Chapter 15 (Section V.C). A significant advantage of hybridization procedures is that a comparison can be made between total genome RNAs, or RNA segments. This contrasts with serological tests, for example, which usually compare antigenic sites on the coat protein, which, for many viruses, represents only about one-tenth or less of the total genome. However, the degree of nucleic acid homology estimated b e t w e e n different strains will depend on various experimental factors such as the stringency of the hybridization conditions and the conditions for enzymatic removal of
746
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
u n p a i r e d nucleotides (see Fig. 17.1). Figure 17.1 also illustrates the use of the dot blot m e t h o d in a semi-quantitative manner. Koenig et al. (1988a) found that quantitative dot blot hybridization provided a very sensitive m e t h o d for distinguishing closely related viruses that could barely be distinguished in serological tests. On the other h a n d , they observed some unexpected cross-hybridization between viruses belonging to different taxonomic groups. These cross-reactions m a y have resulted from real base sequence similarities, as has been found between some viruses in separate groups, and may also have resulted from the use of r a n d o m - p r i m e d cDNAs. iii. Heterogeneity mapping Heterogeneity m a p p i n g is a m e t h o d based on RNA hybridization that allows the detection of single point mutations provided they occur in a significant proportion of the molecules. The m e t h o d takes a d v a n t a g e of the ability of RNase A to recognize and excise single-base mismatches in RNA heteroduplexes (Winter et at., 1985). Labeled (-)-sense RNA probes are transcribed
from cDNA clones of the RNAs. Following hybridization with the test RNA, and digestion by RNase A and T, fragments are separated and sized by PAGE. The m e t h o d has been used to assess the extent and location of heterogeneity a m o n g strains of CMV (Owen and Palukaitis, 1988) and satellite RNAs of this virus (Kurath and Palukaitis, 1989b), and makes possible the assessment of entire populations of RNA molecules for major sites of heterogeneity. By contrast, the sequencing of individual clones gives precise and detailed information on a few molecules that may m a k e up a tiny fraction of the virus population. P o l y m o r p h i s m s between heterologous duplexes from cRNA transcripts from different virus isolates can be recognized by their electrophoretic mobilities (Rosner et al., 1999). This method distinguished isolates of PNRSV. iv. Restriction fragment length polymorphism The cutting of d s D N A with restriction endonucleases can reveal p o l y m o r p h i s m s (RFLPs). d s D N A s can be obtained directly from D N A viruses or as c D N A produced from RNA viruses by reverse transcription. The usual application
Fig. 17.1 Use of the dot blot technique to estimate degree of relationship between strains of a virus. Autoradiograph showing the extent of sequence similarity between MSV-MNM and other MSV strains. Four identical filters were each spotted, on the right with 2 ng of DNA from different MSV strains under test and on the left with doubling dilutions (2 ng ~ 7.8 pg) of MSV-MnM (N) as controls. Filters were hybridized with a full-length nick-translated clone of MSVMnM (N) prior to washing under conditions of different stringency (0.02, 0.1, 0.5 and 2 • SSC at 65~ From Boulton and Markham (1986), with permission.
II.
of this technique is to detect differences between viral genomes (e.g. Hull, 1980; Kruse et al., 1994; Villegas et al., 1997), but it can also be used to show similarities between genomes. v. Single-strand conformation polymorphisms Analysis of single-strand conformation polymorphisms (SSCPs) is a powerful method for detecting differences (or similarities) between genomes. When the (+) and (-) strands of a dsDNA are separated by heat treatment, they attain metastable sequence-specific folded structures that can be detected by their electrophoretic mobilities in non-denaturing polyacrylamide gels (Orita et al., 1989b); even single nucleotide changes are detectable under the right conditions. A non-isotopic variant of SSCP analysis was used for comparing genomes of BNYVV (Koenig et al., 1995). Both SSCP and RFLP can be determined on PCR products after reverse transcription of viral RNAs. vi. Other properties of the genomic nucleic acid
Size of the viral nucleic acid may differ with different virus strains, for example with various TRV isolates (e.g. Cooper and Mayo, 1972). c. Additional genes
The RNA2 of the TCM strain of TRV is considerably larger than that of other strains that have been sequenced, for example PSG. Angenent et al. (1986; 1989b) showed that the greater length was due partly to a repetition of 1099 3' nucleotides from RNA1, which includes a 16kDa ORF, and partly to a 29-kDa ORF that was unique to this RNA2 (see Fig. 6.38). The 29-kDa ORF has no significant homology with the 28.8kDa ORF of PSG RNA1, and its origin and function, if any, are not established. d. Subgroups of strains
When a sufficient number of strains has been examined, nucleotide sequence relationships may be used to delineate subgroups of strains of a virus. For example, based on competition hybridization tests, 30 strains of PSV could be divided into two groups with little homology between them, but extensive homology within
C R I T E R I A FOR T H E R E C O G N I T I O N OF S T R A I N S
747
groups (Diaz-Ruiz and Kaper, 1983). Similarly, CMV strains have been subgrouped (see Section VII.C.4). e. Some limitations concerning base sequence data
A given base substitution, deletion or addition may have very different effects in the protein resulting, depending on a number of circumstances. The following factors may be important: 1. Because the genetic code is degenerate, many base substitutions cause no change in the amino acid being coded for. For example, in the TMV strain LnA, derived from the L strain, there were 10 base substitutions, seven of which occurred in the third position of in-phase codons and did not influence amino acid sequence (Nishiguchi et al., 1985). 2. A given base substitution may result in change to an amino acid of very similar properties, which causes very little change in the protein. Alternatively, the change may be to a very different kind of amino acid (e.g. from an aliphatic side-chain to an aromatic one), giving rise to a viable protein with changed physical properties or to a non-functional mutant that does not survive. 3. A deletion or addition of one or two bases will cause a frameshift mutation with greater or lesser effect depending on whether it is near the beginning or the end of the gene, whether a second change (addition or deletion) brings the reading frame back to the original, and how many proteins are coded for by the section of nucleic acid in question. 4. Amoregeneral probleminusingbase sequence data for classification is that some parts of the genome, and some products, may have multiple functions. Some parts of the genome may code for a single polypeptide, but others may code for more than one; in addition, some polypeptides may have more than one function. Some parts of the genome may have both coding and control or recognition functions; other parts may have just control functions. Furthermore, mutations in one gene may affect the production of another. For example, mutations in the presumed polymerase gene of TMV
748
~7 V A R I A T I O N , E V O L U T I O N A N D O R I G I N S OF P L A N T VIRUSES
mutant LHA cause reduced synthesis of the cell-to-cell m o v e m e n t protein (30 kDa), thus reducing efficiency of movement of this strain (Watanabe et al., 1987b). Thus, even if we knew the full base sequences for the nucleic acids of a set of viral strains, it w o u l d be unwise to use these sequences to establish degrees of relationship without other information. It was once thought that a virus classification scheme based only on nucleotide sequence w o u l d be the ultimate aim (Gibbs, 1969). It is now apparent that the significance to be placed on nucleic acid base sequence data can be judged from a biological point of view only in conjunction with knowledge of the organization of the genome and the functions and interactions of its parts and products. The use of infectious clones and molecular biological techniques is helping in the determination of the association between sequence and biological variation. 2. Structural proteins The coat protein or proteins and other structural proteins found in viruses are very important, both for the viruses and for virologists wishing to delineate viruses and virus strains. The coat proteins of the small RNA viruses must have evolved to give a satisfactory balance between three important functions. 1. The ability to self-assemble around the RNA; mutants are k n o w n in which this function is defective even at normal temperatures. For example, strain PM2 of TMV cannot form virus rods with RNA. The protein aggregates in vitro at pH 5.2 to form long, open, flexuous, helical structures rather than compact rods (Zaitlin and Ferris, 1964); 2. Stability of the intact particle inside the cell, and during transmission to a fresh host plant. 3. The ability to disassemble to the extent necessary to free the RNA for transcription and translation. A variant that could not carry out this function w o u l d not survive in nature. For example, Bancroft et al. (1971) described a m u t a n t of CCMV induced by nitrous acid that was unable to be uncoated in the cell, and was therefore non-infectious
in spite of the fact that RNA isolated by the phenol procedure from the virus was highly infectious. For the small RNA viruses, the coat protein is of particular importance for the delineation of viruses and virus strains. Besides the intrinsic properties of this protein (size, amino acid sequence, and secondary and tertiary structure), many other measurable structural properties of the virus depend largely or entirely on the coat protein. These include serological specificity, architecture of the virus, electrophoretic mobility, cation binding and stability to various agents. Thus, ideas on relationships within groups of virus strains, based on properties dependent on the coat protein, may be rather heavily biased. On the other hand, if mutations in the non-coat protein genes have occurred more or less at the same rate as in the coat protein during the evolution of strains in nature, then such views on relationships may be reasonably well based. The Polyvirus genus will serve to illustrate the use of the properties of coat proteins in the delineation of virus strains. Potyviruses have been one of the most difficult virus groups to study taxonomically. The group contains about one-tenth of all the known plant viruses. The viruses infect a wide range of host plants and exist in nature as many strains or pathotypes differing in biological properties such as host range and disease severity. It has been considered by some workers that strains of potyviruses may form a continuous spectrum between two or more otherwise distinct viruses, making delineation of viruses and groups of s-trains difficult or impossible. However, comparisons between the amino acid sequences of the coat proteins of several viruses and many strains indicate that this approach may provide a useful basis for taxonomy within the group (Shukla and Ward, 1989a,b). Analysis of the 136 possible pairings between a set of viruses and strains revealed a clear-cut bimodal distribution, with distinct viruses having an average sequence homology of 54%, while strains averaged 95% (see Fig. 2.2). These data give no support for the 'continuous spectrum' idea among the potyviruses. Distinct
II. (:RITERIA F O R
PVY-
1. In relative amounts of top component (empty protein shells). For example, a m o n g the tymoviruses, the proportion of empty shells to viral nucleoprotein is usually in the range of 1:2 to 1:5 for TYMV and 10:1 to 15:1 for OkMV. Even quite closely related strains may vary in this property, for example strains of RCMV (Oxelfelt, 1976). For some multipartite viruses the proportion of top component
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIlllllllllIIIII I IIiiIIIII11111 IIIII
D
II
I I II II IIIII-I'I I IIIIIII II III
IIIII III
l[lllllllllllllllllllllg III IIIIIIIIIIIIIII II I'"IIIII"llI llllllnllllllll II
PWV-TB
749
OF STRAINS
4. Proportion of particle classes The proportion of particles with differing sedimentation rates found in purified virus preparations or in crude extracts may vary quite widely with different strains of a virus or members of a virus genus. Variation of three kinds can be distinguished:
••••••••••••••••••••••••••••••••••••••••••••••••••••n•••••••••••••••••••••l•••••j•••••••• IIIIl IIl I IIIIIIIIIIIIIIIIIII I II IllVllll11111
J G M V - JG
RECO(]NITION
3. Non-structural proteins The non-structural virus-coded proteins are not much used in delineating strains. Mayo et al. (1982) could detect no difference in the tryptic peptides obtained from the VPg of different strains of RpRSV or TBRV. Some strains of TMV may differ widely in the amino acid sequence of their 30-kDa proteins, as discussed by Atabekov and Dorokhov (1984). Four structural proteins of the two serotypes of PYDV (SYDV and CYDV) fell into two groups based on peptide mapping. Proteins M and N differed little between the strains, whereas M 2 and G were significantly different (Adam and Hsu, 1984).
viruses showed major differences in length of their coat proteins (Fig. 17.2). Major differences in amino acid sequence were near the N-termini, with high h o m o l o g y in the Cterminal half of the proteins. On the other hand, strains have very similar N-termini. Two exceptions to this pattern appear to reflect the misplacing of certain potyvirus isolates on the basis of previous data. Serological tests suggested that PeMV and PVY were only distantly related, yet the sequence data shown in Fig. 17.2 clearly indicate that PeMV should be considered a strain of PVY. SMV-N and SMV-V, formerly considered to be strains of SMV but, when shown to have a sequence homology of 58%, were considered as two distinct viruses (Shukla and Ward, 1988). In both cases and in others a pragmatic approach has to be taken in assigning species status. High-performance liquid chromatography of tryptic peptides may be useful in differentiating potyviruses and their strains (Shukla et al., 1988a). This technique does not provide the detailed information obtained by amino acid sequencing, but its greatest value may lie in the ease with which the method can be applied. The projecting (P) domain of Tombusvirus coat proteins has a more variable amino acid sequence than the structural (S) d o m a i n (Hearne et al., 1990). Thus, greater variability at the exposed surface may be a feature of both rod-shaped and icosahedral plant viruses.
PPV-D
THE
II I
I
[
I I
II
I III
I
II
llIl I II
IIII 111I11 111111111
llll IIIIII II 111111 II
IIII II IIIIII II[III 11211 IIIII
i "
t'-fll II III I
Pe MV SMV-V
I III
I
I
IIIIIIIIIIIIIII III IIIIIIIIIIIIIIII I I l'lllllllllllllll
J
1
I I I
I I I I
I II
I
III IIII III I
llllllllllllllllllllnllll I IIIllfllllilllllllll IIIIIII I IIIIIIIIIIIIIIIIIIII III ! lllllll III l-llIilIIIIIfl Ill
TVMV
llllllllllllll IIIIIIIII
TEV-NAT
t
.
.
.
.
I 50
.
.
.
.
II IIIIIIII IIIII I'l' II II11IIII11flillllll
I 1OO
.
.
.
.
I
.
.
.
I I I I II I
.
I
150
Amino Acid Residue Number
200 (PPV-
.
.
I I
.
.
III .... I"llflI
I 250
.
.
III
I IIIIIIIIIII Ill
II
.
I
I IIII II
1
II
IIIIII II
I llllll llll I flIII II
.
I
'
'
'
300
D)
Fig. 17.2 Schematic diagram showing the location of amino acid sequence differences between seven distinct members of the Potyvirus genus and PeMV. The sequences were compared with strain D of PVY, the type member. PeMV is very similar to PVY in its coat protein sequence. From Shukla and Ward (1988), with permission.
750
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
has been shown to depend on a function of one RNA species. 2. The proportion of nucleic acid components encapsidated may vary in different strains of viruses with multipartite genomes, for example AMV (van Vloten-Doting et al., 1968). Again, nucleoprotein proportions may be under the control of a particular RNA species. 3. Abnormal particle classes may be produced by particular strains. Thus, Hull (1970a) described an isolate of AMV producing considerable amounts of particles longer than the B component. It should be remembered that the proportion of particle classes can be affected by factors other than the strain of virus. These include: (1) time after infection; (2) host species; (3) environmental conditions; (4) system of culture, for example the proportion of TYMV empty protein shells is higher in infected protoplasts than in whole leaf tissue; and (5) isolation procedure.
structure, which affects the availability of ionizing groups. Mobility is also dependent on the ions present in the buffer used. Isolates of PEMV that differed in aphid transmissibility also differed in electrophoretic behavior (Fig. 2.7) (Hull, 1977b).
c. Stability and density Among the small RNA viruses, differences in stability and density have been used to differentiate virus strains. The RNA content of the virus may vary with strain and thus affect buoyant density in strong salt solutions (e.g. Lot and Kaper, 1976). However, differences in the coat protein most commonly lead to a difference in stability or density.
B. Serological criteria The nature of antigens and antibodies, the basis for serological tests, and their advantages and limitations are discussed in Chapter 15 (Section IV). This section considers the use of serological criteria to delineate viruses and virus strains.
