Advances in Virus Research, Volume 46 (Advances in Virus Research)

Advances in

VIRUS RESEARCH VOLUME 46

ADVISORY BOARD DAVIDBALTIMORE

PAULKAESBERG

ROBERT M. CHANOCK

BERNARD Moss

PETERC. DOHERTY

ERLINGNORRBY

N. FIELDS BERNARD

AKIRAOYA

H. J. GROSS

J. J. SKEHEL

B. D. HARRISON

R. H. SYMONS

M. H. V. VANREGENMORTEL

Advances in VIRUS RESEARCH Edited by

KARL MARAMOROSCH

FREDERICK A. MURPHY

Department of Entomology Rutgers University New Brunswick, New Jersey

School of Veterinary Medicine University of California, Davis Davis, California

AARON J. SHATKIN Center for Advanced Biotechnology and Medicine Piscataway, New Jersey

VOLUME 46

W ACADEMIC PRESS San Diego New York Boston London Sydney Tokyo Toronto

This book is printed on acid-free paper.

@

Copyright 0 1996 by ACADEMIC PRESS, INC. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Academic Press, Inc. A Division of Harcourt Brace 19Company 525 B Street, Suite 1900, San Diego, California 92101-4495 United Kingdom Edition published by Academic Press Limited 24-28 Oval Road, London NWI 7DX

International Standard Serial Number: 0065-3527 International Standard Book Number: 0- 12-039846-X PRINTED IN THE UNITED STATES OF AMERICA 95 96 9 7 9 8 99 0 0 B C 9 8 7 6 5

4

3 2

1

CONTENTS

Poliovirus Assembly and Encapsidation of Genomic RNA

DAVIDANSARDI.DONNAC. PORTER. MARIEJ . ANDERSON. AND CASEYD . MORROW Overview ........................................................ 2 3 Genomic Organization ............................................ Poliovirus Life Cycle ............................................. 6 14 Poliovirus Virion ................................................. 19 Morphogenesis of Poliovirus ...................................... 30 RNA Encapsidation Process ....................................... 34 New Methods to Study Poliovirus Assembly ........................

I. I1. I11. IV . V. VI . VII . VIII . Complementation System to Study Poliovirus Encapsidation ........ IX . Perspectives on Poliovirus Assembly ............................... References .......................................................

39 53 56

Genome Rearrangements of Rotaviruses

I. I1. I11.

IV . V. VI . VII . VIII . IX . X.

ULRICH DESSELBERGER Discovery of Genome Rearrangements ............................. Extent of Genome Rearrangements in Rotaviruses ................. Sequence Data of Rearranged Genes ............................... Genome Rearrangements Generated in Vitro ....................... Mechanisms of Genome Rearrangements .......................... Biophysical Data ................................................. Function of Rearranged Genes and Their Products ................. Genome Rearrangements and Evolution of Rotaviruses ............. Genome Rearrangements in Other Genera of Reouiridae ............ Outlook ......................................................... References .......................................................

71 75 75 79 82 86 86 91 92 92 93

Human ImmunodeficiencyVirus Type 1 Reverse Transcriptase and Early Events in Reverse Transcription

ERICJ . ARTSAND MARKA . WAINBERG I. Introduction .....................................................

I1. Overview of Human Immunodeficiency Virus Type 1 Replication V

....

99 101

vi

CONTENTS

I11. Human Immunodeficiency Virus Type 1 ........................... IV . Human Immunodeficiency Virus Type 1 Reverse Transcription ...... References .......................................................

107 119 146

Hepadnaviruses: Current Models of RNA Encapsidation and Reverse Transription

DOROTHY A . FALLOWS AND STEPHEN P. GOFF I. I1. I11. IV . V. VI .

Introduction ..................................................... Transcription and Translation .................................... RNA Encapsidation .............................................. The Hepadnaviral Polymerase .................................... Reverse Transcription ............................................ Concluding Remarks ............................................. References .......................................................

167 172 176 180 184 192 193

Cell Types Involved in Replication and Distribution of Human Cytomegalovirus

BODOPLACHTER. CHRISTIAN SINZGER. AND GERHARD JAHN I. I1. I11. IV . V. VI . VII .

Introduction ..................................................... Determinants of Human Cytomegalovirus ......................... Organ Tropism of Human Cytomegalovirus ........................ Cells Types Involved in Acute Human Cytomegalovirus Disease ..... Viral Spread and Pathogenesis .................................... Latent Cytomegalovirus Infection ................................. Summary ........................................................ References .......................................................

197 198 216 219 232 236 241 241

Varicella-Zoster Virus: Aspects of Pathogenesis and the Host Response to Natural Infection and Varicella Vaccine

ANN M . ARVIN.JENNIFER F . MOFFAT. AND REBECCA REDMAN I. Introduction ..................................................... 265 I1. The Virus ....................................................... 266

I11. Cell-Associated Viremia in the Pathogenesis of Varicella-Zoster Virus Infection ................................................... IV. The Cell-Mediated Immune Response to Varicella-Zoster Virus ...... V. Summary ........................................................ References .......................................................

267 280 306 307

Anatomy of Viral Persistence: Mechanisms of Persistence and Associated Disease

JUANCARLOS DE

LA

TORREAND MICHAELB . A . OLDSTONE

I . Introduction ..................................................... I1. Requirements for Establishment of Viral Persistence ...............

313 315

CONTENTS

vii

111. Virus-Induced Alterations of Host Cellular Differentiated Functions

in Absence of Cytolysis ........................................... IV . Conclusions ...................................................... References .......................................................

323 338 340

The lridoviruses

TREVOR WILLIAMS I. I1. I11. IV . V. VI . VII .

Introduction ............. ........................... Classification .................................................... Structure ................... ...................... Replication ...................................................... Molecular Biology ................................................ Ecology ...................... ..................... Future Directions for Iridoviruses ................................. References .............. .....................

347 350 366 372 386 391 399 401

Molecular Biology of Luteoviruses

I . Introduction

M . A . MAYOAND V . ZIEGLER-GRAFF .................. ..........................

Mechanisms of Gene Expr ................... Particle Structure ................................................ Location of Luteovirus Replication ....................... Phytopathology .................................................. Taxonomy .............. Concluding Remarks . . . . References .... ...................................

416 417 424 435 444 449 450 453 457 457

INDEX ...........................................................

463

I1. Genome Structure ................................................ 111. Functions of Gene Products .......................................

IV . V. VI . VII . VIII .

IX .

This Page Intentionally Left Blank

t

POLIOVIRUS ASSEMBLY AND ENCAPSIDATION OF GENOMIC RNA David C. Ansardi, Donna C. Porter, Marie J. Anderson, and Casey D. Morrow Department of Microbiology University of Alabama at Birmingham Birmingham, Alabama 35294

I. Overview 11. Genomic Organization 111. Poliovirus Life Cycle A. Virus Entry and Uncoating B. Translation of Viral RNA C. Release of Individual Proteins by Viral Proteases D. Replication of Viral RNA IV. Poliovirus Virion A. Properties of Virion B. Virus Structure C. Myristylation of Poliovirus Capsid Proteins V. Morphogenesis of Poliovirus A. 5s Protomer B. 14s Pentamer C. Empty Capsid D. Provirion VI . RNA Encapsidation Process A. RNA Requirements for Encapsidation B. Poliovirus Defective Interfering Particles C. RNA Encapsidation Signals D. Subcellular Location of Encapsidation VII. New Methods to Study Poliovirus Assembly Process A. Studies of Poliovirus Assembly Process Using Recombinant Vaccinia Viruses B. Expression of Poliovirus P1 and 3CD Using Recombinant Vaccinia Virus Vectors C. Functional Significance of Poliovirus Capsid Myristylation VIII. Complementation System to Study Poliovirus Encapsidation A. Proteolytic Cleavage of Capsid Precursor B. Capsid Mutations Affecting RNA Encapsidation C. Studies on Maturation Cleavage Using Complementation System IX. Perspectives on Poliovirus Assembly References

1 Copyright 0 1996 by Academic Press,Inc. All rights of reproduction in any form resewed.

2

DAVID C. ANSARDI et al.

I. OVERVIEW The biology of poliovirus has been a subject of intense study since the 1950’s. Poliovirus is the causative agent of the paralytic disease poliomyelitis, once a major health problem in the United States that has largely been eradicated since the development of two highly effective vaccines (Sabin and Boulger, 1973; Salk, 1960). Despite control of the disease in industrialized nations, poliomyelitis continues to be a health concern in the undeveloped world. Poliovirus is a member of a family of viruses, the Picornauiridae, that includes members responsible for several diseases of humans, including the human rhinoviruses (common cold), hepatitis type A, and the coxsackieviruses (cardiac infections) (Rueckert, 1990). Other members of the Picornauiridae are responsible for important diseases of livestock, including foot-and-mouth disease virus, bovine enterovirus, and the causative agent of swine vesicular disease. Another group of picornaviruses, the cardioviruses, primarily infect mice and includes members such as mengo virus and encephalomyocarditis virus (EMCV). Poliovirus, like all of the members of the Picornauiridae, is a spherical, single-stranded RNA virus. The viral genome is a n approximately 7500-nucleotide-long RNA molecule of positive polarity (messenger-sense) and is encapsidated within a virion particle that is approximately 30 nm in diameter (Kitamura et al., 1981; Koch and Koch, 1985). The poliovirus genome has been cloned and sequenced (Kitamura et al., 1981; Racaniello and Baltimore, 1981a1, greatly facilitating the analysis of specific proteins and cis-acting regions of the RNA genome in the life cycle of the virus. Three different antigenically distinct serological types of virulent poliovirus have been identified, designated types 1 , 2 , and 3 (Koch and Koch, 1985). The polioviruses are members of the enterovirus genus of the Picornauiridae,and as such primarily inhabit the alimentary canal of the host. Most infections with poliovirus do not result in paralytic disease. However, in some instances, poliovirus spreads from the intestine to the central nervous system; lytic replication of the virus in motor neurons results in paralysis, which can often be fatal (Koch and Koch, 1985). Control of poliovirus infection in modern nations is largely based on the success of a highly effective oral vaccine consisting of live, attenuated strains of poliovirus (Sabin and Boulger, 1973). The attenuated strains given to the vaccine recipient replicate in the intestine, where they stimulate immunity against poliovirus infection, but are incapable of causing paralytic disease. The highly effective nature of the poliovirus vaccines has led to intensive research into the application of poliovirus as a vector for delivering foreign antigens to the

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

3

immune system, opening the possibility that this human pathogen might be harnessed for helpful purposes (Almond and Burke, 1990; Ansardi et al., 1994b; Porter et al., 1993a, 1995). Several developments have made poliovirus an excellent model for studying the molecular processes of viral replication. Poliovirus can be grown in many tissue culture cell lines of human and primate origin (Koch and Koch, 1985). The viral genome has been cloned and sequenced, revealing the nucleotide sequence of the RNA genome and the predicted amino acid sequences for the viral proteins (Kitamura et al., 1981; Racaniello and Baltimore, 1981a). cDNA copies of the poliovirus RNA genome are infectious and result in a productive virus infection on transfection into suitable host cells (Racaniello and Baltimore, 1981b; Semler et al., 1984). The infectivity of poliovirus cDNA has allowed the use of techniques such as site-specific mutagenesis to alter the coding sequence of the virus (Zoller and Smith, 1983). A further advance was made with the finding that positive-sense RNA genomes transcribed in uitro from poliovirus cDNA, under the control of the promoter for bacteriophage T7 RNA polymerase, were highly infectious on transfection into host cells (Van der Werf et al., 1986). In 1985, the three-dimensional structure of poliovirus was solved, providing detailed information about the structure of poliovirus capsid proteins and insight into possible mechanisms of poliovirus morphogenesis (Hogle et al., 1985). The cell surface protein receptor used by poliovirus to gain entry into the host cell has been cloned and sequenced (Mendelsohn et al., 1989). Transgenic mice which express the poliovirus receptor have also been generated, providing an animal model in which the molecular mechanisms of poliovirus pathogenesis can be studied (Ren et al., 1990).

11. GENOMICORGANIZATION The organization of the poliovirus RNA genome and the cascade of the formation of individual viral proteins are presented in Fig. 1. The positive-sense RNA genome of poliovirus is 7441 nucleotides in length (Kitamura et al., 1981; Racaniello and Baltimore, 1981a). The 5' end of the RNA genome is not linked to a 7-methylguanosine cap, but instead is covalently linked to a virus-encoded basic peptide of 22 amino acids, known as VPg (genome-linked protein), through a phosphodiester linkage between the 0 4 hydroxyl group oxygen of a tyrosine residue in VPg and the phosphate of the 5' terminal uridine residue of the RNA genome (Ambros and Baltimore, 1978; Lee et al., 1977; Morrow et al., 1984; Nomoto et al., 1976; Rothberg et al., 1980; Wimmer, 1982).The 3'

4

DAVID C. ANSARDI et al.

VPg-IRESy 33;86 W APSID-NON-C

5,

APSI7370~ 7 4 4 1

AA(AEnAA 3' n=dO

OPEN READING FRAME

POLYPROTEIN

A

A

-PLF 1[yp3lpiq P qm p q

?+N.A.+PO

A

4 m VP4

2BC

3AB

MlzC1ElO38

uncleaved

3C + 3D

(VPg)

A cleavage catalyzed by 2A A cleavage catalyzed by 3CD A cleavage cntdyzd by 3C A ~ ~ ~ ~ ; ~ ~

~

~

~

l

,

w

n

FIG.1. Poliovirus genomic organization and cascade of polyprotein processing. The poliovirus genome is a single-stranded messenger (plus sense) RNA molecule that is approximately 7500 bases in length. The 5' end of the RNA molecule is covalently linked to a small peptide, VPg, and the 3' end contains a genetically encoded polyadenylate tail that is approximately 60 nucleotides long. The first 742 nucleotides at the 5' end of the genome comprise the 5'-N"R, which contains the internal ribosome entry sequence (IRES). The poliovirus genome contains a single open reading frame encoding a 2209amino acid polyprotein precursor. Virus-encoded proteases 2A and 3C catalyze cis-acting cleavages of the polyprotein to initiate the cascade of formation of the individual viral proteins. Further processing of the viral proteins is primarily mediated by 3C, although the 3CD polyprotein catalyzes cleavage of the P1 capsid precursor to VPO, VP3, and VP1. Both 3C and 3CD catalyze cleavages a t glutamine-glycine dipeptides, whereas 2A catalyzes cleavages between tyrosine-glycine amino acid pairs. The final cleavage event occurs at an asparagine-serine amino acid pair on the interior of the virion, resulting in conversion of VPO to VP2 and VP4. The source of this cleavage is unknown, but it is speculated to occur intramolecularly.

end of the genome is polyadenylated, with a tail length of approximately 60 adenine residues. The poly(A) tail is genetically encoded by the virus rather than added by host cell polyadenylation enzymes (Kitamura et al., 1981; Racaniello and Baltimore, 1981a; Spector and Baltimore, 1975; Yogo and Wimmer, 1975). The majority of the viral RNA genome (6627 nucleotides) constitutes a long open reading frame that encodes a single translation product of 2209 amino acids. An unusually long nontranslated region of 742 nucleotides (5'-NTR) pre-

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

5

cedes the open reading frame upstream of the initiation codon used for translation of the genomic polyprotein. The 5'-NTR contains eight AUG triplets prior to the one which actually serves as the initiation codon for translation of viral proteins (Kitamura et al., 1981; Pelletier et al., 1988; Racaniello and Baltimore, 1981a). Translation of poliovirus as well as other picornavirus RNA genomes occurs by a capindependent method, in which ribosome binding occurs at an internal sequence known as the internal ribosome entry sequence (IRES) (Jang et al., 1988; Pelletier et al., 1988; Pelletier and Sonenberg, 1988; Sonenberg, 1990; Trono et al., 1988). The coding portion of the poliovirus genome is subdivided into three distinct regions, designated P1, P2, and P3 (Kitamura et al., 1981; Rueckert and Wimmer, 1984). The P1 region encodes the viral capsid proteins VP1, VP2, VP3, and VP4. The P2 region encodes nonstructural viral proteins including a protease, 2A, and 2B and 2C, which are believed to play roles in replication of the RNA genome. The P3 region encodes nonstructural proteins required for virus replication, including 3Dpo1, the RNA-dependent RNA polymerase, a protease, 3Cpr0, and the VPg protein (also known as 3B). Many of the viral proteins have important functions in polyprotein forms; for example, the membrane-bound 3AB protein is a component of the replication complex (Giachetti and Semler, 1991; Semler et al., 1982), and the 3CD polyprotein catalyzes proteolytic cleavages of the capsid precursor (Jore et al., 1988; Ypma-Wong et al., 1988a). All of the proteolytic cleavages required to liberate individual poliovirus proteins required for replication and encapsidation of the genomic RNA are catalyzed by virus-encoded proteases which cleave the primary translation product both in cis and in trans (Dewalt and Semler, 1989; Hanecak et al., 1982; Harris et al., 1990; Lawson and Semler, 1990; Palmenberg, 1990; Toyoda et al., 1986). The primary cleavage of the genomic polyprotein is an intramolecular event in which the 2A protease processes the peptide bond between a tyrosine-glycine dipeptide, releasing the 97-kDa polyprotein encoded by the P1 region (Toyoda et al., 1986). The P1 protein is a precursor from which the individual capsid proteins of the virus are derived. The virus-encoded protease 3Cpr0, acting in a polyprotein form, 3CD, is responsible for cleavage of the P1 precursor to VPO, VP3, and VP1 (Jore et al., 1988; Ypma-Wong et al., 1988a). Cleavage of VPO to VP2 and VP4 is catalyzed during or after RNA encapsidation and is widely believed to occur intramolecularly (Arnold et al., 1987; Jacobson et al., 1970). The viral proteins encoded in the P2 or P3 regions are released from polyprotein precursors by the protease 3Cpro (Hanecak et al., 1982). These cleavages occur'exclusively at glutamine-glycine dipeptides, although

6

DAVID C. ANSARDI et al.

not every glutamine-glycine dipeptide present in the genomic polyprotein is a substrate for 3C-mediated cleavages. An additional tyrosineglycine dipeptide substrate for 2A~rolies in the 3CD polyprotein, resulting in the production of two proteins, 3C' and 3D'. This cleavage may simply be a fortuitous event as poliovirus mutants without this cleavage site have no apparent growth defects (Lee and Wimmer, 1988).

111. POLIOVIRUS LIFECYCLE Infection of cells by poliovirus is associated with several pronounced cytopathic effects on the host cell, including shrinkage in cell size, an increase in intracellular membranous vesicles, deformation of the nucleus, and changes in the cell cytoskeleton (Koch and Koch, 1985). A schematic representation of the events which take place during a single cycle of poliovirus replication are depicted in Fig. 2. The virus initially attaches to the host cell by binding to a cell-surface glycoprotein molecule. The normal cellular function of the poliovirus receptor is unknown, but the predicted amino acid sequence derived from the cloning of the receptor gene indicates that the molecule belongs to the immunoglobulin-like superfamily of proteins (Mendelsohn et al., 1989). On attachment to the receptor, the virus undergoes conformational changes, and the internal capsid protein VP4 is expelled from the virion (De La Torre et al., 1992; DeSena and Mandel, 1976, 1977; Guttman and Baltimore, 1977a; Rueckert, 1990). The virus is believed to be internalized into the cytoplasm by receptor-mediated endocytosis (Madshus et al., 1984, 1985). The mechanism by which the virus releases its RNA genome across the endosomal membrane and into the cytoplasm is not understood. Once present in the cytoplasm, the messenger-sense viral RNA genome is translated on host ribosomes to yield viral proteins. Translation of the poliovirus genome is an obligatory first step because the virus does not package any of the proteins required to initiate replication of the viral RNA genome. An important consequence of poliovirus infection is the shutoff of translation of host cell mRNA, which occurs primarily as a result of cleavage of the large subunit ( ~ 2 2 0of ) the cap binding complex (eIF-4F) (Etchison et al., 1982). This cleavage is indirectly mediated by the viral protease 2Apro (Bernstein et al., 1985; Krausslich et al., 1987; Lloyd et al., 1988; Wycoff et al., 1990). Once viral proteins are synthesized, RNA synthesis occurs exponentially from approximately 30 min postinfection to 3 hr postinfection, then occurs in a linear fashion until approximately 4.5 hr postinfection followed by a rapid decline in the rate of synthesis (Rueckert, 1990).

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

7

Replication of the viral genome requires that the plus-strand RNA molecule is transcribed first to yield a complementary minus-strand RNA, which is also linked at its 5' end to VPg (Kuhn and Wimmer, 1987; Paul et al., 1987b; Richards and Ehrenfeld, 1990). This minusstrand RNA then serves as the template for synthesis of new plus strands of RNA. Synthesis of plus- and minus-strand RNA molecules is an asymmetric process, with plus strands produced in excess of minus strands by at least 10-fold. Replication of the poliovirus RNA genome occurs in association with smooth membrane vesicles which proliferate on infection, and the combination of these membranes with the viral proteins and RNA template molecules required for RNA replication is referred to as the replication complex (Caliguiri and Tamm, 1970; Ehrenfeld et al., 1970; Kuhn and Wimmer, 1987; Paul et al., 1987b; Richards and Ehrenfeld, 1990). Progeny plus-strand RNA serves as both mRNA for synthesis of additional viral proteins and as the RNA molecule encapsidated in progeny virions. Encapsidated virion RNA is linked to VPg, whereas the VPg protein is removed from the 5' end of plus-strand RNA molecules destined for translation (Hewlett et al., 1976; Nomoto et al., 1977; Petterson et al., 1977). The final aspect of the poliovirus life cycle is the formation of progeny virions. The capsid proteins assemble subviral oligomeric particles, probably prior to interaction with the RNA genome, although the precise pathway of assembly has not been deduced (Putnak and Phillips, 1981a; Rueckert, 1990).Encapsidation of plus-strand VPg-linked RNA may occur by condensation of 12 pentamers of VPO, VP3, and VP1 [(VPO-3-1),] around the RNA molecule or by insertion of VPg-linked RNA into a preformed empty capsid or procapsid consisting of 60 copies of VPO-VP3-VP1 [(VP0-3-1),,1 (Jacobson and Baltimore, 1968; Rueckert, 1990). The encapsidation process is specific for both VPglinked RNA and plus strands as packaging of minus strands does not occur, despite the presence of VPg (Nomoto et al., 1977; Novak and Kirkegaard, 1991; Petterson et al., 1978). At the end of infection, lysis of the cell occurs and virions exit, although mechanisms for active release of virus prior to lysis may exist (Tucker et al., 1993).

A . Virus Entry and Uncoating The mechanism by which poliovirus enters the host cell is poorly understood (Rueckert, 1990). Progress in this field will likely proceed at a faster pace with the identification, cloning, and sequencing of the poliovirus receptor (Mendelsohn et al., 1989). On attachment of poliovirus virions t o the glycoprotein receptor, the virus undergoes conformational changes that are marked by a conver-

8

DAVID C. ANSARDI et al.

FIG.2. Events in a single cycle of poliovirus infection. Poliovirus virions initiate infection by attaching to a glycoprotein receptor on the cell surface. The virus is believed to be internalized into the cell by receptor-mediated endocytosis. On attachment to the receptor and entry into the cell, the capsid undergoes conformational changes, and the messenger-sense RNA genome is released into the cytoplasm in a n unknown manner. The viral RNA genome is translated on host ribosomes to generate proteins required for RNA replication and encapsidation of progeny genomes. RNA replication occurs in virus-induced complexes of viral and host protein(s) that are associated with smooth membrane vesicles. RNA replication proceeds by synthesis of minus-sense RNA followed by synthesis of nascent plus strands, which occurs in excess over minus-strand formation. RNA structures in which several nascent plus strands are simultaneously being synthesized on the same minus-strand template are known as RI or replicative

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

9

sion of the sedimentation coefficient of the virion from 155s to 135s (DeSena and Mandel, 1976, 1977; Everaert et al., 1989; Fricks and Hogle, 1990; Guttman and Baltimore, 1977a;Kaplan et al., 1990).This process is associated with the expulsion of the small, myristylated internal capsid protein, VP4, from the virion. In addition to release of VP4, the conformational changes associated with attachment to the receptor also lead t o exposure of the amino terminus of the viral capsid protein VP1 on the surface of the virion, a location which is far removed from its normal location on the capsid interior (Fricks and Hogle, 1990).Exposure of the amino terminus of VP1 on the surface of the virion increases its hydrophobicity, giving the structurally altered virions the ability to bind to liposomes (Fricks and Hogle, 1990).The amino terminus of VP1 has been modeled as an amphipathic helix (this region of VP1 was unresolved in the X-ray structure), and the hypothetical formation of this structure has led to the proposal that amino-terminal residues of VP1 may be involved in forming a pore through endosomal membranes through which the viral RNA genome can be released into the cytoplasm (Fricks and Hogle, 1990). A role for the amino terminus of VP1 in virus uncoating has been supported by the phenotypes of two temperature-sensitive poliovirus mutants which contain small deletions in the VP1 amino terminus and which are defective in virus uncoating at the nonpermissive temperature (Kirkegaard, 1990; Kirkegaard and Nelson, 1990). Attachment of virus to the cellular receptor is not a guarantee of successful entry into the cell, as this process appears to be largely abortive and is associated with sloughing of a large percentage of attached, altered particles (Mandel, 1965; Rueckert, 1990).Once bound to the receptor, the virus is believed to be internalized through receptor-mediated endocytosis (Madshus et al., 1984, 1985). The process by which RNA is released from the endosomes and into the cytoplasm is not well understood. Acidification of the endosomes might be responsible for conformational changes required for capsid protein fusion with the membrane and release of RNA (Madshus et al., 1984, 1985). A study conducted on a mutant of human rhinovirus, another member of the Picornauiridae, suggested that the conformational changes associated with receptor attachment were not sufficient for RNA release into the cytoplasm, and led to the proposal that the uncoating capsid must form a membrane-associated structure, termed an

intermediate RNA. Plus strands produced in the replication complexes are either encapsidated or translated (following removal of VPg) to generate additional viral proteins. Poliovirus infection results in lysis of the host cell, allowing progeny virions to exit.

10

DAVID C. ANSARDI et al.

infectosome, responsible for injecting the RNA genome into the cytoplasm (Lee et al., 1993).

B . Translation of Viral RNA Poliovirus has evolved a cap-independent method of translation which allows it t o shut off host cell cap-dependent translation by inactivating a component of a translation initiation factor (eIF-4F) which recognizes the capped 5’ ends of host mRNA molecules (Etchison et al., 1982; Pelletier et al., 1988; Pelletier and Sonenberg, 1988; Sonenberg, 1987, 1990; Trono et al., 1988). A host cellular enzyme is believed to unlink the VPg protein from the 5’ end of poliovirus virion RNA prior to translation (Ambros and Baltimore, 1978; Hewlett et al., 1976; Lee et al., 1977; Morrow et al., 1984; Nomoto et al., 1977; Rothberg et al., 1980; Wimmer, 1982). Initiation of translation of poliovirus mRNA does not proceed by the scanning model proposed by Kozak (1989). Ribosome binding to poliovirus RNA occurs in the 5’-NTR, upstream of the initiator AUG codon, and is mediated by an internal sequence of several hundred nucleotides, which has been designated the internal ribosome entry sequence (IRES) (Jang et al., 1988; Pelletier et al., 1988; Pelletier and Sonenberg, 1988; Sonenberg, 1990; Trono et al., 1988). The determinants for recognition of the IRES by host translational machinery have not been elucidated, but secondary RNA structures present in the 5’-NTR between nucleotides 240 and 620 may mediate the internal binding of ribosomes (Sonenberg, 1990).How the IRES operates is not yet clear; possibly the ribosome binds the IRES region and scans the RNA genome until it encounters the initiator AUG codon at position 743 and begins translation. The 5‘-NTR of poliovirus type 1 contains eight AUG codons upstream of the initiating AUG that are not used as initiator codons (Kitamura et al., 1981; Racaniello and Baltimore, 1981a). However, the 100 nucleotides between the 3‘ end of the IRES and the initiator methionine at nucleotide 743 contain no AUG codons. Translation of the single long open reading frame results in the synthesis of a long polyprotein. The actual existence of this translation product in uiuo is doubtful, however, as 2Apro cleaves the P1 portion out of the growing polyprotein intramolecularly and probably cotranslationally (Toyoda et al., 1986). The shutoff of host-cell mRNA translation in poliovirus-infected cells is largely associated with cleavage of the p220 component of the cap-binding complex (eIF-4F)(Etchison et al., 1982).Inactivation of the cap-binding complex prevents binding of the translation initiation factors eIF-4A and eIF-4B to the 5’ end of mRNA molecules, which are

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

11

believed to be required for melting of RNA secondary structure in the 5' ends of mRNA molecules to allow binding of the ribosome (Sonenberg, 1990). Cleavage of the p220 protein is indirectly mediated by the viral protease 2Ap*o in some way associated with eIF-3, although the exact mechanism by which 2Apro induces p220 cleavage is not certain (Bernstein et al., 1985; Krausslich et al., 1987; Lloyd et al., 1988; Wycoff et al., 1990). Generally, 2Apro is not believed to catalyze the cleavage of p220 directly but may somehow activate a latent cellular protease which cleaves p220.

C . Release of Individual Proteins by Viral Proteases The polyprotein organization of the poliovirus RNA genome translation product dictates that proteases required to liberate individual proteins play a critical role in the life cycle of the virus. All cleavages of poliovirus proteins, except for the maturation cleavage of VPO to VP2 and VP4, have been shown to be mediated by virus-encoded proteases (Hanecak et al., 1982; Toyoda et al., 1986). Although not formally proven, the maturation cleavage of VPO to VP2 and VP4 is likely to occur through an intramolecular mechanism subsequent to encapsidation of the genomic RNA (Arnold et al., 1987; Jacobson et al., 1970). A description of the two poliovirus proteases, 2A~r0and 3Cpr0,is given in the following sections. 1 . Protease 2 A p r o

The viral protease 2Apr0 is responsible for two cleavages of poliovirus polyproteins, one which occurs in cis and the other which occurs in trans, and is indirectly involved in the inactivation of the p220 component of eIF-4F, as described in the previous section (Toyoda et al., 1986). The 2Apro protein is speculated to be a member of the serine protease family, but instead of serine the enzyme may use a cysteine residue as the nucleophile in the catalytic active site (Bazan and Fletterick, 1988). These predictions have been substantiated by data showing that mutations of putative members of the catalytic triad of residues inhibit 2A~roactivity (Yuand Lloyd, 1991). The 2Apro protease is responsible for the cotranslational primary cleavage of the poliovirus translation product which occurs in cis at a tyrosine-glycine bond, releasing the P1 capsid precursor protein (Toyoda et al., 1986).The only other confirmed cleavage of poliovirus proteins by ~ A Poccurs ~ o in trans at a tyrosine-glycine dipeptide in the 3CD polyprotein, releasing two proteins designated 3C' and 3D'. These proteins are not required for viral replication, and their formation may simply be the result of a fortuitous processing site (Lee and Wimmer, 1988).

12

DAVID C. ANSARDI et al.

The determinants for substrate recognition by 2Apro have not entirely been identified but clearly involve aspects other than primary sequence, because cleavage occurs at only 2 of 10 tyrosine-glycine dipeptides in the poliovirus polyprotein (Kitamura et al., 1981; Racaniello and Baltimore, 1981a). A study demonstrated that the P2 and P1’ residues relative to the cleavage site (P2 refers to the second residue amino terminal to the scissile bond, and P1’ is the position of the residue immediately carboxyl terminal to the site of cleavage) were important determinants of cleavage site recognition, and that the primary sequence requirements for cleavage site recognition in trans were more stringent than for the cis cleavage at the site between the VP1 protein and 2A protein (Hellen et al., 1992). The requirement for the proteolytic cleavage activity of the enzyme 2A~rohas been shown to be dispensable for replication of a poliovirus replicon containing foreign gene sequences substituted for the capsid gene in vaccinia virusinfected cells (Ansardi and Morrow, 1995; Ghosh and Morrow, 1993). 2 . Protease 3Cpro

The enzyme 3Cpr0 is the viral protease responsible for the majority of poliovirus protein cleavages (Hanecak et al., 1982). The 3Cpr0 enzyme has been predicted to share structural homology with the serine family of proteases, but a cysteine residue is believed to function as the nucleophile in the catalytic triad (Bazan and Fletterick, 1988; Gorbalenya et al., 1989; Ivanoff et al., 1986; Lawson and Semler, 1991). The 3Cpro protease catalyzes proteolytic cleavages at glutamineglycine (QG) dipeptide sites in the poliovirus polyprotein (Hanecak et al., 1982). As with the 2Apro protease, 3Cpro activity is responsible for both cis and trans cleavages of the polyprotein, although the precise pathways of generation of each of the individual polypeptides is still under investigation (Dewalt and Semler, 1989; Hanecak et al., 1984; Harris et al., 1990; Lawson and Semler, 1990,1992; Palmenberg, 1990). Poliovirus is unique among picornaviruses in that all of the cleavages catalyzed by 3Cpro occur at QG bonds. In other picornaviruses, more flexibility in primary sequence at the 3Cpro cleavage sites is evident, primarily at the P1’ residue (Palmenberg, 1990).Information about the tolerance of different substituents at 3Cpro cleavage sites by the poliovirus enzyme is limited, although one study indicated that an alanine substitution for the glycine residue at the QG site between proteins 3C and 3D was compatible with cleavage, whereas more drastic substitutions inhibited cleavage (Kean et al., 1990). Determinants for 3Cpro-mediated cleavages other than a QG primary sequence must exist, because only 8 of the 13 QG dipeptides present in the poliovirus translation product are actually used as cleavage sites (Kitamura et

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

13

al., 1981; Racaniello and Baltimore, 1981a). The selection of QG cleavage sites is at least partially determined by accessibility; for example, unused QG sites present in the P1 capsid precursor protein are located within buried core regions of the capsid proteins and are likely to be inaccessible to the protease (Ypma-Wong et al., 1988b). In addition, it has been reported that cleavage site sequences must be presented in a region of flexible structure, and an alanine at the -4 amino acid position (four residues upstream of the scissile bond) has been shown to be a determinant for efficient cleavage (Blair and Semler, 1991; Mirzayan et al., 1991; Pallai et al., 1989). In addition to cleavage of viral polyproteins, poliovirus 3Cpro has been shown to cleave the transcriptional activator protein, TATA binding factor, at a QG dipeptide, and this cleavage event may be a mediator of the shutoff of host cell transcription which occurs in poliovirus-infected cells (Das and Dasgupta, 1993). An important aspect of 3Cpro-mediated cleavages of the P1 capsid precursor protein is a requirement for the sequences of the 3DpoI protein in addition t o those of 3Cpro in the form of an uncleaved 3CD polyprotein (Jore et al., 1988; Ypma-Wong et al., 1988b). Studies have indicated that 3CD is a stable viral protein and not the precursor to 3Cpro and 3Dpo1; instead, a longer polyprotein, SABCD, is the likely precursor from which 3C and 3D are liberated (Lawson and Semler, 1992; Porter et al., 1993b). The nature of the requirement of the 3D portions of the 3CD protein for cleavage of the P1 precursor have not been defined but are speculated to potentially involve interaction of hydrophobic portions of the 3D domain with hydrophobic regions of the P1 precursor (Harris et al., 1992; Krausslich et al., 1990; Nicklin et al., 1988). A more recent study suggests that a host-cell factor may be involved in a 3CD-P1 processing complex required for efficient P1 precursor cleavage (Blair et al., 1993). The 3D sequences of 3CD are apparently most important for cleavage of P1 between VPO and VP3, as cleavage of P1 between VP3 and VP1 can be catalyzed in uitro by 3Cpro at enzyme concentrations much lower than those required for cleavage between VPO and VP3 (Krausslich et al., 1990; Nicklin et al., 1988).