5. Other structural features a. Architecture of the virus particle
Related viruses will be expected to have very similar size, shape and geometrical arrangement of subunits. However, significant differences in particle morphology have been found within groups of related strains. Differences in rod length are frequent between strains of helical viruses such as TRV (e.g. Cooper and Mayo, 1972) and BSMV (Chiko, 1975). Sometimes the variation in architecture appears to be 'abnormal' even though the strain of virus is a viable one. Thus, the packing of the coat protein of the Dahlemense strain of TMV involves a periodic perturbation of the helix (Caspar and Holmes, 1969). Some AMV strains contain abnormally long particles that have the normal diameter but contain more than one RNA molecule (Hull, 1970a: Heijtink and Jaspars, 1974).
b. Electrophoretic mobility The electrophoretic mobility of a virus depends in the first place on the amino acid composition of the protein and second on the three-dimensional
1. S~me general considerations a. Presence or absence of serological relationship
Serological tests provide a useful criterion for establishing whether two virus isolates are related or not. Any of the tests described in Chapter 15 can be applied, but most commonly some modification of the precipitation reaction or ELISA tests is used. Provided adequate precautions are taken, serological tests can be valuable for placing viruses into groups. If two virus isolates show some degree of serological relationship, it is highly probable that they will have many other properties in common and belong in the same virus group. There are a few unexplained exceptions. Various examples are known of viruses that undoubtedly belong in the same group but that show no serological cross-reactivity, for example TYMV and EMV in the Tymovirus genus. In making tests for serological relationships, there are several potential sources of error: 9 Presence in viral antisera of antibodies reacting with host constituents such as the abun-
II. CRITERIA FOR THE RECOGNITION OF STRAINS 751 d a n t p r o t e i n ribulose l ' , 5 ' - b i s p h o s p h a t e carboxylase. 9 Non-specific precipitation of host materials in crude extracts. 9 Non-specific precipitation of viral antigens, especially at high concentrations. 9 C o n t a m i n a t i o n of antigen preparations with other viruses. 9 Virus altered d u r i n g isolation. It s h o u l d always be borne in m i n d that virus m a y be altered d u r i n g isolation in a w a y that can affect its serological specificity. 9 Non-reciprocal positive reactions. To d e m o n strate that two viruses are serologically unrelated, reactive antisera m u s t be p r e p a r e d against each of the viruses u n d e r test. It m u s t be s h o w n that each reacts w i t h its o w n antiserum, but gives no reaction with the heterologous antiserum. This reciprocal test is necessary because the viruses m i g h t in fact be related, but one m a y occur in too low a concentration in the extracts to give any positive reaction. Negative o n e - w a y tests are of little value. As discussed in the next section, it is preferable to use high-titer antisera to d e m o n strate a lack of serological relationship. 9 Isolates taken from the field m a y be mixtures of several different serotypes (e.g. Dekker et al., 1988). These considerations apply particularly to the use of polyclonal antisera. The use of MAbs avoids several of these problems, but they have
limitations of their o w n as discussed in Section II.B.3.b. b. Degrees of serological relationship i. Among a group of virus strains
A considerable a m o u n t of experimental w o r k has been directed t o w a r d d e t e r m i n i n g degrees of relatedness within groups of strains and in a t t e m p t s to correlate serological p r o p e r t i e s w i t h other biological and chemical characteristics. Delineation of virus strains is a particularly i m p o r t a n t aspect of any p r o g r a m e designed to p r o d u c e resistant varieties of a host species. If two isolates of a virus are identical, they will r e s p o n d identically w h e n cross-reacted w i t h each other's antisera, w h a t e v e r form of serological test is applied. If, however, they are related but distinct, some degree of crossreaction will be observed, at least w i t h polyclonal antisera, a l t h o u g h the reactions will not be identical. Various types of serological test can be used to identify and distinguish virus strains. Examples are given in Table 17.1. W h e n a group of only two or three virus isolates is to be considered, it is a relatively simple matter, p r o v i d e d technical p r e c a u t i o n s are observed, to d e t e r m i n e w h e t h e r the isolates are unrelated serologically, w h e t h e r they are identical, or w h e t h e r they show differing degrees of relationship. Using the same set of isolates and the same antisera, quite reproducible results can be obtained, to indicate, for example, that
Table 17.1 Some serological procedures used for the delineation of viruses and virus strains Procedure
Virus or virus group
Reference
Indirect ELISA
Tymo-, tombus-, como-, tobamo-, potex-, carla-, poty-viruses Carlaviruses Cucumoviruses
Koenig (1981)
PMV
Berger and Toler (1983)
F(ab')2 ELISA Radial double diffusion in agar, ELISA and SSEM Quantitative rocket immunoelectrophoresis Electroblot immuno-assay
Tymo-, tombus-, como-, nepo-, tobamo-, potex-, carla-, poty-viruses Direct or indirect ELISA Ilarviruses and AMV GFLV strains Various ELISA procedures Indirect protein A-sandwich ELISA Tobamoviruses and virus strains Indirect EL1SA MSV isolates Immunocapture-PCR P V Y NTN and CSSV
Adams and Barbara (1982) Rao et al. (1982)
Burgermeister and Koenig (1984) Halk et al. (1984) Huss et aI. (1987) Hughes and Thomas (1988) Dekker et al. (1988) Weidemann and Maiss (1996) Hoffmann et al. (1997)
752
17 VARIATION, EVOLUTION AND O R I G I N S OF PLANT VIRUSES
strains A and B are closely related and that both are more distantly related to strain C. However, w h e n large n u m b e r s of related strains are tested, the situation may become quite complex and less and less meaningful as more strains are considered in relation to one another. ii. Experimental variables There are a number of important experimental variables that can affect the estimated degree of serological relationship between viruses and strains. (1) A major source of experimental variation is the variability in antisera, both in successive bleedings from the same animal and in sera from different individuals. The proportion of cross-reacting antibody present in a series of bleeding taken over a period of months from a single animal may vary widely (Koenig and Bercks, 1968). (2) The extent to which antisera to two virus strains cross-react is usually correlated with the antibody content of the serum. Sera of low titer show lower cross-reactivity, and those with high titers show greater crossreactivity. Thus, to detect serological differences between closely related strains using polyclonal antisera it is preferable to use antisera of fairly low titer. To demonstrate distant serological relationships, it may be necessary to use high-titer antisera. (3) Many virus preparations used for immunization and for antibody assay may contain varying amounts of free coat protein or coat protein in various intermediate states of aggregation or in a denatured state. Coat protein in the intact virus may lose amino acids through proteolysis. Antibodies reactive with coat protein in these various forms may or may not indicate the same sort of relationships as antibodies against intact virus. An example of this sort is discussed in relation to the Potyvirus genus in Section II.B.4. The method used to detect and assay cross-reacting and strain-specific antibodies may affect the apparent degree of relationship. Examples are given in the references listed in Table 17.1. c. The serological differentiation index In spite of all the variables, useful assessment of degrees of serological relationship can be obtained by testing successive bleedings from many animals and pooling the results. Most
quantitative measurements of degrees of serological relationship have been carried out using precipitation titers. In such tests, the extent of serological cross-reactivity can be expressed by a serological differentiation index (SDI) (van Regenmortel and von Wechmar, 1970) (see Chapter 15, Section IV.A.3.d). The SDI values are equal to the difference in those titers expressed as negative log2. For example, such replicated comparisons have been made for sets of tobamoviruses (van Regenmortel, 1975) and tymoviruses (Koenig, 1976). ELISA can also be used to calculate SDIs as a measure of the extent of serological relatedness between viruses or virus strains (Jaegle and van Regenmortel, 1985; Clark and Barbara, 1987). Table 17.2 shows a comparison of the SDIs obtained from ELISA and precipitin tests. There were differences in the reciprocal SDIs found by ELISA for pairs of viruses, and the average of these values did not correspond closely to those found by precipitin SDIs. Nevertheless, both kinds of test show clearly that CGMMV is substantially different from the other tobamoviruses. Clark and Barbara (1987) describe a more refined statistical procedure for calculating SDIs from ELISA tests that is capable of discriminating reliably among virus strains that differ by as little as 0.2 SDI. 2. R~)le of virus components in serological reactions There is no good evidence that plant viral ssRNA can elicit RNA-specific antibodies. Antibodies formed in response to injection with a plant virus react only with the virus protein, either in the intact virus or as various partial degradation products of the intact protein shell. A formal demonstration of the role of the protein was made by Fraenkel-Conrat and Singer (1957). They carried out mixed reconstitution experiments between serologically distinct strains of TMV. The artificial hybrid virus had the serological type of the protein used to coat the RNA, but the progeny following infection had protein of the type from which the RNA was obtained. Nevertheless, the viral RNA may play some secondary role in stimulating production of
II.
( : R I T E R I A FOR T H E R E ( . ' O ( } N I T I O N OF S T R A I N S
753
Table 17.2 Serologicaldifferentiation indices (SDIs) for pairs of tobamoviruses calculated from ELISA and precipitin tests a SDI from ELISA
Tobamoviruses x
y
TMV U2 TMV TMV ToMV U2 RMV U2 TMV ToMV
ToMV ToMV RMV U2 RMV RMV CGMMV CGMMV CGMMV CGMMV
y-anti-x
x-anti-y
Average value
AverageSDI from precipitin test
Sequence similarity in viral coat proteins(%)
1.4 0.6 2.0 2.4 1.9 3.5 7.0 2.6 5.4 5.7
0.5 1.9 1.2 1.9 0.7 2.5 6.4 8 6.9 6.9
0.9 1.3 1.6 2.1 1.3 3.0 6.7 5.3 6.2 6.3
1.2 1.9 2.1 2.7 4 4.5 5 6 6.8 7
82 70 44 74 47 46 30 33 36 33
From Jaegle and van Regenmortel (1985), with kind permission of the copyright holder, 9 Elsevier Science.
antibodies against the viral protein shell. Intact TYMV is substantially more i m m u n o g e n i c than the apparently identical e m p t y protein shell, w h i c h contains no R N A ( M a r b r o o k and Matthews, 1966). This difference was found in rabbits and mice using several routes of injection. The difference persisted t h r o u g h o u t the time course of the p r i m a r y response and was also found in the response to a second injection. Isolated TYMV RNA injected at the same time did not a u g m e n t the i m m u n o g e n i c i t y of the e m p t y protein shells. Artificial e m p t y protein shells p r o d u c e d f r o m the infectious virus in vitro were no more i m m u n o g e n i c than the natural e m p t y shells. Non-infectious TYMV n u c l e o p r o t e i n w a s just as i m m u n o g e n i c as infectious virus. Thus, it was concluded that the enhanced i m m u n o genicity of the nucleoprotein m u s t be due to the physical presence of the RNA inside the particle. Whether this enhanced i m m u n o g e n i c i t y of the viral nucleoprotein is a general feature of plant viruses remains to be determined. TMV appears to be more i m m u n o g e n i c than either p r o t e i n rods or s u b u n i t s ( M a r b r o o k and Matthews, 1966; Loor, 1967). The m e c h a n i s m by which the ssRNA within the virus stimulates i m m u n o g e n i c i t y is not yet understood. dsRNAs can be immunogenic, dsRNA antisera react with dsRNA but not d s D N A or singles t r a n d e d nucleic acids. The antisera lack
specificity for p a r t i c u l a r d o u b l e - s t r a n d e d nucleic acids (see Chapter 15, Section IV.A.2). There are several reasons w h y we w o u l d expect intact viruses (such as TMV and TYMV) and protein subunits or subviral aggregates p r e p a r e d from them to differ in the antigenic sites they possess: 9 Some antibody-combining sites on the intact virus m a y be m a d e up of parts of the exposed surface of two or more subunits. Such a site w o u l d not exist in isolated subunits. 9 Subunits probably have characteristic combining sites, which are m a s k e d w h e n the subunits are packed into the intact shell. 9 Conformational changes occur w h e n the subunits aggregate, so that the configuration of the exposed surface of the packed subunit m a y not be the same as w h e n it exists as a monomer. Examples of all these p h e n o m e n a are k n o w n a m o n g plant viruses and their protein subunits. 3. Procedures used for delineating viruses and strains a. Assay methods
The various serological methods that are used in the detection and assay of viruses are discussed in Chapter 15 (Section IV.B). Most of these procedures have been used for delineating
754
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
viruses and virus strains. Some examples are listed in Table 17.1. This list reflects the fact that ELISA procedures have become the most p o p u l a r for the delineation of viruses and strains.
b. Monoclonal antibodies The advantages and disadvantages of using MAbs for assay, detection and diagnosis of viruses were s u m m a r i z e d in C h a p t e r 15 (Section IV.D). The outstanding value of MAbs in the delineation of virus strains is that their molecular homogeneity ensures that only one antigenic determinant is involved in a particular reaction. The high specificity of this single interaction is not s w a m p e d in a large n u m b e r of other interactions as with a polyclonal antiserum. Provided a MAb can be found that recognizes a small antigenic change b e t w e e n two virus strains, then very fine distinctions can be m a d e in a reproducible manner. A l t h o u g h the ability of different MAbs to distinguish single amino acid exchanges m a y vary widely, some may be able to do so (A1-Moudallal et al., 1982). However, there are several limitations in the use of Mabs. (1) There is usually no i m m u n o precipitation between MAbs and viral protein monomers. (2) MAbs are often sensitive to minor conformation changes in the antigen such as m a y be caused by detergent or by binding of antigen to an ELISA plate (e.g. Dekker et al., 1987). (3) A m o n g a set of virus strains the relative reactivity of different MAbs may vary considerably. For example, TMV strain 06 differs from TMV by residue exchanges at positions 9, 65 and 129 of the coat protein. In the tests s h o w n in Fig. 17.3, this strain reacted like TMV with MAb a, more strongly than TMV with MAb c, and not at all with MAb b. (4) MAbs may be heterospecific, that is, they may frequently react more strongly with other antigens than with the virus used for i m m u n i z a t i o n (see Chapter 15, Section IV.D). The reaction of strain 06 with MAb c, which is stronger than that with TMV, the strain used as the i m m u n o gen, illustrates this p h e n o m e n o n (Fig. 17.3). If, during the selection of h y b r i d o m a clones for the isolation of MAbs, the clones are tested
only with the strain of virus used as i m m u n o gen, MAbs with low affinity for this strain may go undetected and be discarded. A m o n g such MAbs m a y be clones that w o u l d be very useful for the detection of other strains. Thus, w h e n searching for strain-specific MAbs it is important to screen clones against a panel of structural relatives of the i m m u n o g e n .
1I
A
Ioo01,03 TMV
~
V o~
0.6
0.40.2
~ ' ~ - ~ 1 7 6
Y-TIMV 2 R
o
0.8
0.6
Er o u'3
~3 0
0.4
0.2 i D~..---- ?
29
.?
Oe /IY-TAMV
"
~//
C
--?,rU 2 ,HR
os
Y-TAMV
o.81-
9
~
9
OM
0.4 !~ O"21.~~/I...._=.__._._..t 0
0.5
,~O3
1
2
U2,HR
_
4 VIRUSlua/I N--~,
Fig. 17.3 Detection of tobamoviruses by indirect ELISA using three MAbs (a, b and c) obtained using TMV as immunogen. Strains 01, 03 and 06 are orchid strains and 0M is the common Japanese strain of TMV; Y-TAMVis a strain of tomato mosaic virus; U2 and HR are two distinct tobamoviruses. From van Regenmortel (1984c), with kind permission of the copyright holder, 9 Blackwell Science Ltd.
II.
Another potential limitation of MAbs in the delineation of strains can occur if two strains have an identical antigenic determinant in common. If, by chance, the MAb specificity is directed against this determinant, the strains will appear identical even though they may have substantially different determinants elsewhere in the molecule. For example, strain YTAMV is a member of the ToMV group of strains that has an 18% difference in coat protein amino acid sequence compared with TMV. The two viruses are readily distinguished by polyclonal antisera but not by some MAbs (e.g. antibody c in Fig. 17.3). These limitations highlight the importance of generating diverse panels of MAbs for the delineation of viruses and strains. Experiments with ToMV will further illustrate the problem. Strains of this virus are considered to be serologically quite uniform. Ten MAbs raised against the virus reacted in an identical manner with 15 ToMV strains and isolates. However, two of them cross-reacted with TMV and RMV (Dekker et al., 1987). 4. Antigenic sites involved in the serological delineation of viruses and strains Because of the crucial role they play in intersubunit bonding, the sides of protein subunits that make up the shells or rods of a particular virus might be expected to be fairly constrained in the extent to which amino acid replacements would allow the subunit to remain functional. This would be particularly so with rod-shaped viruses (and also certain isometric viruses) in which RNA-protein interactions are also important. One might expect much less constraint on that part of the protein subunit that makes up the surface of the virus, and that would therefore also provide the antigenic sites of the intact virus. This expectation has been confirmed for members of the Potyvirus genus. Biochemical and immunological evidence suggested that the N-terminal 29 amino acids of TEV were hydrophilic and located at or near the virus surface (Allison et al., 1985). Mild proteolysis by trypsin of the particles of six distinct potyviruses showed that the N- and C-terminal regions of the coat proteins are exposed at the particle surface. Trypsinization removed 30-67
C R I T E R I A FOR THE R E C O G N I T I O N OF S T R A I N S
755
residues from the N-terminus and 18-20 from the C-terminus, the length removed depending on the virus (Shukla et al., 1988b). This proteolysis left a fully assembled, infectious virus particle containing protein cores consisting of 216 or 218 amino acids. Electroblot immunoassays with polyclonal antisera showed that potyvirus-specific antigenic sites are located in the trypsin-resistant core protein region. Thus, antibodies to the dissociated core protein should react with most potyviruses. On the other hand, the surface-located N-terminus is the only large region in the coat protein that is unique to a particular potyvirus, and most virus-specific antibodies should react with this region. This fits with the amino acid sequence data, which showed that the N-terminal region is the most variable in potyvirus coat proteins (see Fig. 17.2). It has been known for some time that potyviruses become partly degraded on storage. The use of partially degraded virus as an immunogen or in antigenic analyzes may account for many of the contradictory reports in the literature concerning serological relationships among the potyviruses (Shukla et al., 1988b) (see also Fig. 2.3). Shukla et al. (1989a) developed the following simple procedure to remove cross-reacting group-specific antibodies. The virus-specific N-terminal region of the coat protein of one potyvirus was removed using lysylendopeptidase. The truncated coat protein was then coupled to cyanogen bromide-activated Sepharose. By passing antisera to different potyviruses through such a column, the cross-reacting antibodies were bound. Antibodies that did not bind reacted only with the homologous virus and its strains, as judged by electroblot immunoassays. The practical utility of this procedure has been demonstrated for a group of 17 potyvirus isolates infecting maize, sorghum and sugarcane in Australia and the United States whose taxonomy was in a confused state (Shukla et al., 1989b). The results demonstrated that the 17 strains belong to four distinct potyviruses, for which the names JGMV, MDMV, SCMV and SrMV were proposed. Electroblot immunoassays using native and truncated coat proteins (minus the N-terminus)
756
17 VARIATION, EVOLUTION AND O R I G I N S OF PLANT VIRUSES
can be used to screen MAbs to determine whether they are group specific or virus specific. This procedure was used to distinguish MAbs that were virus specific from those that reacted with two or more, and sometimes all 15, of the potyviruses tested (Hewish and Shukla, reported in Shukla and Ward, 1989b). Using panels of MAbs to ACMV, ICMV and OLCV, epitope profiles have been obtained for various whitefly-transmitted geminiviruses (see Swanson et al., 1992; Konat6 et al., 1995; Harrison et al., 1997). These profiles reveal MAbs that react with several viruses and those that are virus or even strain specific. The epitope profiles of 12 begomoviruses is illustrated in Fig. 17.4. 5. Production of antibodies against defined antigenic determinants Geysen et al. (1984) described a procedure for the rapid concurrent synthesis of hundreds of peptides on solid supports. These had sufficient purity to be used for ELISA. Using sets of such peptides and antisera against a virus, immunologically important amino acid sequences on the virus could be closely defined. This procedure has been applied successfully to the analysis of both polyclonal sera and MAbs raised against potyviruses (Shukla et al., 1989c). This work opened up the possibility of using synthetic peptides corresponding to defined antigenic sites as immunogens to generate group-specific, virus-specific and Virus and source
Host
ACMV, Kenya OLCV, Ivory Coast
Cassava Okra
EACMV, Madagascar TLCV, Burkina Faso TYLCV, Senegal ICMV, India TLCV, India TYLCV, Thailand
Cassava Tobacco Tomato Cassava Tobacco Tomato
CLCuV, Pakistan BGMV, Puerto Rico TGMV, Brazil CLCrV, USA
Cotton Bean Tomato Cotton
perhaps some strain-specific serological probes. Intrinsic and extrinsic factors affecting antigenic sites in relation to the prediction of important amino acid sequences are discussed in Berzofsky (1985). The ability to be able to express single-chain variable fragments from cloned cDNAs on the surface of phage particles (see Chapter 15, Section IV.E) gives another approach to obtaining highly specific antibodies. 6. Antibodies against non-structural proteins Variation in non-structural proteins can sometimes be used to define strains. For instance, Chang et al. (1988) used the serological reactions of nuclear inclusion proteins to study relationships between a set of potyviruses and potyvirus strains. 7. Other uses of strain-specific antisera Besides the use of serological methods for establishing relationships between plant viruses, strain-specific antisera provide very useful reagents for various kinds of experiments. For example, antisera specific for ToMV strains have been used to monitor the effectiveness of the protection given by infection of tomatoes with mild strains of ToMV to superinfection with wild strains (Cassells and Herrick, 1977), and to study the mechanism of crossprotection (Barker and Harrison, 1978). Strain-specific antisera were used to show that, when tobacco leaf protoplasts were
MAb (SCR no.) 12 14 17 18 20 22 23 27 29 32 33 52 53 ,54 55 56 58 60 62 66
0 1 2 3 4
Fig. 17.4 Epitope profiles of 12 begomoviruses from six hosts, illustrating the differences between viruses from the same host in different continents, and similarities among viruses from different hosts in the same continent. Strengths of reaction range from imperceptible (0) to very strong (4). From Harrison and Robinson (1999), with kind permission of the copyright holder, 9 Annual Reviews. www.AnnualReviews.org
II.