D . Replication of Viral RNA The poliovirus virion does not contain any of the viral proteins required for replication of the viral genome, making translation of the genome a prerequisite to RNA replication. Proteins encoded in the P2 and P3 regions of the genome are required for the replication process, whereas proteins encoded in the P1 capsid region of the genome are

14

DAVID C. ANSARDI et al.

dispensable for RNA replication (W.-S. Choi et al., 1991; HadzopoulouCladaras et al., 1989; Hagino-Yamagishi and Nomoto, 1989; Kaplan and Racaniello, 1988). Mutants of many of the proteins encoded in the P2 and P3 regions are not complementable in trans,perhaps reflecting the requirement for some components of the replication machinery to remain associated with the template from which they were translated (Bernstein et al., 1986; Dewalt and Semler, 1987; Hagino-Yamagishi and Nomoto, 1989; Johnson and Sarnow, 1991). Despite years of study, the process of poliovirus RNA replication is not completely understood, although some insights have been gained from both in uitro and in viuo analyses (Kuhn and Wimmer, 1987; Paul et al., 1987b; Richards and Ehrenfeld, 1990). In addition to virally encoded proteins, there are cis-acting features of the poliovirus RNA genome required for replication, likely including the formation of secondary structures at the terminal regions of the RNA genome (Andino et al., 1990; Jacobson et al., 1970). The general strategy for poliovirus RNA replication is to first synthesize a full-length complementary strand (minus-strand RNA) to serve as template RNA molecules for the synthesis of progeny plus-strand RNA genomes. Initiation of synthesis of plus- and minus-strand RNA molecules requires recognition by the replication machinery of different 3' template ends, as the 3' ends of plus strands but not minus strands are polyadenylated (Larsen et al., 1980; Richards and Ehrenfeld, 1980). Replication of poliovirus RNA is associated with the formation of two types of fully double-stranded or partially double-stranded RNA. Completely double-stranded RNA in which a full-length plus strand is hybridized to a full-length minus strand is known as the replicative form (RF). A minus-strand RNA genome partially hybridized to a series of nascent plus strands concurrently being synthesized by different polymerase proteins on a single minus-strand template is a structure known as the replicative intermediate (RI) (see Fig. 2). IV. POLIOVIRUS VIRION A productive poliovirus infection must include synthesis of progeny virions and release of these virions from the host cell. The steps in these processes require formation of a capsid shell with icosahedral symmetry beginning with monomeric subunits and encapsidation of a single copy of a VPg-linked plus-strand RNA genome. The mature poliovirus virion serves many functions: it must protect the RNA genome from nucleases in the environment, it binds to a receptor protein on the surface of the host cell to initiate the infection process, and it

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

15

provides a mechanism for uncoating of the RNA genome on entry into the host cell.

A. Properties of Virion The poliovirus virion is a spherical particle with a diameter of approximately 30 nm (Putnak and Phillips, 1981b). The virion is composed of an icosahedral capsid shell formed by 60 copies of each of the four mature viral capsid proteins, VPl(306 amino acids, 33 kDa), VP2 (272 amino acids, 30 kDa), VP3 (238 amino acids, 26 kDa), and VP4 (69 amino acids, 7.5 kDa), and a single copy of the plus-strand RNA genome (Hogle et al., 1985). One or two copies of VP2 and VP4 may be present in the mature virion in the uncleaved precursor form, VPO (Jacobson et al., 1970), but the biological significance of the uncleaved VPO proteins in the mature virion has not been determined. The interior of the virion contains a single copy of the viral RNA genome linked to VPg (Wimmer, 1982). The poliovirus capsid does not contain sufficient basic amino acid residues to neutralize the negative charges of the RNA backbone and therefore packages numerous cations in addition to the RNA (Koch and Koch, 1985). The cations packaged include approximately 4900 K+ ions, 900 Na+ ions, 110 Mg2+ ions, and a few molecules of the polyamines putrescine and spermidine. Two types of lipid substituents are also present in the mature virion. The amino termini of the VP4 proteins are linked to a single molecule of the fatty acid myristate by an amide linkage (Caliguiri and Tamm, 1968; Chow et al., 1987; Page et al., 1988; Paul et al., 1987a). A second type of lipid, possibly sphingosine, occupies a hydrophobic pocket within the VP1 P-barrel core (Filman et al., 1989). The mature virion is a very stable structure and is resistant to concentrations of sodium dodecyl sulfate (SDS) as high as 1%,high salt concentrations, and exposure to acidic pH (Koch and Koch, 1985). The poliovirus capsid is less permeable than those of most other picornavirus members, a property which is reflected by the lower buoyant density (1.34 g/cm3) of the poliovirus virion in CsCl gradients. Other picornaviruses, such as rhinoviruses and cardioviruses, have higher densities in CsCl (1.4 g/cm3), reflecting their ability to uptake Cs+ ions, whereas the poliovirus virion is impermeable to Cs+ (Burness and Clothier, 1970; Mapoles et al., 1978; Medappa and Rueckert, 1974). The poliovirus virion has a sedimentation coefficient ( s ~ ~of, 155S, ~ ) a value typical for most picornavirus members (Putnak and Phillips, 1981a; Rueckert, 1990). The poliovirus virion also exists in a second type of conformation which is induced on attachment of the viral particle to the host cell glycoprotein receptor (Kaplan et al., 1990; Rueckert,

16

DAVID C . ANSARDI et a2.

1990). The sedimentation coefficient of the virus drops from 155s to approximately 135S, and the VP4 protein is released from the particle. This altered virion is more lipophilic than the 155s virion, possibly as a result of the externalization of the VP1 amino termini (DeSena and Mandel, 1976, 1977; Fricks and Hogle, 1990; Putnak and Philips, 1981b). The conformational changes of the altered virus are also marked by a conversion from a neutral pl in the native particle to a more acidic pl and by a change in the antigenic determinants displayed on the capsid to those resembling heated or denatured virus (Koch and Koch, 1985; Putnak and Philips, 1981a). The precise function of the altered virion in virus entry has not yet been determined.

B . Virus Structure The three-dimensional structure of the type 1poliovirus virion was solved in 1985 at a resolution of 2.9 A (Hogle et al., 1985).Structures of several other picornaviruses have also been solved, including human rhinovirus type 14, mengo virus, foot-and-mouth disease virus, and poliovirus type 3/Sabin (Acharya et al., 1989; Filman et al., 1989; Luo et al., 1987; Rossman et al., 1985). The poliovirus capsid exhibits the symmetrical qualities of a T = 3 icosahedron, with 180 major subunits comprising a complete shell which has five-, three-, and twofold axes of symmetry (Hogle et al., 1985; Rossman and Johnson, 1989). Because the three major proteins which make up the asymmetric unit of the capsid are nonidentical in sequence, the poliovirus virion is said to be a pseudo T = 3 capsid, or a P = 3 capsid. The three major capsid proteins, VP1, VP2, and VP3, have a high degree of structural similarity despite major differences in amino acid sequence. Five copies of VP1 surround each of the twelve fivefold axes of symmetry of the capsid, whereas VP2 and VP3 alternate around each of the twenty threefold axes of symmetry. Each of these proteins forms a P-barrel core structural domain characteristic of the structural proteins of most spherical viruses whose structures have been solved to date (Rossman and Johnson, 1989). This p-barrel structure is formed by eight strands of P-sheet structure arranged in an antiparallel fashion. The P-sheets are named alphabetically from B to I as they occur from the amino to carboxyl termini of the proteins and form a wedge or trapezoidal-shaped structure. The B, I, D, and G P strands are contained within a twisted p-sheet structure that forms the floor and one wall of the P-barrel. The C, H, E, and F strands form a smaller, flatter wall on the opposite side of the barrel from the B-I-D-G wall. The P strands are connected by loop structures that are designated by the two P strands they connect (e.g., the B-C or D-E loops). The P-barrel cores are flanked by two a

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

17

helices. In some cases, these loops extend outward from the core structure of the capsid proteins and are responsible for the surface features of the virus exterior. Whereas the basic core structure of the P barrel is similar for each of the major capsid proteins VP1, VP2, and VP3, the connecting loops differ markedly between the different proteins and account for their unique structures. The VP4 protein is much smaller than the other three capsid proteins and lies entirely on the interior of the viral capsid. The VP4 protein is essentially a continuance of the N-terminal arm of the VP2 protein, and it is released from VP2 either during or after RNA encapsidation. The VP4 protein is involved in the networking of capsid protein termini on the capsid interior and may also be involved in interacting with the RNA genome. The amino and carboxyl termini of capsid proteins VP1, VP2, and VP3 extend outward from the core structure. On the interior of the viral capsid, terminal portions of the capsid proteins form extensive networks responsible for linking the capsid proteins together. The most striking example of this networking occurs on the capsid interior at the fivefold axes of symmetry, where the amino termini from each of the fivefold related copies of VP3 intertwine and form an unusual P-sheet structure, the P annulus, which resembles a twisted tube and represents a conserved structure among the picornaviruses of known structure (Acharya et al., 1989; Hogle et al., 1985; Luo et al., 1987; Rossman et al., 1985; Rossman and Johnson, 1989). The P annulus is flanked by five copies of a short, two-stranded antiparallel P sheet formed by residues 3-8 and 25-29 of VP4. The amino-terminal glycine residue of VP4 is covalently linked to a myristate moiety by an amide linkage (Chow et al., 1987; Paul et al., 1987a). The myristate moieties from the fivefold related VP4 molecules form a hydrophobic cluster which mediates the interaction between the amino termini of VP3 and VP4 (Chow et al., 1987; Filman et al., 1989). A third P strand is formed from portions of the amino-terminal segment of VP1, extending the p-sheet structure toward the capsid interior (Filman et al., 1989). These networking structures likely play key roles in capsid integrity. The virus capsid contains determinants necessary for interacting with the host-cell receptor glycoprotein. The three-dimensional structure of the poliovirus virion revealed a depression or canyon on the surface encircling the fivefold axes of the capsid. This canyon structure is similar t o that seen for human rhinovirus type 14 and is analogous to a pit on the surface of mengo virus (Luo et al., 1987; Rossman et al., 1985). This inaccessibility of the surface of the floor of the canyon to antibody molecules led to the canyon hypothesis which suggested that the canyon floor contained the receptor binding sites for these viruses

18

DAVID C. ANSARDI et al.

(Luo et al., 1987; Rossman et al., 1985). The predominant idea behind this hypothesis was that the canyon was too narrow for the Fab portion of an antibody molecule to bind to residues lining the floor of the canyon, whereas the viral receptor might be a structure of lesser width able t o interact with the floor residues. This strategy would provide a mechanism by which the receptor binding domain of the virus could be maintained and escape immune surveillance. This hypothesis has been confirmed for human rhinovirus, and the structure of human rhinovirus type 16 complexed with extracellular domains of intracellular adhesion molecule (ICAM-11, the host receptor used by rhinovirus, has been determined (Olson et al., 1993). In addition to the receptor binding site, the exterior of the virus capsid contains the major antigenic determinants of the virus (Minor, 1990).Four major antigenic epitopes have been mapped on the surface of the virus by use of escape mutants. Three of the epitopes are present both on the intact virus and on subviral 14s pentamers (Page et al., 19881, whereas a fourth site is formed by the interaction between two pentamers and is present only in the completed, natively antigenic shell (Rombaut et al., 1990a). One of the antigenic sites, site 1, is composed of amino acid sequences in the loop connecting the B and C p strands of the VP1 protein core and has been the site of substitution of foreign antigenic determinants into the capsid to produce antigenic chimeras of poliovirus (Almond and Burke, 1990; Evans et al., 1989; Jenkins et al., 1990; Kitson et al., 1991).

C . Myristylation of Poliovirus Capsid Proteins N-Myristylated proteins are linked cotranslationally to a single molecule of the 14-carbon fatty acid myristate (n-tetradecanoic acid) by the enzyme N-myristoyltransferase (Towler et al., 1987; Wilcox et al., 1987). The myristylation reaction requires a glycine amino terminus, which is generated on the P1 capsid precursor on cleavage of the initiator methionine residue from the polyprotein (Dorner et al., 1982). The myristate moiety is linked to the glycine residue via an amide bond between the a-amino group of the glycine residue and the carbonyl carbon of the myristate molecule. The N-myristylated terminus becomes the amino terminus of the P1 capsid precursor on cotranslationa1 cleavage of the capsid precursor from the genomic polyprotein (Toyoda et al., 1986). Subsequent to cleavage of P1, the N-myristylated glycine is the amino terminus of capsid protein VPO, and finally becomes the amino terminus of VP4 on virion maturation. Electron density associated with the N-linked myristate chains was apparent in structural refinements of the poliovirus capsid (Chow et al., 1987; Filman et al., 1989).The five myristate moieties from within a common

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

19

pentameric subunit were found to cluster together, underneath the p annulus formed by the VP3 N-termini at the fivefold vertices of the capsid. The myristate cluster cradles the p annulus, implicating a role for the myristate moieties in capsid stability. The cotranslational N-myristylation of numerous other cellular and viral proteins has been reported (Schultz et al., 1988; Towler et al., 1988). The determinants for addition of myristate to a nascent peptide include a myristylation signal at the N terminus of the protein. This signal has an absolute requirement for glycine at the amino terminus of the protein and a preference for serine, alanine, or threonine at position 5 relative to the glycine acceptor (Towler et al., 1988). Substitution of the glycine residue with alanine completely abolishes myristate addition to the poliovirus P1 precursor (Krausslich et al., 1990; Marc et al., 1991). In the cases of many other viral and cellular N-myristyl proteins, the myristate moiety has been demonstrated to play a n important role in subcellular localization of the protein by contributing to a targeting signal which directs the protein to the plasmid membrane or to an intracellular membrane (Bryant and Ratner, 1990; Buss et al., 1989; Heuckeroth and Gordon, 1989; Johnson et al., 1990; Rhee and Hunter, 1987; Schult et al., 1988; Schultz and Rein, 1989; Towler et al., 1988). The myristate moiety may also participate in anchoring proteins within a lipid bilayer. The myristate moiety alone, however, is not sufficient to direct intracellular targeting to membranes (Rhee and Hunter, 1990), and several N-myristylated proteins are located in the cytosol (Schultz et al., 1988; Towler et al., 1988). By analogy to the properties of other N-myristylated proteins, the myristic acid moiety of the poliovirus capsid might participate in a targeting signal for directing capsid proteins to intracellular sites of assembly, or it may function as a membrane anchor for the capsid proteins (Chow et al., 1987; Paul et al., 1987a). A final possibility for a role for myristate in the poliovirus life cycle is at the point of uncoating. The hydrophobic myristate moieties might participate in interactions with endosomal membranes required for expulsion of the RNA genome across the membrane and into the cytoplasm. Direct analyses of a role for the myristate moieties of poliovirus in virus entry have been hampered by the complication that poliovirus mutants which do not encode a functional myristylation signal are nonviable (Krausslich et al., 1990; Marc et al., 1989, 1990).

V. MORPHOGENESIS OF POLIOVIRUS The morphogenesis of poliovirus has been a topic of intense study since the 1960’s. Much of the information gathered to date has relied

20

DAVID C. ANSARDI et al.

on studies conducted in uitro and has been largely confined to assembly of empty shells rather than RNA-containing virions because reconstitution of poliovirus virions in uitro from purified components has not been achieved (Putnak and Philips, 1981a; Rombaut et al., 1984,1991). The cumulative information from poliovirus morphogenetic studies has resulted in a hypothetical pathway for assembly, which is depicted in Fig. 3. The proposed steps of assembly include ordered proteolytic cleavages and formation of capsid protein subviral particles prior to RNA encapsidation and maturation of the virion (Putnak and Philips, 1981a; Rueckert, 1990). Briefly, the hypothesized order of these events is as follows: (i)cotranslational release of the 97-kDa P1 capsid precursor from the genomic polyprotein by an intramolecular proteolytic cleavage catalyzed by the 2A protease; (ii) cleavage of the P1 precursor to the individual capsid proteins VPO, VP3, and VP1, catalyzed by the 3CD polyprotein, the form of 3Cpr0 active on the P1 precursor; (iii) assembly of five 5s promoters [(VPO-3-1),] to form a 14s pentamer intermediate [(VP0-3-1),1; (iv) assembly of a 70-80s empty capsid or procapsid consisting of 60 copies of VPO, VP3, and VP1; (v) encapsidation of VPg-plus-strand RNA genome, proceeding either from a 1 4 s intermediate or from an empty capsid, to form a provirion [(VP0-3-1),,1 or immature virion; (vi) maturation of the virion by cleavage of VPO to VP2 and VP4, an event which is probably catalyzed intramolecularly. Each of these steps and pathway intermediates are discussed in the following sections.

A. 5s Protomer The 5s protomer, or (VP0-3-1),, is the smallest identical subunit from which the complete poliovirus capsid is built (VPO is uncleaved VP4 plus VP2). The protomer is derived by proteolytic cleavage of the P1 capsid precursor polyprotein after release from the genomic polyprotein by the 2Apm protease (Toyoda et al., 1986). Cleavage of the P1 precursor occurs at two glutamine-glycine dipeptides in the precursor to generate three proteins, VPO, VP1, and VP3, which have molecular masses of 37.4,33, and 26 kDa, respectively (Koch and Koch, 1985).As reviewed in Section II1,C,1, the glutamine-glycine cleavage sites are substrates for cleavages catalyzed by the virus-encoded enzyme 3Cpro (Hanecak et al., 1982). These cleavages have been demonstrated in uitro to be catalyzed more efficiently by the polyprotein, 3CD, which consists of uncleaved 3Cpm and 3Dpo1(Jore et al., 1988; Ypma-Wong et al., 1988a). Cleavage of P1 by purified 3Cpr0 can still occur in uitro,but cleavage between VPO and VP3 requires high enzyme concentrations (Krausslich et al., 1990; Nicklin et al., 1988).

POLIOVIRUS ASSEMBLY AND RNA ENCAF'SIDATION

21

FIG.3. Proposed pathways of poliovirus assembly. The poliovirus capsid proteins are initially translated as part of the genomic polyprotein and are released by a n autocatalytic cleavage by viral protease 2A as a 97-kDa precursor designated P1. The viral polyprotein 3CD catalyzes cleavage of the P1 precursor at two glutamine-glycine amino acid pairs to generate capsid proteins VPO, VP3, and VP1. The three proteins derived from a common precursor are believed to remain associated, comprising a 5s protomer subunit. Five protomers assemble 14s pentamer subviral particles [(VPO-3-1),1, which are believed to be virion precursors. Twelve pentamers assemble 755 empty capsid (procapsid) particles [(VPO-3-1),,], which some studies suggest are the direct virion precursor, with virion formation proceeding by condensation of twelve pentamers around a nucleating RNA genome. On RNA encapsidation, VPO is cleaved at a n asparagine-serine amino acid pair, releasing VP2 and VP4, a 69-amino acid protein located on the interior of the virion. The mature virion may be directly preceded by a provirion intermediate (not shown) in which the RNA genome has been encapsidated in a complete VPO-3-2 capsid.

The less efficient cleavage reaction at the VPO-VP3 bond has in part been attributed to the primary sequence near the cleavage sites. An alanine residue is present in the -4 position relative t o the scissile

22

DAVID C. ANSARDI et al.

bond at the site between VP3 and VP1, and the presence of alanine at the -4 position has been demonstrated to be a determinant for site recognition by 3Cpro and 3CD (Blair and Semler, 1991; Pallai et al., 1989). In contrast, the -4 position relative to the VPO-VP3 bond is a proline residue. The unfavorable -4 position residue also affects 3CDcatalyzed cleavage at the VPO-VP3 site (Blair et al., 1993). Substitution of this proline residue with an alanine improves cleavage by 3CD in uitro, perhaps by alleviating the requirement for a cellular cofactor to facilitate cleavage at that site (Blair et al., 1993); however, the substitution is lethal for virus growth when introduced into a poliovirus mutant RNA genome. The molecular nature of the requirement of 3D sequences to catalyze efficient cleavage of P1 to VPO, VP3, and VP1 is not understood, but speculations have been made that hydrophobic regions of 3D interact with hydrophobic regions of P1 to promote enzyme-substrate interaction (Harris et al., 1992; Krausslich et al., 1990; Nicklin et al., 1988). This hypothesis is based on the observation that 3CD activity on P1 in uitro is reduced in the presence of nonionic detergent. Speculation that the myristate molecule linked to the amino terminus of VPO might be involved in these hydrophobic interactions arose after separate studies found that 3CD did not efficiently cleave nonmyristylated P1 in uitro (Krausslich et al., 1990; Marc et al., 1989). This question has been addressed by our laboratory using an intracellular system to study proteolytic processing of P1 precursors by the 3CD enzyme (discussed in Section VI1,B). After cleavage of P1 to VPO, VP3, and VP1, the three individual proteins generated from a single precursor most likely remain associated as a 5s protomer subunit (Bruneau et al., 1983). The individual capsid proteins are always found to sediment in sucrose gradients at a 5s position, and free forms of these proteins have not been detected. Formation of the 5s protomer from the uncleaved P1 precursor is likely associated with significant conformational changes in the protomer (Hogle et al., 1985). The amino and carboxyl termini freed from one another by proteolytic cleavage are located on opposite sides of the promoter in the mature virion, indicating that structural rearrangements occur following cleavage. These structural rearrangements may be required to activate the domain responsible for the next step in assembly: formation of the 14s pentamer (Hogle et al., 1985; Rueckert, 1990).

B . 14s Pentamer On formation of the 5s protomer, the capsid subunits rapidly assemble into 14s pentamer structures consisting of five copies of each of the

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

23

individual proteins VPO, VP3, and VP1 (Putnak and Phillips, 1981a; Watanabe et al., 1962). In poliovirus-infected cells incubated with radiolabeled amino acids, incorporation of label into 14s pentamers has been reported to occur as rapidly as within 7-10 min (Putnak and Phillips, 1981b). A point of controversy in picornaviral assembly has been whether 14s pentamer formation occurs prior or subsequent to P1 precursor cleavage (Putnak and Phillips, 1981b).Evidence has been presented for some picornavirus members, including encephalomyocarditis virus (EMCV)and rhinovirus, that P1 precursors assemble to form a 13.4s pentamer precursor prior to proteolytic cleavage (McGreggoret al., 1975; McGreggor and Rueckert, 1977). The 13.4s pentamer might then be converted t o a 14s pentamer structure on cleavage of the pentamerized precursors. It has also been proposed that formation of P1 pentamers precedes proteolytic processing of the P1 precursor for hepatitis A virus (Borovec and Anderson, 1993). Information from the threedimensional structure of various picornaviruses, however, casts doubt on the likelihood of this pathway (Acharya et al., 1989; Hogle et al., 1985; Luo et al., 1987; Rossman et al., 1985). The amino termini of five VP3 proteins within a common pentamer subunit of the capsid interact with one another, forming the @-annulus structure near the 6-fold axes of symmetry. The @-annulusstructure likely provides stabilizing interactions required for pentamer formation. Proteolytic cleavage of the P1 precursor at the site between VPO and VP3 is required to free the VP3 amino termini from the carboxyl end of VPO. Unless the interprotomer interactions that occur within a pentameric subunit of the mature virion are different from those in the subviral pentamer, cleavage of the precursor would appear to be a prerequisite for assembly. In studies of cell-free assembly of in vitro translated EMCV capsid proteins, complete cleavage of the precursor was required for pentamer formation (Palmenberg, 1990; Parks and Palmenberg, 1987). The assembly of a pentamer as a capsid precursor is compatible with the notion that construction of an icosahedral capsid from monomeric subunits requires a stepwise assembly process, with formation of one building block required to activate the domains necessary for assembling the next intermediate structure (Caspar and Klug, 1962; Rossman and Johnson, 1989).A controversy exists, however, about whether the pentamer is the direct precursor to the poliovirus virion (Putnak and Philips, 1981a; Rueckert, 1990). A few lines of evidence suggest that 1 4 s pentamers are the immediate precursor to the virion. In pulse-chase metabolic radiolabeling experiments using poliovirusinfected cells, radiolabel flows from 5s protomers into 14s pentamers and into both empty capsids and virions (Jacobson and Baltimore,

24

DAVID C. ANSARDI et al.

1968). Experiments in which the drug guanidine was used to inhibit poliovirus RNA replication in poliovirus-infected Mi0 cells provided evidence that 14s pentamers are direct precursors to the virion (Ghendon et al., 1972). When RNA replication is halted by guanidine treatment, virion formation also abruptly halts. Under guanidine treatment conditions, capsid protein radioactivity in Mi0 cells was found to accumulate in 14s pentamers. On removal of the guanidine, the 14s pentamer radioactivity was rapidly converted to virions without formation of detectable empty capsids. An additional line of experimental evidence has supported the hypothesis that the 14s pentamer is the direct precursor to the virion (Rombaut et al., 1990b). In cells infected with poliovirus at 30"C, radioactivity in radiolabeled capsid proteins was found to accumulate in 14s pentamers without formation of empty capsids or RNA-containing virions. On shift of temperature to 37"C, a temperature permissive for virion and empty capsid assembly, radioactivity in the 14s pentamer fractions was rapidly chased into mature virions without significant accumulation of an empty capsid intermediate. The investigators could not, however, rule out the possibility that RNA encapsidation was occurring so rapidly that an empty capsid intermediate was obscured. This explanation might also account for the lack of detection of an empty capsid intermediate in the guanidineinhibition studies in Mi0 cells (Ghendon et al., 1972). Electron microscope immunocytochemistry studies have offered further evidence that the 14s pentamer is the direct precursor to the virion (Pfister et al., 1992). These studies were conducted with subcellular fractions containing virus-induced smooth membrane vesicles associated with replication complexes. By using monoclonal antibodies specific for subsets of capsid protein structures, 14s pentamers were detected around the peripheries of the replication complexes in association with the membrane vesicles. In contrast, empty capsids could not be detected in association with the complexes by these methods. The hypothesis was made that 14s pentamers associate with the replication complexes and interact with pools of nascent RNA chains being released from the replication complexes (Troxler et al., 1992). Interestingly, solubilization of membrane-associated replication complexes with nonionic detergents resulted in conversion of 14s pentamers to natively antigenic empty capsids. The investigators suggested that linkage to a membrane support prevents 14s pentamers from coalescing into a capsid until interaction with RNA takes place. When the membranous support is dissolved with nonionic detergents, the 14s pentamers may rapidly assemble empty capsids, opening the possibility that empty capsids previously reported to be associated with replication complexes may actually be artifacts produced on lysis of the host cell and solubilization of membrane-associated pentamers.

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

25

C. Empty Capsid The empty capsid or procapsid has probably been the most controversial intermediate in the proposed assembly pathway. Empty capsids are composed of sixty copies of each of the individual capsid proteins VPO, VP3, and VP1 (Putnak and Philips, 1981a; Rueckert, 1990). In sucrose density gradients, the empty capsid sediments at a rate approximately one-half that of mature virions, and it has a sedimentation coefficient reported t o be between 65s and 80s (Rueckert, 1990). In addition, the empty capsid can exist in two different conformations that are distinguishable by the antigenic epitopes they display and their relative stabilities. One type of empty capsid is very labile and displays the same antigenic epitopes as native virus (N-antigenicity) and is also referred to as the natural empty capsid (Gauntt eta,?.,1981; Maronginu et al., 1981; Putnak and Phillips, 1982; Rombaut et aZ., 1982, 1984; Rueckert, 1990). The other type of empty capsid is much more stable and has antigenicity consistent with heated poliovirus virions, which display a completely different subset of antigenic epitopes (H-antigenic) (Maize1et al., 1967).The labile empty capsid can be dissociated under mild alkaline conditions and is rapidly converted to the H-antigenic form if heated even briefly after extraction from the infected cell (Maronginu et al., 1981; Onodera et al., 1986). An additional difference in the two types of empty capsids is their reported sedimentation velocities. Native antigenic empty capsids are reported to sediment in sucrose gradients at a 65-708 position, whereas the more stable H-antigenic empty capsids sediment at a position of 8 0 s (Putnak and Phillips, 1982; Rombaut et al., 1982). These properties may reflect a more condensed capsid structure in the stable H-antigenic particle (Koch and Koch, 1985). In the course of studies of poliovirus assembly, the empty capsid has been proposed in conflicting hypotheses to be the direct precursor to the virion (Jacobson and Baltimore, 1968),a by-product of assembly in which excess pentamers assemble empty capsids (Koch and Koch, 19851, and an artifact of solubilization methods used to analyze subviral particles by their sedimentation properties (Pfister et al., 1992).A major problem with acceptance of the empty capsid as the direct precursor to the lririon is a conceptual one because envisioning how a 7450-base RNA genome can be tightly wound and threaded into a preformed empty shell is difficult (Putnak and Phillips, 1981a; Rueckert, 1990). Nevertheless, numerous experiments have been presented which suggest that the empty capsid, or procapsid, is the direct precursor to the virion. Poliovirus 14s subunits can self-assemble empty capsids in vztro in the absence of full-length poliovirus RNA (Onodera and Phillips, 1987;

26

DAVID C. ANSARDI et al.

Phillips, 1971; Phillips and Wiemert, 1978; Rombaut et al., 1991).This inherent ability of 14s pentamers to form empty capsids was taken as evidence for their precursor role in virion morphogenesis (Putnak and Phillips, 1981b). In some experiments, the radiolabel in poliovirusinfected cells appeared in empty capsid particles (15-20 min) before appearing in mature virions (20-30 min), and this observation was taken as evidence for a precursor role for empty capsids in virion morphogenesis (Putnak and Phillips, 1981b).In addition, pulse-chase radiolabeling experiments conducted using cells infected with footand-mouth disease virus (FMDV) in the presence of protein synthesis inhibitors indicated a flow of radioactivity from 5s to 14s to empty capsids to virions (Yafal and Palma, 1979).Radiolabeling experiments conducted in the presence of guanidine in poliovirus-infected HeLa cells showed that radiolabel accumulated in empty capsids quickly on inhibition of RNA replication, and the empty capsid-associated radiolabel was rapidly chased into virions on removal of the drug (Fiszman et al., 1972; Jacobson and Baltimore, 1968). This experiment was similar to that conducted in Mi0 cells in which radioactivity accumulated in 14s pentamers in the presence of guanidine was rapidly chased into virions on removal of the inhibitor (Ghendon et al., 1972). Two early studies conclude that poliovirus empty capsids were associated with the viral RNA replication complexes (Caliguiri and Compans, 1973; Yin, 1977);however, those findings have been challenged more recently as being an artifact of solubilization methods used to extract capsid particles from the membranous complexes (Pfister et al., 1992). The inherent ability of 14s pentamers to assemble empty capsids has been studied extensively in uitro (Phillips, 1969, 1971; Phillips et al., 1968, 1980; Phillips and Wiemert, 1978; Putnak and Phillips, 1981a; Rombaut and Boeye, 1991; Rombaut et al., 1984, 1991). On incubation at 37"C, purified 14s pentamers isolated from poliovirusinfected cells assemble empty capsids (Phillips, 1971; Phillips and Wiemert, 1978; Rombaut et al., 1991). Attempts to reconstitute virions in uitro from purified virion RNA and 14s pentamers or empty capsids isolated from poliovirus-infected cells have failed (Putnak and Phillips, 1981b; Rombaut and Boeye, 1991). In in uitro assembly experiments, purified 14s pentamers at sufficient concentrations assemble empty capsids in the absence of additional poliovirus-specific factors, but the empty capsids which form are H-antigenic (Putnak and Phillips, 1982). When 14s pentamers were incubated in the presence of a poliovirus infected-cell extract, however, empty capsids were assembled much more rapidly, and the resulting empty capsids displayed native antigenic epitopes (Phillips, 1969; Putnak and Phillips, 1981b, 1982; Rombaut et al., 1984). These findings led to the search for the morphopoietic factor present

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

27

in poliovirus-infected cells which facilitated assembly of empty capsids from 14s pentamers and which conferred native antigenicity on the in uitro assembled shells. Although the source of the assembly-promoting activity originally was not believed to be simply the endogenous supply of 1 4 s pentamers provided by the infected cell extracts (Phillips, 1969; Putnak and Phillips, 1981b), extracts of cells infected with poliovirus defective interfering particles, which do not encode functional capsid proteins (discussed more extensively in Section VI,B), lacked the assembly-promoting activity (Phillips et al., 1980). The source of the assembly-promoting activity has been identified by Rombaut et al. (19911, who demonstrated a threshold concentration (-1.6 nM) above which purified 14s pentamers rapidly assembled empty capsids in uitro when incubated at 37°C.The ability of infected cell extracts to facilitate the assembly of 14s pentamers at concentrations below the assembly threshold was directly correlated with the supply of 1 4 s pentamers provided in the infected cell extract which brought the final concentration of 14s subunits above the assembly threshold. These observations demonstrated that the assembly-promoting activity was simply the additional 14s pentamers provided by the extract. However, the factor contributing to native antigenicity of empty capsids assembled in the presence of an infected cell extract appears to be different. The VP1 core p barrel contains a hydrophobic pocket normally occupied by an unidentified lipid molecule, probably sphingosine (Filman et al., 1989). This pocket is analogous to the pocket in human rhinovirus VP1 which binds a series of candidate antiviral drugs known as WIN compounds, which act at multiple levels in preventing infection of cells by drug-complexed virions (Badger et al., 1988; Fox et al., 1986; Smith et al., 1986). The WIN compounds have also been demonstrated t o inhibit poliovirus uncoating (Fox et al., 1986) and to protect poliovirus N-antigenic empty capsids from thermal denaturation (Rombaut and Boeye, 1991). Rombaut et al. (1991) found that purified 14s pentamers assembled empty capsids with H-antigenicity in uitro. However, if the drug molecule disoxaril, a WIN compound, was provided in the assembly reactions, the empty capsids which assembled displayed native antigenic epitopes (Rombaut and Boeye, 1991). The authors speculated that the drug mimics a lipid compound provided by the infected cell extracts by binding in the VP1 pocket and promoting native antigenicity of the assembled empty capsid particles. The assembly-enhancing features of infected cell extracts were thus twofold: supply of 1 4 s pentamers to bring concentrations above the threshold required for assembly and provision of some compound, possibly a lipid molecule which was mimicked by the drug disoxaril, to maintain native antigenicity. The existence of different forms of the empty capsid (Nand H-anti-

28

DAVID C. ANSARDI et al.

genic) has led to different interpretations about the role the empty capsid plays in morphogenesis. When extracted from infected cells, empty capsids were originally reported to be very stable, H-antigenic structures (Maize1et al., 1967).Because of this property, empty capsids were thought to not be capable of equilibrating with 14s pentamers. Thus, Jacobson and Baltimore (1968) proposed the procapsid hypothesis in which the viral genome is directly inserted into a procapsid. Subsequent studies have shown, however, that empty capsids in uiuo likely exist in a natively antigenic, dissociable state (Maronginu et al., 1981). The observance of H-antigenic empty capsids probably reflects handling methods, as natively antigenic empty capsids are rapidly thermally denatured (Maronginu et al., 1981). Rapid thermal denaturation of 14s pentamers also occurs in uitro on incubation at 37°C (Rombaut and Boeye, 1991). The lability of the N-antigenic empty capsid suggests that interconversion between 14s and empty capsid forms may occur in uiuo (Rueckert, 1990). This property is consistent with a model in which empty capsids are a storage depot for excess pentamers which can readily dissociate back into 14s pentamers (Maronginu et al., 1981; Rueckert, 1990). The lability of the empty capsid might also reflect, however, a more flexible structure which can uptake RNA (Koch and Koch, 1985). Several methods for how the RNA genome could be inserted into an empty capsid have been proposed. One model suggests that the energy released during synthesis of the RNA genome provides the driving force for inserting the RNA into the capsid (Rueckert, 1990). Such a strategy implies a very close link between RNA synthesis and encapsidation, which is supported by studies which have noted that guanidine inhibition of RNA synthesis is associated with a concurrent inhibition of RNA encapsidation (Caliguiri and Tamm, 1968; Fiszman et al., 1972). Another hypothesis suggests that empty capsids may not have a full complement of capsid subunits and may contain holes available for inserting an RNA genome, a hypothesis primarily based on studies of mengo virus assembly in which empty capsid structures were believed to contain 10 rather than 12 pentameric subunits (Lee and Colter, 1979). Other hypotheses have from time to time been based on more elusive intermediates that may exist in poliovirus-infected cells. A few reports have offered evidence for the transient formation of half-shells, with sedimentation coefficients of approximately 50s (Corrias et al., 1987; Koch and Koch, 1985; Lee et al., 1978;Rombaut et al., 1985). A half-shell might possibly serve as a direct virion precursor, with the RNA being enclosed within the two halves. Interestingly, electron microscopy studies of subcellular fractions containing poliovirus replication complexes have identified structures attached to the

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

29

membrane vesicles which had a half-shell appearance (Pfister et al., 1992).

D . Provirion Whatever the capsid precursor serving as the RNA-binding particle may be, the RNA encapsidation step appears to lead to production of a provirion particle in which the RNA genome has been encapsidated in a completed shell composed of 60 copies of VPO, VP3, and VP1 (Putnak and Phillips, 1981a; Rueckert, 1990). The original evidence for the existence of this intermediate was based on sedimentation studies in which a 125s shoulder was observed on the 155s virion peak on sucrose density gradients (Fernandez-Thomas and Baltimore, 1973; Fernandez-Thomas et al., 1973). The capsid protein composition of this shoulder was found to be enriched for VPO over the 155s peak, in which most if not all VPO had been cleaved to VP2 and VP4. Subsequent studies reported the sedimentation coefficient of these provirions or immature virions to be 150s (Guttman and Baltimore, 197713). These findings led to the speculation that cleavage of VPO to VP2 and VP4 occurred subsequent to RNA encapsidation and possibly by an intramolecular mechanism. l ' b o studies have provided further evidence for the existence of the provirion intermediate. Compton et al. (1990) isolated a temperaturesensitive mutant of poliovirus with a glutamine substitution for arginine at residue 76 of VP2 which accumulated provirions at the nonpermissive temperature. These studies showed that RNA encapsidation and VPO cleavage to VP2 and VP4 could be unlinked. The resulting provirion particles, however, were not infectious. A subsequent study of site-directed mutants of human rhinovirus type 14 has more thoroughly characterized the provirion particle (Lee et al., 1993). Rhinovirus mutants with a threonine substitution for asparagine at the carboxyl terminus of VP4 (at the cleavage site between VP4 and VP2) accumulated provirion particles in cells transfected with an in uitro transcribed RNA genome encoding the substitution. The provision particles were shown to be noninfectious, and the lack of infectivity was traced to a step in the uncoating process of rhinovirus. Provirion particles attached to host receptors normally and underwent the associated conformational changes (155s to 125s conversion). The block with the provirion mutant appeared to occur at the level of RNA release, leading to the hypothesis of an infectosome intermediate in the uncoating pathway in which a membrane-associated virus particle expels its RNA across an endosomal membrane and into the cytosol. Formation of this structure is apparently dependent on cleavage of VPO t o VP2

30

DAVID C. ANSARDI et al.

and VP4. At the other end of the spectrum, why the virus has evolved to delay cleavage of VP4 from VP2 until after RNA encapsidation takes place is not known. Intact VPO may be needed to maintain the required conformations of the 5s and 14s capsid subunits for subsequent assembly events (Koch and Koch, 1985). The maturation cleavage of VPO to VP2 and VP4 is the final proteolytic cleavage in the maturation of poliovirus capsid proteins (Arnold et al., 1987; Hellen and Wimmer, 1992a,b).Following the solution of the three-dimensional structures of several picornaviruses, a potential mechanism for how maturation cleavage might occur was proposed (Arnold et al., 1987). In this autocatalytic model, a serine residue in VP2 (amino acid number 10 in VP2), which forms a hydrogen bond with the carboxyl terminus of VP4 in the mature virion, was believed to be the residue responsible for nucleophilic attack on the peptide bond. Because a nearby histidine residue, which would serve as the proton-abstracting base for the nucleophilic attack, was not present, a nitrogenous base from the RNA molecule was speculated to activate the serine residue for nucleophilic attack. This model provided a convenient explanation for how the maturation cleavage event was dependent on RNA encapsidation since RNA would be required to complete the catalytic triad of the protease. This theory was disproved, however, when Harber et al. (1991) demonstrated that the putative catalytic serine residue could be substituted with other amino acids without affecting the maturation cleavage event. Despite the collapse of this model, a role for the RNA genome in contributing to the catalytic site of the intramolecular protease has not been ruled out.

VI. RNA ENCAPSIDATION PROCESS As discussed in the preceding sections, the precise pathway leading to the formation of poliovirus virions is a major unresolved question in poliovirus morphogenesis. Not only is the identity of the direct capsid precursor to the virion not known, but the mechanisms involved in capsid protein-RNA genome interaction are also not well understood. The two components of this interaction, the capsid protein determinants involved in RNA binding and the regions of RNA specifically recognized by the capsid proteins, have not been identified. The threedimensional structure of poliovirus provided few clues about this interaction, as the encapsidated RNA molecule does not adopt the icosahedral symmetry of the capsid shell (Hogle et al., 1985).Because the VPg-linked RNA molecules exist in multiple conformations within a

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

31

crystal lattice of poliovirus virions, structural determinations for the RNA genome and VPg could not be made.

A . RNA Requirements for Encapsidation Poliovirus virions encapsidate only plus-strand VPg-linked genomic RNA (Lee et al., 1977; Novak and Kirkegaard, 1991; Wimmer, 1982). Beyond these characteristics of encapsidated RNA, few other determinants required for packaging of a poliovirus RNA molecule have been recognized. Most information regarding the RNA requirements for encapsidation has come from studies of defective interfering (DI) particles of poliovirus (Cole, 1975). Studies of the naturally occurring DI genomes, which contain in-frame deletions within the P1 coding portion of the genome, indicate that shorter RNA genomes can be encapsidated and have suggested a minimal size constraint for encapsidation of 80-87% the length of the wild-type genome (Cole et al., 1971; Kuge et al., 1986; Lundquist et al., 1979). A poliovirus RNA genome 108% of the length of the wild-type genome, namely, a genetically engineered dicistronic RNA genome containing an IRES element of EMCV inserted between the P1 and P2 genes, has been demonstrated t o be compatible with virion formation, indicating that the virus can accommodate a lengthier RNA molecule (Molla et al., 1992). Beyond the ability of the poliovirus capsid to accommodate genomes of different sizes, and the requirement for VPg linkage for encapsidation, few other properties of poliovirus RNA necessary for encapsidation have been uncovered.

B . Poliovirus Defective Interfering Particles Cole et al. (1971) were the first to describe the appearance of DI particles within populations of poliovirus passaged at very high multiplicities of infection [>200 pfu (plague-forming units)/celll. The DI particles were first identified by their slower sedimentation properties in sucrose density gradients and were then shown to have lower buoyant densities in CsCl density gradients relative to wild-type virus (1.31-1.325 g/cm3 versus 1.34 g/cm3 for wild-type). The DI particles were found t o exhibit properties of interference with wild-type poliovirus production in mixed infections and were shown to enrich in proportion to wild-type virions on multiple passages (Cole and Baltimore, 1973b,c). Deletions within the DI genomes were mapped to the 5’ region of the genome and were believed to have limitations in minimal size permissible for propagation, with the smallest naturally oc-

32

DAVID C. ANSARDI et al.

curring DI genomes identified having approximately 80%of the length of wild-type genomes (Cole and Baltimore, 1973a; Cole et al., 1971). Subsequently, other investigators reported separate generation of DI particles in populations of poliovirus passaged at high multiplicities of infection (Kajigaya et al., 1985; Lundquist et al., 1979). Electron microscopy studies provided additional evidence that these genomes contained deletions in the 5’ region of the RNA genome in the region believed to encode the capsid proteins (Lundquist et al., 1979). Determination of the nucleic acid sequences of several DI genomes of poliovirus type 1 Sabin confirmed that DI genomes contained deletions in the P1 capsid region which maintained the translational reading frame for the P2 and P3 regions of the genome (Kuge et al., 1986). Poliovirus DI genomes containing deletions in the P2 or P3 regions have never been identified, reflecting the property that replication of genomes encoding mutations of P2 and P3 region proteins is not readily complementable by viral proteins provided in trans (Kuhn and Wimmer, 1987; Page et al., 1988; Paul et al., 1987b; Richards and Ehrenfeld, 1990). The P1 deletions characterized by sequence analysis appeared to have specific boundaries within the P1 gene at both the 5‘ and 3’ ends, as the naturally occurring in-frame deletions were all contained within an internal segment of the P1 gene between nucleotides 1226 and 2705, encompassing much of the VP2 and VP3 genes (Kuge et al., 1986).

C . RNA Encapsidation Signals The finding that portions of the P1 gene were maintained in every isolate of naturally occurring DI genomes of poliovirus type 1 Sabin led to the speculation that portions of the P1 coding region might contain cis elements required for RNA replication and/or encapsidation of the genome (Kuge et al., 1986). Kaplan and Racaniello (1988) generated in uitro transcribed poliovirus RNA genomes which contained genetically engineered deletions in the P1 gene, encompassing all but the final 320 nucleotides of the P1 gene. On transfection into HeLa cells, the deletion-containing genomes replicated normally, indicating that most, if not all, of the P1 gene is dispensable for replication of the RNA genome. The investigators did not report on whether the deletion-containing RNA genomes could be encapsidated if transfected into cells infected with wild-type helper poliovirus. Another report has demonstrated that sequences at the 5’ end of the P1 coding region are dispensable for both RNA replication and encapsidation. These studies demonstrated replication and encapsidation of a poliovirus RNA replicon containing a reporter chloramphenicol

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

33

acetyltransferase (CAT) gene inserted for part of the P1 gene, beginning with the AUG codon for translation initiation (Percy et al., 1992). The CAT gene replaced P1 gene sequences from nucleotides 756 to 1805, indicating that an encapsidation signal does not exist in the 5' portion of the P1 gene. Further evidence that P1 regions are dispensable for RNA replication was provided by W.-S. Choi et al. (1991), who showed that internal regions of the P1 gene could be substituted with foreign gene segments encoding human immunodeficiency virus type 1 (HIV-1) proteins in the same translational reading frame as the poliovirus polyprotein. Experiments have demonstrated that RNA genomes with foreign genes substituted for the complete P1 gene can be encapsidated (Ansardi et al., 199413; Porter et al., 1995). Interestingly, the presence of nucleotides 743-959, which encompass the VP4 gene, appeared to facilitate encapsidation of the replicon RNA, pointing to the possibility that this region of the poliovirus genome might be involved in encapsidation after all (Porter et al., 1995).

D . Subcellular Location

of

Encapsidation

Successful encapsidation of poliovirus RNA might require interaction of the capsid proteins with the RNA at a specific subcellular location. Poliovirus RNA replication occurs in replication complexes associated with smooth intracellular vesicles, and capsid proteins of poliovirus have also been found in association with smooth vesicles as discussed in Section I11 (Caliguiri and Compans, 1973; Caliguiri and Mosser, 1971; Ehrenfeld et al., 1970; Girard et al., 1967; Hewlett et al., 1976). In RNA-labeling experiments conducted using short pulses of incubation with PHluridine, virions associated with the smooth membrane fractions were found to have higher specific activity than those found in other subcellular fractions (Caliguiri and Compans, 1973), implying that the most recently made virions were associated with the smooth membrane complexes. Immunoelectron microscopy studies have demonstrated the presence of capsid-related particles, probably 1 4 s pentamers, associated with the peripheral membrane vesicles of replication complexes isolated from poliovirus-infected cells (Hewlett et al., 1976). Poliovirus capsid proteins, in a precursor form to virions, may possibly be directed to and associate with intracellular membrane vesicles in a location required for interaction with newly synthesized RNA genomes (Hewlett et al., 1976; Koch and Koch, 1985).The mechanisms by which capsid proteins associate with the membranes is not understood. One hypothesis suggests that the myristate molecule linked to the amino terminus of VPO mediates association with intracellular membranes (Chow et al., 1987; Paul et al., 1987a). The lipo-

34

DAVID C. ANSARDI et al.

philic amino terminus of capsid protein VP1 might represent another candidate determinant for capsid protein association with intracellular membranes (Filman et al., 1989).

VII. NEWMETHODS TO STUDY POLIOVIRUS ASSEMBLYPROCESS Until recently, much of the information about picornavirus assembly was gathered from attempts to reconstitute the assembly process in uitro (Putnak and Phillips, 1981a). Although empty capsids assemble from 14s pentamer subunits in uitro, the formation of virions from purified components has not been achieved. Molla et al. (1991)reported on the de nouo synthesis of poliovirus in uitro. This system relied on in uitro translation of poliovirus proteins from full-length genomic RNA in the presence of intracellular membranes which in turn resulted in replication of the RNA genome and encapsidation of RNA to form infectious poliovirions. Other methods of studying the poliovirus assembly process have relied on isolation of temperature-sensitive mutant polioviruses or on the recovery of mutant viruses on transfection of in uitro transcribed RNA genomes containing site-directed mutations (Comptonet al., 1990; Kirkegaard, 1990; Kirkegaard and Nelson, 1990; Marc et al., 1990; Moscufo and Chow, 1992; Moscufo et al., 1991; Reynolds et al., 1992). An inherent problem exists in characterizing these types of poliovirus mutants because the poliovirus replicase, with no known editing capabilities, is prone to error, and reversions of mutations arise with great frequency (De La Torre et al., 1992). RNA genomes encoding capsid mutations replicate normally since capsid proteins are dispensable for replication (W.-S. Choi et al., 1991; Hagino-Yamagishi and Nomoto, 1989; Kaplan and Racaniello, 19881, so opportunity for reversion of mutations in the capsid gene is great. Thus by transfecting in uitro transcribed RNA genomes into cells and recovering mutant viruses, it becomes difficult to assess definitively whether intermediate phenotypes observed are a reflection of populations of revertants (Marc et al., 1990, 1991). In addition, it was also difficult to recover enough material from the transfected cells to characterize the physical features of the subviral particles thoroughly.

A . Studies of Poliouirus Assembly Process Using Recombinant Vaccinia Viruses To understand further the molecular details of poliovirus assembly, it was critical to develop an intracellular system in which the early

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

35

events in the assembly of poliovirus could be studied without having to depend on recovery of mutant polioviruses or on expression of capsid proteins from a replicating template, which had the potential to revert a capsid gene mutation. Vaccinia virus vectors have several features which make them attractive for the expression of poliovirus proteins (Mackett et al., 1985). Among these are the following: the cytoplasmic site of vaccinia virus replication ensures that messenger RNA molecules encoding poliovirus proteins are not exposed to nuclear splicing machinery; the vaccinia virus genome is capable of accepting large amounts of foreign DNA; and generation of recombinant vaccinia viruses is greatly facilitated by recombination plasmids that direct homologous recombination of foreign genes into the thymidine kinase gene of the vaccinia virus genome, thereby providing a mode of selection because the resulting recombinants do not synthesize thymidine kinase. Finally, the recombination plasmid coexpresses P-galactosidase, providing another selection marker for recombinant viruses (Chakrabarti et al., 1985).

B . Expression of Poliovirus PI and 3CD Using Recombinant Vaccinia Virus Vectors Previous studies demonstrated that stable recombinant vaccinia viruses could not be isolated which contained the poliovirus 2A gene (Jewel1 et al., 1990; Turner et al., 1989). The lethal effect of 2A~r0was probably associated with its role in shutting off translation of capped mRNA molecules (Etchison et al., 1982; Krausslich et al., 1987; Lloyd et al., 1988; Wycoff et al., 1990). This observation was important because the carboxyl terminus of the P1 precursor is generated by a cisacting proteolytic cleavage by 2Apr0 (Toyoda et al., 1986). To overcome the need for 2A-mediated cleavage to generate an authentic P1 carboxyl terminus, termination codons were engineered into a recombinant P1 gene downstream of the codon for the authentic tyrosine carobxyl terminal residue (Ansardi et al., 1991). Infection of cells with recombinant vaccinia virus that contains the P1 gene (VVP1) resulted in expression of a 97-kDa protein. Coinfection of cells with VVPl and a second recombinant vaccinia virus, VVP3, which expressed the 3CD protein (Porter et al., 1993b),resulted in expression of both P1 and 3CD in coinfected cells. The P1 precursor was rapidly cleaved to VPO, VP3, and VP1 by the 3CD protease (Fig. 4). In addition, these cleavage products assembled both 14s pentamers and empty capsid particles. The rapidity with which P1 precursors were cleaved to VPO, VP3, and VP1 and assembled subviral particles in VVPl/VVP3-~oinfected cells demonstrates that all of the virally encoded information required

36

DAVID C. ANSARDI et al.

FIG.4. Analysis of poliovirus assembly using recombinant vaccinia viruses which express P1 and 3CD. In previous studies, we described construction and characterization of recombinant vaccinia viruses that express the poliovirus capsid precursor protein P1 and the viral protease 3CD (Ansardi et al., 1991). In this system, cells are coinfected with vaccinia viruses VV-P1 and W-P3. The infection of cells with VV-P1 results in the expression of the poliovirus P1 protein. Expression of 3CD from VV-P3 results in the proteolytic processing of P1 to give the capsid proteins VPO, VP3, and VP1. Once proteolytic processing occurs, the capsid proteins assemble into poliovirus subviral intermediates: 55 protomers, 14s pentamers, and 75s empty capsids. Because no poliovirus RNA is present in this system, the final end point of the assembly is the 75s empty capsid in which VPO is not cleaved to VP4 and VP2.

for these stages in poliovirus assembly is present in the P1 and 3CD proteins and can occur in the complete absence of replicating poliovirus RNA. Poliovirus replication occurs in association with intracellular membranes (Kuhn and Wimmer, 1987; Paul et al., 1987b; Richards and Ehrenfeld, 19901, and more recent studies suggest that poliovirus P2 and P3 proteins required for replication are localized on mem-

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

37

branes via common precursor polyprotein (Lawson and Semler, 1992). In contrast, the P1 capsid precursor is cotranslationally cleaved away from the P2 and P3 proteins by the autocatalytic activity of 2Apr0 (Toyoda et al., 1986) and presumably can diffuse away from the other viral proteins. The results of our studies demonstrate that the proteolytic cleavage and subviral particle assembly steps can occur independently of the replication complexes. On formation of assemblycompetent capsid subunits, the capsid proteins may then be targeted through some unknown mechanism to sites of RNA encapsidation. Such a strategy of partitioning capsid assembly away from sites of RNA replication may ensure that immature or assembly-incompetent capsid subunits are restricted from entering sites of encapsidation. Alternatively, RNA released from the replication complexes may diffuse into the soluble sites of capsid assembly. This mechanism would ensure that capsid proteins, with affinity for viral RNA, would not enter the replication complexes and potentially interfere with RNA synthesis. The rapid assembly of VPO, VP3, and VP1 proteins generated in VVPl/VVPS-coinfected cells suggests that P1 precursors and 3CD proteins might form a processing/assembly complex in which P1 precursors are brought together and can rapidly form 14s pentamers on proteolytic processing. The vast majority of radiolabeled VPO, VP3, and VP1 recovered from VVPl/VVP3-~oinfectedcells was present in 14s pentamer or 75s empty capsid fractions. Although no evidence for assembly of specific oligomeric structures from P1 capsid precursors has been found, it was possible that such structures might be very labile and subject to disruption on lysis of the cells. Formation of labile precursor oligomers might account for the “P1 pentamers” reported in early studies of rhinovirus and EMCV assembly (McGreggor et al., 1975; McGreggor and Rueckert, 1977). Although the threedimensional structure of the P1 precursor is not known, the precursors may have enough affinity for one another to associate prior to cleavage. Formation of stable pentamers is almost certainly dependent on cleavage of P1 to free the amino terminus of VP3, five copies of which intertwine at the fivefold axes of symmetry t o form the p annulus (Acharya et al., 1989; Hogle et al., 1985; Luo et al., 1987; Rossman et al., 1985). The 3CD protein may also help in some unknown way to nucleate the P1 protomers.

C . Functional Significance of Poliovirus Capsid Myristylation To define further the role of myristylation in poliovirus assembly, a recombinant vaccinia virus was constructed that expressed a P1 pre-

38

DAVID C. ANSARDI et al.

cursor with a glycine to alanine substitution at the amino terminus of P1 to prevent myristic acid addition (Ansardi et al., 1992). Using the VVPl/VVP3 coexpression system, the importance for myristylation of the capsid precursor in cleavage by 3CD and assembly of subviral particles was investigated. Previous reports in the literature had indicated that nonmyristylated P1 was not processed as efficiently as myristylated P1 in uitro by 3CD (Krausslich et al., 1990; Marc et al., 1989). Fkports describing intracellular studies of assembly of nonmyristylated capsid proteins expressed on transfection of in uitro transcribed RNA genomes encoding amino acid substitutions that prevented myristylation did not address rates of cleavage of the nonmyristylated precursors, but completely processed nonmyristylated P1 cleavage products were detected (Marc et al., 1990, 1991). Another study had been reported in which mutant polioviruses encoding altered P1 myristylation signals expressed mixed populations of myristylated and nonmyristylated P1 precursors (Moscufo et al., 1991). Interpretation of results from these previous approaches had been difficult, and one author cited reversion of the mutations back to wild type as a complication in data interpretation (Marc et al., 1990, 1991; Moscufo et al., 1991). To circumvent this problem, a nonmyristylated P1 precursor and 3CD protease were coexpressed in the same cell from separate recombinant vaccinia viruses. Surprisingly, the nonmyristylated P1 was completely cleaved to VPO, VP3, and VP1 (Ansardi et al., 1992). The reasons for this contrast with the in uitro cleavage studies is not clear. Enzyme/substrate ratios in the W P l m y r - / VVP3-coinfected cells may have been more favorable for complete proteolytic cleavage than those in the in uitro studies. Alternatively, additional host-cell components may contribute to the cleavage reaction. Evidence has been presented that a host cellular protein factor facilitates cleavage of the P1 precursor by 3CD protease (Blair et al., 1993). Several studies had suggested that myristylation of the P1 precursor was required for assembly of stable poliovirus. By using the P113CD coexpression system, it became clear that a block in assembly of nonmyristylated promoters occurred at the level of 1 4 s pentamer formation. The results of the studies confirmed the speculation based on structural information that the myristate moieties are required to assemble a stable capsid (Chow et al., 1987; Paul et al., 1987a). The use of VVPllVVP3 coexpression system, then, demonstrated that this requirement for myristylation occurs prior to cleavage of VPO to VP2 and VP4, as nonmyristylated subviral particles did not assemble (Ansardi et al., 1992).

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

39

VIII. COMPLEMENTATION SYSTEM TO STUDY POLIOVIRUS ENCAPSIDATION Development of the recombinant vaccinia virus system for intracellular coexpression of the P1 precursor and 3CD protease provided the opportunity to analyze poliovirus capsid mutants for defects in proteolytic cleavage and assembly of subviral particles. The utility of this system was confirmed by analyzing the nonmyristylated P1 precursor expressed by a recombinant vaccinia virus. However, the assembly system at this point did not permit analysis of the RNA encapsidation step of morphogenesis. Information about this stage of assembly is particularly lacking, owing to the inability thus far t o reconstitute poliovirus virions in vitro from purified components (Putnak and Phillips, 1981b; Rombaut and Boeye, 1991),and because of the difficulty in obtaining poliovirus mutants with encapsidation defects. Previous studies by Jewell et al., (1990) had suggested that P1 precursors expressed by recombinant vaccinia viruses could not serve as a supply of capsid proteins in a mixed infection with poliovirus. Because of the large degree of intracellular localization of the processes of poliovirus RNA replication and possibly encapsidation, it was possible that the recombinant P1 precursors expressed by VVPl would be excluded from intracellular poliovirus compartments involved in RNA replication and encapsidation. The first indication that this was not the case came from experiments which repeated the recombinant vaccinia virus/ poliovirus coinfection experiments of Jewell et al. (1990). In contrast to these previous studies, in cells coinfected with VVPl and type 1poliovirus, the P1 precursor expressed by VVPl was proteolytically cleaved in trans by poliovirus protease 3CD, resulting in production of VPO, VP3, and VP1. In addition, mature virion protein VP2 derived from recombinant precursors was observed, strongly suggesting that vacciniaexpressed capsid proteins were incorporated into the poliovirus encapsidation pathway. Analyzing the incorporation of recombinant vaccinia-expressed capsid proteins into poliovirus virions in mixed infections with wild-type poliovirus provided one method for determining whether a mutant P1 precursor had defects at the encapsidation stage of assembly (Ansardi et al., 1992). However, this system suffered from an inherent complication because mutant capsid proteins expressed by the recombinant vaccinia virus were synthesized in the presence of wild-type capsid proteins expressed by the coinfecting poliovirus. Assembly defects of the mutant capsid subunits might then be overshadowed if they were incorporated into mixed wild-type and mutant particles. This appeared

40

DAVID C. ANSARDI et al.

to be the case with the nonmyristylated capsid proteins expressed by a recombinant vaccinia virus which did not assemble in VVPlmyr-/ VVP3 coinfected cells, but which were incorporated into poliovirus virions to some degree in cells coinfected with VVPlmyr- and type 1 poliovirus (Ansardi et al., 1992). To establish a trans-complementation system in which the only source of functional capsid proteins was the recombinant vaccinia virus, a poliovirus defective interfering (DI) genome was used as the source of a poliovirus replicon which did not express functional capsid proteins (Cole, 1975; Hagino-Yamagishi and Nomoto, 1989; Kuge et al., 1986). The DI genome we used for these studies had been previously described and could be generated in uitro by transcription of a cDNA copy of the genome contained in plasmid pSMl(T7)l under the control of a promoter for bacteriophage T7 RNA polymerase (HaginoYamagishi and Nomoto, 1989). The DI genome contained a deletion of 816 nucleotides of the P1 gene (-31% of the P1 gene) and had been constructed in vitro by ligating a cDNA copy of a segment of the deletion-containing P1 gene from a naturally occurring DI genome of poliovirus type 1Sabin into a type 1Mahoney cDNA background. The in-frame deletion encompasses sequences from portions of the VP2 and VP3 genes. In their studies, Hagino-Yamagishi and Nomoto (1989) demonstrated that the in uitro transcribed DI genome replicated on transfection into poliovirus-infected cells and was encapsidated by capsid proteins provided in trans by helper wild-type poliovirus. Furthermore, the genetically engineered DI genome was maintained in serial passage in a mixed stock of wild-type and DI viruses. To establish a complementation system, Ansardi et al. (1993) transfected the RNA derived from the in uitro transcription of the DI cDNA into cells previously infected with VVP1. The DI RNA replicated and was encapsidated by the P1 provided in trans by VVP1; the encapsidated defective genome was referred to as PVdefSM (Fig. 5). This was the first demonstration of trans-complementation of a defective poliovirus genome and generation of a homogeneous population of defective poliovirus particles free of contaminating wild-type helper virus. By serially passaging PVdefSM in the presence of VVP1, stocks of PVdefSM were generated that could be used subsequently as a means of delivering the capsid gene-deficient replicon to every cell in a monolayer, overcoming limitations of transfection efficiencies. These studies established that vaccinia virus vectors expressing P1 capsid precursors could be used as the exclusive source of capsid proteins for a capsid gene-deficient poliovirus replicon, providing the opportunity to analyze the poliovirus assembly process in all of its stages, including

FIG.5. Complementation system in which to study poliovirus assembly (Ansardi et al., 1993). We utilize a cDNA clone of a poliovirus defective interfering genome (DI) (Hagino-Yamagishi and Nomoto, 1989). Cells are first infected with VV-P1, followed by transfection with the DI RNA obtained from in uitro transcription. The genome is defective because i t lacks a complete coding region for the poliovirus P1 protein. Transfection of this RNA into cells results in the complete replication cycle of poliovirus because the DI genome encodes all of the necessary proteins for RNA replication. The first step following transfection is translation of the DI RNA, which results in the production of poliovirus proteins required for replication including 3CD, which processes the poliovirus P1 protein expressed from VV-P1, resulting i n cleavage and assembly of subviral intermediates. In parallel, the viral proteins replicate the defective viral genome, resulting in the synthesis of multiple copies of the plus-stranded RNA. The plus-stranded RNA genome interacts with subviral intermediates, resulting in encapsidation. The encapsidated RNA is then released from the cells by a n as yet undetermined mechanism. Serial passage of the encapsidated RNA (referred to as a replicon) in the presence of VVP1 results in amplification; following extended serial passage of greater than 20 or more, stocks of the encapsidated replicons can be obtained. Removal of residual VV-P1 from the stocks can be achieved using centrifugation in combination with anti-vaccinia antibodies. The resulting stock of encapsidated replicons is devoid of any VV-P1. The stock of encapsidated replicons can be used in combination with VV-P1 variants containing defined mutations in the poliovirus capsid genes to assess the effects of the mutations on poliovirus assembly and encapsidation (Ansardi and Morrow, 1993, 1995; Ansardi et al., 1994a).