doubly infected with two TMV strains, some progeny rods contained a mixture of both coat proteins (Otsuki and Takebe, 1976b). Antibodies specific for TMV strains were used to study the conditions under which phenotypically mixed rods of TMV could be formed in vivo and in vitro (Atabekova et al., 1975; Taliansky et al., 1977). Purcifull et al. (1973) used strain-specific antisera for several potyviruses to show that the protein found in the inclusion bodies induced by each strain was distinct, unrelated to the viral coat protein, and independent of host species in which the virus was grown. The site of initiation and direction of TMV assembly were elegantly confirmed by Otsuki et al. (1977) using strain-specific antibody.
C. Biological criteria 1. Symptoms a. Macroscopic symptoms As noted above, s y m p t o m differences are of prime importance in the recognition of mutant strains. However, the extent of differences in disease symptoms may be a quite unreliable measure of the degree of relatedness between different members of a group of strains. S y m p t o m s p r o d u c e d by different virus strains in the same species and variety of host plant may range from the symptomless 'carrier' state, through mosaic diseases of varying degrees of severity, to lethal necrotic disease. Figure 17.5 illustrates the range of systemic s y m p t o m types produced by four strains of TSV in tobacco. The strains are sufficiently closely related that experimental reassortment experiments are possible between them. The diseases produced by a given set of strains in one host plant may not be correlated at all with the kinds of disease produced in another host species. Most viruses, including many of widespread occurrences such as TMV, PVX, PVY, AMV and CMV, occur as numerous strains in nature. Many 'new' viruses have been described primarily based on symptoms and other biological properties, which have turned out later to be a strain of one of these corn-
C R I T E R I A FOR T t t E R E f : O G N I W I O N
OF S T R A I N S
757
monly occurring viruses. Some viruses appear to have given rise to relatively few strains as judged by symptoms, for example PLRV in potato varieties. A set of defined cultivars that give differential local lesion responses may provide a particularly useful and rapid method for delineating strains a m o n g field isolates of a virus. However, the important influence of environmental conditions on local lesion responses must be controlled. A virus causing severe disease is often said to be more 'virulent' than one causing mild disease. From what has been said in other sections, it should be apparent that the description can be applied only to a given strain of the virus inoculated into a particular variety of host plant in a specific manner and growing under particular environmental conditions. A named variety of host plant, especially a long-established one, may come to vary considerably in its reaction to a given strain of virus, due, for example, to the fact that seed merchants in different localities may make different selections for propagation. This may add a further complication to the identification of strains by means of s y m p t o m s produced on named cultivars. Nevertheless, a systematic study of symptoms produced on several host species or varieties under standard conditions may help considerably to delineate strains among large numbers of field isolates of a virus. b. Cytological effects
The cytological changes induced by different strains of a virus are often readily distinguished. Differences are of three kinds: (1) in the effects on cell organelles; (2) in the virusinduced structures within the cell; or (3) in the distribution or aggregation state of virus particles within the cell. Such differences may be of increasing importance in the delineation of viruses and virus strains. However, other factors may cause variation in the extent of differences between strains. For example, various strains in the stock culture of TYMV have markedly different effects on chloroplasts in cells of systemically infected leaves (see Fig. 9.18), but these differences may be much less
758
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
Fig. 17.5 Control of disease expression by the viral genome. Variation in chronic disease s y m p t o m type caused by four TSV isolates in tobacco. (a) The 'Standard' North American strain. Tobaccos became more or less symptomless. (b) A strain causing toothed margins on the leaves. (c) A strain in which tobaccos continue to show mosaic and necrotic symptoms. (d) A strain causing severe chronic stunting. These s y m p t o m types can be artificially re-assorted by making crosses between top, middle and bottom components of the various strains (see Fulton, 1975). (Courtesy of R.W. Fulton.)
From Matthews (1991). marked or non-existent in the infected cells of local lesions. Different strains of TuMV show differences in the morphology of their cylindrical inclusions (McDonald and Hiebert, 1975). Ultrastructural changes in both nucleus and cytoplasm of oat cells infected with BYDV strains differed between strains that were specific for a particular aphid vector and those that were not (Gill and Chong, 1979). Different strains of AMV m a y differ markedly in the way in which virus particles form aggregates within infected cells (e.g. Hull et al., 1970; Wilcoxson et at., 1975). The characteristic viroplasms found in cells infected with caulimoviruses (see Fig. 8.23) may vary with
different strains (Givord et al., 1984; Stratford et al., 1988). The variation may be associated with differences in gene II and the proportions of the products of genes II and IV. Mixed infections with two variants of BYDV in oats gave rise to altered patterns of effects in vascular tissue, including a predisposition for the xylem to become infected (Gill and Chong, 1981). 2. Host range and host plant genotype Host ranges of viruses generally are discussed in Chapter 3 (Section V). Many strains of a virus may have very similar host ranges, but others may differ considerably. Similar responses of a set of host plant genotypes to two viruses may provide good evi-
II.
dence that they are related strains (e.g. Schroeder and Provvidenti, 1971). On the other hand, a loss in ability to infect a particular host may be brought about by a single mutation. Dahl and Knight (1963) studied 12 mutants isolated from ToMV that had been treated with nitrous acid. One of these strains had lost the capacity to infect tomato. Strains of a virus that have different host ranges often produce different disease symptoms on some c o m m o n host. This is not always so. For example, four strains of TMV that were not clearly distinguishable by symptoms on N. tabacum or on c o m m o n varieties of Lycopersicon esculentum could be differentiated by their host ranges on a set of Lycopersicon hosts, including two varieties of L. esculentum and three selections of L. peruvianum (McRitchie and Alexander, 1963). Strains of PVX have been grouped according to their reactions to a range of host plant genotypes (Cockerham, 1970). 3. Methods of transmission Different arthropod vector species or different races of a single species may differ in their transmission of various strains of the same virus. Differences may be of the following kinds. 9 In the percentage of successful transmissions, for example MDMV strains by aphid species (Louie and Knoke, 1975). 9 In m i n i m u m acquisition time by the vector, for example MDMV in a p h i d vectors (Thongmeearkom et al., 1976). 9 In the length of the latent period, for example strains of PEMV in its aphid vector (Bath and Tsai, 1969). 9 In the time that the vectors remain infective (e.g. T h o n g m e e a r k o m et al., 1976). 9 Some strains may not be transmitted at all by particular vectors, for example strains of PYDV and leafhopper species (Black, 1941). Patterns of transmissibility by three aphid species have allowed large numbers of field variants of BYDV found in North America to be placed into five main groups (Rochow, 1979), which are now recognized as different species. The quite stable groupings have facilitated
C R I T E R I A FOR THE R E C O G N I T I O N OF S T R A I N S
759
studies on the distribution of virus variants both geographically and in successive seasons. If one strain of a virus is transmissible by mechanical means all others usually are too. However, there are reports of marked variation in mechanical transmissibility depending on both host clone and virus strain, for example AMV in alfalfa (Frosheiser, 1969). Defective strains may occur in which the RNA is not coated or is incompletely coated with protein. Such strains will not be mechanically transmissible except under conditions where they are protected from attack by nucleases. 4. Cross-protection The mechanism of cross-protection is discussed in Chapter 10 (Section V.D). Early observations on the interactions between virus strains led to the development of the concept of cross-protection. It was s h o w n by M c K i n n e y (1929) that tobacco plants infected with a green mosaic virus (TMV) developed no further symptoms w h e n inoculated with a yellow mosaic virus. Salaman (1933) found that tobaccos inoculated with a mild strain of PVX were i m m u n e from s u b s e q u e n t inoculation with severe strains of the virus, even if inoculated after only 5 days. They were not i m m u n e to infection with the unrelated viruses, TMV and PVY. This phenomenon, which has been variously called cross-protection, antagonism or interference, was soon found to occur very c o m m o n l y among related virus strains. It is most readily demonstrated w h e n the first strain inoculated causes a fairly mild systemic disease and the second strain causes severe s y m p t o m s or necrotic local lesions. Development of such lesions can be readily observed and a quantitative assessment can be made. Interference between related strains can also be demonstrated by mixing the two viruses in the same inoculum and inoculating to a host that gives distinctive lesions for one or both of the two viruses or strains. Interference by type TMV with the formation of yellow local lesions by another strain is shown in Fig. 10.22. For a time, cross-protection tests were given considerable weight in establishing whether two virus isolates were related strains or not,
760
17 VARIATION, EVOLUTION AND O R I G I N S OF PLANT VIRUSES
but subsequent d e v e l o p m e n t s have indicated the need for caution. A m o n g a group of strains that on other grounds are u n d o u b t e d l y related, some m a y give c o m p l e t e cross-protection, while with other combinations protection m a y be only partial. This is illustrated in Fig. 17.6 for strains of PVX in Datura. Some virus strains do not appear to crossprotect at all. Thus, none of the strains of BCTV protects against the others in water pimpernel Samolus parviflorus (Raf) (Bennett, 1955). Within a set of isolates that are u n d o u b t e d l y related strains, all possibilities may exist: reciprocal cross-protection of varying degrees of completeness, unilateral cross-protection and no cross-protection, as was found for strains of TSV in tobacco (Fulton, 1978). The other factor that m a y make cross-protection tests ambiguous is that there can be quite strong interference between some unrelated viruses (Bos, 1970). Most experiments on cross-protection have been carried out using mechanical transmission. Cross-protection m a y also occur in the plant with viruses transmitted in a persistent m a n n e r by insect vectors. Thus, Harrison (1958) found that infection with a mild strain of PLRV protected plants against infection with a severe
strain introduced by the aphid vector Myzus persicae (Sulz). Cross-protection also occurs in viroids (Chapter 14, Section I.D.7). 5. Productivity Different strains of a virus m a y vary widely in the a m o u n t of virus p r o d u c e d in a given host u n d e r standard conditions. For example, the c o m m o n strain of TMV was the most productive, and other naturally occurring strains varied over a range d o w n to about one-tenth that of c o m m o n TMV w h e n productivity was measured as the n u m b e r of local lesions produced from inocula m a d e from extracts of single local lesions produced in N. tabacum cultivar Xanthi nc (Veldee and Fraenkel-Conrat, 1962). Chemically i n d u c e d m u t a n t s also varied widely in productivity, and all were less productive than c o m m o n TMV. Some of these strains caused severe s y m p t o m s in certain hosts, but there was no correlation between severity of disease and productivity. Productivity appeared to be a genetically stable character as it remained fairly constant for a given m u t a n t w h e n tested after successive transfers. Chemical mutation quite frequently increased the severity of disease produced, but
Fig. 17.6 Cross-protection by strains of PVX in Datura tatula. {A) Healthy leaf inoculated with a strain giving necrotic local lesions. (B) Leaf previously systemically infected with a very mild strain of the virus, and showing complete protection against inoculation with the necrotic strain. {C} Leaf previously systemically infected with a mottling strain, and showing only partial protection. From Matthews (1949b), with permission.
II.
rarely if ever increased the productivity. From a type culture of TMV, B. Kassanis (quoted in Matthews, 1991) isolated strains causing slowly spreading bright yellow local lesions, usually without systemic spread, in White Burley tobacco (see Fig. 10.22). Virus content of these yellow lesions was extremely low. Such strains are difficult to maintain in the laboratory and would never survive in nature. 6. Specific infectivity Bawden and Pirie (1956) showed that infectivity per unit weight of purified type TMV was greater than that of a Datura strain when they were tested in N. glutinosa. There is some evidence suggesting that the protein coat of a virus may be involved in differences in specific infectivity at least between different viruses. Thus, Fraenkel-Conrat and Singer (1957) found that RMV had only about 5% of the specific infectivity of common TMV. However, when RMV RNA was reconstituted with common TMV protein, the specific infectivity was about four times higher than the RMV preparation that provided the RNA. Reconstituted TMV usually has a lower specific infectivity than the intact virus. The reason for the increase when the RMV RNA was coated with type TMV protein is not known, but might be due to the relative ease with which intact RMV and the RMV RNA reconstituted with type protein are uncoated in vivo. 7. Genome compatibility The possibility of carrying out viability tests with mixtures of components from different isolates of viruses with multi-partite genomes provides a functional biological test of relationship. Such tests were carried out with TRV strains by Sanger (1969). Only two of the 20 combinations he tested gave a functional interaction. Members of the Nepovirus genus show various degrees of compatibility in genetic reassortment experiments (Randles et al., 1977). Rao and Francki (1981) found that the RNAs 1, 2 and 3 of three strains of CMV were interchangeable in all combinations. However, with TAV, a distinct virus in the Cucumovirus genus, only RNA3 could be exchanged with those of the CMV strains. Similarly, only RNA3 could be
C;RITERIA FOR T t t E RE(;OC3NITION C3F S T R A I N S
761
successfully exchanged between two members of the B romovirus genus, BMV and CCMV (Allison et al., 1988). The incompatibility of RNAs 1 and 2 of these viruses is presumably due to the way in which their gene products interact (see Chapter 8, Section IV). Genome compatibility can be tested in a more direct fashion when the gene products can be isolated and their function is known. For example, Goldbach and Krijt (1982) showed that the protease coded for by CPMV did not process the primary translation products of other comoviruses. The transcriptase activities found in the particles of two rhabdoviruses (LNYV and BNYV) did not carry out transcription with the heterologous virus (Toriyama and Peters, 1981). 8. Activation of satellites Particular isolates of TNV will support the replication of some STNVs but not others (Uyemoto and Gilmer, 1972). Similarly, among the cucumoviruses and the small satellite RNAs found in association with them, some viruses support the replication of particular satellites while others do not (Chapter 14, Section II).
D. Discussion In considering use of the various possible criteria for the delineation of virus strains, we must bear in mind that, from a strictly genetic point of view, complete nucleotide sequence data would be sufficient to establish relationships between strains. Nevertheless, small changes in nucleotide sequence could have very different phenotypic effects. At one extreme a single base change in the coat protein gene could give rise to changes in several of the phenotypic properties noted earlier. On the other hand, several base changes might give rise to no phenotypic effects at all. For practical purposes, phenotypic characters such as host range, disease symptoms and insect vectors must usually be given some weight in delineating and grouping virus strains. One further consideration in the delimitation of virus strains is how to differentiate them from virus isolates. A common mistake is to name different virus isolates as strains when
762
~r V A R I A T I O N , E V O L U T I O N A N D O R I G I N S OF P L A N T V I R U S E S
there are no real differences between them. This is discussed further in Section III.
III. ISOLATION OF STRAINS The property of a new strain that first allows it to be distinguished from other known strains of a virus has been almost always biological m usually a difference in disease symptoms in some particular host. There are several ways in which new strains may be obtained. Where the virus is mechanically transmissible, it is usual to pass new isolates through several successive single local lesion cultures, if a suitable host is available. This is done to ensure as far as possible that a single strain is being dealt with, and that no unrelated contaminating virus is present in the culture. There is good evidence that a single virus particle or infection unit can give rise to a local lesion (see Chapter 12, Section II.E). On the other hand, there is ample evidence that new mutants soon appear. As discussed in Section I.A, a virus culture actually consists of a large range of variants with usually one being dominant. M a n y mutants arise even during the development of a single local lesion. For example, such mutants have been found in U1 TMV passaged through single lesions (Garcia-Arenal et al., 1984). The various chemical and physical m e t h o d s described in this chapter give useful information only because they are not sufficiently sensitive to detect the small proportion of any particular mutant or variant present in a culture.
A. Strains occurring naturally in particular hosts Different strains of a virus frequently occur in nature, either in particular host species or varieties or in particular locations. These can be cultured in appropriate host plants in the greenhouse.
B. Isolation from systemically infected plants Plants systemically infected with a virus frequently show atypical areas of tissue that contain strains differing from the major strain in
the culture. These areas of tissue may be different parts of a mosaic disease pattern (see Chapter 9, Section IV.D) or they may be merely small necrotic or yellow spots in systemically infected leaves, for example in tobacco plants infected with mild mottling strains of PVX. When such areas or spots are dissected out and inoculated to fresh plants, they may be shown to contain distinctive strains. On occasions, routine passage of a virus through a host either by mechanical inoculation or by vector transmission can result in symptom variants. An example of this is the separation of variants of RTBV during routine transmission in rice with the leafhopper vector, Nephotettix virescens (Cabauatan et al., 1995). Four s y m p t o m variants (named strains) were isolated and their genomes have subsequently been sequenced, revealing various nucleotide substitutions, insertions and deletions (Cabauatan et al., 1999). The preparation of protoplasts from systemically infected leaves, even w h e n these are showing apparently uniform symptoms, offers the possibility of a very fine 'dissection' of the leaves. Natural m u t a n t s m a y be revealed among such protoplasts either by regenerating plants from them (Shepard, 1975) or by inoculating test plants with virus from a single cell (Fraser and Matthews, 1979a).
C. Selection by particular hosts or conditions of growth A particular strain may multiply and move ahead of others in a certain plant. Such a host can be used to isolate the strain. Similarly, strains may differ in the rate at which they multiply and move at different temperatures in a given host. Holmes (1934) found that tomato and tobacco stems inoculated with a severe strain of TMV and incubated at 35~ subsequently contained mild strains, which were able to multiply readily at 35~ although the original severe strain was not. This result has since been amply confirmed by others. For example, Lebeurier and Hirth (1966) used serial passage in tobacco at successively higher temperatures to isolate a strain of TMV that grew effectively at 36~
III. ISOLATION OF STRAINS
Low temperatures have also been used to isolate strains (McGovern and Kuhn, 1984). It seems reasonable to suppose that selection of strains by particular hosts or conditions of growth may sometimes involve selection of a strain with a cell-to-cell movement protein that is better adapted to the host or conditions than the previously dominant strain. A small subpopulation of virus exists in a U1 TMV culture that can cause the hypersensitive reaction in Nicotiana sylvestris. This subpopulation moves upward rapidly and is selected for in the upper parts of the plant during rapid growth (bolting) of the shoot axis (Khan and Jones, 1989). The new concepts of host defense systems based on gene silencing and the ability of viruses to suppress this defense system (see Chapter 10, Section IV) raise issues of strains differing in suppression ability. These issues have yet to be considered. D . I s o l a t i o n by m e a n s of v e c t o r s Vectors may be used in three ways to isolate strains. First, by using short feeding periods on the plants to be infected, only one strain out of a mixture may be transmitted. This occurred with BCTV transmitted by a leafhopper vector following a 15-minute infection feeding (Thomas, 1970). Second, particular vectors may preferentially transmit certain strains of a virus (see Chapter 11, Section III.E.4). Third, inoculation of insect vectors with diluted inoculum followed by repeated selection of infected plants for a particular type of s y m p t o m can result in the isolation of a variant virus (Kimura et al., 1987). E. I s o l a t i o n of a r t i f i c i a l l y i n d u c e d mutants Experimentally induced mutants have been used for two important kinds of investigation in plant virology. Nitrous acid-induced mutations in the coat protein gene of TMV were of considerable importance in determining the nature of the genetic code and in confirming the nature of mutation. Nitrous acid has also been used to induce temperature-sensitive (ts) mutants, mainly in TMV, to aid in the delin-
763
eation of the in vivo functions carried by the viral genome. Nitrous acid mutants have been used as a source of mild virus strains to give disease control by cross-protection. There is a spontaneous background mutation rate giving the quasi-species described in Section I.A, but many of the natural mutants will be suppressed in inoculated leaves by a dominant strain or master sequence. Mutagens such as H N O 2 inactivate much of the treated virus. Thus, to show that a mutagen is increasing the mutation rate rather than selecting for a minor sequence, it is necessary for treated and untreated virus to be assayed at fairly high dilutions that give about the same number of total infections for treated and control samples (e.g. Melchers, 1968). 1. Coat protein and other mutants of TMV To isolate mutants, the reactions of different hosts containing the N or N N genes (see Chapter 10, Sections III.D and III.E) have been exploited. One convenient method for isolating mutants is to inoculate a necrotic local lesion host under conditions where most lesions will have arisen from infection with single virus particles. Sometimes mutants give a recognizably different necrotic local lesion, very frequently smaller than normal (Siegel, 1960). To detect other symptom differences, single lesions are dissected out and inoculated to hosts giving systemic symptoms. Mutants may then be recognized by the different symptoms that they produce. Another method for isolating mutants and estimating their frequency is to use a parental strain of virus in a host that gives systemic symptoms without necrotic or other conspicuous local lesions. One then looks for mutants producing necrotic, yellow or other characteristic local lesions in the host. This method selects one class of mutants, but it is sometimes difficult to eliminate the parental virus from the culture, even by repeated single local lesion culture (see Fig. 10.22). A third method for isolating mutants depends on diluting the inoculum to a point where only about one-half (or less) of the plants inoculated become infected. A host giving systemic infec, tion is used. Under these conditions, it can be assumed that most of the infected plants were
764
~7
V A R I A T I O N , E V O L U T I O N A N D O R I G I N S OF P L A N T VIRUSES
infected by a single virus particle or, for multicomponent viruses, a single infection unit. 2. ts Mutants
In isolating ts mutants following treatment of virus with nitrous acid, the objective is to obtain a range of independent mutants that are defective in one virus function at the nonpermissive temperature. The procedure used to select for such m u t a n t s should not select against mutants in any particular cistron. Early methods selected for ,s mutants with amino acid substitutions giving coat proteins that were defective at the non-permissive temperature (Jockusch, 1964; W i t t m a n n and Wittmann-Liebold, 1966). Dawson and Jones (1976) described a selection procedure for ts mutants of TMV based on the idea that all ts mutants should infect and begin replication more slowly at the nonpermissive temperature (32~ U1 TMV replicates in Xanthi nc leaves at 32~ but no necrotic lesions appear. On return to 25~ necrotic lesions develop rapidly. If an infection was due to a ,s mutant, development of necrosis on return to 25~ would be significantly delayed. Using this procedure, D a w s o n and Jones screened approximately 50 000 lesions formed by nitrous acid-treated TMV. They eventually obtained 25 ts mutants. The reversion to wild type by m a n y of these mutants is a sufficiently rare event that they can be used for biochemical experiments at 25~ (Jones and Dawson, 1978). In an attempt to obtain ts mutants of TRV with the mutation located in the long-rod RNA rather than the coat protein, Robinson (1973) added a large excess of untreated short rods to nitrous acid-treated virus. The mixture was inoculated to a local lesion host at 20~ Single lesions were isolated and inoculated to a local lesion host at 20~ and 30~ as a preliminary screen for mutants.
E Isolation of strains by molecular cloning For m a n y viruses, infectious genomes of D N A viruses and cDNAs of RNA viruses have been cloned. The process of cloning means that individual genome molecules of the quasi-species
population are separated and thus infection arises from just one component of that population. This is the ultimate way of separating strains. An example of this is the separation of two distinct s y m p t o m variants of MSV from plants infected with the Nigerian strain (MSVN) (Boulton et al., 1991a). One variant (MSV-Nm) gave narrow, mildly chlorotic, discontinuous streaks after agro-inoculation to maize whereas the other variant (MSV-Ns) gave wide, severely chlorotic streaks. Some molecular aspects of these variants are described in Chapter 10 (Section III.O.l.d).
IV. THE MOLECULAR BASIS OF VARIATION A. Mutation (nucleotide changes) 1. Chemical mutagens Mutations involving single nucleotides consist of the replacement of one base by another at a particular site, or the deletion or the addition of a nucleotide. Single base changes occurring in a coding region may lead to: (1) replacement of one amino acid by another in the protein product; (2) the introduction of a new stop codon that results in early termination of translation and a shorter polypeptide; or (3) replacement of a codon that has either greater or lesser usage in the particular host. Deletion or addition of a single nucleotide in a coding region will give rise to a frameshift, with consequent amino acid changes downstream of the deletion or addition. Such deletions or additions will usually be lethal unless compensated for by a second change (addition or deletion) that restores the original reading frame. Nucleotide changes in non-coding regions will vary in their effects depending on the regulatory or recognition functions of the sequence involved. Treatments that cause mutation also inactivate viruses. Quantitative studies on inactivation and on the appearance of necrotic local lesions in a culture normally giving chlorotic lesions fitted quite well with a theoretical curve based on the assumptions that a single chemical event can result in inactivation and that a single event can result in mutation.
IV. THE MOLECULAR BASIS OF VARIATION
Gierer and Mundry (1958) first demonstrated the high efficiency of nitrous acid as an in vitro mutagen for TMV. The mutagenic action was considered to be through deamination of cytosine to give uracil, and deamination of adenine to give hypoxanthine. The hypoxanthine acts like guanine during replication of the treated RNA and base-pairs with a cytosine. In the next copying event this will pair with a guanine and, thus, an adenine is replaced by a guanine at the original site. The amino acid exchanges found in the coat protein of mutants of TMV induced by nitrous acid confirmed these deaminations as the basis for the induced mutations. Such studies on TMV mutants made an important contribution to our understanding of the genetic code. They confirmed that the code is non-overlapping and degenerate, and gave strong support to the idea that the code is universal (Wittmann and Wittmann-Liebold, 1966; Sarkar, 1986). The single base changes brought about by nitrous acid have also been amply confirmed by later amino acid and nucleotide sequence data (e.g. Rees and Short, 1982; Knorr and Dawson, 1988). 5-Fluorouracil is an analog of uracil that is incorporated into the RNA of viruses that are replicating in plants supplied with the analog. It replaces uracil residues in the RNA and can lead to the changes uracil-* cytosine and adenine-* guanine (Wittmann and WittmannLeibold, 1966). Some other chemicals have less clearly defined mutagenic effects on RNA viruses. For the experimental induction of mutations, nitrous acid will usually be the most useful mutagen. 2. X-irradiation and ultraviolet irradiation Ionizing radiation and UV irradiation inactivate viruses containing dsDNA, and these agents also cause mutations in such viruses as well as in ssRNA viruses. For instance, temperature-sensitive mutants of AMV have been isolated following irradiation of purified Tb component of AMV with UV light (van VlotenDoting et al., 1980). 3. Raised temperature Several workers have noted that an increased number of variant strains could be isolated
765
when plants were grown at higher temperatures (e.g. Mundry, 1957). However, there is good evidence that the multiplication of and invasion by particular strains may be favored by certain temperatures. It seems probable that a major effect of heat treatment is to favor certain types of spontaneous mutant or a minor member of the quasi-species population. 4. Natural mutations There is no doubt that a single base change giving rise to a single amino acid substitution in the protein concerned is a frequent source of virus variability under natural conditions in vivo. Thus, of 16 spontaneous TMV mutants listed by Wittmann and Wittmann-Liebold (1966), six had one amino acid exchange in the coat protein, one had two, and one had three such exchanges. Many of the coat protein exchanges represented base changes different from those induced by nitrous acid. Eight of the mutants had normal coat proteins and, therefore, must have had one or more base changes in the RNA outside the coat protein cistron. Most of the 15 amino acid substitutions found between the coat proteins of two naturally occurring strains of AMV could be explained by single point mutations (Castel et al., 1979). Presumably, most of these mutants would have arisen from copying errors made by the RNA-dependent RNA polymerase during viral RNA replication. The differences between naturally occurring sequence variants of viroids usually consist mainly of a series of base substitutions (e.g. Visvader and Symons, 1985), although additions and deletions of nucleotides also occur. Natural mutation is discussed in more detail in Chapter 8 (Section IX.A).
B. Recombination For many years, it was considered that recombination was a genetic mechanism confined almost entirely to organisms with DNA as their genetic material. It is now known that recombination occurs in plant viruses with genomes consisting of either DNA or RNA. Recombination in both DNA and RNA viruses is discussed in detail in Chapter 8 (Section IX.B).
766
17 V A R I A T I O N , E V O L U T I O N A N D O R I G I N S OF P L A N T V I R U S E S
C. Deletions and additions
D. Nucleotide sequence re-arrangement
Recombinational events can lead to deletions in the genomes of both DNA and RNA viruses. These often give rise to defective or defective interfering nucleic acids, which are discussed in Chapter 8 (Section IX.C). In the literature, m a n y examples are described of mutant viruses with additional base sequences that have been generated by in vitro modification of recombinant D N A plasmids. Naturally occurring examples of such additions are much less common. In a group of naturally occurring variants, it may sometimes be difficult to establish whether a difference in length is due to an addition or a deletion of nucleotides. Kimura et al. (1987) repeatedly selected rice plants for severe s y m p t o m s after inoculation with RDV by leafhoppers injected with a dilute inoculum of the stock culture (strain 0). This procedure allowed the isolation of a severe strain (S). The fourth largest RNA of strain S had an M about 2 0 k D a greater than that of strain 0. The M r of the protein corresponding to the 43-kDa protein of strain 0 was 44 kDa in strain S. It has not been demonstrated that S was derived from 0 by an addition of nucleotides. Strain 0 could have been derived from S by a deletion event, with the parent strain maintained at a low level in the culture. A m o n g the brornoviruses, the CCMV RNA3 5' non-coding sequence contains a clearly demarcated 111-base insertion not present in BMV, which must represent a sequence rearrangement in one of the two viruses (Allison et al., 1989). Repetition of blocks of sequences is k n o w n in some RNA viruses. For example, sequences of 56 and 75 nucleotides are duplicated in the leaders of the RNAs 3 of strain S and strain L of AMV (Langereis et al., 1986). The duplications are next to one another in the leader sequences and do not appear to be essential for replication. Langereis et al. suggested that these duplications may have been generated by polymerase molecules releasing prematurely from a (-)-strand RNA3 template and re-initiating again on the same template with the nascent strand still attached.
An example of nucleotide re-arrangement has been described a m o n g satellite RNAs of TRSV (Buzayan et al., 1987). STRSV RNAs from strains 62L and NC-87 of TRSV have the same 360-residue sequence. The budblight satellite RNA with 359 residues differs from these mainly in the nucleotides from 100 to 140. The differences in sequences in this region are consistent with r e - a r r a n g e m e n t s of blocks of nucleotide residues, as indicated in Fig. 17.7.
E. Re-assortment of multi-particle genomes Since the classical experiments with the long and short rods of strains of TRV (Lister, 1966, 1968; Frost ei al., 1967), genetic assortment experiments have been carried out with most of the known multi-partite viruses. These experiments have demonstrated beyond doubt that new variants can arise by re-assortment of the pre-existing segments of the viral genome, both in the laboratory and in nature (Fulton, 1980). Successful re-assortment may not always be m u t u a l l y effective. For example, RaG and Hiruki (1987) found with two strains of RCNMV that a mixture of RNA1 of strain TpM34 plus RNA2 of strain TpM48 was infectious, while the reciprocal mixture was not. A l t h o u g h BMV and CCMV have been regarded as distinct bromoviruses, their individual RNA components will complement one another in certain combinations (Allison et al., 1988). Capped in vitro transcripts were made from complete cDNA copies of genomic RNAs G UUGU GCCU CG UGGAGGUGG GAUGC C ACCUCGU GGAGCAGCCUUC
/
..J_--
I
i~------7-- -2_ ~
I
GUAGGGGUC UGC U ACCUCGU UGGAGGUGG A GA UUGU A GCCU UC
Fig. 17.7 Possible nucleotide sequence re-arrangements between two satellite RNAs of TobRSV. Upper line: strain 62L, nucleotides 100-145. Lower line: budblight strain, nucleotides 100-143. Heavy underlines indicate blocks of sequences that may have been re-arranged. Connecting lines indicate identical sequences in the two strains. (Courtesy of G. Bruening.) From Matthews (1991).