42

DAVID C. ANSARDI et al.

encapsidation, without having to rely on expression of mutant capsid proteins from a replicating poliovirus genome. To characterize the complementation system, the role myristylation of P1 has on processing, assembly, and encapsidation of RNA was reexamined. In assembly experiments conducted by using cells coinfected with VVPlmyr- and PVdefSM, results markedly different than those obtained with the VVPlmyr- /VVP3-coinfected cells were observed (Ansardi et al., 1993). Low levels of the VPO, VP3, and VP1 proteins recovered from cells coinfected with VVPlmyr- and PVdefSM had sedimentation properties on sucrose density gradients consistent with empty capsids. Importantly, this difference in assembly phenotypes of mutant capsid proteins between the two systems was not a general property of all of the assembly-defective mutants because in most cases in which assembly of mutant capsid proteins was analyzed in both systems, those that did not assemble in cells coinfected with the mutant P1-expressing recombinant vaccinia virus and VVP3 also did not assemble in cells coinfected with PVdefSM and the mutant P1expressing recombinants. The difference in assembly phenotypes indicated that some factor associated with VVPlmyr-/PVdefSM-coinfectedcells, but not with VVPlmyr- /VVP3-coinfected cells, facilitated assembly of nonmyristylated protomers. One explanation might be that cleavage of the nonmyristylated P1 precursor occurred much more rapidly in cells coinfected with PVdefSM, increasing pools of nonmyristylated 5s protomers available for assembly. This would seem to be a plausible explanation since the processed proteins generated in VVPlmyr-/VVP3-~oinfected cells were unstable; more efficient cleavage of nonmyristylated P1 might then allow greater concentrations of nonmyristylated protomers to build prior to their degradation. However, no delays were observed in complete cleavage of the nonmyristylated precursor in comparison to the wild-type precursor in cells coinfected with VVP3, and expression levels of the 3CD protease in cells infected with 20 pfdcell of VVP3 appear to be comparable to those expressed by the defective poliovirus genomes when introduced at levels sufficient to infect every cell in a monolayer. A second difference associated with the VVPlmyr- /PVdefSMcoinfected cells was the presence of the nonfunctional capsid precursor with an internal deletion of 272 amino acids expressed by the defective poliovirus genome. This precursor retains the myristylation signal at the amino terminus and is presumably cotranslationally modified by myristate addition. We have observed that this deletion-containing P1 protein is unstable. The possibility that the myristylated deletioncontaining P1 protein is facilitating assembly of nonmyristylated cap-

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

43

sid proteins seems unlikely. The deletion-containing P1 precursor would also not be able to supply a myristylated VPO protein through phenotypic mixing with the individual proteins derived from the recombinant nonmyristylated precursor, as the deleted region of the precursor encompasses portions of VPO. Other components present during a poliovirus infection may be contributing to the assembly of nonmyristylated protomers. Although clearly poliovirus P1 and 3CD proteins are the only poliovirus proteins required for assembly of subviral particles, this does not rule out the possibility that another poliovirus protein plays a facilitory role in capsid assembly. A more intriguing possibility is that the viral RNA genome plays a nucleating role in virus assembly, and capsid protein interactions with the RNA genome might play a stabilizing role in formation of the capsid. If the viral RNA plays some active role in nucleating the capsid protomers at an early stage, then interaction of nonmyristylated subunits with the viral RNA might facilitate their assembly into capsid particles. Clearly, a pronounced block exists in forming nonmyristylated RNA-containing virions; therefore, this type of assembly model for poliovirus suggests that at least some populations of subviral particles may be derived from the dissociation of unstable ribonucleoprotein complexes on extraction from the infected cell. Some evidence in the literature suggests that empty capsids may be a degradation product of an unstable ribonucleoprotein complex that easily dissociates into empty capsids and RNA on extraction from the cell (Koch and Koch, 1985; Maronginu et al., 1981).The absence of myristate form V P O may prevent completion of the RNA encapsidation event and condensation of a mature virion, resulting in a byproduct of empty capsids. In Section VII, a hypothesis was presented that the events of poliovirus assembly and encapsidation may be sequestered from the replication complexes. If nascent RNA chains diffuse away from the replication complexes to be encapsidated, this might allow genomic RNA to participate in the assembly process and facilitate formation of nonmyristylated subviral particles. Alternatively, poliovirus capsid proteins may be targeted to membranous sites of RNA encapsidation which are adjacent to replication complexes (Caliguiri and Compans, 1973; Pfister et al., 1992).The subcellular localization of nonmyristylated and myristylated P1 precursors expressed by recombinant vaccinia viruses (cells infected with VVPl or VVPlmyr- alone) was analyzed by performing crude separations of cytosol from intracellular membranes (D. A. Ansardi and C. D. Morrow, unpublished results, 1991). The vast majority of myristylated and nonmyristylated P1 precursors partitioned in the soluble fractions. The finding that myristylated P1 is a cytosolic protein strongly

44

DAVID C. ANSARDI et al.

suggests that cleavage of this protein by 3CD is not associated with intracellular membranes. The rapid assembly of cleavage products of myristylated P1 suggested that any targeting to membranous locations of RNA encapsidation occurs subsequent to pentamer formation.

A . Proteolytic Cleavage of Capsid Precursor Although the favored pathway for poliovirus morphogenesis indicates that complete cleavage of the capsid precursor is required prior to assembly of protomers into subviral particles or virions, this question had not been addressed previously by using intracellular assembly systems (Rueckert, 1990). In vitro assembly studies of EMCV capsid precursors with cleavage site defects indicated that complete proteolytic cleavage of the capsid precursor was required for assembly of EMCV subviral particles (Palmenberg, 1982; Parks and Palmenberg, 1987). Previous assembly studies of different picornaviruses, however, had suggested that even uncleaved P1 precursors from rhinovirus, EMCV, and poliovirus might assemble pentamer precursors (McGreggor et al., 1975; McGreggor and Rueckert, 1977). The most convincing arguments that cleavage of the P1 precursor is required for poliovirus capsid assembly were made from studies of the three-dimensional structure of picornavirus virions: formation of the p-annulus structure at the fivefold vertices of the virion requires cleavage between VPO and VP3 to free the amino termini of VP3 that form this structure (Acharya et al., 1989; Hogle et al., 1985; Luo et al., 1987; Rossman et al., 1985). Discrepancies between the early assembly studies and assembly pathways deduced from structural information suggested that an investigation of whether cleavage intermediates might assemble precursor subviral particles was warranted. As a first step in determining whether complete cleavage of P1 was required for assembly of subviral particles, P1 precursors were generated with valine substitutions for glycine at the P1’ position of the QG cleavage sites in the precursor with the hope that these substitutions would prevent cleavage at the altered sites (Ansardi and Morrow, 1993). The valine substitutions were chosen so that cleavage would be prevented by making the most conserved substitution possible; previous studies of the cleavage of the QG bond between the 3C and 3D proteins of poliovirus found that an alanine substitution for glycine still permitted cleavage, whereas a valine substitution inhibited cleavage at the site (Kean et al., 1990). However, studies by Kirkegaard suggested that glutaminemethionine could serve as a functional cleavage site between VP3 and VP1, although the use of this cleavage site had not been confirmed

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

45

(Kirkegaard, 1990). Parks and Palmenberg had determined that a QV dipeptide was not a functional cleavage site for proteolytic processing of the EMCV capsid precursor in uitro a t the site between VP3 and VP1; the wild-type cleavage site at that location in the EMCV P1 precursor is QG as in the case of poliovirus (Parks and Palmenberg, 1987). Finally, all of the cleavage sites in poliovirus polyproteins processed by poliovirus 3Cpr0 or 3CD are QG (Kitamura et d,1981; Racaniello and Baltimore, 1981a). This strict conservation of cleavage site primary sequence is unique among picornaviruses, suggesting that the poliovirus enzyme had a particularly stringent requirement for a QG cleavage site sequence (Palmenberg, 1990). Analysis of P1 precursors which had QV cleavage sites at either the VPO-VP3 cleavage site or the VP3-VP1 cleavage site revealed that cleavage at the altered sites occurred, although less rapidly than at the QG sites (Ansardi and Morrow, 1993). Complete cleavage of these precursors prevented a thorough analysis of assembly phenotypes of incompletely cleaved precursors. However, in unpublished experiments conducted using cells coinfected with recombinant vaccinia viruses that expressed a mutant precursor and VVP3, we were able to determine the sedimentation properties of the P1 cleavage intermediates on sucrose density gradients (Ansardi and Morrow, unpublished observations, 1992). Unlike the completely cleaved proteins, the cleavage intermediates displayed nonspecific sedimentation properties without accumulating in distinct peak fractions and were more or less evenly distributed among fractions of the gradient above the 14s pentamer peak. The reasons for the nonspecific sedimentation pattern are unclear but might reflect an aggregation of cleavage intermediates into nonfunctional oligomers. Because the metabolically radiolabeled, partially cleaved precursors were presumably present in the same cell with unlabeled completely cleaved precursors, drawing definitive conclusions about the oligomeric state of P1 cleavage intermediates is difficult. Evidence from pulse-chase radiolabeling experiments indicated that P1 cleavage intermediates were degraded in the infected cells, further suggesting that they did not assemble stable subviral particles. Instability of the cleavage intermediates is compatible with the observation that assembly-defective protomers are rapidly degraded in cells. On the basis of pulse-chase experiments it was clear that not all of the partially cleaved intermediates were chased into completely cleaved proteins, and at least a portion of the partially cleaved proteins were degraded. Successful complete cleavage of the capsid precursor might require that the events occur in a single interaction event with the enzyme, with partially cleaved proteins not serving as substrates for the protease.

46

DAVID C. ANSARDI et al.

A question that has not been resolved in the proteolytic cleavage step of capsid formation is whether the 3CD-mediated cleavages of the P1 precursor occur in a defined order. Evidence from previous studies suggested that the P1 precursor was cleaved first at the site between VP3 and VP1, generating an uncleaved VPO-VP3 protein, and second at the site between VPO and VP3 (Reynolds et al., 1992). This order of cleavage had been suggested because a VPO-VP3 cleavage intermediate is typically detected in lysates of poliovirus-infected cells, whereas little VP3-VP1 is detected. The P1 precursor with the QV cleavage site between VP3 and VP1 was delayed in processing at the altered site, allowing confirmation that cleavage at the site between VPO and VP3 could occur without prior cleavage between VP3 and VP1. Thus, cleavage between VP3 and VP1 is not a prerequisite for cleavage between VPO and VP3. A second component of the ordered cleavage hypothesis was that a defined processing order series of cleavages might be required to generate functional, assembly-competent capsid proteins. Interestingly, capsid protomers derived from the precursor with the altered cleavage site between VPO and VP3 (VP3-G001V, which gave rise to increased amounts of VPO-VP3 and VP1) failed to assemble subviral particles. In contrast, capsid protomers generated from the precursor with the altered cleavage site between VP3 and VP1 (VP1-G001V)were capable of assembly. Clearly an argument can be made based on these results that the previously predicted order of P1 cleavage, in which VP1 is released followed by VPO and VP3, is not required to generate assembly-competent capsid proteins. The results may even suggest that the order of cleavage required to generate assembly-competent protomers is cleavage at the VPO-VP3 site followed by cleavage at the site between VP3 and VP1. Further studies will be required to substantiate this claim, because the capsid proteins derived from the P1 precursor with the QV cleavage site between VPO and VP3 have a valine substitution at the amino terminus of VP3, and the amino termini of VP3 form the p-annulus structure responsible for interlocking common protomers within a pentamer (Acharya et al., 1989; Hogle et al., 1985; Luo et al., 1987; Rossman etal., 1985). To distinguish between these two possibilities, new mutants can be constructed with substitutions at the Q position of the QG bond, thus altering the carboxyl terminus of VPO rather than the amino terminus of VP3. If Q-substituted mutants with slower cleavage kinetics at the cleavage site between VPO and VP3 can be isolated, this might allow a resolution of these two possibilities. Secondary consequences of the valine substitutions introduced at the amino termini of both VP3 and VP1 were manifested in defects of the

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

47

completely cleaved capsid protomers in assembly and RNA encapsidation (Ansardi and Morrow, 1993). The secondary defects of these mutants indicate that maintenance of the QG cleavage site primary sequence in the poliovirus P1 capsid precursor is required for proper function of the capsid proteins. This property is likely to be one reason why poliovirus 3Cpro (and 3CD) cleavage sites appear to have less flexibility in primary sequence than the 3Cpro cleavage sites of other picornaviruses (Palmenberg, 19901, a t least in the case of the capsid precursor.

B . Capsid Mutations Affecting R N A Encapsidation Encapsidation of a viral RNA genome requires specific recognition of the genome by the virus capsid to ensure packaging of viral RNA without nonspecific packaging of host cellular mRNA molecules. The protein-RNA interactions required for genome encapsidation consist of two elements: capsid protein determinants that specifically interact with the viral RNA, and cis-acting elements of the viral RNA genome that are recognized by the virus capsid and distinguish it from nongenomic RNA molecules. The virus capsid may also contain interior features that interact with RNA in a less specific manner not dependent on nucleotide sequence. In studies using the recombinant vaccinia virus expression systems, mutants with changes at the amino terminus of VP1 were shown to have defects in the RNA encapsidation step of assembly (Ansardi and Morrow, 1993).Two mutants were analyzed: one with a valine substitution for the glycine residue at the amino terminus of VP1 (VP1G001V) and a second in which the first four amino acids of VP1 were deleted (VPlAl-4). The VP1 deletion mutant had been previously described in the literature and was found t o have a delayed kinetics of encapsidation at 39.5"C (Kirkegaard, 1990). Surprisingly, the encapsidation defect was more pronounced in the mutant with the valine substitution for the amino-terminal glycine than in the deletion mutant. The results of these experiments raise the possibility that the amino-terminal portion of VP1 is one of the poliovirus capsid determinants involved in capsid protein-RNA interaction. The involvement of an amino-terminal arm from a viral capsid protein in interaction with nucleic acid has precedent in the literature (Geigenmuller-Gnirke et al., 1993; Rossman et al., 1985). In the cases of some icosahedral RNA viruses, including several plant viruses, the terminal extensions of capsid proteins not associated with the @-barrelcore are disordered in the three-dimensional structures and point toward the interior (Rossman et al., 1985). These amino-terminal structures are likely in con-

48

DAVID C. ANSARDI et al.

tact with RNA and may play important roles in capsid-RNA interaction. Often these terminal extensions contain a large proportion of basic residues, but this is not the case with the amino terminus of poliovirus VP1. The disordered amino-terminal segment of the Sindbis capsid protein, the core of which does not follow the classic p-barrel fold of most icosahedral viruses (H. K. Choi et al., 19911, has been demonstrated to contain a segment of 32 amino acids critical for interaction with the RNA genome (Geigenmuller-Gnirke et al., 1993). This segment contains a highly conserved stretch of 10 amino acids containing three lysine and two arginine residues. In addition to the role for the VP1 amino terminus in RNA encapsidation, this region has been implicated from both genetic and biochemical studies to be involved in the processes of virus entry and release (Fricks and Hogle, 1990; Kirkegaard, 1990). Although the first 20 amino acids of VP1 were disordered in the three-dimensional structure determined for type 1 poliovirus (Hogle et al., 1988, this region was proposed by Fricks and Hogle (1990) to form an amphipathic helix structure. Even though the amino acid sequence homology is limited at the amino terminus of VP1 among different picornavirus members, Fricks and Hogle demonstrated that the first 18-23 residues of VP1 amino termini from several enteroviruses and from human rhinovirus type 14 could be modeled on an amphipathic helical wheel. This structural feature is not likely to be shared among the aphthovirus and cardiovirus members of the Picornauiridm, however, which have shorter VP1 amino termini lacking the potential structural homology of the VP1 amino termini of enteroviruses and rhinoviruses (Acharya et al., 1989; Luo et al., 1987). Further refinements of the structure of poliovirus type 3 Sabin found a partially ordered stretch of five amino acids from an unidentified portion of the VP1 amino terminus that formed a short segment of p-sheet structure along with portions of VP4. In the studies of Fricks and Hogle (19901, externalization of the amino terminus of VP1 after binding of the virus to the receptor was shown to be responsible for binding of the altered virus to liposomes, and they proposed that the amino-terminal portion of VP1 in concert with VP4, which is also extruded from the virus after receptor attachment (Everaert et al., 1989), played a role in disrupting the endosomal membrane to allow RNA to be released into the cytosol. The linkage of VP4 to the short segment of VP1 was suggested to provide a way for externalization of these two capsid features together. If the amino terminus of VP1 is also involved in RNA interaction, a mechanism might be envisioned in which extrusion of VP4 and the VP1 amino terminus triggers the release of the RNA genome from the interior of the capsid. So far, however, biochemical evidence that this region of

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

49

VP1 interacts specifically with the RNA genome has not been reported, either with purified VP1 or with synthetic peptides corresponding to the amino-terminal portions of the VP1 protein. Future experiments might be aimed at addressing these questions. Although the VP1 amino terminus may directly interact with the poliovirus RNA genome, other regions of the virus capsid are likely to interact with the RNA genome as well. The three-dimensional structure of poliovirus type 1 did not reveal electron density that could be interpreted as RNA, but in refined structures of poliovirus type 3, some electron density was attributed to RNA base ring structures stacking with the aromatic side chains of the tryptophan-38 and phenylalanine-41 residues of VP2 (Filman et al., 1989). Although significant regions of encapsidated picornaviral RNA molecules do not adopt the regular pattern of capsid binding required for nucleic acid visualization in the crystal structures, some insight into viral capsid-RNA interactions has been gained from RNA viruses and single-stranded DNA viruses in which segments of the viral genomes do associate in a sequence-independent manner with regions of the capsid interior (Chen et al., 1989; Fisher and Johnson, 1993; Larson et al., 1993; McKenna et al., 1992; Tsao et al., 1991). In the case of bean-pod mottle virus (a comovirus), a single-stranded RNA virus with a bipartite RNA genome encapsidated in separate particles, nearly 20% of the viral RNA genome binds to the capsid interior in a symmetric fashion (Chen et al., 1989). The ordered RNA is single-stranded and associates with a hydrophilic pocket around the 3-fold symmetry axes of the capsid. The binding of the RNA around the threefold axes results in the formation of a trefoil-shaped cluster at each of the twenty three-fold axes of the capsid, with each cluster consisting of 33 ribonucleotides. The binding of RNA in this manner indicates that the protein-RNA interactions are not sequence specific as an exactly repeating set of bases in the RNA genome is not present. The interactions between the RNA segment and residues lining the binding pocket are primarily van der Waals and electrostatic interactions. RNA-protein interactions have been characterized for a number of viral and nonviral proteins (reviewed by Frankel et al., 1991; Mattaj, 1993). An example of one extensively studied RNA-protein interaction is the binding of the HIV-1 Tat protein to a segment of RNA designated TAR (Calnan et al., 1991; Tao and Frankel, 1992). The Tat protein is a transcriptional activator which binds to TAR, a bulged stem-loop structure present at the 5' end of viral mRNA molecules. Tat contains a 9-amino acid region of basic amino acids that are required for the recognition of TAR. Arginine-rich motifs (ARMS) are conserved among many RNA binding proteins, including the capsid

50

DAVID C. ANSARDI et al.

proteins of some RNA viruses (Frankel et al., 1991; Lazinski et al., 1989; Mattaj, 1993). The ARM sequences may be involved in recognizing specific RNA secondary structures as well, and this function has been confirmed for recognition of RNA hairpins by some bacteriophage antiterminator proteins (Lazinski et al., 1989). Although not all RNAbinding proteins contain linear ARM sequences (poliovirus does not), sequence-specific RNA binding motifs formed from clusters of arginine residues brought together in the folded protein might be envisioned as a method of RNA recognition (Calnan et al., 1991). By using the recombinant vaccinia virus systems, we analyzed two arginine residues associated with a cavity on the poliovirus capsid interior (Ansardi et al., 1994a). The dimensions of the cavity, approximately 10 A wide and 5 A deep, are sufficient to accommodate a helical segment of RNA. The cavity displays similarity to those of some of the icosahedral viruses in which nucleic acid interaction with the capsid was visible in the three-dimensional structure (canine parvovirus and +X174), including the presence of several basic amino acid residues (McKenna et al., 1992; Tsao et al., 1991). Most of the basic residues of the poliovirus cavity are well-conserved in amino acid sequence alignments of capsid proteins from various picornaviruses. 'Ibo arginine residues associated with this depression were analyzed by sitedirected mutagenesis for their functional role in capsid assembly and RNA encapsidation. One of the arginine residues (VP1-R129) is wellconserved among different picornaviruses in capsid sequence alignments, and substitution of this residue with lysine or glutamine disrupted assembly of the capsid. The second cavity-associated arginine residue targeted for mutagenesis, VP4-RO34, is not well conserved in sequence alignments of picornavirus capsid proteins. The lysine substitution for arginine at this position had no observable effects on assembly or encapsidation. In contrast, substitution of glutamine for this arginine residue rendered the mutant capsid defective for virion formation, especially at 395°C. The encapsidation defect for this mutant could not be separated in these studies from a primary defect in capsid assembly that was noted even in the absence of the defective RNA genome. Other amino acid residues within the depression provide potential targets for future mutagenesis studies. A lysine residue at VP3-041 and an arginine residue at VP1-267 both have side chains well exposed to the interior of the virus. Surprisingly, substitution of an arginine residue buried at a protomer-protomer interface (VP3-R223) affected RNA encapsidation. At 37"C,capsid proteins derived from cleavage of the VP3-R223K mutant precursor did not assemble subviral particles or virions; at 33"C, however, capsid proteins derived from the mutant precursor were

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

51

found to cosediment on sucrose density gradients with poliovirus empty capsids. RNA-containing virions derived from the mutant were not observed at either temperature. The ability of the mutant to assemble empty capsids but not virions at 33°C suggested that this substitution might have destroyed an encapsidation determinant. The side chain of this arginine residue is not well exposed on the interior surface of the capsid in the mature virion, however, and almost certainly could not make direct contact with the RNA genome. The lysine substitution for arginine at this location might have resulted in secondary structural effects that rendered the capsid incapable of forming mature virions. The VP3-R223 residue potentially forms a hydrogen bond with the side chain of threonine VP3-031. The cavity-associated residue VP1R129 potentially forms a hydrogen bond with the main-chain oxygen of this same residue. Thus, substitution of the buried residue might exert secondary defects on the cavity region. The phenotype of this mutant provides some preliminary evidence that this region of the capsid may be involved in RNA interaction. Poliovirus capsid protein-RNA binding may include nonsequencespecific interactions similar to those described for the icosahedral viruses in which nucleic acid was observed in the X-ray structure. The poliovirus capsid may also contain determinants required for recognizing a sequence-specific RNA secondary structure which acts in cis as an encapsidation signal. Just as information about poliovirus capsid protein determinants required for RNA encapsidation is limited, little is known about the cis elements of the poliovirus genome required for encapsidation. Whatever the RNA encapsidation signal of poliovirus may be, its identification is likely to be difficult. Important information has been gained from in uitro analyses of protein-RNA interactions in other systems (Calnan et al., 1991; Geigenmuller-Gnirke et al., 1993; Gott et al., 1991; Fbmaniuk et al., 1987). Unfortunately, in uitro encapsidation systems for poliovirus have not been developed. In uitro interactions between poliovirus RNA and capsid proteins are difficult to study for several reasons. Encapsidated RNA molecules are always linked to VPg (Wimmer, 1982), and linkage of VPg to the RNA genome requires that the processes of poliovirus RNA replication take place (Kuhn and Wimmer, 1987; Paul et al., 1987b; Richards and Ehrenfeld, 1990). Thus, encapsidatable RNA genomes likely cannot be generated by in uitro transcription from a cDNA copy because RNA molecules synthesized by this method are not linked to VPg. Furthermore, construction of deletions in poliovirus RNA outside of the P1 region are incompatible with RNA replication, because the replication functions of many of the P2 and P3 proteins cannot be complemented in trans (Bernstein et al., 1986; Johnson and Sarnow, 1991). For these

52

DAVID C. ANSARDI et al.

same reasons, naturally occurring DI genomes of poliovirus contain deletions only in the P1 region, because propagation of these genomes requires that they retain the capacities for both replication and encapsidation. The system developed by Ansardi et al. (1993) provides a method for trans-encapsidation of a poliovirus subgenomic replicon. An important feature of this trans-encapsidation system is that it separates the supply of capsid proteins away from the subgenomic replicon. An exciting potential use for this system would be to force adaptation of the RNA genome to a mutant capsid protein. In other words, if mutant capsid proteins are identified with defects at the encapsidation stage, serial passage of the subgenomic replicon with the recombinant vaccinia virus expressing a continuous supply of the mutant capsid might result in adaptation of the replicon to be encapsidated. This strategy might provide a method for identifying otherwise elusive cis elements of the poliovirus RNA genome required for encapsidation. In fact, serial passage of PVdefSM with the recombinant vaccinia virus which expresses the VP1-G001V precursor and the VP3-R223K precursor at 33°C results in low levels of encapsidation, suggesting that the PVdefSM genome can “adapt” to this mutant capsid protein (D. C. Ansardi and C. D. Morrow, unpublished, 1994).

C . Studies on Maturation Cleavage Using Complementation System The assembly of an infectious poliovirus virion requires the proteolytic cleavage between an asparagine-serine amino acid pair in VPO after encapsidation of the viral genomic RNA. This cleavage, which results in the processing of VPO to VP4 and VP2, has been termed the maturation cleavage and is believed to occur via an intramolecular event (Arnold et al., 1987). It has been difficult to study the features of this cleavage as well as the generation of infectious poliovirions owing to the rapid cleavage of VPO and maturation into infectious virus. Studies using the recombinant vaccinia virus systems have described mutants in which a glutamine-glycine amino acid pair (VP4-QG)and a threonine-serine amino acid pair (VP4-TS)were substituted for the asparagine-serine amino acid pair in the maturation cleavage site (Ansardi and Morrow, 1995). The mutations in which a glutamine-glycine amino acid pair were substituted in the maturation cleavage site resulted in a capsid protein that could be proteolytically processed and assembled into subviral intermediates including an empty capsid-like structure in the presence of PVdefSM. However, no PVdefSM was encapsidated in virions containing a QG. In contrast, the threonine-serine substitution for the asparagine-serine at the

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

53

maturation cleavage site resulted in a capsid protein which, on proteolytic processing, could assemble into subviral intermediates and encapsidate the PVdefSM RNA. The maturation cleavage was significantly delayed compared to wild type. Interestingly, the cleavage event in these particles occurred in vitro as well, and this cleavage could be accelerated by incubation at physiological temperatures. The results of these studies support the concept that a series of conformational changes occur during the maturation cleavage of VPO. The mechanism by which this occurs throughout the entire poliovirion (i.e., the 60 copies of VPO to be cleaved) is not clear. It is possible that a cooperative effort exists between the subunits during the maturation cleavage, as has been suggested for nodaviruses (Zlotnick et al., 1994). Whether this is the case for the maturation cleavage of poliovirus is currently under investigation utilizing a different set of mutants in combination with the complementation system. IX. PERSPECTIVES ON POLIOVIRUS ASSEMBLY

A proposed ordered pathway for poliovirus morphogenesis was discussed in earlier sections. The phenotypes of mutants generated from our studies have afforded some valuable tools with which to assess the validity of this pathway (Table I). The first step in the proposed pathway of poliovirus morphogenesis is cleavage of the precursor protein P1. The cleavage site mutants described were not completely blocked in processing at the altered (QV) sites, making it difficult to determine whether P1 precursors cleaved at only one site would assemble. However, no evidence was found for assembly of specific stable structures from P1 cleavage intermediates. Using sucrose density gradients, no evidence was found that uncleaved P1 precursor assembled stable subviral particles, as most uncleaved P1 was localized in fractions of the gradients above the 14s pentamer fractions. It is possible that uncleaved precursors are associated together as labile oligomers, and the rapid assembly of P1 cleavage products suggests that some type of P1-3CD assembly complex may exist. A significant proportion of the mutants analyzed failed to assemble subviral particles (Pl-myr-, VP3-G001V, VP3-R223K at 37"C, VP1R129K, VP1-R129Q). A common feature of all of these mutants was that they were stable in the precursor form but on proteolytic cleavage were degraded. A similar observation was made with FMDV capsid proteins expressed by a recombinant vaccinia virus (Belsham et al., 1991) and with an assembly-defective attenuated mutant of poliovirus

54

DAVID C. ANSARDI et al. TABLE I ASSEMBLY PHENOTYPES OF POLIOVIRUS CAPSIDMUTANTSEXPRESSED BY RECOMBINANT VACCINIA VIRUSES

Precursor P1 (wild type) PlmyrVP3-GOOlV VP1-G001V VP1-A1-4 VP4-RO34K VP4-RO34Q VP3-R223K VP1-R129K VP1-R129Q VP4-ND69T VP4-NO69Q VP2-SOlG

Cleavage by 3CD.

14s VVP3b

EC VVP3c

EC PVdefSMd

RNA Encapsidatione

++++f

++++

++++

++++ ++

++++

++++ ++++ ++++

+++ ++++

++++ ++ ++ +++ ++++ ++++

++++ ++++ ++++ ++++ ++++

-

++++ ND

++++ +++ -

37°C

NDg ND ND

++++ +++ - 37°C

++ 33°C

++ 33°C

ND

ND ND

++++ +

++++ -

-

-

+

- 37°C

+ + + 33°C

+*

+++ 37°C + 39.5”C - 37°C - 33OC*

-

++++ ++

Cleavage of precursor in cells coinfected with VVP3. Assembly of 1 4 s pentamers in cells coinfected with VVP3. Assembly of 75s empty capsids in cells coinfected with VVP3. d Assembly of 755 empty capsids in cells coinfected with PVdefSM. Formation of mature RNA-containing virions in cells coinfected PVdefSM; reduced yields may result from a defect at a n earlier step. Mutants with defects a t the encapsidation step are marked (*). f + + + +, wild type; + + +, 50-70% wild type; + +, 25-50% wild type; + ,lo-25% wild type; -, none detected. 8 ND, Not determined. Delayed cleavage of VPO resulting in accumulation of poliovirions. a

type 3 (Macadam et al., 1991). In the P1 precursor form, the recombinant vaccinia virus-expressed FMDV capsid proteins were stable, but on cleavage of the precursor, the capsid proteins, which failed to assemble subviral particles, were rapidly turned over. Together, these results suggest that picornaviral capsid precursors are recognized by the cell as correctly folded proteins, but on cleavage of the precursor, the capsid proteins must rapidly assemble into their oligomeric forms or are targeted for degradation. The stability of the P1 precursors may also provide further evidence that P1 precursors or P1‘and 3CD proteins assemble protein complexes prior to cleavage. P1 or P1-3CD oligomers may be recognized by the cell as “correctly folded.” However, on cleavage, mutant capsid proteins that fail to assemble stable 14s pentamers may dissociate and be recognized as “misfolded.” In preliminary experiments the P1 capsid precursor has been found to interact

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

55

with the hsp 72/73 members of the 70-kDa family (Hsp 70) of heatshock proteins (Beckman et al., 1990; Pelham, 1988; D. C. Ansardi and C. D. Morrow, unpublished, 1993).A published report has described the interaction of poliovirus capsid proteins with proteins of the hsp 70 family of chaperones (Macejak and Sarnow, 1992). It is not clear if the interaction with hsp 72/73 plays a role in targeting assembly-defective mutant capsid protomers for degradation (Beckman et al., 1990; Pelham, 1988). In most cases in which cleaved poliovirus capsid proteins failed to assemble in cells coinfected with the mutant P1-expressing recombinant vaccinia and VVP3, they also failed to assemble in cells coinfected with the P1-expressing recombinant and PVdefSM. These results might be taken as evidence that the ability to assemble subviral particles is a prerequisite for interaction with the RNA genome. A potential deviance from this idea was noted by the different assembly phenotypes of the capsid proteins derived from the nonmyristylated precursor and the precursor containing a QG substitution at the VP4/VP2 junction. In cells coexpressing 3CD as the only other poliovirus component, the cleavage products derived from these precursors failed to assemble. However, in the presence of the defective poliovirus genomic RNA, cleavage products of these precursors assembled low levels of subviral particles. We have speculated that these particles may represent uncondensed capsids assembled around a nucleating RNA genome. The phenotype of these mutants were unique among those studied and provide preliminary evidence that the RNA genome might play a nucleation role in assembly and facilitate interactions among capsid protomer subunits. Clearly, extensive further studies are needed t o address these issues. Both 14s pentamers and empty capsids have been proposed to be the direct precursor of the poliovirus virion. In our studies, three capsid mutants, nonmyristylated P1 in VVPlmyr-/PVdefSM-coinfected cells, the VP4-QG mutant in VV-VP4QG/PVdefSM-coinfected cells, and VP3-R223K at 33"C, assembled structures consistent with empty capsids but did not assemble RNA-containing virions at detectable levels. The phenotypes of the mutants suggest that empty capsid formation is not sufficient to ensure assembly of RNA-containing virions. In the course of these studies, one mutant, VP1-GOOlV, was identified which assembled empty capsids in excess over virions, whereas a second mutant, VP4-R034Q, assembled virions over empty capsids at a higher ratio than normal. The assembly phenotype of the VP4-RO34Q mutant can be traced to a diminished pool of assembly-competent protomers in comparison to wild type, suggesting that empty capsids and virions assemble separately from common pools of capsid subunits.

56

DAVID C. ANSARDI et al.