V. CONSTRAINTS ON VARIATION
1, 2 and 3 of each virus. No viral replication was detected with any heterologous combination of RNAs 1 and 2, which code for proteins involved in RNA replication (discussed in Section II.C.7). By contrast, heterologous RNA3 was viable in both combinations, replicating in protoplasts and giving local lesions in Chenopodium. However, neither hybrid virus systemically infected the natural parental host plants.
F. The origin of strains in nature All the kinds of molecular change noted above will contribute to the evolution of strains in nature. A single base change resulting in a single amino acid change in a protein is probably one of the most common events giving rise to natural variation. The primary structure of the coat proteins of naturally occurring strains of TMV s u p p o r t s this view (Wittmann and Wittmann-Liebold, 1966). Viruses appear to vary quite widely in the rate at which they give rise to new strains. Large numbers of TMV strains are known, while only a few have been isolated for TBSV. Different strains of a virus may also vary quite markedly in the rates at which they give rise to mutants of a particular s y m p t o m type. Thus, some strains of PVX producing chlorotic local lesions in tobacco frequently gave rise to ring spot local lesion strains, and other strains gave none at all (Matthews, 1949a). Strains of TNV that produced white lesions on cowpea frequently gave rise to strains giving red lesions (Fulton, 1952). These red lesions always first appeared as spots resembling sectors in association with a white lesion. Various white strains produced red mutants at different rates over a 5-fold range. The reverse process (red strains giving rise to white) was not observed. These various differences in the rate of appearance of strains may not necessarily be due to differences in actual mutation rate. Some viruses or strains may produce a much higher proportion of defective or completely nonviable mutants than others. Some of the apparent differences probably reflect our ability to detect mutants. We can envisage strains diverging further
767
and further from the parent type as changes brought about by various mechanisms accumulate in the various proteins specified by the virus. The survival and spread of such strains will often depend on their competitive advantage within the host in which they happen to arise, and in others that they subsequently infect. The survival of new strains may not always d e p e n d on immediate selective advantage, although continued survival would of course require an adequate combination of properties. For example, Reddy and Black (1977) pointed out how deletion variants of WTV could by chance come to dominate in particular shoots of a growing clover plant. If the deletion event occurred in a cell near the apical meristem where the virus concentration is low, the mutant may come to replace the parent virus entirely in a particular shoot. They point out that the process can be regarded as an example of evolution by geographical isolation in miniature, with the branches of the plant providing the geographical isolation. The evolution of virus strains is further discussed in Section IX.
V. C O N S T R A I N T S O N V A R I A T I O N As discussed in Chapter 8 (Section IX.A), in most viral genomes there is no proofreading on replication; this leads to much of the variation described in previous sections in this chapter. In light of this high level of variation, there must be some constraints that control the preservation of the identity of a virus species or strain. It is likely that most mutations in a viral genome will be either neutral or deleterious. Mutations that lead to loss of a critical function w o u l d not be p r o p a g a t e d in a p o p u l a t i o n unless they were c o m p l e m e n t e d by other members of that population. However, mutations that caused only a decrease in 'fitness' would be more likely to be retained.
A. Muller's ratchet Theoretical considerations indicate that high mutation rates can have significant impact on populations, especially if they are small in size.
768
17 V A R I A T I O N , E V O L U T I O N A N D O R I G I N S OF PLANT VIRUSES
The concept of 'Muller's ratchet' is that, if the average mutation is deleterious, there will be a drift to decrease of population fitness leading to 'mutational meltdown' (see Muller, 1964; Lynch et al., 1993). Muller's ratchet is particularly effective in small populations, and for many viruses transmission and infection forms a bottleneck in which the population is small (Duarte et aI., 1993; Bergstrom et al., 1999). Back mutations at the specific site of a deleterious mutation or compensatory mutations are likely to occur at a lower rate than forward mutations (Haigh, 1978; Maynard Smith, 1978). In populations of higher organisms, this drift is limited by sex which re-creates, t h r o u g h genetic exchange, genomes with fewer or no mutations. Obviously, this process does not occur in the conventional sense in viruses but it is likely that recombination or genetic reassortment within the quasi-species could play a part in controlling the effects of Muller's ratchet (see Chao, 1997).
B. Does Muller's ratchet operate with plant viruses? In a study on the tobamoviruses that infected Nicotiana glauca in Australia over a 100-year period, Fraile et al. (1997a) found an example that could be interpreted as being due to Muller's ratchet. The tobamoviruses in herbarium samples and living samples of New South Wales N. glauca covering a period from 1899 to 1993 were analyzed. Before 1950, many plants were infected with both TMV and TMGMV but after that date only TMGMV was found. In experimental joint infections of N. glauca TMV accumulated to about 10% of the level of that in single inoculations; the level of TMGMV was not affected. Fraile et al. (1997a) concluded that TMV colonized N. glauca in N e w South Wales earlier or faster than TMGMV, but in joint infections the latter virus caused a decrease of the TMV population below the threshold at which deleterious mu.tations were eliminated. However, n e m a t o d e transmission of TRV serves as a bottleneck to clear the virus population of DI RNAs (Visser et al., 1999a). These DI RNAs, derived from RNA2 of TRV isolate PpK20 (see Fig. 6.38) have a modified coat pro-
tein gene and interfere with viral replication. It was suggested that TRV RNA2 and the DI RNA are encapsidated in cis by their encoded coat proteins which are, respectively, functional and non-functional in transmission, eliminating the DI RNA at the transmission bottleneck. This contrasts with heterologous encapsidation of potyvirus and luteovirus genomes (see Chapter 8, Section X) leading to isolates defective in vector transmission becoming transmissible. Similarly, aphid transmission often sorts the populations of genomic RNA variants and DRNAs present in CTV isolates (Albiach-Marti et al., 2000b).
VI. VIRUS STRAINS IN THE PLANT In the previous sections, we have considered ways of isolating virus strains, the molecular mechanisms by which they originate, and the criteria that can be used for distinguishing them. Here certain activities of strains in the infected plant are discussed.
A. Cross-protection It is generally accepted that protection of a plant by one strain of a virus against infection with a second depends on the presence of the protecting virus in the protected tissue. This is discussed in detail in Section II.C.4. If, for any reason, a plant or part of a plant becomes freed of the protecting virus, it is often susceptible to re-infection with the first strain or other strains. Exceptions may occur with systemic spread of the response to the host general defense system (see Chapter 10, Section IV.E).
B. Selective survival in specific hosts When a virus culture that has been maintained in an apparently stable state in one host species is transferred to another species and then inoculated back to the original host, it is sometimes found that the dominant strain in the culture has been changed. Carsner (1925) showed that a culture of BCTV could be altered by transmission to Chenopodium murale. W h e n the virus was
VI.
returned to beet from this plant, it produced only mild symptoms. Lackey (1932) found that the change was reversible and that the virus culture could be returned to its original condition, with respect to symptoms in beet, by passage through Stellaria media. According to Salaman (1938), a strain of PVX, which caused ring spot symptoms, when inoculated into seedling beet, produced small necrotic rings only. When virus from the local lesions was returned to tobacco, only a faint mottle developed. A reverse situation was described by Matthews (1949c). When PVX cultures giving a mild mottle in tobacco were passed through Cyphomandra betacea and re-inoculation was made to tobacco, only local and systemic ring spot-type disease was produced. Cyphomandra was the only one of 19 solanaceous species tested to cause this change. The mild cultures that were used contained a small proportion of ring spot strains that had presumably arisen by mutation. It was suggested that these ring spot strains multiplied more effectively in Cyphomandra than did the mild strains and that they thus came to dominate in this host. Ring spot strains alone reached several times the virus concentration in Cyphomandra c o m p a r e d with tobacco. Mild strains were four to eight times as concentrated in tobacco as the ring spot strains in this host. This is an example of the host selecting a different member of a quasi-species population to become the dominant member. Johnson (1947) found that passage of the ordinary severe-type TMV through sea holly (Eryngium aquaticum) resulted in mild symptoms on tobacco. He showed that severe strains moved more slowly in sea holly, thus accounting for the 'filtering' action of this host. Inoculation of a culture of CMV that did not cause systemic infection in Blackeye cowpea to the variety Catjang led to the appearance of some abnormally large local lesions. Inoculation from these large local lesions to the variety Blackeye was followed by systemic necrotic disease (Lakshman et al., 1985). These various selection p h e n o m e n a may well involve differences in the cell-to-cell movement protein coded for by strains of the virus or possibly in the interaction that the virus has with the host defense system. For
VIRUS S T R A I N S IN THE PLANT
769
example, the mild cultures of PVX in tobacco just described may have been giving rise continually to mutants that could not compete in tobacco with the parent strain based on their cell-to-cell m o v e m e n t proteins. However, the movement protein of some of these mutants may have been better adapted to systemic movement in Cyphomandra than the parental strain. The back mutation rate must be very low since the strain or strains selected in Cyphomandra appeared quite stable when cultured in tobacco. N o w that the nucleotide sequences of the genomic RNAs are k n o w n it should be possible to establish whether strain selection by particular hosts involves mutations in the m o v e m e n t protein. Satellite RNAs such as those associated with CMV m a y undergo differential replication in particular hosts. This may provide another basis for variation in s y m p t o m s following culture of a virus in a given species (Waterworth et al., 1978). For example, CARNA5 exists as a series of closely related sequences (Richards et al., 1978), which could provide for a rapid response to changed conditions for replication.
C. Loss of infectivity for one host following passage through another Loss of infectivity for one host may develop following repeated passage t h r o u g h another species. For example, several strains of PVX lost infectivity for potato during continued propagation in tobacco (Matthews, 1949a). No change in symptoms produced in tobacco, N. glutinosa or Datura tatuta could be observed over the period that the AP strain lost its infectivity for potato. The nature of the change is not understood, but possibly reflects gradual selection of a strain or strains better adapted to growth in tobacco. Alternatively, a minor member of the quasi-species in potato might have been able to multiply only in that host through complementation by the master sequence or other more major sequence. An immediate loss of infectivity for one host following passage through another can be due to a virus inhibitor that is effective only in certain hosts, rather than to any change in the virus itself.
770
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
D. Double infections in vivo The existence of p h e n o t y p i c mixing (see Chapter 8, Section X) suggests that two virus strains can replicate together in the same cell. The following evidence confirms this view: 9 Various observations have demonstrated the presence of two strains of a virus in the same cell in intact plants. Thus, Hull and Plaskitt (1970) could recognize characteristic aggregates of particles of two AMV isolates in the same cell. 9 When the ts mutant of TMV, Nil18, was inoculated in a mixture to tobacco with c o m m o n TMV and grown at 35~ some mutant RNA was found to be coated with common strain protein (Takebe, 1977). 9 Protoplasts inoculated with TMV show a multitarget response to inactivation by UV light (Takebe, 1977), indicating that more than one particle can initiate infection in the same cell. Thus, while proof is lacking, the weight of evidence suggests that two strains can infect and replicate simultaneously in the same cell in the intact plant. The quasi-species concept (Section I.A) with a master sequence and a multitude of minor ones suggests that this is an unusual situation. However, on some occasions it is likely that there are two or more master sequences, which would reflect the situation with co-infection by two strains.
E. Selective multiplication under different environmental conditions Temperature is probably the environmental factor that most commonly influences the survival or predominance of strains occurring in nature. Experimental use of temperature to isolate strains is noted in Sections III.C and III.E.2. In parts of the world with hot climates, strains of viruses surviving at these temperatures are selected naturally. Some understanding of the basis for the effects of temperature on naturally occurring strains can be gained from the results of the studies on artificially induced ts mutants (see Section III.E.2).
VII. CORRELATIONS BETWEEN CRITERIA FOR CHARACTERIZING VIRUSES A N D VIRUS STRAINS In the preceding sections, we have surveyed the various criteria that can be used to delineate variation among virus isolates. H o w can we use these criteria to decide whether a particular isolate is identical to another isolate or a related variant or strain, or whether it is a distinct virus? This is a question of considerable practical importance, because the recognition and identification of virus strains may be most important for effective virus control. In addition, the virological literature is cluttered with inadequate descriptions of virus isolates. These are frequently described as new viruses or new strains, particularly if they are found in a new host or a different country, when adequate study might show they were very probably identical to some virus already described. The definition of a virus species is given in Chapter 2 (Section I.C).
A. Criteria for identity There is only one criterion that will establish that two virus isolates are identical: the identity of the complete base sequence of their genome nucleic acids. In spite of recent rapid advances in nucleic acid sequencing techniques, it is usually impractical for this to become a routine procedure. For most practical purposes, the following criteria would be sufficient to establish provisional identity of two virus isolates: (1) identity of size, shape and any substructure of the virus particle as revealed by appropriate electron microscopy; (2) serological identity in adequate tests; (3) identical disease symptoms and host ranges for a set of indicator hosts and genotypes; and (4) identical transmission, especially with respect to any arthropod, nematode or fungal vectors. The presumption of identity would be greatly reinforced by information on some aspect of nucleic acid sequence, for example identical sequences in a particular region of the genome or identity as judged by heterogeneity mapping.
Vll.
C O R R E L A T I O N S BETWEEN C R I T E R I A FOR C H A R A C T E R I Z I N G V I R U S E S A N D VIRUS S T R A I N S
B. Strains and viruses The broad questions of virus classification are dealt with in Chapter 2. Here we will consider the problems involved in using the various properties outlined earlier in this chapter to define virus strains, to group them and to decide whether an isolate is a strain or a different virus. One method is to take a quantitatively determined set of characteristics such as the amino acid composition of the coat protein. Statistical procedures and computer analysis are then used to derive a classification with degrees of relationship indicated. Computer analysis is particularly useful for handling large amounts of numerical data as was used, for example, to derive Fig. 2.3 from amino acid sequences. However, a classification based on computer analysis is no more objective than other ways of making a classification. It will depend on the personal judgments and selections made by the taxonomist providing the data. In the adansonian approach, all the k n o w n characters are given weight in determining groupings. This approach has become popular with the widespread availability of computers, but there are significant limitations. For example, as noted earlier, many of the fairly easily measured characters of a small virus depend on properties of the coat protein. Thus, differences in the coat protein may be given undue weight. Similarly, s y m p t o m differences between two strains could be emphasized, merely by recording differences on an extended host range. On the other hand, the hierarchical system involves m a k i n g arbitrary decisions about which characters are the most important. There are serious objections to applying such a system, w i t h o u t some modification, especially when we are considering classification within a group of related viruses. The most useful characters will be different within different virus groups. Thus, the coat protein of STNV, being the only gene product of this virus, should be given more substantial weight than would the coat protein of a virus with, say, 10 genes. Similarly, particle morphology may be most useful for those groups such as the Rhabdoviridae that possess a complex structure.
771
The best course is probably the pragmatic one of considering all k n o w n properties within a group of variants and weighting them in a commonsense manner in relation to the overall properties of the group in question. When strains arise in the stock culture of a virus in the laboratory, as they do with such viruses as TMV and TYMV, we can be reasonably sure that they will be closely related to the parental strainmusually arising from a single mutation. Phenotypic differences in most properties will usually be small, but may sometimes be large as with the TMV strains, such as PM1, that produce defective coat proteins and no intact virus. Virus isolates collected in the field, perhaps from different host species in different countries, may appear to be related on the basis of some properties and unrelated on others. The only generalization that can be made at present is that closely related strains will differ in only a few properties, while distantly related strains will differ in many. The extent to which different properties show correlations varies widely in the different groups of viruses.
C. Correlations for various criteria From a purely genetic point of view, the relationships between a set of virus strains can be assessed precisely if we know the differences in nucleotide sequences between their genomes. However, from the virological point of view, other factors must be taken into consideration. For example: (1) nucleotide changes that are silent, that is, lead to no change in the structure or function of the virus, are usually of little interest. However, it must be remembered that, with accumulating knowledge, what are considered as non-functional nucleotides at one time may in the future be highly significant. (2) Particular gene functions may be of particular ecological and therefore practical significance, for example mutations in a viral gene that affects insect vector specificity. (3) When large numbers of field isolates have to be typed over a short time interval, only rapid diagnostic methods are practicable. The confidence with which particular criteria can be used depends in part on the extent to which they correlate
772
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
with other criteria. This section gives a brief overview of these problems. 1. Host responses Where a group of strains are fairly closely related, host responses may provide the best, or even only practicable, criteria for establishing strain types. For example, Mosch et al. (1973) found that 18 isolates of TMV from glasshouse tomato crops could be placed in three groups depending on their pathogenicity for a set of Lycopersicon esculentum clones. There were no differences in certain physical properties (buoyant densities and S20,w) and only small individual differences in coat protein composition. These did not correlate with the pathogenicity groups. When a virus of economic importance such as AMV is highly variable, the classification of large numbers of field isolates must usually depend primarily on symptoms and host range on a standard set of indicators (e.g. Crill et al., 1971; Hajimorad and Francki,
988). 2. Vector transmission Among three isolates of BYDV there was a correlation between closeness of serological relationship and transmission by aphid vectors (Aapola and Rochow, 1971). Pead and Torrance (1989) found that MAbs could be used to type the three major vector-specific strain groups (now species) of BYDV. On the other hand, an isolate of PLRV that was poorly transmitted by aphids was indistinguishable serologically from readily transmitted isolates (Tamada et al., 1984). There was no correlation between serological relatedness and the ability of English populations of Longidorus attenuatus to transmit different isolates of TBRV (Brown et al., 1989b). 3. Multi-partite genomes The ability of multi-particle viruses to complement one another successfully provides a powerful functional, criterion indicating relationship. However, this property may not correlate closely with the physical properties of the virus particle or other properties of the virus. For example, certain viruses that have been considered as strains of CPMV (Swaans and
van Kammen, 1973) did not successfully complement one another in mixed infection experiments (van Kammen, 1968). Successful complementation has been shown to occur not only between already well-recognized strains but also between viruses thought to be distinct members of the same group. Such results further complicate the use of complementation tests as a criterion of relationship. For example, Bancroft (1972) demonstrated successful complementation between BMV and CCMV. These are both in the Bromovirus genus, although they have almost totally different host ranges and appear unrelated serologically. At present two genera of viruses are known with a tripartite genome and a separately encapsulated coat protein cistron: the Ilarvirus and Alfamovirus genera. With members of both these groups, if the three-genome RNAs are used for infection, the coat protein RNA, or some coat protein itself, is required for infectivity (discussed in Chapter 8, Section IV.G). The coat protein or the coat protein RNA of some ilarviruses, for example TSV, will activate the RNAs of AMV. The reverse combination is also active. However, mixtures of the three-genome RNAs from the two viruses do not complement one another (van Vloten-Doting, 1975; Gonsalves and Fulton, 1977), and there is no sequence similarity between the corresponding RNA segments. Transgenic tobacco plants expressing the TSV coat protein gene were resistant to infection with TSV but susceptible to AMV. They could be infected with AMV RNAs 1, 2 and 3, demonstrating that the endogenously produced TSV coat protein can activate the AMV genome, even though it does not protect against this virus (van Dun et al., 1988b). The coat protein of AMV nucleoproteins is specifically removed by the addition of AMV RNA. Similarly, the nucleoproteins of ilarviruses may lose their protein when free viral RNA is added. There is reciprocity in this reaction between certain ilarviruses and AMV (van Vloten-Doting, 1975; Gonsalves and Fulton, 1977). These results have led some workers to suggest that AMV should be placed in the Ilarvirus genus (see Chapter 2, Section III.J.4).
VII.
(X~RREI.ATIONS BETWEEN C'RITERIA FOR (.'HARACTERIZ1N(} VIRUSES AN[) VIRUS S T R A I N S
4. General nucleotide sequence similarities Using hybridization techniques, there may be complete lack of detectable base sequence homology between viruses that on other grounds, such as morphology of the particle and serology, are certainly related (Zaitlin et aI., 1977). At the other extreme, Bol et al. (1975) described four strains of AMV with well-characterized differences in biological tests that were virtually indistinguishable in nucleic acid hybridization tests. Strains of CMV are divided into two subgroups, based on serology and nucleic acid hybridization, as discussed by Rizzo and Palukaitis (1988). Of 39 strains examined by nucleic acid hybridization, 30 belong to subgroup I and nine to subgroup II. RNAs belonging to the two subgroups can be re-assorted to yield viable recombinants. The RNAs I and 2 of representatives of the two groups have been sequenced and compared (Rizzo and Palukaitis, 1988, 1989). Different regions of the RNAs varied in the extent of sequence homology (from 62% to 81 To). Strains within the two subgroups cannot be distinguished by the usual nucleic acid hybridization techniques. However, Owen and Palukaitis (1988) used molecular heterogeneity mapping to place 13 of the CMV strains into two groups based on their ability to hybridize to two representative strains. Molecular heterogeneity mapping could distinguish strains within the two groups. 5. 3' Non-coding nucleotide sequences Another approach for discriminating between distinct potyviruses and strains has been explored by Frenkel et al. (1989). They compared the 3' non-coding nucleotide sequences of 13 potyviruses and found that viruses that were distinct on other grounds had 3' noncoding sequences of different lengths (189-475 nucleotides). The degree of sequence similarity ranged from 39% to 53%. Such values are comparable to that obtained when the 3'-untranslated regions from unrelated potyviruses are compared, and they are probably in the range expected for chance matching. By contrast, the 3'-untranslated regions of sets of viruses recognized on other criteria as related strains were very similar in length and in nucleotide
773
sequence homology (83-99%). WMV-2 and SGMV-N were found to have 78% homology and on this basis were considered to be strains of the same virus. 6. Serological relationships Relationships determined by serological methods might, by chance, correlate quite well with any other properties. However, it is reasonable to expect that they may show some correlation with those criteria that also depend on some property of the coat protein. Correlations have been reported between degree of relatedness, measured by crossprotection tests, and serological relatedness (e.g. PVX strains--Matthews, 1949b; BYDV isolatesmAapola and Rochow, 1971). On the other hand, there was no correlation between serological relatedness within a group of TNV isolates and their ability to support the replication of three differing isolates of STNV (Kassanis and Phillips, 1970), nor was there any correlation between serological relatedness and symptoms in tobacco for TRSV (Gooding, 1970). MAbs raised against strains of PVX reacted in a complex manner with the strains in different groups based on the reaction of host varieties (Torrance et al., 1986a). Nevertheless, a resistance-breaking strain could be identified. A panel of 10 MAbs raised against PLRV failed to differentiate between strains that caused different symptoms in indicator hosts (Massalski and Harrison, 1987). Traditionally, viruses and strains within the Potyvirus genus have been very difficult to delineate. This has been not only because of the large number of viruses involved but because different tests for relationship gave different answers. The work of Shukla and colleagues has gone some way toward establishing a sound basis for classification of virus isolates belonging to this group. For example, earlier work suggested that cross-protection did not distinguish some isolates that, on other criteria such as serology, were considered to be separate viruses. Shukla et al. (1989b), using antibodies directed against the N-terminal part of the coat protein, showed that potyviruses infecting maize, sorghum and sugarcane in Australia and the United States comprised four
774
17 VARIATION, EVOLUTION AND O R I G I N S OF PLANT VIRUSES
distinct viruses. The earlier cross-protection tests fell neatly into place on this basis (Shukla and Ward, 1989a). Similarly, the kind of cytoplasmic inclusions found with these isolates supported the idea of four distinct viruses. Difficulties remain, however, because some unexpected serological cross-reactions occur between viruses that on other soundly based criteria are regarded as distant members of the Potyvirus genus. The antigenic site for these cross-reactions may reside with a few common amino acids close to the N-terminus of the coat protein (Shukla and Ward, 1989a,b). 7. Non-structural proteins Yeh and Gonsalves (1984) used antisera raised against the inclusion body proteins of two potyviruses to confirm that they were related strains of one virus rather than two distinct viruses. Thornbury and Pirone (1983) showed that the helper component protein of two different potyviruses were serologically distinct. There was no serological relationship between the 35-kDa protein coded for by AMV and the corresponding proteins of three other viruses with a tripartite genome (van Tol and van Vloten-Doting, 1981).
VIII. DISCUSSION A N D SUMMARY The study of variability is one of the most important aspects of plant virology. It is important from the practical point of view because strains vary in the severity of disease they cause in the field, and because strains can mutate to break crop plant resistance to a virus. It is important also for developing our understanding of how viruses have evolved in the past, and how they are evolving at present. A range of procedures is available for isolating virus variants either from nature or following some form of mutagenesis or other manipulation outside the plant. Mutants of the ts type have been particularly useful in studying various aspects ofvirus structure and replication. Because of the very high mutation rate, it is probable that all cultures of plant viruses con-
sist of a mixture of numerous strains even after single lesion passage. However, a master genome sequence will usually dominate in the culture and m a n y variants will not be detected. The selection of the master sequence will depend on m a n y factors, including the host genotype and the environment in which it is growing. The molecular mechanisms by which variation within a virus population is produced are like m a n y of those found in cellular organisms, except that for m a n y plant virus groups the material upon which variation operates is RNA rather than DNA. Mechanisms include mutations involving single nucleotide changes or the addition or deletion of one or a few nucleotides, recombination, deletions or additions of blocks of nucleotide sequences, rea r r a n g e m e n t of nucleotide sequences, and re-assortment among multi-partite genomes. A range of structural, serological and biological criteria is available for delineating viruses and virus strains within a group or family of viruses. The kind of criteria to be used will depend on the purpose of the study. If we are studying evolutionary relationships within a virus group or family, or among the variants of a single virus, then the full nucleotide sequences of the viruses concerned will be of prime importance, but knowledge of the functional products of the viral genome will often be needed as well. If the full nucleotide sequences are known for representative viruses, then other methods, such as various forms of nucleic acid hybridization or PCR, can be usefully interpreted for additional viruses and strains. If we are interested in developing methods for reliably and rapidly diagnosing viruses and virus strains from the field, then other methods will be appropriate. Dot blot serological assays using some form of ELISA are an important type of test. Polyclonal antisera of wide specificity or MAbs of very narrow specificity can be used in such tests as appropriate. Biological criteria such as disease symptoms, host range, methods of transmission and cross-protection may be important in defining viruses and virus strains. The extent to which virus species have been clearly delineated varies widely among the different genera and families of viruses. There are
IX. SPECULATIONS ON ORI(~INS AND EVOLUTION
dangers in formalizing virus species or virus groups before a sufficient number and diversity of strains have been investigated. For example, at a stage when only about seven tymoviruses were known, two subgroups were suggested on the basis of serological relationships and RNA base composition (Gibbs, 1969; Harrison et al., 1971). Since then, further tymoviruses have been discovered with i n t e r m e d i a t e characteristics (Koenig and Givord, 1974). For some groups, such as the potyviruses, 'a common set or pattern of correlating stable properties' has emerged that can allow the grouping of virus strains into species with some degree of confidence. The relative importance, or weight, to be placed on different properties of a virus for purposes of classification remains a difficult problem. An adequate understanding of the significance to be placed on the various properties may come only w h e n we have a detailed knowledge of the structure of the viral genome, the polypeptides it codes for and their functions, and the regulatory or other roles of any translated or u n t r a n s l a t e d regions in the genome. Even with such knowledge difficulties will remain. For example: 9 Disease induction, which is a complex process, has been shown for some viruses to depend on the functions of two or more viral genes. 9 Various possible m e c h a n i s m s are n o w k n o w n whereby a single mutation could have effects on two or more functions. 9 A single gene product may have two or more functions, differing in importance, for the virus infection cycle. Thus from a practical point of view it may be an oversimplification to establish relationships between viruses and strains within a family or group solely on the basis of nucleotide sequences.
IX. SPECULATIONS ON ORIGINS AND EVOLUTION The most fundamental single property of an organism is the size of its genome. In this respect, viruses infecting all kinds of organisms
775
span a range of almost three orders of magnitude (see Fig. 1.4). The genome of the smallest virus, STNV, consists of a monocistronic m R N A coding for the coat protein of the virus shell. The largest viruses (infecting animals) have genomes about as large as those of the smallest cells. Where and w h e n did this great diversity of agents arise? W h e n and h o w did they evolve? Although much relevant information has become available from studies on the structure and replication of viruses and on the molecular biology of normal cells, the origin and evolution of viruses is only now beginning to emerge from the realm of speculation. Nevertheless, the topic is one of general interest, and one that will be relevant to the problem of classification. For a meaningful discussion of these topics, we must also consider examples from other groups of viruses besides those infecting plants. In discussing evolution of viruses, we must recognize that it is distinct from the evolution of virus diseases. A new disease may be the consequence of the 'evolution' of the causal virus, but can also result from no change in the virus (Nathanson et al., 1995). For instance, a new disease can result from the m o v e m e n t of a n ' o l d ' virus into a new situation. It is likely that the epidemics of swollen shoot in cacao in west Africa and of tungro in rice in south-east Asia were due to the spread of the viruses from asymptomatic natural hosts into either a new species in that area or a changed agronomic situation. When considering virus origins, it must be remembered that there are three basic types of plant viral genome: those that replicate RNA --* RNA, those that replicate D N A --, RNA --* D N A (reverse transcribing) and those that replicate D N A -~ DNA. It is likely that these three g e n o m e types have different e v o l u t i o n a r y pathways but it is not k n o w n whether they originate from the same type of macromolecule. Similarities in polymerase structure (see Chapter 8, Section IV.B) suggest a possible c o m m o n basic origin, although they could be the result of convergent evolution. Thus, there is no compelling reason to suppose that all viruses arose in the same way. Furthermore, it
776
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
is possible that viruses that originated in one major group of organisms may now exist primarily or solely as agents infecting another group. The origin and evolution of plant viruses is reviewed in Gibbs (1999b) and that of viruses in general by Domingo et al. (1999b).
X. TYPES OF E V O L U T I O N A. Microevolution and macroevolution In earlier sections of this chapter, I discussed the variation of plant viruses and its molecular basis. This variation gives the material on which selection pressures can act which results in virus evolution. The different forms of variation have different importance in the level or type of evolution (Fig. 17.8). Thus, strains are differentiated mainly by mutations and small insertions or deletions that are selected, changing the master sequence in the quasi-species cloud. This can be termed microevolution. Larger and more radical changes caused by recombination a n d / o r acquisition of new genes, termed macroevolution, lead to the generation of new genera or families. This essentially starts a new quasi-species cloud that
Evolution
Mutation
Recombination Re-assortment (pseudo-recombination de novo gene acquisition Family
Genus ('group')
Species
Fig. 17.8 The apparent relative importance of some sources of genetic n o v e l t y that influence virus evolution. From Gibbs and Keese (1995), with kind permission of the copyright holder, 9 The Cambridge University Press.
is selected u p o n in further microevolutionary diversification. Because of the population structure of a virus isolate, microevolution can be a continuous process. However, as discussed by Holland and Domingo (1998), extremely high mutation rates do not necessitate rapid evolution. As described below (Section X.D), there is evidence that some plant virus genera appear to be diversifying rapidly at present whereas others appear to be more stable. It is likely that, as with the evolution of higher organisms, viruses go through a stage of relatively rapid diversification w h e n presented with a changing environment and then enter a relatively stable phase where they have adapted to the new environment(s). The selection pressures are discussed in Section XIII. Macroevolution is a much rarer and a stepwise process. The great majority of recombination events between viral sequences or those leading to the acquisition of new genes will be lethal or deleterious to the virus. However, it is the very rare event that leads to the formation of a viral genome with new properties that enable it to be more successful than its progenitors. Examples of this will be discussed later in this section. Microevolution and macroevolution are not necessarily independent evolutionary systems. It is quite possible that microevolutionary changes of one virus could lead to the formation of 'hotspots' for recombination with sequences in other viruses. Similarly, not all recombination events will lead to major changes. Recombination between near homologous sequences will create new sequence combinations that have only minor differences to the parental sequences. Thus, microevolution and macroevolution are all a matter of degree.
B. Sequence divergence or convergence Sequence similarity between two genes does not necessarily indicate evolutionary relationship (homology). Without other evidence, it may be impossible to establish w h e t h e r sequence similarity between two genes is due to a c o m m o n evolutionary origin or to convergence. The lysozyme enzymes of foregut
x. TYpEs OF EVOLUTION
fermenters constitute a good example of sequence convergence (Stewart and Wilson, 1987). Foregut fermentation has evolved independently twice in placental mammals, first in ruminants and later in colobine monkeys. In both instances, lysozyme C was recruited to function in the true stomach. About half the amino acid replacements along the langur lysozyme lineage were in parallel or were convergent with those that had evolved earlier along the cow stomach lysozyme lineage. This convergence was driven by selection for adaptation to the acidic pepsin-containing environment of the stomach. Similar convergence has probably occurred from time to time during the evolution of viral genes. However, with such genes it is difficult to obtain independent evidence demonstrating convergence, as has been possible with the stomach lysozymes. Amino acid sequence similarities that are sufficient to lead to a serological cross-reaction may sometimes arise by chance, as appears to have happened between the TMV coat protein and the large subunit of ribulose bisphosphate carboxylase (Dietzgen and Zaitlin, 1986). Zanotto et al. (1996) argue for the convergent evolution of R N A - d e p e n d e n t RNA polymerases. However, this is debated by others (see Koonin and Dolja, 1993; Gorbalenya, 1995), who suggest that the evolution of this gene is divergent.
777
the modules had to co-evolve within a single genomic unit. The essentials of modular evolution include: 1. The product of evolution is a favorable combination of modules selected to work optimally individually and together to fill a particular niche. 2. Joint infection of the host by two or more viruses is essential for the assembly of new combinations of modules. The viruses do not necessarily have to give full systemic infection of the host; they just have to replicate in the same cell. This can lead to changes in virus host ranges. 3. Viruses in the same 'interbreeding' population can differ widely in any characteristic as these are aspects of the function of individual modules. 4. Evolution acts primarily at the level of the individual module and not at the level of the intact virus. Selection upon modules is for a good execution of function, retention of the appropriate regulatory sequences and functional compatibility with most, if not all, other modules in that genome. There is increasing evidence for a modular mechanism of macroevolution of plant viruses, which will be illustrated below. The question of where the modules came from is discussed in Section XI.