However, it is possible that assembly ofempty capsids with this mutant was a rate-limiting step in virion formation and that, once this step occurred, RNA encapsidation proceeded rapidly. The identity of the direct precursor to the poliovirus virion thus remains an open question. The RNA encapsidation step of poliovirus assembly remains the most elusive in viral morphogenesis. It is clear that the amino terminus of VP1 plays a role in this process, giving merit to the future analysis of this region of the capsid as a possible determinant for encapsidation. The arrangement of basic amino acid side chains within an interior depression in the capsid makes this region of the capsid an attractive candidate for further analyses. Finally, mutations in the maturation cleavage site have profound effects on the capacity of the processed viral proteins to assemble and encapsidate genomic RNA. The existence of the provirion as an assembly intermediate was supported from the analysis of a capsid mutation containing a threonineserine mutation cleavage site. In summary, the use of recombinant vaccinia virus vectors for analyzing the processes of poliovirus capsid assembly and RNA encapsidation overcomes limitations of previous intracellular assembly analyses which required isolation of mutant polioviruses subject to the potential for reversion. The systems described offer the dual capability of analyzing P1 capsid precursor cleavage and subviral particle formation separately from the encapsidation step. Information from the three-dimensional structure of poliovirus and its picornavirus relatives provides a rational basis for targeting regions of the capsid for mutagenesis studies. The use of recombinant vaccinia viruses with defined mutations in P1 in combination with PVdefSM might allow for scaleup and recovery of sufficient virus particles for structural analysis. Generation of additional poliovirus capsid mutants with defined mutations to be analyzed by these new methods will, it is hoped, further an understanding of the molecular mechanisms of poliovirus morphogenesis.

ACKNOWLEDGMENTS We thank Dee Martin for the preparation of the manuscript. M.J.A. was supported by a National Institutes of Health training grant (T32-A1 07150). Research was supported by a grant from the National Institutes of Health, National Institute for Allergy and Infectious Disease (A1 25005). to C.D.M.

REFERENCES Acharya, R., Fry, E., Stuart, G., Fox, G., Rowlands, D., and Brown, F. (1989). The threedimensional structure of foot-and-mouth disease virus at 2.9 A resolution. Nature (London) 337,709-716.

POLIOVIRUS ASSEMBLY AND RNA ENCAF'SIDATION

57

Almond, J. W., and Burke, K. L. (1990). Poliovirus as a vector for the presentation of foreign antigens. Sem. Virol. 1, 11-20. Ambros. V., and Baltimore, D. (1978). Protein is linked to the 5' end of poliovirus RNA by a phosphodiester linkage to tyrosine. J. Biol. Chem. 253, 922-928. Andino, R., Rieckhof, G. E., and Baltimore, D. (1990). A functional ribonucleoprotein complex forms around the 5' end of poliovirus RNA. Cell (Cambridge,Mass.) 63,369380. Ansardi, D. C., and Morrow, C. D. (1993). Poliovirus capsid proteins derived from P1 precursors with glutamine-valine cleavage sites have defects in assembly and RNA encapsidation. J. Virol. 67, 7284-7297. Ansardi, D. C., and Morrow, C. D. (1995). Amino acid substitutions in the poliovirus maturation cleavage site affect assembly and results in accumulation of provirions. J. Virol. 69, 1540-1547. Ansardi, D. A., Porter, D. C., and Morrow, C. D. (1991). Coinfection with recombinant vaccinia viruses expressing poliovirus P1 and P3 proteins results in polyprotein processing and formation of empty capsid structures. J. Virol. 65, 2088-2092. Ansardi, D. C., Porter, D. C., and Morrow, C. D. (1992). Myristylation of poliovirus capsid precursor P1 is required for assembly of subviral particles. J. Virol. 66, 4556-4563. Ansardi, D. C., Porter, D. C., and Morrow, C. D. (1993). Complementation of a poliovirus defective genome by a recombinant vaccinia virus which provides P1 capsid precursor in trans. J . Virol. 67, 3684-3690. Ansardi, D. C., Luo, M., and Morrow, C. D. (1994a). Mutations in the poliovirus P1 capsid precursor at arginine residues VP4-ARG34, VP3-ARG223, and VP1-ARG129 affect virus assembly and encapsidation of genomic RNA. Virology 199, 20-34. Ansardi, D. C., Moldoveanu, Z., Porter, D. C., Walker, D. P., Conry, R. M., LoBuglio, A. F., McPherson, S., and Morrow, C. D. (1994b). Characterization of poliovirus replicons encoding carcinoembryonic antigens. Cancer Res. 54,6359-6363. Arnold, E., Luo, M., Vriend, G., Rossman, M. G., Palmenberg, A. C., Parks, G. D., Nicklin, M. J. H., and Wimmer, E. (1987). Implications of the picornavirus capsid structure for polyprotein processing. Proc. Natl. Acad. Sci. U.S.A.84, 21-25. Badger, J., Minor, I., Oliveria, M. J. M. A., Smith, T.J., Griffith, J. P., Guerin, D. M. A., Krishnaswamy, S., Luo, M., Rossman, M. G., McKinlay, M. A., Diana, G. D., Dutko, F. J., Francher, M., Rueckert, R. R., and Heinz, B. A. (1988). Structural analysis of a series of antiviral agents complexed with human rhinovirus 14. Proc. Natl. Acad. Sci. U S A . 85, 3304-3308. Bazan, J. F., and Fletterick, F. J. (1988). Viral cysteine proteases are homologous to the trypsin-like family of serine proteases: Structural and functional implications. Proc. Natl. Acad. Sci. U.S.A.85, 7872-7876. Beckman, R. P., Mizzen, L. A., and Welch, W. J. (1990). Interaction of hsp7O with newly synthesized proteins: Implications for protein folding and assembly. Science 284,850854. Belsham, G. J., Abrams, C. C., King, A. M. Q., Roosien, J., and Vlak, J. M. (1991). Myristoylation of foot-and-mouth disease virus capsid protein precursors is independent of other viral proteins and occurs in both mammalian and insect cells. J. Gen. Virol. 72, 747-751. Bernstein, H. D., Sonenberg, N., and Baltimore, D. (1985). Poliovirus mutant that does not selectively inhibit host cell protein synthesis. Mol. Cell. Biol. 5, 2913-2923. Bernstein, H. D., Sarnow, P., and Baltimore, D. (1986). Genetic complementation among poliovirus mutants derived from an infectious cDNA. J. Virol. 60, 1040-1049. Blair, W. S., and Semler, B. L. (1991). Role for the P4 amino acid residue in substrate utilization by the poliovirus 3CD proteinase. J. Virol. 65, 6111-6123.

58

DAVID C. ANSARDI et al.

Blair, W. S., Li, X., and Semler, B. L. (1993). A cellular cofactor facilitates efficient 3CD ' cleavage of the poliovirus P1 precursor. J. Virol. 67, 2336-2343. Borovec, S. V., and Anderson, D. A. (1993). Synthesis and assembly of hepatitis A virusspecific proteins in BSC-1 cells. J . Virol. 67, 3095-3102. Bruneau, P., Blondel, B., Crainic, R., Horodniceanu, F., and Girard, M. (1983).Poliovirus type 1 capsid polypeptides: Absence of a free form in the cytoplasm of infected HeLa cells. Ann. Virol. Znst. Pasteur 134E, 151-164. Bryant, M., and Ratner, L. (1990). Myristoylation-dependent replication and assembly of human immunodeficiency virus 1. Proc. Natl. Acad. Sci. U.S.A.87, 523-527. Burness, A. T. H., and Clothier, F. W. (1970). Particle weight and other biophysical properties of encephalomyocarditis virus. J. Gen. Virol. 6, 381-393. Buss, J. E., Solski, P. A., Schaeffer, J. P., MacDonald, M. J., and Der, C. J. (1989). Activation of the cellular proto-oncogene product p21 by addition of a myristylation signal. Science 243, 1600-1603. Caliguiri, L. A., and Compans, R. W. (1973). The formation of poliovirus particles in association with the RNA replication complexes. J. Gen. Virol. 21, 99-108. Caliguiri, L. A., and Mosser, A. G. (1971). Protein associated with the poliovirus RNA replication complex. Virology 46, 375-386. Caliguiri, L. A., and Tamm, I. (1968). Action of guanidine on the replication of poliovirus RNA. Virology 35, 408-417. Caliguiri, L. A., and Tamm, I. (1970). The role of cytoplasmic membranes in poliovirus biosynthesis. Virology 42, 100-111. Calnan, B. J., Tidor, B., Biancalana, D., Hudson, D., and Frankel, A. D. (1991).Argininemediated RNA recognition: The arginine fork. Science 252, 1167-1171. Caspar, D. L. D., and Klug, A. (1962). Physical principles in the construction of regular viruses. Cold Spring Harbor Symp. Quant. Biol. 27, 1-24. Chakrabarti, S., Brechling, K., and Moss, B. (1985).Vaccinia virus expression vector: Coexpression of p-galactosidase provides visual screening of recombinant virus plaques. Mol. Cell. Biol. 5, 3403-3409. Chen, Z., Stauffacher, C., Li, Y., Schmidt, T., Bomu, W., Kamer, G., Shanks, M., Lomonossoff, G., and Johnson, J. E. (1989). Protein-RNA interactions in a n icosahedral virus at 3.0 A resolution. Science 245, 154-159. Choi, H. K., Tong, L., Minor, W., Dumas, P., Boege, U., Rossman, M. G., and Wengler, G. (1991). Structure of Sindbis virus core protein reveals a chymostrypsin-like serine proteinase and the organization of the virion. Nature (London) 354, 37-43. Choi, W.-S., Pal-Ghosh, R., and MOROW,C. D. (1991). Expression of human immunodeficiency virus type 1 HIV-1 gag, pol and env proteins from chimeric HIV-1-poliovirus minireplicons. J. Virol. 65, 2875-2883. Chow, M., Newman, J. F. E., Filman, D. J., Hogle, J. M., Rowlands, D. J., and Brown, F. (1987). Myristylation of picornavirus capsid protein VP4 and its structural significance. Nature (London) 327,482-486. Cole, C. N. (1975). Defective interfering (DI) particles of poliovirus. Prog. Med. Virol. 20, 180-207. Cole, C. N., and Baltimore, D. (1973a). Defective interfering particles of poliovirus. 11. Nature of the defect. J. Mol. Biol. 76, 325-343. Cole, C. N., and Baltimore, D. (1973b). Defective interfering particles of poliovirus. 111. Interference and enrichment. J. Mol. Biol. 76, 345-361. Cole, C. N., and Baltimore, D. (1973~).Defective interfering particles of poliovirus. IV. Mechanisms of enrichment. J. Virol. 12, 1414-1426. Cole, C. N., Smoler, D., Wimmer, E., and Baltimore, D. (1971). Defective interfering particles of poliovirus. I. Isolation and physical properties. J. Virol. 7 , 478-485.

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

59

Compton, S. R., Nelsen, B., and Kirkegaard, K. (1990). Temperature-sensitive poliovirus mutant fails to cleave VPO and accumulates provirions. J. Virol. 64, 4067-4075. Corrias, M. V., Flore, O., Broi, E., Marongiu, M. E., Pani, Torelli, S., and LaColla, P. (1987). Characterization and role in morphogenesis of a new subviral particle (555) isolated from poliovirus-infected cells. J . Virol. 61, 561-569. Das, S., and Dasgupta, A. (1993). Identification of the cleavage site and determinants required for poliovirus 3Cpro-catalyzed cleavage of human TATA-binding transcription factor TBP. J. Virol. 67, 3326-3331. De La Torre, J. C., Giachetti, C., Semler, B. L., and Holland, J. J. (1992).High frequency of single-base transitions and extreme frequency of precise multiple-base reversion mutations in poliovirus. Proc. Natl. Acad. Sci. U S A . 89, 2531-2535. DeSena, J., and Mandel, B. (1976).Studies on the in uitro uncoating of poliovirus. I. Characterization of the modifying factor and the modifying reaction. Virology 70,470-483. DeSena, J., and Mandel, B. (1977). Studies on the in uitro uncoating of poliovirus. 11. Characterization of the membrane-modified particle. Virology 78, 554-566. Dewalt, P. G., and Semler, B. L. (1987). Site-directed mutagenesis of proteinase 3C results in a poliovirus deficient in synthesis of viral RNA polymerase. J. Virol. 61, 2162-2170. Dewalt, P. G., and Semler, B. L. (1989). Molecular biology and genetics of poliovirus protein processing. In “Molecular Aspects of Picornavirus Infection and Detection” (B. L. Semler and E. Ehrenfeld, eds.), pp. 73-93. American Society for Microbiology, Washington, D. C. Dorner, A. J., Dorner, L. F., Larsen, G. R., Wimmer, E., and Anderson, C. W. (1982). Identification of the initiation site of poliovirus polyprotein synthesis. J. Virol. 42, 1017-1028. Ehrenfeld, E. J., Maizel, J. V., and Summer, D. F. (1970). Soluble RNA polymerase complex from poliovirus-infected HeLa cells. Virology 40, 840-846. Etchison, D., Milburn, S. C., Edery, I., Sonenberg, N., and Hershey, J. W. B. (1982). Inhibition of HeLa cell protein synthesis following poliovirus infection correlates with the proteolysis of a 220,000-dalton polypeptide associated with eukaryotic initiation factor 3 and a cap binding protein complex. J. Biol. Chem. 257, 14806-14810. Evans, D. J., McKeating, J., Meredith, K. L., Burke, K. L., Katrack, K., John, A., Ferguson, P. D., Minor, P. D., Weiss, R. A., and Almond, J. W. (1989). An engineered poliovirus chimera elicits broadly reactive HIV-1 neutralizing antibodies. Nature (London) 339, 385-388. Everaert, L., Vrijsen, R., and Boeye, A. (1989). Eclipse products of poliovirus after coldsynchronized infection of HeLa cells. Virology 171, 76-82. Fernandez-Thomas, C. B., and Baltimore, D. (1973). Morphogenesis of poliovirus. 11. Demonstration of new intermediates, the provirion. J. Virol. 12, 1122-1130. Fernandez-Thomas, C. B., Guttman, N., and Baltimore, D. (1973). Morphogenesis of poliovirus. 111. Formation of provirion in cell-free extracts. J. Virol. 12, 1181-1183. Filman, D. J., Syed, R., Chow, M., Macadam, A. J., Minor, P. D., and Hogle, J. M. (1989). Structural factors that control conformational transitions and serotype specificity in type 3 poliovirus. EMBO J . 8, 1567-1579. Fisher, A. J., and Johnson, J. E. (1993).Ordered duplex RNA controls capsid architecture in an icosahedral animal virus. Nature (London) 361, 176-179. Fiszman, M., Reynier, M., Bucchini, D., and Girard, M. (1972). Thermosensitive block of the Sabin strain of poliovirus type 1. J . Virol. 10, 1143-1151. Fox, M. P., Otto, J. J., and McKinlay, M. A. (1986). The prevention of rhinovirus and poliovirus uncoating by WIN 51711: A new antiviral drug. Antimicrob. Agents Chemother. 30, 110-116.

60

DAVID C. ANSARDI et al.

Frankel, A. D., Mattaj, I. W., and Rio, D. C. (1991). RNA-protein interactions. Cell (Cambridge, Mass.) 67, 1041-1046. Fricks, C. E., and Hogle, J. M. (1990). Cell-induced conformational change in poliovirus: Externalization of the amino-terminus of VP1 is responsible for liposome binding. J . Virol. 64, 1934-1945. Gauntt, C., Gilbert, S. F., Grieves, J.,Anderegg, J.,and Rueckert, R. (1981).A neutralizing monoclonal antibody against poliovirus and its reaction with related antigens. Virology 115, 211-215. Geigenmuller-Gnirke, U., Nitschko, H., and Schlesinger, S. (1993). Deletion analysis of the capsid protein of Sindbis virus: Identification of the RNA binding protein. J. Virol. 67, 1620-1626. Ghendon, Y., Yakobson, E., and Mikhejeva, A. (1972). Study of some stages of poliovirus morphogenesis in Mi0 cells. J. Virol. 10, 261-266. Ghosh, R. P., and Morrow, C. D. (1993). A poliovirus mini-replicon containing an inactive 2A proteinase is expressed in vaccinia virus infected cells. J. Virol. 67, 46214629. Giachetti, C., and Semler, B. L. (1991). Role of a viral membrane polypeptide in strandspecific initiation of poliovirus RNA synthesis. J. Virol. 66, 2647-2654. Girard, M., Baltimore, D., and Darnell, J. E. (1967). The poliovirus replication complex: Site for synthesis of poliovirus RNA. J. Mol. Biol. 24, 59-74. Gorbalenya, A., Donchenko, A. P., Blinov, M., and Koonin, E. V. (1989). Cysteine proteases of positive strand RNA viruses and chymotrypsin-like serine proteases: A distinct protein superfamily with a common structural fold. FEBS Lett. 243, 103114. Gott, J. M., Wilhelm, L. J., and Uhlenbeck, 0. C. (1991). RNA binding properties of the coat protein from bacteriophage GA. Nucleic Acids Res. 19, 6499-6503. Guttman, N., and Baltimore, D. (1977a). A plasma membrane component able to bind and alter virions of poliovirus type 1: Studies on cell-free alteration using a simplified assay. Virology 82, 25-36. Guttman, N., and Baltimore, D. (1977b). Morphogenesis of poliovirus. IV. Existence of particles sedimenting at 150s and having the properties of provirion. J. Virol. 23, 363-367. Hadzopoulou-Cladaras, M., Felber, B. K., Cladaras, C., Atganas-Sopoulos, A., Tse, A., and Pavlakis, G. N. (1989).The rev (trs/art)protein of human immuodeficiency virus type 1affects viral mRNA and protein expression via a cis-acting sequence in the env region. J. Virol. 63, 1265-1274. Hagino-Yamagishi, K., and Nomoto, A. (1989). In uitro construction of poliovirus defective interfering particles. J. Virol. 63, 5386-5392. Hanecak, R., Semler, B. L., Anderson, C. W., and Wimmer, E. (1982). Proteolytic processing of poliovirus polypeptides: Antibodies to polypeptide P3-7c inhibit cleavage of glutamine-glycine pairs. Proc. Natl. Acad. Sci. U.S.A.79, 3973-3977. Hanecak, R., Semler, B. L., Ariga, H., Anderson, C. W., and Wimmer, E. (1984). Expression of a cloned gene segment of poliovirus in E . coli: Evidence for autocatalytic production of the viral proteinase. Cell (Cambridge, Mass.) 37, 1063-1073. Harber, J. J., Bradley, J., Anderson, C. W., and Wimmer, E. (1991). Catalysis of poliovirus VPO maturation cleavage is not mediated by serine 10 of VP2. J. Virol. 65, 326-334. Harris, K. S., Hellen, C. U. T., and Wimmer, E. (1990). Proteolytic processing in the replication of picornaviruses. Semin. Virol. 1, 323-333, Harris, K. S., Reddigari, S. R., Nicklin, J. H., and Wimmer, E. (1992). Purification and characterization of poliovirus polypeptide 3CD, a proteinase and a precursor for RNA polymerase. J. Virol. 66, 7481-7489.

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

61

Hellen, C. U. T., and Wimmer, E. (1992a). Maturation of poliovirus capsid proteins. Virology 187, 391-397. Hellen, C. U. T., and Wimmer, E. (1992b). The role of proteolytic processing in the morphogenesis of virus particles. Erperientia 48, 201-215. Hellen, C. U. T., Lee, C. K., and Wimmer, E. (1992). Determinants of substrate recognition by poliovirus 2A proteinase. J. Virol. 66,3330-3338. Heuckeroth, R. O., and Gordon, J. I. (1989). Altered membrane association of p60 v-src and a murine 63 kDa N-myristoylprotein after incorporation of an oxygen-substituted analog of myristic acid. Proc. Natl. Acad. Sci.U S A . 86, 5262-5266. Hewlett, M. J., Rose, J. K., and Baltimore, D. (1976). 5’-Terminal structure of poliovirus polyribosomal RNA is pup. Proc. Natl. Acad. Sci. U.S.A. 73, 327-330. Hogle, J. M., Chow, M., and Filman, D. J. (1985).Three-dimensional structure of poliovirus at 2.9 A resolution. Science 229, 1358-1365. Ivanoff, L. A., Towari, T., Ray, J., Korant, B. D., and Petteway, S. R., Jr. (1986). Expression and site-specific mutagenesis of the poliovirus 3C protease in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A.83, 5392-5396. Jacobson, M. F., and Baltimore, D. (1968). Morphogenesis of poliovirus. I. Association of the viral RNA with the coat protein. J. Mol. B i d 33, 369-378. Jacobson, M. F., Asso, J., and Baltimore, D. (1970). Further evidence on the formation of poliovirus proteins. J. Mol. Biol. 49, 657-669. Jang, S. K., Krausslich, H. G., Hicklin, M. J. H., Duke, G. M,. Palmenberg, A. C., and Wimmer, E. (1988). A segment of the 5’nontranslated region of encephalomyocarditis virus RNA directs internal entry of ribosomes during in vitro translation. J. Virol. 62, 2636-2643. Jenkins, O., Cason, J., Burke, K. L., Lunney, D., Gillen, A., Patel, D., McCance, D. J., and Almond, J. W. (1990). An antigen chimera of poliovirus induces antibodies against human papillomavirus type 16. J. Virol. 64, 1201-1206. Jewell, J. E., Ball, L. A., and Rueckert, R. (1990). Limited expression of poliovirus by vaccinia virus recombinants due to inhibition of the vector by proteinase 2A. J. Virol. 64, 1388-1393. Johnson, D. R., Cox, A. D., Solski, P. A., Devadas, B., Adams, S. P., Leimgruber, R. M., Heuckeroth, R. O., Buss, J. E., and Gordon, J. L. (1990). Functional analysis of protein N-myristoylation: Metabolic labeling studies using three oxygen-substituted analogs of myristic acid and cultured mammalian cells provide evidence for protein-sequencespecific incorporation and analogspecific redistribution. Proc. Natl. Acad. Sci.U.S.A. 87, 8511-8515. Johnson, K. L., and Sarnow, P. (1991). Three poliovirus 2B mutants exhibit noncomplementable defects in viral RNA amplification and display dosage-dependent dominance over wild-type poliovirus. J. Virol. 65, 4341-4349. Jore, J., DeGeus, B., Jackson, R. J., Pouwels, P. H., and Enger-Valk, B. E. (1988). Poliovirus protein 3CD is the active protease for processing of the precursor protein P1 in vitro. J. Gen. Virol. 69, 1627-1636. Kajigaya, S., Arakawa, H., Kuge, S., Koi, T., Imura, N., and Nomoto, A. (1985). Isolation and characterization of defective-interfering particles of poliovirus Sabin 1 strain. Virology 142, 307-316. Kaplan, G., and Racaniello, V. R. (1988). Construction and characterization of poliovirus subgenomic replicons. J. Virol. 62, 1687-1696. Kaplan, G., Freistadt, M. S., and Racaniello, V. R. (1990). Neutralization of poliovirus by cell receptors expressed in insect cells. J . Virol. 64,4697-4702. Kean, K. M., Teterina, N., and Girard, M. (1990). Cleavage specificity of the poliovirus 3C protease in not restricted to Gln-Gly a t the 3C/3D junction. J. Gen. Virol. 71, 2553-2563.

62

DAVID C. ANSARDI et al.

Kirkegaard, K. (1990). Mutations in VP1 of poliovirus specifically affect both encapsidation and release of viral RNA. J. Virol. 64, 195-206. Kirkegaard, K., and Nelson, B. (1990). Conditional poliovirus mutants made by random deletion mutagenesis of infectious cDNA. J. Virol. 64, 185-194. Kitamura, N., Semler, B. L., Rothberg, P. G., Larsen, G. R., Adler, C. J., Dorner, A. J., Emini, E. A., Hanecak, R., Lee, J. J., van der Werf, S., Anderson, C. W., and Wimmer, E. (1981). Primary structure, gene organization and polypeptide expression of poliovirus RNA. Nature (London) 291, 547-553. Kitson, J. D. A., Burke, K. L., Pullen, L. A., Belsham, G. J., and Almond, J. W. (1991). Chimeric polioviruses that include sequences derived from two independent antigenic sites of foot-and-mouth disease virus (FMDV) induce neutralizing antibodies against FMDV in guinea pigs. J. Virol. 65, 3068-3075. Koch, F., and Koch, G. (1985). “The Molecular Biology of Poliovirus.” Springer-Verlag, Vienna. Kozak, M. (1989). The scanning model for translation: An update. J. Cell Biol. 198,228241. Krausslich, H. G., Nicklin, M. J. H., Toyoda, H., Etchison, D., and Wimmer, E. (1987). Poliovirus proteinase 2A induces cleavage of eucaryotic initiation factor 4F polypeptide p220. J. Virol. 61, 2711-2718. Krausslich, H. G., Holscher, C., Reuter, Q., Harber, J., and Wimmer, E. (1990). Myristoylation of the poliovirus polyprotein is required for proteolytic processing of the capsid and for viral infectivity. J. Virol. 64, 2433-2436. Kuge, S., Saito, I., and Nomoto, A. (1986). Primary structure of poliovirus defectiveinterfering particle genomes and possible generation mechanisms of the particles. J . Mol. Biol. 192, 473-487. Kuhn, R. J., and Wimmer, E. (1987). The replication of picornaviruses. In “Molecular Biology of Positive Strand RNA Viruses” (D. J. Rowlands, M. A. Maho, and B. W. J. Mahy, eds.), pp. 17-51. Academic Press, Orlando, Florida. Larsen, G. R., Forner, A. J., Harris, T. J. R., and Wimmer, E. (1980). The structure of poliovirus replicative form. Nucleic Acids Res. 8, 1217-1229. Larson, S. B., Koszelak, S., Day, J., Greenwood, A., Dodds, J. A., and McPherson, A. (1993). Double helical RNA in satellite tobacco mosaic virus. Nature (London) 361, 179-182. Lawson, M. A., and Semler, B. L. (1990). Picornavirus protein processing-enzymes, substrates, and genetic regulation. Curr. Top. Micobiol. Immunol. 161, 49-87. Lawson, M. A., and Semler, B. L. (1991). Poliovirus thiol proteinase 3C can utilize a serine nucleophile within the putative catalytic triad. Proc. Natl. Acad. Sci. U S A . 88,9919-9923. Lawson, M. A., and Semler, B. L. (1992). Alternate poliovirus nonstructural protein processing cascades generated by primary sites of 3C proteinase cleavage. Virology 191,309-320. Lazinski, D., Grzadzielska, E., and Das, A. (1989). Sequence-specific recognition of RNA hairpins by bacteriophage antiterminators requires a conserved arginine-rich motif. Cell (Cambridge, Mass.) 59, 207-218. Lee, C., and Colter, J. S. (1979). Further characterization of mengo subviral particles: A new hypothesis for picornavirus assembly. Virology 97, 266-274. Lee, C., and Wimmer, E. (1988). Proteolytic processing of poliovirus polyprotein: Elimination of 2A pro mediated, alternative cleavage of polypeptide 3CD by in uitro mutagenesis. Virology 166, 404-414. Lee, C., Paucha, E., and Colter, J. S. (1978). Identification and partial characterization of a new (50s) particle in mengo virus-infected L cells. Virology 85, 286-295.

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

63

Lee, W.-M., Monroe, S. S., and Rueckert, R. R. (1993). Role of maturation cleavage in infectivity of picornaviruses: Activation of an infectosome. J. Virol. 67, 2110-2122. Lee, Y. F., Nomoto, A., Detjen, B. M., and Wimmer, E. (1977).A protein covalently linked to poliovirus genome RNA. Proc. Natl. Acad. Sci. U S A . 74, 59-63. Lloyd, R. E., Grubman, M. J., and Ehrenfeld, E. (1988). Relationship of p220 cleavage during picornavirus infection to 2A protease sequences. J. Virol. 62,4216-4223. Lundquist, R. E., Sullivan, M., and Maizer, J. V. (1979). Characterization of a new isolate of poliovirus defective-interfering particles. Cell (Cambridge, Mass.) 18, 759-769. Luo, M. G., Vriend, G., Kamer, G., Minor, I., Arnold, E., Rossman, M. G., Boege, U., Scraba, D. G., Duke, G. M., and Palmenberg, A. C. (1987). The atomic structure of mengo virus a t 3.0 A resolution. Science 235, 182-191. Macadam, A. J., Fergurson, G., Arnold, C., and Minor, P. D. (1991). An assembly defect as a result of an attenuating mutation in the capsid proteins of the poliovirus type 3 vaccine strain. J. Virol. 65, 5225-5231. Macejak, D. G., and Sarnow, P. (1992).Association of heat shock protein 70 with enterovirus capsid precursor P1 in infected human cells. J. Virol. 66, 1520-1527. McGreggor, S., and Rueckert, R. R. (1977). Picornaviral capsid assembly: Similarity of rhinovirus and enterovirus precursor subunits. J. Virol. 21, 548-553. McGreggor, S., Hall, L., and Rueckert, R. R. (1975). Evidence for the existence of protomers in the assembly of encephalomyocarditis virus. J. Virol. 15, 1107-1120. McKenna, R., Xia, D., Willingmann, P., Ilag, L. L., Krishmaswamy, S., Rossman, M. G., Olson, N. H., Baker, T.S., and Incaradona, N. L. (1992). Atomic structure of singlestranded DNA bacteriophage +X174 and its functional implication. Nature (London) 355, 137-143. Mackett, M. G., Smith, L., and Moss, B. (1985). The construction and characterization of vaccinia virus recombinants expressing foreign genes. In “DNA Cloning” (D. Glover, ed.), Vol. 2, pp. 191-212. IRL Press, Oxford. Madshus, I. H., Olsmes, S., and Sandvig, K. (1984). Requirements for entry of poliovirus RNA into cells at low pH. EMBO J. 3, 1945-1950. Madshus, I. H., Olsmes, S., and Sandvig, K. (1985). Mechanism of entry into the cytosol of poliovirus type I: Requirement for low pH. J . Cell Biol. 98, 1194-1200. Maizel, J. V., Phillips, B. A., and Summers, D. F. (1967). Composition of artificially produced and naturally occurring empty capsids of poliovirus type 1. Virology 32, 692-699. Mandel, B. (1965). The fate of the inoculum in HeLa cells infected with poliovirus. Virology 25, 152-154. Mapoles, J. E., Anderegg, J. W., and Rueckert, R. R. (1978). Properties of poliovirus propagated in medium containing cesium chloride: Implications for picornaviral structure. Virology 90, 103-111. Marc, D., Drugeon, G., Haenni, A. L., Girard, M., and van der Werf, S. (1989). Role of myristoylation of poliovirus capsid protein VP4 as determined by site-directed mutagenesis of its N-terminal sequence. EMBO J. 8, 2661-2668. Marc, D., Masson, G., Girard, M., and van der Werf, S. (1990). Lack of myristoylation of poliovirus capsid polypeptide VPO prevents the formation of virions or results in the assembly of noninfectious virus particles. J. Virol. 64, 4099-4107. Marc, D., Girard, M., and van der Werf, S. (1991). A Gly-1 to Ala substitution in poliovirus capsid protein VPO blocks its myristoylation and prevents viral assembly. J.Gen. Virol. 72, 1151-1157. Maronginu, M. E., Pani, A., Corrias, M. V., Sau, M., and La Colla, P. (1981). Poliovirus morphogenesis. I. Identification of 80s dissociable particles and evidence for the artifactual production of procapsids. J . Virol. 39, 341-347.

64

DAVID C. ANSARDI et al.

Mattaj, I. (1993). RNA recognition: A family matter? Cell (Cambridge, Mass.) 73,837840. Medappa, K. D., and Rueckert, R. R. (1974). Binding of cesium atoms to human rhinovirus 14. Abstr. Annu. Meet. Soc. Microbiol. V43, 207. Mendelsohn, C. L., Wimmer, E., and Racaniello, V. R. (1989). Cellular receptor for poliovirus: Molecular cloning, nucleotide sequence, and expression of a new member of the immunoglobulin superfamily. Cell (Cambridge, Mass.) 56, 855-865. Minor, P. D. (1990). Antigenic structure of picornaviruses. Curr. Top. Microbiol. Zmmunol. 161, 121-153. Mirzayan, C., Ingraham, R., and Wimmer, E. (1991). Specificity of the poliovirus proteinase 3C towards genetically engineered cleavage sites in the viral capsid. J. Gen. Virol. 137, 1159-1163. Molla, A., Paul, V., and Wimmer, E. (1991). Cell-free, de nouo synthesis of poliovirus. Science 254, 1647-1650. Molla, A., Jang, S., Paul, A. B., Reuer, Q.,and Wimmer, E. (1992). Cardioviral internal ribosome entry site is functional in a genetically engineered dicistronic poliovirus. Nature (London) 356, 255-257. Morrow, C. D., Hocko, J., Navab, M., and Dasgupta, A. (1984).ATP required for initiation of poliovirus RNA synthesis: In uitro demonstration of tyrosine-phosphate linkage between in uitro synthesized RNA and genome-linked protein. J. Virol. 50,515-523. Moscufo, N., and Chow, M. (1992). Myristate-protein interactions in poliovirus: Interactions of VP4 threonine 28 contribute to the structure conformation of assembly intermediates and the stability of assembled virions. J. Virol. 66, 6849-5857. Moscufo, N., Simons, J., and Chow, M. (1991). Myristoylation is important at multiple stages in poliovirus assembly. J. Virol. 65, 2372-2380. Nicklin, M. J. H., Harris, K. S., Pallai, P. V., and Wimmer, E. (1988). Poliovirus proteinase 3C: Large-scale expression, purification and specific cleavage activity on natural and synthetic substrates in uitro. J. Virol. 62, 4586-4593. Nomoto, A., Lee, Y.F., and Wimmer, E. (1976). The 5' end of poliovirus mRNA is not capped with m7G(5')ppp(5')Np. Proc. Natl. Acad. Sci. U.S.A.73, 375-380. Nomoto, A., Kitamura, N., Golini, F., and Wimmer, E. (1977). The 5'4erminal structures of poliovirus RNA and poliovirus mRNA differ only in the genome-linked protein VPg. Proc. Natl. Acad. Sci. U S A . 74, 5345-5349. Novak, J. E., and Kirkegaard, K. (1991). Improved method for detecting poliovirus negative strands used to demonstrate specificity of positive-strand encapsidation and the ratio of positive to negative stand in infected cells. J. Virol. 65, 3384-3387. Olson, N. H., Kolatkar, P. R., Oliveria, M. A., Cheng, R. H., Greve, J. M., McClelland, A,, Baker, T.S., and Rossman, M. G. (1993). Structure of human rhinovirus complexed with its receptor molecule. Proc. Natl. Acad. Sci. U.S.A. 90, 507-511. Onodera, S., and Phillips, B. A. (1987). A novel method for obtaining poliovirus 1 4 s pentamers from procapsids and their self-assembly into virus-like shells. Virology 159,278-287. Onodera, S., Cardamone, J. J., Jr., and Phillips, B. A. (1986). Biological activity and electron microscopy of poliovirus 1 4 s particles obtained from alkali-dissociated procapsids. J. Virol. 58, 610-618. Page, G. S., Mosser, A. G., Hogle, J. M., Filman, D. J., Rueckert, R. R., and Chow, M. (1988). Three-dimensional structure of poliovirus serotype 1 neutralizing determinants. J. Virol. 62, 1781-1794. Pallai, P. V., Burkhardt, F., Skoog, M., Schreiner, K., Bax, P., Cohen, K. A., Hansen, G., Palladino, D. E. H., Harris, K. S., Nicklin, M. J:, and Wimmer, E. (1989). Cleavage of synthetic peptides by purified poliovirus 3C proteinase. J. Biol. Chem. 264, 97389741.

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

65

Palmenberg, A. C. (1982). In vitro synthesis and assembly of picornaviral capsid intermediate structures. J.Virol. 44,900-906. Palmenberg, A. C. (1990). Proteolytic processing of picornaviral polyprotein. Annu. Rev. Microbiol. 44,603-623. Parks, G. D., and Palmenberg, A. C. (1987). Site-specific mutations at a picornavirus VP3/VP1 cleavage site disrupt in vitro processing and assembly of capsid precursors. J. Virol. 61, 3680-3687. Paul, A. V., Schultz, A., Pincus, S. E., Oroszlan, S., and Wimmer, E. (1987a). Capsid protein VP4 of poliovirus is N-myristoylated. Proc. Natl. Acad. Sci. U.S.A.84,78277831. Paul, A. V., Yang, C. F., Jang, S. K., Kuhn, R. J., Tada, H., Nicklin, M., Krausslich, H. G., Lee, C. K., and Wimmer, E. (1987b). Molecular events leading to poliovirus genome replication. Cold Spring Harbor Symp. Quant. Biol. 52, 343-352. Pelham, H. (1988).Heat shock proteins: Coming in from the cold. Nature (London) 332, 776-777. Pelletier, J., and Sonenberg, N. (1988). Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature.(London) 334, 320-325. Pelletier, J., Kaplan, G., Racaniello, V. R., and Sonenberg, N. (1988). Cap-independent translation of poliovirus mRNA is conferred by sequence elements within the 5’ noncoding region. Mol. Cell. Biol. 8, 1103-1112. Percy, N., Barclay, W. S., Sullivan, M., and Almond, J. W. (1992). A poliovirus replicon containing the chloramphenicol acetyltransferase gene can be used to study the replication and encapsidation of poliovirus RNA. J. Virol. 66, 5040-5046. Petterson, R. F., Flangan, J. B., Rose, J. K., and Baltimore, D. (1977). 5’-Terminal nucleotide sequence of poliovirus polyribosomal RNA and virion RNA are identical. Nature (London) 268,270. Petterson, R. F., Ambros, V., and Baltimore, D. (1978). Identification of a protein linked to nascent poliovirus RNA and to the polyuridylic acid of negative-strand RNA. J . Virol. 27, 357-365. Pfister, T., Pasamontes, L., Troxler, M., Egger, D., and Bienz, K. (1992). Immunocytochemical localization of capsid-related particles in subcellular fractions of poliovirus-infected cells. Virology 188, 676-684. Phillips, B. A. (1969). In vitro assembly of poliovirus. I. Kinetics of assembly of empty capsids and the role of extracts from infected cells. Virology 39, 811-821. Phillips, B. A. (1971).In vitro assembly of polioviruses. 11. Evidence for the self-assembly of 1 4 s particles into empty capsids. Virology 44,307-316. Phillips, B. A,, and Wiemert, S. (1978). In uitro assembly of poliovirus. V. Evidence that the self-assembly activity of 1 4 s particles is independent of extract assembly factorb) and host proteins. Virology 88, 92-104. Phillips, B. A., Summers, D. F., and Maizel, J. V. (1968). I n vitro assembly of poliovirus related particles. Virology 35, 216-226. Phillips, B. A,, Lundquist, R. E., and Maizel, J. V. (1980). Absence of subviral particles and assembly activity in HeLa cells infected with defective-interferring (DI) particles of poliovirus. Virology 100, 116-124. Porter, D. C., Ansardi, D. C., Choi, W. S., and Morrow, C. D. (1993a). Encapsidation of genetically engineered poliovirus mini-replicons which express HIV-1 gag and pol proteins upon infection. J. Virol. 67, 3712-3719. Porter, D. C., Lentz, M. R., and Morrow, C. D. (1993b). Expression of poliovirus P3 proteins using a recombinant vaccinia virus results in proteolytically active 3CD precursor proteins without further processing to 3Cpro and 3D~ol.Virus Res. 29,241-254. Porter, D. C., Ansardi, D. C., and Morrow, C. D. (1995). Encapsidation of poliovirus

66

DAVID C. ANSARDI et al.

replicons encoding the complete human immunodeficiency virus type 1 gag gene using a complementation system which provides the P1 capsid protein in trans. J . Virol. 69, 1548-1555. Putnak, J. R., and Phillips, B. A. (1981a). Picornaviral structural and assembly. Microbwl. Rev. 45, 287-315. Putnak, J. R., and Phillips, B. A. (1981b). Differences between poliovirus empty capsids formed in uiuo and those formed in uitro: A role for the morphopoietic factor. J. Virol. 40, 173-198. Putnak, J. R., and Phillips, B. A. (1982). Poliovirus empty capsid morphogenesis: Evidence for conformational differences between self- and extract-assembled empty capsids. J. Virol. 41, 792-800. Racaniello, V. R., and Baltimore, D. (1981a). Molecular cloning of poliovirus DNA and determination of the complete nucleotide sequence of the viral genome. Proc. Natl. Acad. Sci. U.S.A. 78, 4887-4891. Racaniello, V., and Baltimore, D. (1981b). Cloned poliovirus complementary DNA is infectious in mammalian cells. Science 214, 916-919. Ren, R., Costantini, F. C., Gorgacz, E. J., Lee, J. J., and Racaniello, V. R. (1990). Transgenic mice expressing a human poliovirus receptor: A new model for poliomyelitis. Cell (Cambridge, Mass.) 63, 353-362. Reynolds, C., Birnby, D., and Chow, M. (1992). Folding and processing of the capsid protein precursor P1 is kinetically retarded in neutralization site 3B mutants of poliovirus. J. Virol. 66, 1641-1648. Rhee, S. S., and Hunter, E. (1987). Myristylation is required for intracellular transport but not for assembly of D-type retrovirus capsids. J. Virol. 61, 1045-1053. Rhee, S. S., and Hunter, E. (1990). Structural role of matrix protein of type D retroviruses in gag polyprotein stability and capsid assembly. J. Virol. 64,4383-4389. Richards, 0. C., and Ehrenfeld, E. (1980). Heterogeneity of the 3’ end of minus strand RNA in poliovirus replicative form. J. Virol. 36, 387-394. Richards, 0. C., and Ehrenfeld, E. (1990). Poliovirus RNA replication. Curr. Top. Microbiol. Immunol. 161, 89-119. Romaniuk, P. J., Lowary, P., Wu, H. U., Stormo, G., and Uhlenbeck, 0. C. (1987). RNA binding site of R17 coat protein. Biochemistry 26, 1563-1568. Rombaut, B., and Boeye, A. (1991). In uitro assembly of poliovirus 14s subunits: Disoxaril stabilization as a model for the antigenicity conferring activity of infected cell extracts. Virology 180, 788-792. Rombaut, B., Vrijsen, R., Brioen, P., and Boeye, A. (1982). A pH-dependent antigenic conversion of empty capsids of poliovirus studied with the aid of monoclonal antibodies to N and H antigen. Virology 122, 215-218. Rombaut, B., Vrijsen, R., and Boeye, A. (1984). In uitro assembly of poliovirus empty capsids: Antigenic consequences and immunological assay of the morphopoietic factor. Virology 135, 546-550. Rombaut, B., Vrijsen, R., Delgadillo, R., Vanden Berghe, D., and Boeye, A. (1985). Characterization and assembly of poliovirus-related 45s particles. Virology 146, 111-1 19. Rombaut, B., Boeye, A., Ferguson, M., Minor, P. D., Mosser, A., and Rueckert, R. (1990a). Creation of a n antigenic site in poliovirus type 1 by assembly of 14s subunits. Virology 174,305-307. Rombaut, B., Vrijsen, R., and Boeye, A. (1990b). New evidence for the precursor role of 145 subunits in poliovirus morphogenesis. Virology 177, 411-414. Rombaut, B., Foriers, A., and Boeye, A. (1991). In uitro assembly of poliovirus 14s subunits: Identification of the assembly promoting activity of infected cell extracts. Virology 180, 781-787.

POLIOVIRUS ASSEMBLY AND RNA ENCAPSIDATION

67

Rossman, M. G., and Johnson, J. E. (1989).Icosahedral RNA virus structure. Annu. Rev. Biochem. 58, 533-573. Rossman, M.G., Arnold, E., Erickson, J. W., Frankenberger, E. A., Griffith, J. P., Hecht, H. J., Johnson, J. E., Kamer, G., Luo, M., Mosser, A. G., Rueckert, R. R., Sherry, B., and Vriend, G. (1985).Structure of a human common cold virus and functional relationship to other picornaviruses. Nature (London) 317, 145-153. Rothberg, P. G., Harris, J. J. R., Nomoto, A,, and Wimmer, E. (1980).The genome-linked protein of picornaviruses. V. 0-4-(5’-UridylyI)-tyrosineis the bond between the genome-linked protein and the RNA of poliovirus. Proc. Natl. Acad. Sci. U.S.A. 75, 4868-4872. Rueckert, R. R. (1990).Picronaviridae and their replication. In “Virology” (B. N. Fields, D. Knipe, et al., eds.), pp. 507-547. Raven, New York. Rueckert, R., and Wimmer, E. (1984).Systematic nomenclature for picornavirus proteins. J. Virol. 50, 957-959. Sabin, A. B., and Boulger, L.R. (1973).History of Sabin attenuated poliovirus oral live vaccine strains. J. Biol. Stand. 1, 115-118. Salk, J. E. (1960).Persistence of immunity after administration of formalin-treated poliovirus vaccine. Lancet 2,715-723. Schultz, A. M., and Rein, A. (1989).Unmyristylated Moloney murine leukemia virus Pr65 gag is excluded from virus assembly and maturation events. J. Virol. 63,23702373. Schultz, A. M., Henderson, L. E., and Oroszlan, S. (1988).Fatty acylation of proteins. Annu. Rev. Cell Biol. 4,611-647. Semler, B. L.,Anderson, C. W., Hanecak, R., Dorner, F., and Wimmer, E. (1982).A membrane-associated precursor to poliovirus VPg identified by immunoprecipitation with antibodies directed against a synthetic heptapeptide. Cell (Cambridge, Mass.) 28, 405-412. Semler, B. L., Dorner, A. J., and Wimmer, E. (1984).Production of infectious poliovirus from cloned cDNA is dramatically increased by SV40 transcription and replication signals. Nucleic Acids Res. 12, 5123-5141. Smith, T. J., Kremer, M. J., Luo, M., Vriend, G., Arnold, E., Kamer, G., Rossmann, M. G., McKinlay, M. A,, Diana, G. D., and Otto, M. J. (1986).The site of attachment in human rhinovirus 14 for antiviral agents that inhibit uncoating. Science 233, 12861293. Sonenberg, N. (1987).Regulation of translation by poliovirus. Adv. Virus Res. 33, 175204. Sonenberg, N. (1990). Poliovirus translation. Curr. Top. Microbiol. Immunol. 161,23-47. Spector, D. H., and Baltimore, D. (1975).Polyadenylic acid on poliovirus RNA. 11.Poly(A) on intracellular RNAs. J. Virol. 15,1418-143. Tao, J., and Frankel, A. D. (1992).Specific binding of arginine to TAR RNA. Proc. Natl. Acad. Sci. U.S.A.89,2723-2726. Towler, D. A., Adams, S. P., Eubanks, S. R., Towery, D. S., Jackson-Machelky, E., Glaser, L., and Gordon, J. I. (1987).Purification and characterization of yeast myristoy1 CoA:protein N-myristoyltransferase. Proc. Natl. Acad. Sci. U.S.A. 84, 27082712. Towler, D. A., Gordon, J. I., Adams, S. P., and Glaser, L. (1988).The biology and enzymology of eukaryotic protein acylation. Annu. Rev. Biochem. 57,69-99. Toyoda, H., Nicklin, J. W.,Murray, M. G., Anderson, C. W., Dunn, J. J., Studies, F. W., and Wimmer, E. (1986).A second virus-encoded proteinase involved in proteolytic processing of poliovirus polyprotein. Cell (Cambridge, Mass.) 45,761-770. Trono, D., Pelletier, J., Sonenberg, N., and Baltimore, D. (1988).Translation in mam-

68

DAVID C. ANSARDI et ul.

malian cells of a gene linked to the poliovirus 5’ noncoding region. Science 241,445448. Troxler, M., Egger, D., Pfistner, T., and Bienz, K. (1992). Intracellular localization of poliovirus RNA by in situ hybridization a t the ultrastructural level using singlestranded riboprobes. Virology 191, 687-697. Tsao, J., Chapman, S., Agbandje, M., Keller, W., Smith, K., Wu, H., Luo, M., Smith, J., Rassmann, M. G., Compans, R. W., and Pai-rish, C. R. (1991). The three-dimensional structure of canine parvovirus and its functional implications. Science 251, 14561464. Tucker, S. P., Thornton, C. L., Wimmer, E., and Compans, R. W. (1993). Vectorial release of poliovirus from polarized human intestinal epithelial cells. J. Virol. 67,4274-4282. Turner, P. C., Young, D. C., Flanegan, J. B., and Moyer, R. W. (1989).’Interference with vaccinia virus growth caused by insertion of the coding sequence for poliovirus protease 2A. Virology 173, 509-521. Van der Werf, S., Bradley, J., Wimmer, E., Studier, F. W., and Dunn, J. J. (1986). Synthesis of infectious poliovirus RNA by purified T7 RNA polymerase. Proc. Natl. Acud. Sci. U S A . 83,2330-2334. Watanabe, Y., Watanabe, K., and Hinuma, Y. (1962). Synthesis of poliovirus-specific proteins in HeLa cells. Biochim. Biophys. Actu 61, 976-977. Wilcox, C., Hu, J. S., and Olson, E. N. (1987). Acylation of proteins with myristic acid occurs contranslationally. Science 238, 1275-1278. Wimmer, E. (1982). Genome-linked proteins of viruses. Cell (Cambridge, Muss.) 28,199201. Wycoff, E. E., Hershey, J. W. B., and Ehrenfeld, E. (1990). Eukaryotic initiation factor 3 is required for poliovirus 2A protease-induced cleavage of the p220 component of eukaryotic initiation factor 4F. Proc. Natl. Acad. Sci. U.S.A.87, 9529-9533. Yafal, A. G., and Palma, E. L. (1979). Morphogenesis of foot-and-mouth disease virus. I. Role of procapsids as virion precursors. J. Virol. 30, 643-649. Yin, F. H. (1977). Involvement of viral procapsid in the RNA synthesis and maturation of poliovirus. Virology 83, 299-307. Yogo, Y., and Wimmer, E. (1975). Sequence studies of poliovirus RNA. 111. Polyuridylic acid and polyadenylic acid as components of purified poliovirus replicative intermediate. J . Mol. Biol. 92, 467-477. Ypma-Wong, M. F., Dewalt, P. G., Johnson, V. H., Lamb, J. G., and Semler, B. L. (1988a). Protein 3CD is the major poliovirus proteinase responsible for cleavage of the P1 capsid precursor. Virology 166, 265-270. Ypma-Wong, M. F., Filman, D. J., Hogle, J. M., and Semler, B. L. (1988b3. Structural domains of the poliovirus polyprotein are major determinants for proteolytic cleavage at Gln-Gly pairs. J. Biol. Chem. 263, 17846-17856. Yu, S. F., and Lloyd, R. E. (1991). Identification of essential amino acid residues in the functional activity of poliovirus 2A protease. Virology 182, 615-625. Zlotnick, A., Reddy, V. S., Dasgupta, R., Schneemann, A., Ray, W., Jr., Rueckert, R. R., and Johnson, J. E. (1994). Capsid assembly in a family of animal viruses primes an autoproteolytic maturation that depends on a single aspartic acid residue. J. Biol. Chem. 269,13680-13684. Zoller, M. J., and Smith, M. (1983). Oligonucleotide-directed mutagenesis of DNA fragments cloned into M13 vectors. In “Methods in Enzymology” (R. Wu, L. Grossman, and K. Moldave, eds.), Vol. 100, pp. 468-500. Academic Press, New York.

ADVANCES IN VIRUS RESEARCH, VOL. 46

GENOME REARRANGEMENTS OF ROTAVIRUSES

Ulrich Desselberger Clinical Microbiology and Public Health Laboratory Addenbrooke’s Hospital Cambridge CB2 2QW, England

I. 11. 111. IV.

V.

VI. VII. VIII. IX. X.

Discovery of Genome Rearrangements Extent of Genome Rearrangements in Rotaviruses Sequence Data of Rearranged Genes Genome Rearrangements Generated in Vitro in Cultured Cells Mechanisms of Genome Rearrangements Biophysical Data Function of Rearranged Genes and Their Products Genome Rearrangements and Evolution of Rotaviruses Genome Rearrangements in Other Genera of Reoviridae Outlook References

I. DISCOVERY OF GENOME REARRANGEMENTS Rotaviruses are the main cause of viral gastroenteritis in infants and young children and in the young of a large variety of animal species (Kapikian and Chanock, 1990).There are at least five different groups, named A-E (Pedley et al., 1986). Group A rotaviruses are responsible for the vast majority of human infections. Rotaviruses have a genome consisting of 11 segments of double-stranded RNA (dsRNA) of approximately 18,500 nucleotide pairs in total size (Estes, 1990). The RNA segments can be easily extracted from virus particles, separated by polyacrylamide gel electrophoresis (PAGE),and visualized by silver staining, ethidium bromide staining, or radiolabeling. Typical RNA profiles show four size classes (I,segments 1-4; 11, segments 5 and 6; 111, segments 7-9; and IV, segments 10 and 11)(Estes, 1990). However, these profiles are not always seen. Pedley et al. (1984) investigated rotaviruses isolated from chronically infected children with severe combined immunodeficiency (SCID). Rotavirus infections in the immunocompetent host are normally overcome within 1 week, but in SCID children rotaviruses and many other viruses establish chronic infections that result in virus shedding over many weeks, months, and even several years (Saulsbury et al., 1980; Booth et al., 1982; Chrystie et al., 1982). Rotaviruses obtained from serial fecal specimens of such children produced abnormal RNA profiles: normal 69

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

70

ULRICH DESSELBERGER

RNA segments were decreased in their relative concentration or even completely lost from the profiles, but additional bands of RNA were seen migrating between RNA segments 1and 7 (Fig. 1). The intensity of the additional bands varied. It was quickly established that the additional bands consisted of dsRNA and also that they had not arisen by noncovalent linkage (Pedley et al., 1984). Northern blots of the RNAs were probed with segment-specific radiolabeled cDNA clones of bovine rotaviruses under conditions to give segment-specific reactions in controls. Blots of atypical profiles often showed multiple hybridizations: that of the homologous RNA segment of standard size and several bands of dsRNA which always migrated higher up in the gel. This hybridization pattern was maintained when blots of RNA separated under denaturing conditions (Bailey and Davidson, 1976) were probed. An example for segment 9-specific hybridization is given in Fig. 2. This indicated that the additional bands of dsRNA contained segmentspecific sequences in the form of covalently bonded concatemers.

FJG. 1. RNA profiles of serial rotavirus specimens obtained from chronically infected patient A.K. (dates of specimens are indicated a t top). All specimens were 3’-endlabeled with [32PlpCp as described by Clarke and McCrae (1981),separated on a nondenaturing polyacrylamide gel, and autoradiographed. Bovine rotavirus and human rotavirus obtained from acute infections served as controls. Order numbers of segments are indicated on both sides, and additional bands are marked by arrowheads. From Pedley et al. (1984), with permission of the authors and publisher.

GENOME REARRANGEMENTS OF ROTAVIRUSES

71

FIG.2. Hybridization of rotavirus RNA samples of patient A.K. (dates of specimens are indicated at top) and of human and bovine control RNAs on DPT paper blots to RNA segment 9-specific radiolabeled cDNA probe. 3'-End-labeled bovine rotavirus RNA (L bovine) and unlabeled bovine and human rotavirus RNAs served as controls. Autoradiogram. From Pedley et al. (1984), with permission of the authors and publisher.

When several specimens sequentially obtained from the same person were subjected to such investigation, extra bands were found over a wide range of the profile. These bands varied in intensity and appeared and disappeared on passing through the chronological series (Pedley et al., 1984;Fig. 3).Where segment derivation could be established, the molecular weights of the additional bands were not simple multiple integers of the segments from which they were derived. The variable intensity of the additional bands and of some of the normal RNA segments lead to the hypothesis that either parts of the RNA genome occurred in abnormal configuration in single virus particles or that subpopulations of viruses possessing normal and abnormal genomes coexisted and cocirculated (Pedley et al., 1984). The question also arose whether viruses possessing such genomes were defective interfering (DI) particles (Holland et al., 1980).By contrast to DI RNAs that are characterized by internal deletions (Davis et al., 19801,the larger size and migrational pattern of the additional rotavirus bands, which were maintained under denaturing conditions, excluded such a possibility (Pedley et al., 1984).Mosaic structures as

72

ULRICH DESSELBERGER

FIG.3. RNA profiles of sequential rotavirus samples (dates are indicated at top) of patient U.H. Cenomic dsRNA was extracted, separated on a 2.8% polyacrylamide-6 M urea gel, and stained with silver. Bovine rotavirus RNA served as an internal control. Numbers of segments and positions of additional bands (arrowheads) are indicated a t right. From Pedley et al. (1984), with permission of the authors and publisher.

described by Fields and Winter (1982) remained a possibility. When RNA segments were separated by PAGE for a short period, in no case were additional bands of RNA found migrating faster than the smallest RNA segment (U. Desselberger, 1985, unpublished results). The discovery of group A rotaviruses with abnormal RNA profiles also raised the question of whether the dictum of “ atypical” RNA profiles in other rotavirus groups (B-E)could be maintained (Pedley et al., 1984).

GENOME REARRANGEMENTS OF ROTAVIRUSES

73

11. EXTENT OF GENOME REARRANGEMENTS IN ROTAVIRUSES Since the original discovery genome rearrangements have been described by several independent groups t o occur not only in human rotaviruses but also in rotaviruses of a variety of animal species (humans: Albert, 1985; Dolan et al., 1985; Eiden et al., 1985; Matsuno et al., 1985; Besselaar et al., 1986; Hundley et al., 1987; Matsui et al., 1990; Mendez et al., 1992; Gault-FrBre et al., 1995; calves: Pocock, 1987; Paul et al., 1988; Scott et al., 1989; Tian et al., 1993; rabbits: Thouless et al., 1986; Tanaka et al., 1988; piglets: Bellinzoni et al., 1987; Mattion et al., 1988; Lambs: Shen et al., 1994). Whereas the initial observation was in immunodeficient children, the observations in animals and some of those in humans were in immunocompetent hosts. In a South African hospital viruses with genome rearrangements circulated for several months, infecting apparently healthy children (Besselaar et al., 1986). DATAOF REARRANGED GENES 111. SEQUENCE Nucleotide sequences of rearranged genes of several group A rotavirus strains of different origin have been obtained, and references and nucleotide sequence accession numbers are summarized in Table I. In most cases the genome rearrangement consists of a partial duplication of sequences of the open reading frame (ORF) starting beyond the termination codon and extending then to the 3‘ end of the normal gene. This is diagrammatically shown in Fig. 4 for rearranged RNA 10 of a human rotavirus isolate (Ballard et al., 1992); similar changes were also found for rearrangements of other RNAs 10 (Matsui et al., 1990),for RNAs 11(Gonzalez et al., 1989; Gorziglia et al., 1989; Scott et al., 19891, and for one RNA 5 (Hua and Patton, 1994). In most rearranged genes, the sequence runs from a normal 5’ untranslated region (UTR) and through a normal ORF. At various nucleotide positions after the termination codon (0-23; Table I), the duplication starts reinitiating from various places within the ORF but downstream of the initiation codon and then reads through a duplicated termination codon and toward a normal 3’ UTR. As the duplication of the sequence normally starts beyond the initiation codon, it remains silent as a whole, and the resulting genes have enormously long 3’ UTRs, up to 1800-1900 bp (McIntyre et al., 1987; Hua and Patton, 19941, in contrast to the relatively short 3’ UTRs (17-185 nucleotides) of the standard length genes (Estes, 1990; Desselberger and McCrae, 1994).

TABLE I SEQUENCED GENOMEREARRANGEMENTS OF ROTAVIRUSES

RNA segment=

5 5 6 7 10

10 11 11 11

Strain brv E brv A Lp 14 H 57 A64

VMFU C71183 C60 X1 Alabama

Origin

Start of reiteration in relation to termination codon

Bovine Bovine Ovine Human Human Human Bovine Pig Lapine

-596 -52 23 0 2 0 0 6 4

Number of point mutations compared to standard geneb ND 16 6 ND 11

23 NA 33 NA

GenbankIEMBL accession number Standard gene

Rearranged gene

224735 L12248 L11596 NAc DO1146 NA NA NA NA

212108 L11575 L11595 NA DO1145 NA NA NA NA

Refs. Tian et al. (1993) Hua and Patton (1994) Shen et al. (1994) Mendez et a1. (1992) Ballard et al. (1992) Matsui et al. (1990) Scott et al. (1989) Gonzalez et al. (1989) Gorziglia et al. (1989)

a Genome rearrangements have also been observed in segment 6 of a human strain (Pedley et al., 1984) and in segments 7,8, and 9 of human strains (coding for NSP2 and NSP3) (Pedley et al., 1984; Hundley et al., 1987; Gault-Frere et al., 1995), but they have not been sequenced so far. b ND, Not determined NA, not applicable. c Partial sequence (junction region).

75

GENOME REARRANGEMENTS OF ROTAVIRUSES Normal gene 10 I4181

569

I "

751 I I

I

4 Met

141 I

'*,

'' '',56? '\

t

stop

572

'

")f

/'

3'

'' '' '' '

''

'. ', '

1242 I

'\

5'

I

'I3'

Met

stop

Rearranged gene 10 FIG.4. Structures of normal and rearranged genes 10 of a human rotavirus (isolates A28 and A64, respectively).The solid bar represents the complete ORF, and the open bar symbolizes the duplicated part of the ORF of the normal gene (untranslated).Solid lines indicate 5' and 3' untranslated regions as well as sequences between the normal and duplicated ORFs. From Ballard et al. (19921, with permission of the authors and publisher.

However, RNA segment 5 of a bovine rotavirus (brv) variant was rearranged in several different ways with variants being called brv A and brv E (Tian et al., 1993). In the case of segment 5 of the brv E variant, the duplication had started before the termination codon, and an extended ORF ensued encoding segment-5-specific amino acids as the reiteration had started in frame (Fig. 5). The extended ORF codes for a protein VP5E of 728 amino acids, which was verified by PAGE of [36Slmethionine-labeled proteins of brv E-infected cells (Hundley et al., 1985; Tian et al., 1993). In contrast, rearranged segment 5 of the brv A variant possesses a different structure. The reiteration starts 52 nucleotides before the stop codon (in position 1454), but one of several additional point mutations changes the picture further: a mutation in position 808 results in a new termination codon (TAG) allowing an ORF of only 258 amino acids k e . , of 31 kDa size, slightly more than half the size of the normal product of 491 amino acids, i.e., 58 kDa). Thus a gene of 2693 bp results in only 774 (positions 33-806, i.e., 28.7%) coding for a protein! The abnormal product (Fig. 6) was detected by Hua and Patton (1994) after it had escaped screening by Hundley et al. (1985) and Tian et al. (1993), apparently because it comigrates with cellular gene products. A point mutation in the ORF of a rearranged gene 6 also had profound consequences for protein stability (see below).

76

ULRICH DESSELBERGER

FIG. 5. Structures of normal and rearranged forms of RNA segment 5 of bovine rotavirus (brv UKtc and brv E, respectively). The junction sequence is spelled out at the bottom, showing 6 amino acids on either side. From Tian et al. (19931,with permission of the authors and publisher.

Some time ago it was found that RNA segment 10 of a “short” electropherotype human group A rotavirus codes for a protein which corresponds to the product of RNA 11 of “long” electropherotype rotaviruses (Dyall-Smith and Holmes, 1981). The observations by Matsui et al. (1990) on gene 11 equivalents of rotavirus genomes yielding

FIG. 6. Diagram of standard gene 5 of bovine rotavirus and of gene A of the brv A variant. Gene duplication in gene A starts 2 positions after the termination codon. The point mutation in position 808 giving rise to an additional termination codon in gene A is indicated. The ORFs of the gene products are also shown. From Hua and Patton (1994), with permission of the authors and publisher.

GENOME REARRANGEMENTS OF ROTAVIRUSES

77

“short” and “supershort” PAGE profiles were most intriguing: whereas “supershort” strain VMRI clearly contained a partial duplication at its 3’ end, the RNA segments 10 of “short” strain DS-1 and of “supershort” strain M69 have sequences at their 3’ ends that were similar to one another but not related to any other available rotavirus gene sequence. Finally, it is remarkable that direct repeats of nucleotide sequences were observed closely upstream of the start of the duplications in a number of cases (Gorziglia et al., 1989; Ballard et al., 1992; Shen et al., 1994), but not in others (Scott et al., 1989; Matsui et al., 1990). The numbers of point mutations in the rearranged compared to the normal genes varied widely: between 6 and 33 have been counted (Table I). No genome rearrangement has been described so far which had resulted in a mosaic of sequences donated from several different RNA segments, in contrast to the DI mosaic structures of influenza viruses described by Fields and Winter (1982). IV. GENOME REARRANGEMENTS GENERATED in Vitro IN CULTURED CELLS Before nucleotide sequences of rearranged genes and biological properties of the viruses carrying them were known (see below), the phenomenon of genome rearrangements appeared to be related to that of the formation of DI RNAs. As serial passage in uitro of virus at high multiplicity of infection (MOI) has been found to be the most efficient way to generate viruses with DI genomes, this method was used to propagate rotaviruses (Hundley et al., 1985).Surprisingly, viruses with genome rearrangements (i.e., partial duplications) but not genome deletions emerged (Fig. 7). Bovine rotavirus with a standard genome transformed into brv variants with rearranged RNA segments 5 , among others variants brv A and brv E (Fig. 7, lanes 2 and 3; see also Section 111). As in virus cultures with DI RNAs, yields in virus increased and decreased in a periodic manner, and the absolute yields in viral infectivity were inversely correlated with the ratios of numbers of virus particles (nop) over infectivity [nvp/pfu (plaque-forming units)]. A t passages 7-8, viruses with genome rearrangements appeared and overgrew the virus with standard genome. This was a reproducible phenomenon and was obtainable after repeated plaquepurifications of standard virus (Fig. 8). The outcome of repeat experiments, however, was not identical in that RNA 5 equivalents with apparently different forms of rearrangements were found. The in uitro generation of viruses with rearranged genomes was reproduced with Chinese lamb rotaviruses by Shen and Bai (1990).

78

ULRICH DESSELBERGER

FIG. 7. RVA profiles of plaque-purified bovine rotaviruses obtained after serial passage at high MOI. RNA segment numbers are indicated on the right-hand side. Open arrowheads denote missing RNA segments, closed arrowheads additional RNA bands. Lanes 1and 6 show standard bovine rotavirus; lane 2, brv A; lane 3, brv E; lane 4, brv F; lane 5; brv G/H (likely to be a mixture). Analysis in 2.8% polyacrylamide-6 M urea gel, stained with silver. From Hundley et al. (1989, with permission of the authors and publisher.