D. Evidence for virus evolution C. Modular evolution A process of modular evolution was proposed by Botstein (1980) for DNA bacteriophage but is now considered to also apply to RNA viruses. It is suggested that viruses have evolved by recombinational re-arrangements or re-assortments of interchangeable elements or modules. Modules are defined as interchangeable genetic elements, each of which carries out a particular biological function; examples are replicase proteins, capsid proteins and regulatory systems. This interchange enables independent evolution of the modules under a wide variety of selective conditions. Such modular mobility can overcome the evolutionary constraints that would occur if all
No fossil viruses have yet been discovered. Some may await discovery, for example in insects preserved in amber or PCR of longpreserved material. In the meantime, evidence for virus evolution must come from the study of present-day viruses in present-day hosts. Most of the current ideas and concepts about viral evolution are based on molecular data ('molecular fossils') that are accumulating at a considerable rate. It is the comparisons of genome organizations and the details of sequences that are giving evidence for the evolutionary pathways of many of the plant viruses. Below I describe how the data are analyzed and the evidence for six virus groups, which illustrates some current ideas on plant virus evolution.
778
iv
VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
1. Phylogenetic analyzes There is a variety of methods used for inferring phylogenetic or other relationships of sequences (reviewed by Weiller et at., 1995). The main aim of these methods is to analyze the nucleic acid or amino acid sequence data in such a way as to reveal relationships. Thus, important steps are in the recording and presentation of the data for analysis and in the interpretation of the analysis. The theory, methods and practice of analysing molecular sequence information is the subject of several monographs (e.g. Waterman, 1988; Doolittle, 1990; Gribskov and Devereux, 1991).
1887 1750 -
/.,
1500 1250 CoYMV P216
/
1000 -
/
The simplest m e t h o d for c o m p a r i n g two sequences to determine whether there are regions of sequence similarity is the dot plot (dot diagram) (Gibbs and McIntyre, 1971). In this, the two sequences are placed at right angles to form the adjacent areas of a rectangular matrix and a dot is placed in the matrix wherever a row and column with the same sequence element intersect (Fig. 17.9). In practice, overlapping w i n d o w s of a pre-determined size of sequence are compared and the value above which the similarity is recorded as a dot is selected beforehand to reduce the 'background' dots. Sequence similarities appear as a diagonal run of dots that can easily be seen by the eye. Deviations in this dot matrix reveal features of the two sequences being compared. For instance, mutations give gaps in the diagonal run, insertions or deletions cause the run to change from one diagonal to another, and sequence repetitions give parallel diagonal runs of dots. Sequences can also be compared by using various alignment c o m p u t e r programs. These attempt to give the o p t i m u m sequence alignment by putting blanks in one or other of the sequences. This is best used for pairs of sequences that a dot plot has shown to be related. There are two types of method for aligning sequences. Global methods attempt to find an 'optimal' alignment throughout the length of the sequence, and local methods only attempt to identify short regions of similarity and do not
|
..
750 -
.
5O0 ,.
250
@ :
a. Comparison of sequences
,.,,
/
,
I .....
250
,
500
.
7;0
.
.
1000
.
.
12;0
.
9
-
1500 1676
RTBV 194
Fig. 17.9 (A) Comparison of RTBV P194 and CoYMV, P216 by DIAGON. Three major regions of homology were identified as indicated. (B) Comparison of the spatial distribution of various functional domains in
RTBV P194 and CoYMV P216. RB, RNA-binding domain; PR, proteinase domain; RT, reverse transcriptase domain; RH, RNase H domain; aa, amino acids. Also shown are the regions of homology identified in (A). From Hay et al. (1991), with kind permission of the copyright holder, 9 The Oxford University Press.
attempt to align the sequences between these regions. There is a wide variety of sequence alignment programs (see McClure et al., 1993) and a useful one color-codes amino acids according to their properties. This enables the manual alignment of amino acids on their properties (e.g. basic or acidic or polar or non-polar) without the need to align specific residues. From multiple alignments of amino acid sequences, sequence motifs that are characteristic of specific proteins such as enzymes can be
X. TYPES OF EVOLUTION
identified. Various motifs found in viral proteins are described in previous chapters (e.g. those involved in RNA replication; Chapter 8, Section IV.B). Identification of sequence motifs and comparison of the sequence(s) with databases can enable the function of the protein to be identified and the closeness of relationship to known proteins to be determined. Although such searches up to now have not revealed functions for viral proteins in many cases, the rapid increase in accessions to databases, especially of plant genome sequences, coupled with advanced computer programs that enable reiterative searches, will reveal more relationships.
b. Reconstructing phylogeny Multiple alignments of sequences a n d / o r lists of numbers giving details of sequence similarities do not enable the easy interpretation of evolutionary relationships or pathways. There are two groups of methods for reconstructing phylogenies. The phenetic methods use phenotypic data in attempts to reconstruct phylogeny without completely understanding the evolutionary pathway. The cladistic methods concentrate on the evolutionary pathway by attempting to predict the ancestors. Initially, it would appear that the cladistic approach is superior but the criteria used to predict ancestors may be invalid and lead to wrong conclusions. In practice, one has to use a pragmatic approach, and methods that involve a combination of phenetic and cladistic approaches can be equally useful. Phylogenetic relationships are often presented as a structure resembling a tree or dendrogram made up of various parts termed root, stem, branch, node and leaf. There are various styles of dendrograms (see Fig. 17.10) and various methods for their construction. These are discussed in detail in Weiller et al. (1995) and their advantages and disadvantages are beyond the scope of this book. However, in determining evolutionary relationships, it is important to recognize the limitations of the various types of dendrogram; it is also important to test the statistical significance of the branching within the tree. These statistical methods are discussed by Weiller et al. (1995), who comment that no one method of tree con-
779
struction is superior to others and advise the use of more than one method for each data set and then comparison of the results. 2. Bromoviridae
Members of the B romoviridae have genomes that are divided between three (+)-strand ssRNA segments (see Chapter 6, Section VIII.A). There is evidence for both genome reassortment and recombination in the evolution of members of this family. A phylogenetic analysis of cucumoviruses using aligned amino acids revealed different relationships among species when the three genome segments were compared (White et al., 1995). This suggested that re-assortment events have given rise to the current isolates. Furthermore, an interspecies pseudo-recombinant between CMV and PSV was found. However, a study in which 217 field isolates of CMV from 11 natural populations were typed by RNase protection assay (Fraile et al., 1997b) provided evidence of selection pressure against re-assortment. CMV is divided into two subgroups based on serological data, and sequence data suggest a further subdivision of subgroup I (Palukaitis et al., 1992; C h a u m p l u k et al., 1996). Recombination has been found within RNA3 of two cucumoviruses, CMV and TAV, co-inoculated to tobacco plants and grown under minimal selection pressure (Aaziz and Tepfer, 1999a). Alignment of the 5' non-translated regions of RNA3 of 26 isolates identified possible re-arrangements, deletions and insertions in this region that may have been the precursors of the subsequent radiation of each subgroup (Roossinck et al., 1999). Phylogenetic analyzes indicated that the three subgroups evolved radially, each from a single origin. Roossinck (2001) describes CMV as a model for RNA virus evolution. 3. Closteroviruses (reviewed by Karasev, 2000) The genome organizations of members of the Closteroviridae are described in Chapter 6 (Section VIII.F) and are illustrated in Figs 6.33 and 6.34. These viruses have the largest (+)strand ssRNA genomes among plant viruses. Dolja et al. (1994) describe the modular organization of closterovirus genomes, identifying
780
17 VARIATION, EVOLUTION AND O R I G I N S OF PLANT VIRUSES
BeD
Be
A
E
A
E
l(a) AB
CD
E
A
A B CD
D
3 (a)
E
A
E
1 R
A B CD
2(c) D
A
3 (b)
D
,3 ....... E
1(c)
A B CD
2(b)
B C~ 1 E
A/
l (b)
2(a)
I
C
o
A
2(d) D
/ E
Y
3(c)
E
A
2(e)
D
3 (d)
B CDE
A
D
3(e)
Fig. 17.10 Styles of dendrograms. Series 1 (a-c) represents unrooted trees (networks) as undirected graphs: l(a) is unweighted and merely indicates relationships; l(b) is weighted and gives the relative edge lengths; 1(c) also shows the possible root 'R', which is at the midpoint of the path between the most dissimilar OTUs (operational taxonomic units, a generic term that can represent many types of comparable taxa), A and D. Series 2 (unweighted) and 3 (weighted) are rooted dendrograms representing 1(c); (a)s are phenograms, (b)s are cladograms, (c)s are curvograms, (d)s are eurograms and (e)s are swoopograms. From Weiller et al. (1995), with kind permission of the copyright holder, 9 The Cambridge University Press. four modules. (At that time of writing, some viruses were classified as closteroviruses but have n o w been moved to other genera; therefore, these comments apply to closteroviruses sensu stricto.) For BYV the core module consists of the key domains for RNA replication (the methyl transferase (MTR), the helicase (HEL) and the R N A - d e p e n d e n t RNA p o l y m e r a s e (POL) (in ORFs 1 a and lb; Fig. 6.33) and belongs to the alphavirus s u p e r g r o u p (see Chapter 8, Section IV.B.1). The u p s t r e a m accessory m o d u l e is the p a p a i n protease (P-PRO) at the Nterminus of ORF l a. The chaperone module, which is separated from the core m o d u l e by a short intergenic spacer, combines a small h y d r o p h o b i c protein (ORF 2), the HSP70 homolog (ORF 3) and the 64-kDa protein with similarities to HSP90 (ORF 4). The fourth module contains the 3' ORFs 5, 6, 7 and 8, which include the major and minor coat proteins.
Obviously, there w o u l d be a slightly different c o m p o s i t i o n of each m o d u l e for CTV (see Table 6.5). Dolja et al (1994) suggest that the closterovirus genome arose from a c o m m o n ancestor, with re-arrangement of that genome and acquisition of other modules by recombination. The progenitor of the alphavirus supergroup has been proposed to comprise a complex of genes encoding MTR, HEL, P-PRO, POL and capsid protein. They propose that the evolutionary p a t h w a y was as shown in Fig. 17.11. This evolutionary p a t h w a y comprised the following steps: 1. Deletion of P-PRO from the core of the replication genes. 2. Substitution of the postulated alphavirustype capsid protein by a capsid protein capable of forming elongated virions.
X. TYPES OF EVOLUTION
3. Invention of the frameshift m e c h a n i s m of POL expression (see C h a p t e r 7, Section V.B.10.b). 4. Acquisition of the HSP70 from the cellular genome. 5. Duplication of the capsid protein and the functional switch for one of the t a n d e m copies to facilitate aphid transmission.
781
6. Insertion of long coding sequences between the MET and HEL cistrons. 7. Secondary acquisition of the leader P-PRO, perhaps from a potyvirus or a related virus. 8. Additional diversification and acquisition of the 3'-terminal genes. 9. Split of the genome into two c o m p o n e n t s giving the crinivirus genome organization.
Fig. 17.11 A tentative scenario for the evolution of closteroviruses. Cpe designates an elongated particle capsid protein (a possible ancestor of capsid protein of both rod-shaped and filamentous viruses). CPH, capsid protein homolog; HEL 1, RNA helicase of superfamily 1; HSP70r, HSP70-related protein; MTR 1, methyltransferase of type 1; POL 3, polymerase of supergroup 3; P-PRO, papain-like proteinase. From Dolja et al. (1994), with kind permission of the copyright holder, 9 Annual Reviews. www.AnnualReviews.org
782
17 V A R I A T I O N , E V O L U T I O N A N D O R I G I N S OF PLANT VIRUSES
It is suggested that steps 1 and 2 occurred early in evolution and gave a common ancestor of the whole tobamovirus cluster. The order of the other events is rather arbitrary. Recombination is an essential part of this proposed evolutionary pathway. Analyzes of multiple species of CTV defective RNAs show that they appear to have arisen by recombination of a subgenomic RNA (sgRNA) with distant parts from the 5' end of the CTV genome (Bar-Joseph et al., 1997). It is suggested that closteroviruses are able to use the sgRNA a n d / o r their promoter signals for modular exchange and rearrangement of their genomes. In spite of the potential for variation, a study showed that mild CTV isolates maintained in different citrus hosts, from several geographical locations (Spain, Taiwan, Colombia, Florida and California) and isolated at different times were remarkably similar (Albiach-Marti et al., 2000a). This indicated a high degree of evolutionary stasis in some CTV populations. 4. Luteoviruses The organization of the (+)-sense ssRNA genomes of the three genera of the Luteoviridae is described in Chapter 6 (Section VIII.G) and shown in Fig. 6.35. Gibbs (1995) recognized a 'supergroup' of small icosahedral viruses that have (+)-sense ssRNA genomes, which now comprise the Luteoviridae, Tombusviriclae and Sobemovirus taxa. The genome organization of these viruses comprises two basic modules: the 5' replicase proteins and the 3' proteins, which include the virion coat protein. Phylogenetic analyzes of the RdRp and the coat protein (Fig. 17.12) suggest that there have been gene transfer events between the modules of members of the supergroup. The phylogenetic analyzes indicated that this supergroup fell into two main clusters termed the 'enamo' and 'carmo' clusters. Likely gene transfers were recognized both between and within these main clusters. The current classification of the Luteoviridae recognizes three genera: the Luteovirus, which falls in the 'carmo' cluster, and the Polerovirus and the Enamovirus, which are placed in the 'enamo' cluster (Fig. 17.12).
Various models have been proposed for the evolution of the luteovirus and polerovirus genome organizations. There is a clear relatedness among the P3 sequences of both genera, and the organizations of ORFs 3, 4 and 5 are obviously similar (see Fig. 6.35). However, the arrangement and composition of the 5' ORFs of the two genera are clearly very different. The most likely model for the origin of the genomes of the two genera is shown in Fig. 17.13 (Miller et al., 1997). In this model, it is suggested that recombination arose by strand switching at subgenomic RNA start sites during RNA replication in cells jointly infected by the two parental viruses. For the derivation of the luteovirus genome, the sgRNA start site on diantho-like viruses has homology to that of poleroviruses. Recombination at this site would create a hybrid virus with dianthovirus polymerase and polerovirus coat protein and neighboring genes. A recombination event at the sgRNA start site downstream of ORF 5 would give the complete luteovirus genome organization. A single recombination between the 5' region of a sobemovirus and the 3' part of a luteovirus followed by premature termination would give the polerovirus genome organization (Fig. 17.13). There is further considerable evidence for recombination within the Luteoviridae. One can distinguish three types of event: recombination within a gene, recombination of large parts of the genome within a genus, and recombination between large parts of the genome between genera. An example of recombination within a gene can be found in the analysis the luteovirus readthrough proteins (ORFs 3 and 5) (Gibbs and Cooper, 1995). There appeared to be a recombinational event between the ancestors of CABYV and PEMV that led to the transfer of the RNA encoding the 5' part of this region to CABYV. As a result of the second type of event, BMYV appears to be a chimera between two members of the genus Polerovirus, with the 5' sequence encompassing ORFs 0, 1 and 2 resembling CABV and the 3' region covering ORFs 3, 4 and 5 being similar to BWYV (Guilley et al., 1995). For at least two viruses, there is evidence for intergeneric recombina-
X. TYPES OF EVOLUTION
Dendrogram A RNA-dlrected RNA polymerase
H
I
9
,
BWYV PLRV BYDV-RPV PEMV SBMV CarMV
9
I
I .
,
....
!
I
TCV MNSV MCMV TNV-A TBSV CyRSV CNV TNV-D BYDV-PAV BYDV-MAV RCNMV
O t
d,, ,,,,
....
i I f(?')
.....
I , _ , :_.
II~I
I
,~ I'l,d~ ............
Dendrogram
/"
B
I I
r .....
//
-r -- --~-'
I ~ "---I
..............
,.,
I
I
i
l
/
I",',
,
.....
! i I i ~ Ill
~lll I'
ii/i il a
..........................
I i
-J
l;t
'l
" 'I
----_ "
0
,
,
,,
,: .
t"~l
, .... V r ......... " 0 . . . . .
.
.
.
BWYV BYDV-RPV PLRV BYDV-PAV BYDV-MAV PEMV CarMV TCV TBSV
s
.
".
enamo
. . " "
'.
9c a r m o
"
9 ..
Vector'
I ~
tll~
9-
|!
Ii I
~
r-----!
[
/
Vlrlon p r o t e i n shell d o m a in
783
9. ". 9 enamo ... ..
CyRSV
-:
9
CNV
.
MNSV RCNMV SBMV
TNV-A TNV-D
..... : . L -
MCMV
carmo 9
. "
aphid aphid aphid aphid aphid aphid mechanical beetle mechanical / soil mechanical / soil Olpidium sp. OIpidium sp. soil beetle Olpidium sp. Olpidium sp. beetle
Fig. 17.12 Classification of RNA-directed RNA polymerase and virion protein shell domain amino acid sequences from 17 sequenced members of the luteovirus 'supergroup' (dendrograms A and B respectively). Dotted lines linking the two dendrograms represent likely gene transfer events between the two main clusters of the 'supergroup'. The dendrograms were deduced by the neighbor-joining method from estimates of the evolutionary distance between each pair of sequences calculated after the progressive alignment of each complete set of sequences. The pattern of transfer events was assessed from the incongruity of lineages given by different genes. The links are drawn arbitrarily at the midpoints of lines joining the nodes where incongruities are evident, as it is not possible to assess at what point the transfer event occurred within an incongruent lineage. The lineages leading to PLRV and BWYV are assumed to be congruent as this minimizes a rate anomaly between the polymerase and virion protein trees within this subcluster. The lineages leading to CarMV, TCV, MNSV and TNV-D are assumed to be congruent within the carom cluster. The means of transmission of each virus is indicated next to the virion protein tree. From Gibbs (1995), with kind permission of the copyright holder, 9 The Cambridge University Press. tion b e t w e e n l u t e o v i r u s e s and poteroviruses. The g e n o m e a r r a n g e m e n t of S b D V (Rathjen et al., 1994) s u g g e s t s that it s h o u l d be classified in the g e n u s Luteovirus, as (1) there is no ORF 0; (2) ORF 1 is relatively small and overlaps ORF 2 by only a f e w nucleotides; (3) the RdRp is carmovirus-like; and (4) the 5' and 3' UTRs contain s e q u e n c e s typical of that genus. H o w e v e r , the p r o d u c t s of S b D V ORFs 3, 4 and 5 are m o r e typical of the g e n u s Polerovirus. Thus, the S b D V g e n o m e resembles a hybrid b e t w e e n those of the two genera. Phylogenetic analysis of the genomic n u c l e o t i d e s e q u e n c e of ScYLV s h o w s that ORFs 1 a n d 2 m o s t c l o s e l y r e s e m b l e polerovirus counterparts, ORFs 3 and 4 are m o s t closely related to counterparts in the
luteovirus g e n o m e s , and ORF 5 is related to the read-through protein g e n e of the only k n o w n e n a m o v i r u s ( M o o n a n et al., 2000; Smith et al., 2000a). The t w o r e c o m b i n a t i o n sites o n the SCYLV g e n o m e m a p to the k n o w n sites for the transcription of subgenomic RNAs. As described in Chapter 11 (Section III.H.l.a), l u t e o v i r u s e s are i n v o l v e d in close biological a s s o c i a t i o n s w i t h other v i r u s e s , e s p e c i a l l y umbraviruses. In t h e s e complexes, the l u t e o v i r u s p r o v i d e s the coat p r o t e i n that enables a p h i d t r a n s m i s s i o n of the u m b r a v i r u s together w i t h the luteovirus. Gibbs (1995) suggested that close associations of viruses such as these c o m p l e x e s could give a greater chance of interviral recombination.
784
~7
V A R I A T I O N , E V O L U T I O N A N D O R I G I N S OF PLANT VIRUSES
Thus, r e c o m b i n a t i o n a p p e a r s to be r a m p a n t b o t h w i t h i n the L u t e o v i r i d a e a n d b e t w e e n m e m b e r s of this family a n d those of s o m e other g r o u p s of small s s R N A viruses. There is also evidence that PLRV can r e c o m b i n e w i t h host sequences (Mayo a n d Jolly, 1991); the 5'-terminal 119 nucleotides of s o m e R N A s of a Scottish
isolate of PLRV are v e r y similar to an exon of tobacco chloroplast D N A . 5. Potyviruses Potyviruses have a (+)-strand ssRNA genome, w h i c h e n c o d e s a p o l y p r o t e i n that is processed to the final gene p r o d u c t s (see C h a p t e r 7,
Fig. 17.13 (see Plate 17.1) Model for origin of luteovirus subgroups. Solid black lines represent viral genomic RNA. Dashed lines indicate subgenomic RNAs. Boxes indicate genes. Blue shading, genes with sequence similarity to umbra-, diantho- and carom-viruses; green, sequence similarity to sobemoviruses. Grey boxes represent putative origins of replication and subgenomic mRNA promoters. POL, RNA-dependent RNA polymerase; PRO?, putative protease; CP, coat protein; MP?, putative movement protein; AT, read-through domain of the coat protein gene, possibly required for aphid transmission. Pink line shows the proposed path of the replicase as it switched strands during copying of viral RNAs in a mixed infection. From Miller et al. (1997b), with kind permission of the copyright holder, 9 The American Phytopathological Society.
• vYpEs oF EvoLuvIoN 785 Section V.B.I.b). Phylogenetic relationships have been assessed using the coat protein sequence (reviewed by Ward et al., 1995). The coat protein comprises three domains: a Nterminal region, a conserved core, and a much smaller C-terminal region. Various phylogenetic analyzes have been performed using the coat protein core sequences (Fig. 17.14). In both dendrograms shown in Fig. 17.14, the basal node is on the branch connecting the bymovirus, BaYMV, coat protein to the others. The second branch is the tritimovirus WSMV and the third branch connects all the aphidtransmitted potyviruses. It is suggested that these taxonomic features correlate with the taxonomy of the hosts and with the vector specificities. The bymoviruses and tritimoviruses are found naturally only in monocotyledonous species, with bymoviruses being transmitted by plasmodiophoraceous fungi and tritimoviruses by eriophyid mites. These are separated by the dicotyledonous plantinfecting viruses that are aphid-transmitted. However, a more recent analysis (Fig. 17.15) (Berger et al., 2000) does not fully support this, with dicot-infecting ipomoviruses and both dicot- and monocot-infecting macluraviruses being relatively close to the tritimovirus and bymovirus branches respectively. The phylogenetic tree shown in Fig. 17.15 does place the dicot-infecting aphid-transmitted viruses together (except for macluraviruses) and separates these from the virus genera with other vectors. Analysis of the potyviral coat protein Nterminal domain sequences reveals evidence for partial gene duplication. For instance, two strains of SCMV have core domains that are 92% identical but N-terminal domains that differ in length and are only 22% identical (Frenkel et al., 1991). Comparison of the Nterminal domain sequences show that they appear to come from different sources and also that there is duplication in the larger sequence. These N-terminal sequences are on the exposed surface of the coat protein and changes in them would not affect the basic coat protein structure important for assembly of rod-shaped particles. The DAG recognition site for the aphid-transmission helper protein
(see Chapter 11, Section III.E.7.b) is in this domain. Recombination also appears to have played a role in the evolution of YMV (Bousalem et al., 2000). Analysis of a region covering the Cterminal part of the NIb, the coat protein and the 3' untranslated region of 27 YMV isolates showed that it had the most variable coat protein compared with eight other potyviruses. There was no correlation between the coat protein and the 3'-UTR diversities and phylogenies. There was a geographical distinction between isolates from the Caribbean, South America and Africa, with most variation occurring in the African isolates. There was also evidence for host selection of particular isolates. Bousalem et al. (2000) hypothesize an African origin of YMV in the Dioscorea cayenensis-D, rotunda complex and independent transfers to D. alata and D. trifida. Geographical distinction has also been recognized between isolates of the peanut stripe strain of BCMV collected from China and Indonesia (Higgins et al., 1999), with variation in the coat protein nucleotide sequence reaching up to 3.5% within and 7.3% between geographical groups. It is suggested that this strain appears to have arisen independently in different locations. 6. Tobamoviruses Tobamoviruses are among the most studied of plant viruses. Relationships have been assessed from a wide variety of data including nucleic acid sequences, amino acid sequences, peptide mapping, serological relationships and biological properties. The evolution of this virus genus has been reviewed by Fraile et al. (1995) and Gibbs (1999b). Estimates of the average number of amino acid substitutions per site between pairs of tobamoviruses for four coding regions (Table 17.3) showed that the 54-kDa region (ORF 2; see Fig. 6.36) was the most conserved and that the 30-kDa movement protein was the most divergent. Most of the nucleotide substitutions were synonymous (not causing a change in amino acid) and did not differ between the four ORFs. Thus, the whole genome appears to have diverged as an entity, with little evidence of recombination playing a significant part in
786
iv VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
diversification. One possible exception is SHMV, which appears to have acquired its 3' non-coding region from a tymo-like virus. Various studies on t o b a m o v i r u s e s h a v e
revealed very stable populations. For example, RNA fingerprinting of 26 isolates of PMMoV from epidemic outbreaks in Spain and Sicily in the 1980s revealed high stability (RodriguezBaYMV
(a)
WSMV TEV-HAT, TEV-NAT SMV-V TVMV Tam MV PRSV-P, -W PVY-10, -D, - 18. -43, -Zhou. -1
............. mm
PVY-N. -GO16 PeMV
...........
LMV
?.
PPV-D, -R, -NAT. -AT TuMV-Kong. -Tremby J
BYMV-CS, -G, -S
JJ H
BYMV-30, CYVV-NZ PS-BMV Ornith MV PWV-TB, -M, -S, -K
S
m I
BCMV-NL3, -NLS, -NL8 I
PsIV, BCMV-NL4 WMV-2,-2FC, SMV-N PWV-SA ZYMV-F, -C, -NAT MDMV-KSI, JGMV MDMV-B, SCMV-SC MDMV-A Sorg MV
(~)
"
BaYMV
"
WSMV
TVMV TEV-HAT, TEV-NAT SMV-V Tam MV PVY-10,-D,-18.-43,-Zhou,-N,-I.-GO16,
PeMV
LMV PS-BMV PPV-D, -R. -NAT, -AT TuMV-Kong, -Tremby PRSV-P, .W
Ornith MV MOMV-B. SCMV-SC MDMV-A
r
Sorg MV BYMV-CS, -G. -S BYMV-30. CYVV-NZ PWV-TB, -M, -S, -K BCMV-NL3, -NL5, -NL8 PWV-SA
PsIV, BCMV-1~.4 WMV-2,-2FC, SMV-N ZYMV-F. -C. -NAT
m,
MDMV-KSI. JGMV
Fig. 17.14 Neighbor-joining trees showing the relationships of potyviruses calculated from the '% non-identity' (panel (a)) or 'FJD distance' (Feng et al., 1985) (panel (b)) of the aligned core domains of their coat proteins. From Ward et at. (1995), with kind permission of the copyright holder, 9 The Cambridge University Press.
X. TYPES OF EVOLUTION 7 8 7
Cerezo et al., 1989). Although there was variation within the population, presumably reflecting its quasi-species status, the master sequence was relatively constant in different geographical sites and over a period of time. The genetic divergence between five populations of TMGMV naturally infecting Nicotiana glauca in southern Spain was very small (Moya et al., 1993), as was that from isolates of this virus from N. glauca plants from Australia, California, Spain and the east Mediterranean Basin (Fraile et aI., 1996). In vitro studies on TMV show that homologous recombination can occur frequently. For instance, repeated sequences inserted into cloned cDNA are eliminated rapidly on infection of plants (see Dawson et al., 1989; Beck and
Dawson, 1990). However, under natural conditions there is little evidence for recombination. Similarly, quantification of mutational error rates of TMV replication, using bacterial genes inserted into the viral genome, give values of 10 -3 to 10 -5, which are very similar to those of other RNA viruses (Donson et al., 1991; Kearney et al., 1999). However, the passage of cloned TMV DNA through seven plant host species over a period of 413-515 days revealed a small mutation rate of 3.1 x 10-4 nucleotide substitutions per base-year (Kearney et al., 1999). Thus, the inherent variability of tobamoviruses is similar to that of other RNA viruses and the sequence conservation in the natural situation must reflect considerable constraints on the accumulation of variants.
Bymovirus
Potyvirus U.r
Macluravirus
%
SCMV-JG
sPea~v Ipomovirus D.. .... t,,
1 " PVY
2 = PeSMV 3 = PepMoV
Rymovirus
~
lg~tk~er Tritimovirus
Fig. 17.15 Phylogenetic tree of the family Potyviridae using amino acid relationships of the coat protein. Inference based on the Fitch and Margoliash (1967) least-squares method. The sequences were aligned using PILEUP (Devereux et al., 1984). Branch lengths are proportional to sequence distances. The dendrogram was bootstrapped 100 times (percentage scores shown at nodes) and are not rooted. Many branches are condensed and may contain multiple viruses. BCMV+, for example, includes BCNMV, CABMV, DMV, PWV, SAPV, SMV, WMV and ZYMV, as well as all the viruses contained in the BCMV subgroup (AzMV, B1CMV, DeMV and PStV). Similarly, the branch labeled BYMV + includes CYVV, PMV, PSbMV and PeMotV. From Berger et al. (2000), with permission.
788
17 VARIATION, EVOLUTION AND ORIGINS OF PLANT VIRUSES
Table 17.3 Estimates of the average number of amino acid substitutions per site, d for different encoded proteins a, and estimates of the nucleotide substitution per site for the 3' ncr b Gene 126K TMV ToMV TMGMV PMMV 54K TMV ToMV TMGMV PMMV MP TMV To MV TMGMV PMMV ORSV SHMV CP TMV To MV TM G MV PMMV ORSV SHMV CGMMV 3'ncr TMV To MV TMGMV P MMV ORSV SHMV
ToMV
TMGMV
PMMV
ORSV
SHMV
CGMMV
0.0929
0.4246 0.4108
0.3064 0.2871 0.4480
-
-
0.7967 0.7824 0.7899 0.8007
0.0622
0.1999 0.2992
0.2211 0.1961 0.2985
-
-
0.5236 0.5477 0.5667 0.5327
0.2384
0.5307 0.4979
0.3729 0.4054 0.4214
0.4397 0.4761 0.5017 0.3539
1.1865 1.0303 1.0026 1.0671 0.9899
0.9848 0.9822 0.9822 0.9636 0.9899 0.9004
0.1785
0.3415 0.3264
0.3558 0.3289 0.3835
0.3136 0.2792 0.3249 0.3585
0.6424 0.8077 0.7722 0.8092 0.8518
0.7820 0.9618 0.9618 0.8625 0.8625 0.7910
0.3199
1.0578 1.0692
0.6785 0.6833 1.0333
0.8080 0.8317 1.1382 0.8863
1.3837 1.381 7 1.7285 1.7461 1.6126
0.8084 0.8996 0.9948 0.9192 0.7687 1.7319
RMV
0.7459 0.7657 0.6868 0.7668 0.7744 0.8692 0.9556
d = - l o g e n / n , where n = average number of amino acids in the protein, n~ = number of identical amino acids in the two proteins (Nei, 1987). b Nucleotide substitutions per site in the 3' non-coding region (ncr) calculated by Kimura's two-parameter method (Kimura, 1980). 126k and 54k, 126- and 54-kDa parts of the replicase; MP, cell-to-cell m o v e m e n t protein; CP, coat protein. From Fraile et al. (1995), with kind permission of the copyright holder, ~ The Cambridge University Press.
7. O e m i n i v i r u s e s T h e D N A -+ D N A r e p l i c a t i o n of g e m i n i v i r u s e s s h o u l d r e s u l t in less m u t a g e n i c v a r i a t i o n t h a n R N A ~, R N A r e p l i c a t i o n ( s e e F i g . 8 . 2 9 ) . H o w e v e r , t h e r e is e v i d e n c e for s i g n i f i c a n t g e n o m i c v a r i a t i o n in at l e a s t t w o of t h e g e m i n i v i r u s g e n e r a , t h e b e g o m o v i r u s e s a n d t h e m a s t r e v i r u s e s . T h e r e is widespread genomic and serological variation in the begomoviruses (reviewed by Harrison and R o b i n s o n , 1999), m u c h of w h i c h r e p r e s e n t s
g e o g r a p h i c a l l y r e l a t e d l i n e a g e s t h a t h a v e little r e l a t i o n to h o s t r a n g e (Fig. 17.16). Using short acquisition and inoculation feedi n g p e r i o d s of s i n g l e Cicadulina mbila leafhopper vector, transmission from an initial M S V i n f e c t i o n of a w i l d p e r e n n i a l h o s t a n d serial passage on almost completely resistant cultivars yielded three MSV isolates (Isnard et al., 1998). S e q u e n c e a n a l y s i s of t h e s e i s o l a t e s indicated that the original infection had a
x. TYPES OF EVOLUTION q u a s i - s p e c i e s s t r u c t u r e w i t h m u t a t i o n s distribu t e d t h r o u g h o u t the g e n o m e . M u t a t i o n freq u e n c i e s w e r e e s t i m a t e d to be b e t w e e n 3.8 • 1-4 a n d 10.5 • 10-4, levels similar to t h o s e found with RNA * RNA replication. This v a r i a t i o n w a s h i g h e s t in s y n o n y m o u s p o s i t i o n s
789
in AC1 a n d A V 1 0 R F s of C L C u V (Sanz et al., 1999). T h e ratio of n o n - s y n o n y m o u s to s y n o n y m o u s s u b s t i t u t i o n s v a r i e d for t h e d i f f e r e n t ORFs, b e i n g h i g h e r for AV1 t h a n for A C 1 a n d l o w e r still in the A C 4 a n d A V 2 0 R F s . It is sugg e s t e d t h a t the e v o l u t i o n of the A C 4 a n d AV2
86 I'~
tgmv
bctv bgmv
65r.r olc311
I ~ L_clca i
94'
tylcvis sacmv tylcvni tylcvsa eacmvt acmvk tblcvc tlcvau tylcvth 979.[---sqlcvc
I
99 - ~ 57
L.j 2i
.
(a)
.
.
.
.
.
I
~clcv72b clcv26 icmv itmlcv 69j'- tlcrv IL.- bdmv 98 ~ - - abmv I ; tomov L-pymv
L___ bgmv
L._ clc26
86 I !~ ' J (b) |
| i
,
acmvk clc804a
.... .. clc72b oyvm201
oyvm301
tylcvsa icmv eacmvk ~ - [ - . - . - . - tylcvis tlcvp 55-~.- tlcvau
I L--tlcvp
89 [ I
,
.
tgmv ~bgmv 9.~
icmv
tylcvth
loo~[--- tlcvp Mr-- clc72b 84 ~ F clca I
57~_ oyvm301
95
100} clc804a