Viruses possessing genome rearrangements could be plaque-purified very easily, and six times plaque-to-plaque purified virus grew perfectly well without showing the appearance of virus with standard ge-

GENOME REARRANGEMENTS OF ROTAVIRUSES RNAsegmentS+ RNAbandA-

+

+ +

-

- - -

t

+

+

+

- - -

(+I

t

t

79

I+)

-

-

-

-

-

-

+

t

+

+

+

+

+

106 J n.v.p./p.f.u.

I

9

!'

f I ' I

'

1J 2

4

6

8

10

12

14

16

Passage number FIG.8. Plot of yield of infectious virus (pfuhl) and of nvp/pfu in harvested tissue culture fluids against passage number (at high MOI). The presence or absence of RNA segment 5 and RNA band A (Fig. 7, lane 2) is indicated at the top. From Hundley et al. (1985),with permission of the authors and publisher.

nomes and remained genetically stable. These and other experiments (see below) proved that bovine rotaviruses with rearranged genes are not DI viruses. The nvp/pfu ratios were equally low for brv standard and brv A viruses (Table 11). It was also shown that genome rearrangements were a continuous phenomenon. When six times plaque-purified brv A was again serially propagated a t high MOI, second generation

80

ULRICH DESSELBERGER TABLE I1 INFECTIVITY, CONCENTRATION OF VIRUSPARTICLES, AND nvplpfu STOCKS OF STANDARD BOVINEROTAVIRUS.AND RATIOOF CLONED brv A VARIANT WITH REARRANGED GENOME~,~ Stock and preparation Standard brva 1 2 3

Infectivity (pfulml, x 107)

4 15 10 10 +. 5 brv A with rearranged genomeb 1 3 10 15 6 2 3 6 3 6 k 4

Concentration (nvpiml, x 106)

nvpipfu

1.2 9 11 7 t 4

3 6 11 7 2 3

8 5 8 10 6 2 4 4 6?3

27 5 5 16 30 7 7 13 14 9

*

aN=3. bN=8. c From Hundley et al. (1985), with permission of the authors and publisher. The brv A had a genome missing RNA segment 5 and possessing RNA band A. The arithmetic means +- standard deviation are indicated. The corresponding arithmetic means of standard brv and of brv A with rearranged genome did not differ significantly one from another ( t test, p < 0.05).

rearrangements resulted (viruses brv K and brv L; Fig. 9). When cells were infected with brv standard and brv A at different MOIs, the outcome depended on whether passage was at low or high MOI: in the first case, standard brv overgrew; in the latter, the brvA variant (Hundley et al., 1985). The effect of genome rearrangements on growth in cell culture will be discussed below.

V. MECHANISMS OF GENOMEREARRANGEMENTS The sequence data available (see Section 111) allow a formal description of genome rearrangements as partial duplications (concatemer formation) with varying consequences relative to their expression. Start of the duplication after the termination codon (excluding the

GENOME REARRANGEMENTS OF ROTAVIRUSES

81

FIG.9. RNA profiles of viruses with second generation genome rearrangements, brv K and brv L, obtained after repassage of plaque-purifiedbrv A at high MOI.(A) Analysis in a 10%polyacrylamide gel, stained with ethidium bromide. (B,C) Autoradiograms of Northern blots probed with 32P-labeled cDNA produced from RNA segment 5 (B)or RNA band L (C) according to Hundley eb al. (1987).

initiation codon) leads to long 3' UTRs, whereas start of the duplication before the termination codon leads to longer than normal ORFs and normal 3' UTRs. It is not clear, however, at which step of the replication cycle the duplication event occurs. It has been suggested that the RNAdependent RNA polymerase of rotaviruses (associated with the particle core and coded for by RNA 1; Estes, 1990) can fall back on its template at various steps of transcription (plus strand synthesis) and reinitiate and retranscribe from that template a t different places (Fig. 10A). Messenger RNAs of the rearranged size are transcribed in uitro from particles containing rearranged genes (Hundley et al., 1985), and rotaviruses with genome rearrangements are genetically stable (see Section IV). Alternatively, the primary duplication event could occur at the level of replication (negative strand synthesis) (Fig. 10B).Whereas occurrence of duplication at the transcription stage would mean that the abnormal mRNA is packaged and a strand of negative sense replicated from it in the new precore particles (Gallegos and Patton, 1989), occurrence of rearrangements at the replication stage would imply that a rearranged negative strand forms a heterohybrid with the normal positive strand, and that at the next round of infection this rearranged negative strand is transcribed at full length. As packaging of the RNA genome is very tightly controlled, a virus particle will contain only one form of one segment each.

ULRICH DESSELBERGER

82

A

B

+5’

3

-3

5

+5’,

3 +

5

U

U

U +5’---

+5’< n.o 3

5

-3

5’

3+

--. -.-. -. --5’ I.

U

(5‘OV

---------- ---

3+

--5-

------I-

U

FIG.10. Possible mechanisms for emergence of genome rearrangements (A) during plus strand synthesis and (B) during minus strand synthesis (in precore particles). Bold lines represent the minus strand; fine lines, plus strand; dashed lines, newly synthesized strands; open circle, RNA-dependent RNA polymerase; arrows, direction of synthesis.

GENOME REARRANGEMENTS OF ROTAVIRUSES

83

Most genome rearrangements that have been sequenced can be described as intramolecular recombination events, and direct repeats close to the recombination site are often but not always found. This is similar to what has been observed in the phi (4) 6 system (Mindich et al., 1992; Onodera et al., 1993).In the case of poliovirus recombination, it was shown that recombination favored the step of secondary transcription (from the negative strand of the replicative intermediate, or RI) (Kirkegaard and Baltimore, 1986).In the phi 6 system where intermolecular recombination between the three different segments of dsRNA of the viral genome can be observed, it was shown very elegantly that recombination also occurs at the step of negative strand synthesis (Onodera et al., 1993). However, under the special conditions of those experiments only phi 6 recombinants would survive and would therefore be positively selected for in the surviving viruses. It is of interest to note that direct repeats favor genome rearrangements of rotaviruses although they do not seem to be an absolute requirement. They were found in sequenced genes by Gorziglia et al. (1989), Ballard et al. (19921, and Shen et al. (19941, but not by Scott et al. (1989) or Matsui et al. (1990). Onodera et al. (1993) in their system show very nicely that sequence identity of the landing pad for the donor strand-polymerase complex compared to the lift off point is not a prerequisite but is preferred. Although the mechanism of recombination in genome rearrangements of rotaviruses has not been elucidated, the data are consistent with the copy choice model (Kirkegaard and Baltimore, 1986; Lai, 1992) in which specific sequence homologies or secondary structures are involved in directing the switch of the polymerase (Romanova et al., 1986). Mechanisms of genome rearrangements should be explored further in in uitro transcription (Cohen et al., 1979) and replication (Chen et al., 1994) systems. Some electron microscopy data on rotaviruses are of interest in this context. Using the Kleinschmidt technique the lengths of rotavirus RNA segments have been determined, and the measurements were very precise when compared to the length obtained by sequence data (Rixon et al., 1984). Whereas viruses with standard genomes show less than 2%RNA molecules which are larger than RNA 1,viruses with rearranged genomes show about 15%RNA concatemers longer than RNA 1 and of varying length (U. Desselberger and F.Rixon, 1985, unpublished data). This suggests that rearrangements which are amplified to amounts of normal segments (see below) are only part of numerous other recombination events which did not survive. Some of the results obtained by Matsui et al. (1990) are difficult to explain; these workers obtained long 3’ UTRs of RNA segments of

84

ULRICH DESSELBERGER

larger than normal size without the evidence of an intramolecular duplication. These sequences could have “mutated away” from original duplications (being under no functional constraint) or could have been picked up from as yet unidentified cellular sequences (Qian et al., 1991),or identifiable cellular sequences as found for influenza viruses (Khatchikian et al., 1989). VI. BIOPHYSICAL DATA Once it became possible to grow human rotaviruses with genome rearrangements (Hundley et al., 1987;see below), various variants with different combinations of genome rearrangements were found. The viruses had between 450 and 1790 bp of additional RNA packaged, amounting to 1.4 to 9.6% of the standard genome size. By electron microscopy such particles were indistinguishable in size or shape from viruses possessing a standard genome (Hundley et al., 1987;McIntyre et al., 1987).Examples of the RNA profiles of such viruses are given in Fig. 11A;rotaviruses of such RNA profiles had 450,1070,1570,and 1790 bp additional RNA packaged. The viruses differed in density as determined by analytical ultracentrifugation (Fig. llB),and the differences in density were directly proportional to the number of additionally packaged base pairs (Fig. 11C;McIntyre et al., 1987).Thus, packaging of rotavirus genomes is flexible in terms of the size of packaged segments, and additionally packaged base pairs amounting to up to 10% of the total genome size were tolerated in the variant with rearrangements. Particles of viruses with up to 10% additional base pairs packaged were morphologically indistinguishable from standard rotavirus (Hundley et al., 1987).In recombinants of the phi 6 system up to 16.7% of the genome size were additionally packaged without apparent effect on the procapsid (L. Mindich, 1994,personal communication). VII. FUNCTION OF REARRANGED GENESAND THEIRPRODUCTS For some time after the first observation of viruses with genome rearrangements it was not possible to grow them in tissue culture. Therefore, it was not clear whether those viruses were functionally defective, possibly due to rearrangements of genes. Secondary rhesus monkey kidney (RMK) cell cultures infected with a human rotavirus isolate (U.H.) showing genome rearrangements by PAGE (Pedley et al., 1984)showed no cytopathic effect, although common primary group A rotavirus isolates grow very well on RMK cells (Ward et al., 1984). However, on superinfection with the tissue culture-adapted bovine ro-

FIG.11. (A) RNA profiles of bovine rotavirus and of human rotavirus variants with rearranged genomes of genotypes 2,3,7,and 9 (Hundley et al., 1987). Segment numbers (1-11) are indicated on the left-hand side, and the position and origin of rearranged bands identified on the right-hand side (bands a and f were derived from RNA 8, band d from RNA 10, and bands c and e from RNA 11; Hundley et al., 1987). The number of additionally packaged base pairs is indicated at the bottom. Analysis in 2.8% polyacrylamide-6 M urea gel, stained with silver. (B) Scans after analytical equilibrium centrifugation in CsCl of mixtures of single-shelled particles containing RNA of standard bovine rotavirus and of human rotaviruses with rearranged genomes of genotypes 2,3, and 9. Numbers of additionally packaged base pairs are indicated below scans. (C) Plot of difference in density as determined from data shown in B against number of additionally packaged base pairs. From McIntyre et al. (1987), with permission of the authors and publisher.

86

ULRICH DESSELBERGER

taviruses (brv, UK Compton strain), some of the rearranged RNA segments of the U.H. virus were disproportionally amplified (Fig. 12, lane M; Allen and Desselberger, 1985). Plaques obtained from the original yield of such cultures were grown, and the RNA was extracted and analyzed by PAGE. Extensive reassortment had taken place (Fig. 12; Allen and Desselberger, 1985) occurring between standard segments of the brv and U.H. virus (no segments 5 and 6; segment 6 not shown) but also the rearranged RNAs. Standard length RNA segment 11 of brv was replaced by RNA bands F (not shown) or G and standard RNA

FIG. 12. RNA profiles of rotaviruses grown from 12 individual plaques derived from a mixed infection of bovine rotavirus and human rotavirus U.H. with rearranged genome. The RNA profile of the direct yield of the mixed infection (M)is shown in the right-hand lane. Segments 1-11 of brv and segment 5 (h5) and several rearranged bands (D,G) of the human rotavirus are denoted at right. Analysis in 2.8% polyacrylamide-6 M urea gel, stained with silver. From Allen and Desselberger (1989, with permission of the authors and publisher.

87

GENOME REARRANGEMENTS OF ROTAVIRUSES

segment 9 by bands B (not shown) or D. The segmental origin of the rearranged bands was confirmed by Northern blotting followed by hybridization of segment-specific radiolabeled probes (Allen and Desselberger, 1985). The PAGE profiles of proteins from infected cells demonstrated that the rearranged RNA bands produced normal-sized length virus-coded proteins (Allen and Desselberger, 1985), indicating that the rearranged RNAs replaced the normal RNA segments structurally and functionally. The reassortants grew well on their own in uitro, could be plaque-to-plaque purified multiple times, and remained genetically stable. Rotaviruses with genome rearrangements that had arisen after serial passage a t high MOI in uitro were equally able to reassort with human rotaviruses carrying a standard genome (Biryahwaho et al., 1987). In contrast, in cases when the normal ORF was extended (brv variant E; Hundley et ul,., 1985; Tian et al., 1993), abrogated (brv variant A; Hua and Patton, 1994), or mutated (Chinese lamb rotavirus; Shen et al., 19941, functional changes were observed. Bovine rotavirus variants E and A showed 9- to 60-fold lower yields, respectively, in single-step growth experiments and produced smaller plaques, with brv E giving plaques 40%and brv A 2% the size of plaques of standard brv (Table 111; Tian et al., 1993). The analysis of rearrangements of RNA segment 5 were of particular interest as both extension (brv E) and abrogation (brv A) of the normal ORF were found. Bovine rotavirus variant A had a truncated VP5 of 258 amino acids (due to a termination codon at positions 806808) instead of the authentic size of 491 amino acids, but was viable, nondefective, and genetically stable (Hundley et ul., 1985; Hua and

TABLE I11 IN VITROGROWTH PROPERTIES OF STANDARD BOVINE F~OTAVIRUS AND VARIANTS brv E AND brv Aa Log pfu/ml at time postinfectionb Virus

30 hr

46 hr

Plaque diameter (mm) at 7 days (mean ? SD)

Mean plaque size (mm2)

Standard brv brv E brv A

8.9 7.8 7.2

8.5 7.5 7.1

7.6 2 0.8 ( n = 9) 4.8 & 0.7 ( n = 16) 1.0 2 0.7 ( n = 50)

45.4

~~

~

18.1

0.4

~

From Tian et al. (1993), with permission of the authors and publisher. Single-step growth experiments were carried out in MA-104 cells infected at an MOI of 10 pfu per cell. Q

88

ULRICH DESSELBERGER

Patton, 1994). It was also found to be associated with the cytoskeleton of the infected cell like its normal size counterpart, demonstrating that the carboxyl-terminal half of VP5 (NSP1, NS53) is not required for rotavirus replication in vitro (Hua and Patton, 1994). Tian et al. (1993) described the even more drastically deleted VP5 gene product of rotavirus P9A5 (originally isolated from a foal) in which a deletion occurred between nucleotides 460 and 768 of the normal gene sequence. This deletion then caused a frameshift such that a stop codon was introduced 8 amino acids downstream of the deletion point, giving a predicted size of the gene product of 150 amino acids instead of the authentic size of 491 amino acids. Taniguchi et al. (1994,1995) recently described deleted VP5 genes of bovine rotavirus isolates from Thailand which had additional termination codons predicting ORFs of only 40-50 amino acids in length. The predicted protein products have so far not been found. The overall requirement of VP5 for rotavirus replication is under discussion. In a Chinese lamb rotavirus, rearrangement of RNA segment 6, the gene coding for the inner capsid protein VP6, was observed in a similar way (Shen et al., 1994) as shown by Ballard et al. (1992)for segment 10. However, the rearranged RNA6 was found to be accompanied by a point mutation in nucleotide position 949 (within the normal ORF), giving rise to a change in amino acid position 309 (from a proline to a glutamine) as the only amino acid difference compared to the VP6 of the standard genome virus, which was also available from the same lamb isolate. Proline in position 309 of VP6 is highly conserved in all group A rotavirus strains. The amino acid difference in position 309 occurred in a region of VP6 previously implicated as being important for trimerization and the formation of single-shelled particles (Clapp and Patton, 1991).The VP6 protein carrying the 309 mutation was found to be less stable than the corresponding standard VP6. Under mild denaturing conditions it did not separate on gels as a trimer (Sabara et al., 1987)but as a monomer, and it was less stable toward acid pH by almost a whole pH unit compared to the standard VP6 (Shen et al., 1994). The nvp/pfu ratio of virus possessing normal VP6 was significantly lower than that of virus carrying the mutated VP6 (Shen et al., 1994). Analysis of over 500 plaque isolates of a reassortant mixture of human viruses with genome rearrangements and standard bovine rotaviruses showed that reassortment was nonrandom, that there was linkage of occurrence of certain genes (i.e., RNA segments 5,9, and 11) in reassortants, and that the host cells on which plaque isolates were obtained (MA104 or BSC-1 cells) influenced the frequencies with which certain reassortants were recovered (Graham et al., 1987).These findings were not different from those established for other viruses

GENOME REARRANGEMENTS OF ROTAVIRUSES

89

with segmented genomes (reoviruses; Wenske et al., 1985; influenza viruses, Lubeck et al., 19791, and RNA segments with rearrangements participated in this process like standard RNA segments (Graham et al., 1987). VIII. GENOMEREARRANGEMENTS AND EVOLUTION OF ROTAVIRUSES Initially, genome rearrangements of rotaviruses were seen only in rare cases of immunodeficient human hosts (Pedley et al., 1984; Albert, 1985; Eiden et al., 1985; Dolan et al., 1985) and were thought to be more a curiosity than of particular significance. However, when rotaviruses with genome rearrangements were found to circulate for months in immunocompetent children as a nosocomial infection (Besselaar et al., 19861, and also freely circulating in a variety of otherwise healthy animal hosts (rabbits: Thouless et al., 1986; Tanaka et al., 1988; calves: Pocock, 1987; pigs: Bellinzoni et al., 19871, it became clear that the phenomenon was more frequent than originally anticipated. The various forms of rearrangements occurred mainly in genes coding for nonstructural proteins [RNA segments 5 , 8, and 9 (depending on strain), 10, and 111,but were also found for gene 6 (Pedley et al., 1984; Shen et al., 1994). These rearrangements produced RNA profiles of great diversity that were highly atypical for group A rotaviruses (Fig. 13; Desselberger, 1989). The data presented in Section VII demonstrated that genome rearrangements alone (or combined with point mutations) were able to change the structure and function of encoded proteins. It had been shown that within a single individual various forms of genome rearrangements (e.g., affecting RNA segments 8, 10, and 11) and various combinations thereof in plaque-purified viruses coexisted. At least 12 subpopulations were identified in one isolate (Fig. 14; Hundley et al., 1987) and changed in relative prevalence when observed over time in chronically infected hosts (Pedley et al., 1984; Hundley et al., 1987).Thus, multiple rearrangement variants coexisted in a constantly varying (dynamic) equilibrium, fulfilling the criteria for the presence of a quasispecies as has been described for the coexistence of various point mutants for a number of RNA viruses (Holland et al., 1982; Holland, 1984; Doming0 et al., 1985).In summary, it is therefore proposed that genome rearrangements, besides genetic point mutations (Sabara et al., 1982; Desselberger et al., 1986) and a reassortment continuum (Palese, 1984), are a third principle of the evolution of rotaviruses and can contribute to the diversity of rotaviruses in the field (Hundley et al., 1987; Desselberger, 1989; Tian et al., 1993).

90

ULRICH DESSELBERGER Group A Rotaviruses

1-

23 4-

5-

~

B

-

C

D

E

-

-

6 -

-4

1 1 0 11

-+

Rotaviruses of Groups

'Atypical'

'TvDical' human

-

- -

4

'long' ' s h o d

4 a

- -

,I b

c

d

d

1

! e

-

e

-

GI f

f

FIG.13. Diagram of RNA profiles of various group A rotaviruses with genome rearrangements and of typical RNA profiles of group A and group B-E rotaviruses. Open arrowheads denote missing normal RNA segments: closed arrowheads show various rearranged equivalents in viruses a-f. From Desselberger (1989),with permission of the publisher.

IX. GENOME REARRANGEMENTS IN OTHERGENERA OF Reouiridue Genome rearrangements have also been found involving different RNA segments of several genotypes of bluetongue virus, members of the orbivirus family (Ramig et al., 1985; Eaton and Gould, 19871, and Joklik, 1992, personal comare likely to occur in orthoreoviruses (W. munication). Thus, this mechanism of genome change seems to be possible for most animal dsRNA viruses although much less is known for viruses other than rotaviruses.

X. OUTLOOK Since the original observation of genome rearrangements in rotaviruses, much has been learned about the detailed structure of rearranged genes and their products, their functions, and their significance for the overall diversity of rotaviruses. There are still gaps in our knowledge about the exact mechanismb) by which these genome forms emerge, and it remains to be seen to what extent they occur in other double-stranded RNA viruses.

91

GENOME REARRANGEMENTS OF ROTAVIRUSES

- I - =- - -- --- --- - -- - - - - -

l o - - - - - - 11

-

-

FIG. 14. Diagram of 12 subpopulations (lanes 1-12)of human rotaviruses with various forms of genome rearrangements isolated from a single individual with chronic infection. The bovine rotavirus standard genome is shown for comparison. RNA segments (1-11) are denoted on the left-hand side, as are rearranged bands (bands c and e derived from RNA 11,band d from RNA 10,and bands a, b, f, and g from RNA 8).From Hundley et al. (1987),with permission of the authors and publisher.

ACKNOWLEDGMENTS The author thanks M. K. Estes, H. Greenberg, M. A. McCrae, L. Mindich, and J. Patton for stimulating discussions and critical reading of the manuscript.

REFERENCES Albert, M. J. (1985).Detection of human rotaviruses with a ‘super-short’ RNA pattern. Acta Paediutr. Scand 74,975-976. Allen, A. M., and Desselberger, U. (1985).Reassortment of human rotaviruses carrying rearranged genomes with bovine rotavirus. J. Gen. Virol. 66,2703-2714. Bailey, J. M.,and Davidson, N. (1976).Methylmercury as a reversible denaturing agent for agarose gel electrophoresis. Anal. Biochem. 70, 75-85. Ballard, A., McCrae, M. A., and Desselberger, U. (1992).Nucleotide sequences of normal and rearranged RNA segments 10 of human rotaviruses. J. Gen. Virol.73,633-638.

92

ULRICH DESSELBERGER

Bellinzoni, R. C., Mattion, N. M., Burrone, O., Gonzalez, A., La Torre, J. L., and Scodeller, E. A. (1987),Isolation of group A swine rotaviruses displaying atypical electropherotypes. J. Clin. Microbiol. 25,952-954. Besselaar, T. G., Rosenblatt, A,, and Kidd, A. H. (1986).Atypical rotavirus from South African neonates. Arch. Virol. 87, 327-330. Biryahwaho, B., Hundley, F., and Desselberger, U.(1987).Bovine rotavirus with rearranged genome reassorts with human rotavirus. Arch. Virol. 96, 257-264. Booth, I. W., Chrystie, I. L., Levinsky, R. J.,Marshall, W. C., Pincott, J.,and Harries, J. T. (1982).Protracted diarrhoea, immunodeficiency and viruses. Eur. J. Paediatr. 138, 271-272. Chen, D., Zeng, C. Q. Y., Wentz, M. J., Gorziglia, M., Estes, M. K., and Ramig, R. F. (1994).Template-dependent, in uitro replication of rotavirus RNA. J. Virol. 68,70307039. Chrystie, I. L., Booth, I. W., Kidd, A. H., Marshall, W. C., and Banatvala, J. E. (1982). Multiple faecal virus excretion in immunodeficiency. Lancet 1, 282. Clapp, L. L., and Patton, J. T. (1991).Rotavirus morphogenesis: Domains in the major inner capsid protein essential for binding to single-shelled particles and for trimerization. Virology 180,697-708. Clarke, I. N., and McCrae, M. A. (1981).A rapid and sensitive method for analysing the genome profiles of field isolates of rotaviruses. J. Virol. Methods 2, 203-209. Cohen, J., Laporte, J., Charpilienne, A., and Schemer, R. (1979).Activation of rotavirus RNA polymerase by calcium chelation. Arch. Virol. 60,177-186. Davis, A. R., Hiti, A. L., and Nayak, D. P. (1980).Influenza defective interfering RNA is formed by internal deletion of genomic RNA. Proc. Natl. Acad. Sci.U.S.A. 77, 215219. Desselberger, U. (1989).Molecular epidemiology of rotaviruses. In “Viruses and the Gut” (M. J. G. Farthing, ed.), pp. 55-65. Swan Press, London. Desselberger, U., and McCrae, M. A. (1994).The rotavirus genome. Curr. Top. Microbiol. Immunol. 185,31-66. Desselberger, U., Hung, T., and Follett, E. A. C. (1986).Genome analysis of human rotaviruses by oligonucleotide mapping of isolated RNA segments. Virus Res. 4,357368. Dolan, K. T., ‘hist, E. M., Horton-Slight, P., Forrer, C., Bell, L. M., Plotkin, S. A., and Clark, H. F. (1985).Epidemiology of rotavirus electropherotypes determined by a simplified diagnostic technique 4 t h RNA analysis. J. -Clin. Microbiol. 21, 753758. Domingo, E., Martinez-Salas, E., Sobrino, F., De la Torre, J. L., Portela, A., Ortin, J. Lopez-Galindez, C., Perez-Brena, P., Villanueva, M., Najera, R., VandePol, S., Steinhauer, D., dePolo, N., and Holland, J. (1985).The quasispecies (extremely heterogeneous) nature of viral RNA genome populations: Biological relevance-a review, Gene 40, 1-8. Dyall-Smith, M. L., and Holmes, I. H. (1981).Gene-coding assignments of rotavirus double-stranded RNA segments 10 and 11. J. Virol. 38, 1099-1103. Eaton, B. T.,and Gould, A. R. (1987).Isolation and characterization of orbivirus genotypic variants. Virus Res. 6, 363-382. Eiden, J., Losonski, G. A., Johnson, J., and Yolken, R. (1985).Rotavirus RNA variation during chronic infection of immunocompromised children. Pediatr. Infect. Dis. 4,632637. Estes, M. K. (1990).Rotaviruses and their replication. In “Virology” (B. N. Fields, D. Knipe, R. M. Chanock, M. S. Hirsch, J. L. Melnick, T. P. Monath, and B. Roizman, eds.), pp. 1329-1352. Raven, New York. Fields, S., and Winter, G. (1982).Nucleotide sequences of influenza virus segments 1 and

GENOME REARRANGEMENTS O F ROTAVIRUSES

93

3 reveal mosaic structure of a small viral RNA segment. Cell (Cambridge, Mass.) 28, 303-313. Gallegos, C. O., and Patton, J . T. (1989). Characterization of rotavirus intermediates: A model for the assembly of single-shelled particles. Virology 172, 616-627. Gault-Frere, E., Cassel-Beraud, A. M., and Garbarg-Chenon, A. (1995). Study of a human group A rotavirus with rearranged RNA segments isolated from an immunodeficient child: Preliminary results. Fifth International Symposium on dsRNA Viruses, Djerba; Abstract P1. Gonzalez, S. A., Mattion, N. M., Bellinzoni, R., and Burrone, 0. R. (1989). Structure of rearranged genome segment 11 in two different rotavirus strains generated by a similar mechanism. J. Gen. Virol. 70, 1329-1336. Gorziglia, M., Nishikawa, K., and Fukuhara, N. (1989). Evidence of duplication and deletion in super short segment 11 of rabbit rotavirus Alabama strain. Virology 170, 587-590. Graham, A., Kudesia, G., Allen, A. M., and Desselberger, U. (1987). Reassortment of human rotavirus possessing genome rearrangements with bovine rotavirus: Evidence for host cell selection. J. Gen. Virol. 68, 115-122. Holland, J . J . (1984). Continuum of change in RNA virus genomes. In “Concepts in Viral Pathogenesis” (A. L. Notkins and M. B. A. Oldstone, eds.), pp. 137-143. SpringerVerlag, New York. Holland, J. J., Kennedy, S. I. T., Semler, B. T., Jones, C. L., Row, L., and Grabau, E. A. (1980). Defective interfering RNA viruses and the host-cell response. In “Comprehensive Virology” (H. Fraenkel-Conrat and R. R. Wagner, eds.), Vol. 16, pp. 137-192. Plenum, New York. Holland, J., Spindler, K., Horodyski, F., Grabau, E., Nichol, S., and Vandepol, S. (1982). Rapid evolution of RNA genomes. Science 215, 1577-1585. Hua, J., and Patton, J . T. (1994). The carboxyl-half of the rotavirus nonstructural protein NS53 (NSP1) is not required for virus replication. Virology 198, 567-576. Hundley, F., Biryahwaho, B., Gow, M., and Desselberger, U. (1985). Genome rearrangements of bovine rotavirus after serial passage a t high multiplicity of infection. Virology 143,88-103. Hundley, F., McIntyre, M., Clark, B., Beards, G., Wood, D., Chrystie, I., and Desselberger, U. (1987). Heterogeneity of genome rearrangements in rotaviruses isolated from a chronically infected, immunodeficient child. J. Virol. 61, 3365-3372. Kapikian, A. L., and Chanock, R. M. (1990). Rotaviruses. In “Virology” (B. N. Fields, D. Knipe, R. M. Chanock, M. S. Hirsch, J. L. Melnick, T. P. Monath, and B. Roizman, eds.), pp. 1353-1404. Raven, New York. Khatchikian, D., Orlich, M., and Rott, R. (1989). Increased viral pathogenicity after insertion of a 285 ribosomal RNA sequence into the haemagglutinin gene of a n influenza virus. Nature (London) 340, 156-157. Kirkegaard, K., and Baltimore, D. (1986). The mechanism of RNA recombination in poliovirus. Cell (Cambridge, Mass.) 47, 433-443. Lai, M. M. C. (1992). RNA recombination in animal and plant viruses. Microbial. Rev. 56,61-79. Lubeck, M. D., Palese, P., and Schulman, J. L. (1979). Non-random association of parental genes in influenza A virus recombinants. Virology 95, 269-274. McIntyre, M., Rosenbaum, V., Rappold, W., Desselberger, M., Wood, D., and Desselberger, U. (1987). Biophysical characterization of rotavirus particles containing rearranged genomes. J. Gen. Virol. 68, 2961-2966. Matsui, S. M., Mackow, E. R., Matsuno, S., Paul, P. S., and Greenberg, H. B. (1990). Sequence analysis of gene 11 equivalents from “short” and “supershort” strains of rotavirus. J. Virol. 64, 120-124.

94

ULRICH DESSELBERGER

Matsuno, S., Hasegawa, A., Mukoyama, A,, and Inouye, S. (1985).A candidate for a new serotype of human rotavirus. J. Virol. 54, 623-624. Mattion, N., Gonzalez, S. A., Burrone, O., Bellinzoni, R., La Torre, J. L., and Scodeller, E. A. (1988).Rearrangement of genomic segment 11 in two swine rotavirus strains. J. Gen. Virol. 69,695-698. Mendez, E.,Arias, C. F., and Lopez, S. (1992).Genomic rearrangements in human rotavirus strain Wa: Analysis of rearranged RNA segment 7.Arch. Virol. 125, 331338. Mindich, L., Qiao, X., Onodera, S., Gottlieb, P., and Strassman, J. (1992).Heterologous recombination in the double-stranded RNA bacteriophage phi 6.J. Virol. 66,26052610. Onodera, S., Qiao, X., Gottlieb, P., Strassman, J., Frilander, M., and Mindich, L. (1993). RNA structure and heterologous recombination in the double-stranded RNA phage phi 6.J. Virol. 67, 4914-4922. Palese, P. (1984).Reassortment continuum. In “Concepts in Viral Pathogenesis” (A. L. Notkins and M. B. A. Oldstone, eds.), pp. 144-151.Springer-Verlag, New York. Paul, P. S., Young, S. L., Woode, G. N., Zheng, S., Greenberg, H. B., Matsui, S., Schwartz, K. J., and Hill, H. T. (1988).Isolation of a bovine rotavirus with a “super-short” RNA electrophoretic pattern from a calf with diarrhoea. J. Clin. Microbiol. 26,2139-2143. Pedley, S., Hundley, F., Chrystie, I., McCrae, M. A., and Desselberger, U. (1984).The genomes of rotaviruses isolated from chronically infected immunodeficient children. J. Gen. Virol. 65, 1141-1150. Pedley, S.,Bridger, J. C., Chasey, D., and McCrae, M. A. (1986).Molecular definition of two new groups of atypical rotaviruses. J. Gen. Virol. 67, 131-137. Pocock, D. H. (1987).Isolation and characterization of two group A rotaviruses with unusual genome profiles. J. Gen. Virol. 68, 653-660. Qian, Y., Jiang, B., Saif, L. F., Kang, S. Y., Ojeh, C. K., and Green, K. Y. (1991). Molecular analysis of the gene 6 from a porcine group C rotavirus that encodes the NS34 equivalent of group A rotaviruses. Virology 184, 752-757. Ramig, R. F.,Samal, S. K., and McConnell, S. (1985).Genome RNAs of virulent and attenuated strains of bluetongue virus serotypes 10, 11, 13,and 17.In “Bluetongue and Related Orbiviruses” (T. L. Barger and M. M. Yochim, eds.), pp. 389-396. Alan R. Liss, New York. Rixon, F., Taylor, P., and Desselberger, U. (1984).Rotavirus RNA segments sized by electron microscopy. J. Gen. Virol. 66,233-239. Romanova, L. I., Blinov, V. M., Tolskaya, E. A., Victorova, E. G., Kolesnikova, M. S., Guseva, E. A., and Agol, V. I. (1986).The primary structure of crossover regions of intertypic poliovirus recombinants: A model of recombination between RNA genomes. Virology 155,202-213. Sabara, M., Deregt, D., Babiuk, L. A., and Misra, V. (1982).Genetic heterogeneity within individual bovine rotavirus isolates. J. Virol. 44,813-822. Sabara, M., Ready, K. F. M., Frenchick, P. J., and Babiuk, L. A. (1987).Biochemical evidence for the oligomeric arrangement of bovine rotavirus nucleocapsid protein and its possible significance in the immunogenicity of this protein. J. Gen. Virol. 68, 123133. Saulsbury, F. T., Winkelstein, J. A,, and Yolken, R. J. (1980).Chronic rotavirus infection in immunodeficiency. J. Pediatr. 97,61-65. Scott, G. E.,Tarlow, D., and McCrae, M. A. (1989).Detailed structural analysis of a genome rearrangement in bovine rotavirus. Virus Res. 14,119-128. Shen, S., and Bai, Z. S. (1990).Genome variation and rearrangements of a lamb rotavirus after 96 passages in cell culture a t high m.0.i. Chin. J. Epidemiol. ll, 110114.

GENOME REARRANGEMENTS OF ROTAVIRUSES

95

Shen, S., Burke, B., and Desselberger, U. (1994). Rearrangements of the VP6 gene of a group A rotavirus in combination with a point mutation affecting trimer stability. J . Virol. 68, 1682-1688. Tanaka, T. N., Conner, M. E., Graham, D. Y.,and Estes, M. K. (1988). Molecular characterization of three rabbit rotavirus strains. Arch. Virol. 98, 253-265. Taniguchi, K. (1995). “Sequence analysis of VP5 genes of porcine rotaviruses from Thailand.” International Symposium on Viral Gastroenteritis, Sapporo, Japan. [Abstract] Taniguchi, K., Kojima, K., Kobayashi, N., Urasawa, T., and Urasawa, S. (1994). F’roperties of a bovine rotavirus variant with gene 5 having a deletion of 500 base pairs. ‘henty-eighth Joint Working Conference on Viral Diseases, Japan-US Cooperative Medical Science Program, Tokyo, Japan. [Abstract] Thouless, M. E., DiGiacomo, R. F., and Neuman, D. S. (1986). Isolation of two lapine rotaviruses: Characterization of their subgroup, serotype and RNA electropherotypes. Arch. Virol. 89, 161-170. Tian, Y.,Tarlow, O., Ballard, A., Desselberger, U., and McCrae, M. A. (1993). Genomic concatemerization/deletion in rotaviruses: A new mechanism for generating rapid genetic change of potential epidemiological importance. J. Virol. 67, 6625-6632. Ward, R. L., Knowlton, D. R., and Pierce, M. J. (1984). Efficiency of human rotavirus propagation in cell culture. J. Clin. Microbiol. 19, 748-753. Wenske, E. A., Chanock, S. J., Krata, L., and Fields, B. N. (1985). Genetic reassortment of mammalian reoviruses in mice. J. Virol. 56, 613-616.

This Page Intentionally Left Blank

ADVANCES IN VIRUS RESEARCH, VOL. 46

HUMAN IMMUNODEFICIENCY VIRUS TYPE 1 REVERSE TRANSCRIPTASE AND EARLY EVENTS IN REVERSE TRANSCRIPTION Eric J. Arts and Mark A. Wainberg McGill University AIDS Centre The Sir Mortirner B. Davis-Jewish General Hospital Montreal, Quebec H3T 1E2, Canada

I. Introduction 11. Overview of Human Immunodeficiency Virus Qpe 1 Replication A. Initial Events in HIV-1 Replication B. Virus Assembly and Maturation of HIV-1 111. Human Immunodeficiency Virus Q p e 1 Reverse Transcriptase A. Structure of HIV-1 Reverse Transcriptase B. Interaction of HIV-1 Reverse Transcriptase with Primer and Template C. Polymerase Active Site and Deoxynucleoside 5'-Triphosphate Binding Site of HIV-1 Reverse Transcriptase IV. Human Immunodeficiency Virus Q p e 1 Reverse Transcription A. Overview of Reverse Transcription Scheme of Retroviruses B. Origin of HIV-1 Reverse Transcription C. Host tRNALys3 Primer in HIV-1 Reverse Transcription D. RNA- and DNA-Dependent DNA Polymerization E. Fidelity of Polymerization by HIV-1 Reverse Transcriptase F. Ribonucleases of HIV-1 Reverse Transcriptase G. First Template Switch References

I. INTRODUCTION The first cases of acquired immunodeficiency syndrome (AIDS) were reported in 1981 (Gottlieb et al., 1981; Masur et al., 1981).By this date, the disease had already rapidly spread in the homosexual community and among intravenous drug users. The causative agent thought to be responsible for this syndrome showed a pattern of blood-borne transmission, but it was not until 1983 that the pathogen was isolated and partially characterized (Barre-Sinoussi et al., 1983; Popovic et al., 1984). A unique human retrovirus, lymphadenopathy-associated virus or human T-cell lymphotropic virus type I11 [later termed the human 97

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

98

ERIC J. ARTS AND MARK A. WAINBERG

immunodeficiency virus type 1 (HIV-111, was the proposed etiological agent of AIDS. There are now five known human retroviral species: Human T-cell lymphotropic virus (HTLV) type I, HTLV type 11, HIV type 1,HIV type 2, and the human foamy virus. The human foamy virus belongs to the subfamily Spumavirinae and is not a human pathogen (Hotta and Loh, 1987), whereas HTLV-I and -11 are classified as Oncouirinae and are shown to cause chronic and sometimes fatal leukemia (Wong-Stahland Gallo, 1985). Infections with HIV-1 or HIV-2 showed the same latent progression to disease as occurs with animal lentiviruses such as Visna and equine infectious anemia viruses (EIAV) (reviewed by Cheevers and McGuire, 1985; Davis et al., 1987).The latter viruses are classified as Lentiuirinae and differ from classic retroviruses in genomic organization. Lentiviruses and HTLV/bovine leukemia viruses contain small open reading frames (ORF) found 3’ of the gag and pol genes and surrounding the enu gene. Many of these ORFs encode for viral accessory proteins, some of which are essential for virus replication. In the case of HIV-1, there are at least six ORFs in addition to the gag, pol, and enu genes (reviewed by Cullen, 1991). These accessory genes vary in position and code for proteins that differ in structure and function in different retroviruses. However, the pol genes of all retroviruses encode three enzymes, reverse transcriptase, integrase, and an aspartic proteinase, all of which are highly specific in function. The pol gene of EIAV and possibly other lentiviruses such as Visna virus and caprine arthritis-encephalitis virus (CAEV)contains a ORF encoding a deoxyuridine triphosphatase, similar to that found in herpesvirus (Threadgill et al., 1993). The reverse transcriptase (RT) enzyme was discovered independently by Howard Temin and David Baltimore (Temin and Mizutani, 1970; Baltimore, 1970). The discovery of an RNA-dependent DNA polymerase challenged a central scientific dogma which stated that the key to reproduction of any entity was limited to progression from DNA to RNA to protein. Since the discovery of RT in RNA tumor viruses, RNA-dependent DNA polymerization activities have been characterized in the telomeric DNA of nearly all eukaryotes (Greider and Blackburn, 1985), in retrotransposons (Boeke et al., 1985; Temin, 1985), in cauliflower mosaic virus (Guilley et al., 1983; Pfeiffer and Hohn, 1983), in bacteria such as Myxococcus xanthus and Escherichia coli (Dhundale et al., 1987; Inouye et al., 1989), and in hepadnaviruses (Summers and Mason, 1982).This review focuses on the properties and inhibition of early events in HIV-1 reverse transcription.

HIV-1 REVERSE TRANSCRIPTASE

99

II. OVERVIEW OF HUMAN IMMUNODEFICIENCY VIRUSTYPE 1REPLICATION

A . Initial Events in HIV-1 Replication The outer envelope glycoprotein (gp120) of HIV-1 specifically interacts with the extrinsic CD4 protein on the host cell plasma membrane (Maddon et al., 1986; McDougal et al., 1986). The CD4 receptor is expressed on macrophages, monocytes, and a subset of T lymphocytes (Thelper cells), but the V3 domain of gp120 may control the tropism of viral entry into the latter cells (Hwang et al., 1991). On CD4-gp120 binding, the viral and cellular membranes fuse in a pH-independent manner (Stein et al., 1987). This fusion is thought to be facilitated by the viral gp41 transmembrane protein, which has a hydrophobic amino-terminal domain that bears a high degree of amino acid sequence similarity to the amino termini of the fusion and hemagglutinin proteins of paramyxoviruses and orthomyxoviruses, respectively (Bosh1 et al., 1989). Virus-host cell membrane fusion or endocytosis permits HIV-1 core entry into the cytoplasm. The fate of the viral core remains uncertain, but only partial dissolution is required for deoxynucleotide 5’-triphosphate (dNTP) entry and initiation of reverse transcription (Zhang et al., 1993). In HIV-1 reverse transcription, proviral double-stranded DNA (dsDNA) is synthesized from the ( +) RNA genome by the viral RT enzyme (Di Marzo Veronese et al., 1986; reviewed by Skalka and Goff, 1993). As will be described, reverse transcription may be initiated in virions prior to host cell entry (Arts et aZ., 1994a; Lori et al., 1992; Trono, 1992; H. Zhang et al., 1993, 1994),but more recent results suggest that this virion DNA may not be required for infection. In addition, completion of HIV-1 reverse transcription in a quiescent CD4+ lymphocyte may require cell activation (Zack et al., 1990). This stall in HIV-1 reverse transcription during infection of a quiescent host cell may be an initial step leading to latent viral infection. The 5’ ends of both strands in double-stranded proviral DNA are subjected to endonucleolytic dinucleotide cleavage by HIV-1 integrase in the nucleus or during transport of the viral DNA to the nucleus (Bushman et al., 1990; Engelman et al., 1991).Integrase, possibly associated with a nucleoprotein complex (HIV-1 nucleocapsid and matrix proteins) (Bukrinsky et al., 1993131, remains bound to the ends of proviral DNA, thus permitting a nucleophilic attack at a single, nonspecific site on host genomic DNA (Engelman et aZ., 1991).Integration of proviral DNA of nearly all retroviruses is only possible during mitosis of a cycling cell (Lewis et al., 1992; Roe et al., 1993; Peters et al., 1977).

<

Integrated HIV Genomfc DNA

packaging of (+) RNA. by p55QaP and p160WPpo'

4. budding of immature virion

FIG.1. Assembly and maturation of virions during human immunodeficiency virus replication.

102

ERIC J. ARTS AND MARK A. WAINBERG

However, HIV-1 proviral DNA was found recombined with the cellular DNA of HIV-1 exposed, quiescent CD4+ peripheral blood lymphocytes that remained in Go phase (Stevenson et al., 1990). HIV-1 proviral integration in quiescent CD4+ lymphocytes may be controlled by active transport of the preintegration complex to the nucleus. Mutations in the nuclear localization signal of the HIV-1 matrix protein (MApl7; Fig. 1) results in mitosis-dependent integration of HIV-1 proviral DNA, typical of other retroviruses (Bukrinsky et al., 1993a).It has also been proposed that a latent HIV-1 infection may be related to the site of proviral integration in the host cell genome (Winslow et al., 1993). Integration that does not disrupt transcription of an essential host gene will ultimately lead to productive viral infection. As in all retroviruses, transcription of the HIV-1 genome begins at the repeat (R) region of the long terminal repeat (LTR) (Fig. 1).The LTR of HIV-1 contains a number of binding sites for host transcriptional enhancers and promoters such as NF-KB,NFAT-1, USF, AP-1, Spl, and glucocorticoid receptor (reviewed by Cullen, 1991).Enhancement of RNA polymerase I1 activity by the latter factors is required for the initiation of HIV-1 transcription from the LTR (Cullen, 1991). As a result of two splicing events, early HIV mRNA only encodes for the HIV-1 regulatory proteins, namely, tat, rev, and nef (Cullen, 1991). Low concentrations of host transcriptional promoters and enhancers could result in another form of HIV-1 latency, because the HIV-1 tat protein, encoded by the viral RNA that is trans-activated by the latter host factors, is responsible for high levels of HIV-1 RNA transcription (Arya et al., 1985; Sodroski et al., 1984). The tat interaction with the trans-activation response (TAR)element found in the R region of viral RNA increases the presence of HIV RNA transcripts by either tat-induced stabilization of an elongating RNA transcript (Kao et al., 1987; Laspia et al., 1989) or trans-activation of RNA polymerase I1 in the process of initiation of a new HIV RNA transcript (Kashanchi et al., 1994; Sharp and Mariniak, 1989). The eventual buildup of rev protein in the nucleus permits unspliced and singly spliced HIV RNA to exit the nucleus for translation of HIV structural proteins (Fig. 1) (Pomerantz et al., 1990; Sodroski et al., 1986).The rev protein interacts with the rev responsive element (RRE) in the env gene (Hadzopoulos-Cladaras et al., 1989; Malim et al., 1989) and has been shown to interfere with the splicing events in HIV-1 RNA (Chang and Sharp, 1989) and the transport of unspliced and singly spliced transcript out of the nucleus (Malim et al., 1989). The singly spliced transcript encoding for the envelope and upu proteins is translated by endoplasmic reticulum-bound ribosomes. The envelope precursor protein (gp160) is glycosylated, folded, and dimer-

HIV-1 REVERSE TRANSCRIPTASE

103

ized in the endoplasmic reticulum (Earl et al., 1991), then cleaved to gp41 and gp120 by the cellular furin proteinase in the Golgi complex (Hallenberger et al., 1992). The extrinsic gp120 remains bound to the transmembrane gp41 via noncovalent links (Schneider et al., 1986) and is transported to the plasma membrane prior to virus budding (Earl et al., 1991) (Fig. 1).Unspliced viral RNA is translated by free ribosomes from the first AUG (+789). At least 90% of all translation events terminate at the UAA stop codon (+2289) and result in the synthesis of the gag polyprotein (p55gag) (Fig. 1)(reviewed by Levin et al., 1993). An infrequent - 1ribosomal frameshift (-10% o f gag translation) at a stretch of uridine bases (+2083 to +2089) results in readthrough of the latter UAA stop codon and in the synthesis of the gag-pol precursor protein (pl60gag-po’)(Fig. 1)(Jacks et al., 1988). Most retroviruses, including HIV-1, require an RNA stem-loop or pseudoknot structure 3‘ of the purine-rich sequence for efficient ribosomal frameshifting (Fig. 1)(Jacks et al., 1987, 1988; Parkin et al., 1992).

B . Virus Assembly and Maturation of HIV-1 The N terminus of both pl60gag-pol and p 5 @ are ~ myristylated during translation (Gottlinger et al., 1989) (Fig. 1). Only the attached myristate and the amino-terminal sequences of ~ 5 % -and pl6OgW-pol are necessary for the transport and attachment of the precursors to the plasma membrane, and for viral budding (Wang et al., 1993; Yuan et al., 1993). Deletion of the myristate addition signal on pl6oBag-po2, in the presence of wild-type p5!9--myristate, did not appreciably affect virus particle production, suggesting that specific p16Og--pol-p5%W interactions may be required for proper virus assembly (Park and Morrow, 1992; Smith et al., 1993). The gag and pol precursors may also be required for specific sequestration and encapsidation of viral genomic RNA (Berkowitz et al., 1993),the tRNALy*isoacceptor species (Jiang et al., 1993; Mak et al., 19941, and HIV-1 upr protein (Lavallee et al., 1994) into budding virus particles. Deletions in p5@- outside of the N-terminal and myristate addition signals resulted in some endoplasmic reticulum localization, a reduction of gp120/gp41 on mature virus particles, and a loss of viral infectivity (Fache et al., 1993; Yu et al., 1992). Therefore, p55 gag may coordinate virus assembly on the plasma membrane (von Poblotzki et al., 19931, but the actual process of viral budding is still poorly understood (Fig. 1).It has been shown that upu and uif are essential for virus budding in some HIV-l-infected cell lines (Gottlinger et al., 1993; Sakai et al., 1991, 1993). The HIV-1 virion is classified as a type D retrovirus. In type D retrovirions, proteolytic processing and mature core formation do not gener-

104

ERIC J. ARTS A N D MARK A. WAINBERG

ally occur until virus is released from the infected cell (Fig. 1). Proteolytic processing in HIV-1 begins with autoproteolysis of the proteinase, the first pol enzyme in the gag-pol fusion protein derived from the p160 gag-& precursor (Zybarth et al., 1994; Peng et al., 1989). The free aspartic proteinase (PRpll) then dimerizes and enacts cleavage at two and five sites in p55g~gand pl6Ogag-~ol, respectively (Kohl et al., 1988). No amino acid consensus sequence exists among the five sites, except that a string of at least six hydrophobic amino acids (i.e., proline, tryptophan) is required for strong binding in the hydrophobic pocket of the HIV-1 proteinase dimer (Fitzgerald et al., 1990; Kohl et al., 1988; LeGrice et al., 1988).The initial products of HIV-1 proteinase cleavage are the matrix protein (MApl7), the capsid protein (CAp24), and the nucleocapsid protein precursor (NCpl5) from the p5Wg precursor, along with MApl7, CAp24, NCpl5, PRpll, the reverse transcriptase monomer (p66), and the integrase protein (INp32) from the pl6Og~g-polprecusor. As will be described in more detail, the p66 RT monomer can be further cleaved by proteinase to form p51 and p15 subunits (Di Marzo Veronese et al., 1986).The ~ 6 6 1 ~ heterodimer 51 is the fully functional, mature form of HIV-1 RT (reviewed by JacoboMolina and Arnold, 1991; Larder et al., 1987a). In addition, the NCpl5 protein is further cleaved in the virion to the NCp7, p l , p2, and p6 proteins (Henderson et al., 1992). The N terminus of the matrix protein is still myristylated and remains attached to the viral membrane or envelope (Yuet al., 1992). The capsid protein forms a bullet-shaped core surrounding the dimerized genomic RNA, the tRNALys3 isoacceptor species, mature HIV-1 RT (p66/p51), NCpl5 protein, upr protein and possibly other hiral components, e.g., PRpll homodimers and INp32 (reviewed by Haseltine, 1991).The precursor NCpl5 and the proteinase-cleaved NCp7 contain two functionally distinct zinc fingers capable of binding and annealing viral RNA (i.e., tRNALys and viral genomic RNA) (Barat et al., 1993; De Rocquigny et al., 1992; South et al., 1990).The NCp7 protein likely coats the viral RNA, thus protecting the RNA from nucleases and promoting reverse transcription (De Rocquigny et al., 1992). Association of tRNALys3with genomic RNA and dimerization of HIV-1 RNA may be enhanced by specific binding of the NC proteins to the anticodon loop of tRNALys species and the proposed dimerization signal region (between the primer binding sequence and the AUG start codon of gag), respectively (Barat et al., 1993; Darlix et al., 1990; Sakaguchi et aZ., 1993). However, preferential incorporation of tRNALys species and specific encapsidation of viral genomic RNA must occur prior to virus release and subsequent proteolytic digestion of precursors (Fig. 1).Therefore, the RNA binding domains in NCpl5 may also be present in p55gw and p160gw-p01precursors.

HIV-1 REVERSE TRANSCRIPTASE

105

111. HUMANIMMUNODE~~CIENCY VIRUSTYPE1 REVERSE TRANSCRIPTASE RNA-dependent DNA polymerization (RDDP) activity in the supernatant of cultured cells from a n AIDS-afflicted individual was key in the discovery and characterization of a new pathogenic human retrovirus (Barre-Sinoussi et al., 1983; Popovic et al., 1984). The RT enzyme isolated from the HIV-1 core functionally resembled that of many other retroviruses, including Moloney murine leukemia virus (MLV) and avian myeloblastosis virus (AMV) (Roth et al., 1985; reviewed by Weiss et al., 1985). However, HIV was characterized as a lentivirus due to its slow, progressive infection, and thus its RT could also be compared to other lentiviral RTs such as that of EIAV (Borroto-Esoda and Boone, 1991; LeGrice et al., 1991b; Wohrl et al., 1994). In the 1980s, research on the RTs of HIV-1 and HIV-2 surpassed that of all other cellular or viral polymerases. The main objective of this work was the development of potent antiviral compounds to disrupt the unique retroviral process of reverse transcription. To date, the most successful chemotherapies in the treatment of AIDS or HIV infections are anti-RT drugs categorized as nucleoside analogs, for example, 3'-azido-3'-deoxythymidine(zidovudine or AZT) (Mitsuya et al., 19851, and nonnucleoside antagonists of RT, such as dipyridodiazepinones (e.g., nevirapine) (Kroup et al., 1991). Research into HIV RT and reverse transcription has also resulted in the biochemical and molecular characterization of several enzymatic processes common to many RNA and DNA polymerases. HIV-1 RT is capable of several enzymatic functions which include RDDP, DNA-dependent DNA polymerization (DDDP), DNA-RNA duplex-dependent ribonuclease activity (RNase HI, and RNA-RNA duplex-dependent ribonuclease activity (RNase D or RNase H*). The HIV-1 RT enzyme was initially isolated and identified in virus particles (Di Marzo Veronese et al., 1986). Sera from HIV-1-infected individuals recognized proteins of 66 and 51 kDa (p66 and p51) of unknown identity among the core and envelope antigens of HIV-1. The p66 and p51 proteins were purified from virions by immunoaffinity chromatography, then sequenced or assayed for enzymatic activity. It was shown that these proteins were coded by the same region of the pol gene, and both possessed RNA-dependent DNA polymerization activity (Di Marzo Veronese et al., 1986). The pol gene is translated due to a ribosomal frameshift during gag translation of the full-length (+) RNA (Fig. 1) (Jacks et al., 1988). The p16Wg-Pol protein of HIV-1 is processed during virus maturation by HIV-1 proteinase (Kohl et aZ., 1988). The RT protein (~661,flanked by the proteinase and integrase proteins in the p 1 6 0 g ~ -precursor, ~2 is always found in association

106

ERIC J. ARTS AND MARK A. WAINBERG

with p51 in the virion (Di Marzo Veronese et al., 1986; Lightfoote et al., 1986). Expression of the cloned pol gene in E . coli resulted in the synthesis not only of p66 but also of pi51 (Tanese et al., 1986; Mous et al., 1988). Thus, HIV-1 proteinase, expressed along with p66 from the pol gene is required for proteolytic cleavage of p66 to form p51 (Farmerie et al., 1987; Mous et al., 1988). Cleavage of p66 by HIV-1 proteinase occurs between F440 and Y441 (LeGrice et al., 1989; Mizrahi et al., 1989), resulting in the RNase H-deficient product, p51 (Schatz et al., 1989; Tisdale et al., 19881, and the inactive RNase H domain product, p15 (Becera et al., 1990; Hansen et al., 1988; Starnes and Chang, 1989). Recombinant enzyme has been utilized for many functional and structural studies on HIV-1 RT. Recombinant HIV-1 RT has been expressed in a variety of cell types such as bacteria [i.e., E . coli (Farmerie et al., 1987; Larder et al., 1987b) and Bacillus subtilis (LeGrice et al., 1987)], Saccharomyces cereuisiae (Barr et al., 1987), and insect cells (Kawa et al., 19931, using expression vectors containing the entire pol gene or separate genes for the p66 and p51 sequences. The p66/p51 or p66 RT protein can be purified from cell lysates by size exclusion chromatography (Restle et al., 19901, immunoaffinity chromatography (Furman et al., 19911, or metal chelate affinity chromatography (Chattopadhyay et al., 1992; LeGrice and Gruninger-Leitch, 1990; LeGrice et al., 1994). Rapid RT purification from crude lysates by immobilized metal chelate affinity chromatography (IMAC) required a short polyhistidine affinity label added to the amino terminus of the p66 coding sequence (LeGrice and Gruninger-Leitch, 1990; LeGrice et al., 1994). For rapid analysis of mutated forms of HIV-1 RT, enzyme activity can be measured directly in lysates of E . coli trp- mutants grown on tryptophan-deficient media and transformed with a trpE-HIV-1 pol fusion vector (Prasad and Goff, 1989a; Tanese et al., 1986). Different methods of RT expression and purification can often result in significant differences in enzymatic activity (L. Martin and S. F. J. LeGrice, personal communications, 1995; Craven et al., 1992). The HIV-1 RT, like HIV-2 FtT, avian myeloblastosis virus (AMV) RT, and b u s sarcoma virus (RSV) RT, is found in the virion as a stable heterodimer (Lowe et al., 1988; Lightfoote et al., 1986). The association constant for heterodimer formation was significantly greater than that for homodimer formation (i.e., 109 M - 1 for p66/p51, 2.3 x 105 M - 1 for ~ 6 6 1 ~ 6and 6 , 1.3 X lo3 M - l for p51/p51), but the constants were dependent on initial monomer concentrations (Becerra et al., 1991; Restle et al., 1990). In addition, the RT heterodimer was by far the most stable, showing no measurable dissociation after 1000 h r while the half-lives of the p66 and p51 homodimers were 19 and 3 h r at

LEGEND Subdornalns 01 HIV-1 reverse tranbcrlutase I Fingers I Palm I Thumb I Connection I polymerase actlve slte I dNTP binding site Secondary

--

A.

7-24 2844 49-77

7&83

86-1 12 114-127 128-147 155-174 178-191 195-212 214-242 255-31 1 316-358 364-382 366-391 395-404 408.430

FIG.2 Crystal structure of the p66 and p51 subunits of HIV-1 reverse transcriptase. Modified from the 2.4 A resolution of the crystal structure of HIV-1 KT complexed with dsDNA (JacoboMolina et al., 1993). The and a represent the predicted P-sheet and a-helix secondary structures in the subunits.

P

This Page Intentionally Left Blank

HIV-1 REVERSE TRANSCRIPTASE

107

O'C, respectively (Restle et al., 1990). All three RT dimers and the p l 6 @ ~ - precursor ~0~ possessed polymerase activity (Bavand et al., 1993; Gottlinger et al., 1989; Lowe et al., 1988; Peng et al., 1991; Restle et al., 1990),but only the p66 homodimer and p66/p51 were efficient for both RDDP and RNase H digestion (Hansen et al., 1988; Starnes and Cheng, 1989). The p66/p51 RT heterodimer is the only form of RT found in HIV-1 virions and may suffice for complete proviral DNA synthesis (Lowe et al., 1988).

A . Structure of HIV-1 Reverse Transcriptase The structure of HIV-1 RT has now been determined t o a resolution of 2.9 to 3.5 h; by electron density mapping of crystals of HIV-1 p66/p51 complexed with nevirapine (Kohlstaedt et al., 1992; Smerdon et al., 1994), or p66/p51 complexed with dsDNA (Jacobo-Molina et al., 1993) and p15 RNase H (Davies et al., 1991). The size and quaternary structure of p66/p51 was also determined by a neutron small-angle scattering technique (Lederer et al., 1992). The p66 domain (110 x 30 x 45 h;), known to contain the functional catalytic domains for polymerase and RNase H activities, resembles a human hand grasping the primer-template complex. The p66 hand structure has been further subdivided into five subdomains (Fig. 2A and Table I) (Kohlstaedt et al., 1992). The fingers subdomain (amino acids 1-84 and 120-150) is composed of a mixed five-stranded P-pleated sheet with three a helices facing the primer-template binding pocket. Attached to the fingers subdomain is the palm subdomain (amino acids 85-119 and 151-244) containing the catalytic site for polymerization and consisting of a five-stranded P sheet mixed with 3 (Y helices. The thumb subdomain (amino acids 245-322) may recognize and bind the primer through its bundle of four (Y helices. The polymerase region of RT (fingers, palm, and thumb) and the RNase H subdomain (amino acids 438-5601, a structure of a five-stranded P sheet mixed with four a helices (Davies et al., 1991), are linked together via a 14-stranded P sheet segment called the connection subdomain (amino acids 323-437) (Fig. 2A and Table I) (Jacobo-Molina et al., 1993; Kohlstaedt et al., 1992). The p51 subunit in the crystal structure of p66/p51 RT differs from the p66 subunit in subdomain positioning in the tertiary structure but not in the a-helix and P-pleated sheet formation in the secondary structure (Fig. 2B) (Jacobo-Molina et al., 1993; Kohlstaedt et al., 1992). In p51, the connection domain is pulled toward the palm, filling in the cleft necessary for polymerization. Displacement of the connection subdomain in p51 of the RT heterodimer, as compared with p66, signif-

TABLE I

FUNCTIONAL SUBDOMAINS OF HIV-1 REVERSETRANSCRIPTASE Subdomain of R T a and amino acid residues Fingers 1-62

63-84 and 120-150

Palm 85-119

151-244

Palm and thumb 195-244and245-322

Functional amino acids (form of analysis)

Role or function in RT

Ref.

L26, 131, P25, T27, K46, P55, N57, T58 (cassette substitutions)

Necessary for efficient polymerization

Boyer et al., (1992a)

1. K73 (photoaffinity labeling with rmethyl-3HI dTTP) 2. 163, K64, K65 (cassette substitutions) 3. 65-70 (MAb inhibition)

dNTP binding site*

1. Cheng et al. (1993)

D110, D113, A114, Y115 (substitutions)

Part of

Q155, K154, Y181, Y183, M184, D185, D186 (substitutions)

Polymerization active sitec

Larder et al. (1987a,b, 1989a,b), Boyer et al. (1992b), Wakefield et al. (1992)

1. 195-244 (photoaffinity labeling and UV crosslinking with poly(dT) primer)

1. Primer binding domain

1. Sobol et al. (1991); Kumar et al. (1993); Basu et 01. (1989)

2 . Boyer et al., (1992b) 3. Wu et al., (1993)

polymerization active sitec

Larder et al., (1987a,b, 1989a,b), Boyer et al. (1992b)

Connection 323-437

2. 283-310 (series of leucine residues) 3. C280 (substitution)

2. Proposed dimerization

2. Becerra et al. (1991)

sequence 3. Involved in RNase H activity

3. Loya et al. (1992; Hizi et al. (1993)

1. 398-414 (series of tryptophan residues) 2. 400-426 (deletion)

1. Dimerization signal

1. Becerra et al. (1991)

2. Necessary for p15 RNase H activity 3. Necessary for primertemplate binding in RNase H domain

2. Smith and Fbth (1993)

RNase H active site

Schatz et al. (1989); Tisdale et al. (1991); Wohrl et al. (1991); Cirino et al. (1993)

3. 436-440 (deletion)

RNase H 438-560 F

0

co

a 6

534-539 (conserved sequence for all RT RNase H); H539, E478

Subdomains of HIV-1 RT are shown schematically in Fig. 2. The dNTP binding site (pink in Fig. 2). Polymerization active site (light blue in Fig. 2).

3. Cirino et al. (1993)

110

ERIC J. AIETS AND MARK A. WAINBERG

icantly reduces the solvent-accessible surface area and buries several hydrophobic regions (Wang et al., 1994).The thumb subdomain in p66, necessary for primer binding, is found perpendicular to the connection subdomain. However, in p51 it is pulled back 155”,placing it almost parallel with the p sheets of the connection subdomain (Kohlstaedt et al., 1992). It is hypothesized that the tertiary structure of one of the p66 subunits in a p66/p66 homodimer may resemble the p51 conformation found in the RT heterodimer V‘p51-like”) (Davies et al., 1991; Sharma et al., 1994; Wang et al., 1994).Separation of imidazole-treated p66/p66 homodimer by high performance liquid chromatography (HPLC) revealed two stable peaks; one peak displayed wild-type RDDP and RNase H activities, whereas the other displayed reduced RNase H activity but wild-type RDDP activity. Reduction in RNase H activity of the p66 subunit found in the latter peak suggests that this p66 subunit assumed a conformation similar in subdomain positioning as the p51 subunit found in the ~ 6 6 1 ~ heterodimer 51 (Sharma et al., 1994).Interestingly, the stability of p66 from both peaks contradicts the assumption that the “p51-like” p66 subunit possessed a lower energy conformation or that the p66 and p51 monomers would favor a “p51-like” configuration (Wang et al., 1994).Nevertheless, the asymmetric conformation of the homodimer may be necessary for HIV-1 proteinase cleavage at site 440 and the release of the RNase H domain (p15) from the polymerase region (Davies et al., 1991; Hostomoska et al., 1991). The proposed dimerization sites for RT are found in two regions. The first is a series of leucines between residues 283 and 310 (Table I) (aJ in Fig. 2B). This sequence is thought to form a leucine zipper between p66 and p51 (Becerra et al., 1991). However, the p66 and p51 do not contact one another at this sequence in the aJ helix of the thumb subdomains (Jacobo-Moline et al., 1993; Kohlstaedt et al., 1992). The other dimerization sequence is a series of tryptophans between residues 399 and 414 in the connection subdomain (Table I) (Becerra et al., 1991). Although this sequence is adjacent to the RNase H subdomain, the tryptophan residues were not important for RNase H function (Smith and Roth, 1993). In the crystal structure of FtT complexed with dsDNA, the tryptophan repeat sequence (aL and p20; Fig. 2) in p66 and p51 were in close association, thus supporting their proposed role in dimerization (Jacobo-Molina et al., 1993; Kohlstaedt et al., 1992). X-Ray crystallography studies of HIV-1 RT also revealed that contacts between the connection subdomains of p66 and p51 constitute nearly one-third of all subunit interactions (Jacobo-Molina et al., 1993; Kohlstaedt et al., 1992; Sharma et al., 1994) and that the connection subdomains likely play a central role in dimer formation (Wang et al.,

HIV-1 REVERSE TRANSCRIPTASE

111

1994). In fact, synthetic peptides derived from the connection subdomain are capable of inhibiting dimerization and, as a result, all RT enzymatic functions (Divita et al., 1994). The most effective dimerization-inhibiting peptide was homologous t o residues 389 t o 407 of RT, that is, the tryptophan repeat sequence (Divita et al., 1994).

B . Interaction of HIV-1 Reverse Transcriptase with Primer and Template The polymerase active site is found in the palm subdomain and is separated from the RNase H active site by at least 50 A (Lederer et al., 1992) or 15 to 16 bases of dsDNA (Arnold et al., 1992). However, the predicted distance between the two sites was 14 to 19 nucleotides when comparing the initiation site of RDDP and RNase H on an RNA template (DeStefano et al., 1991; Furfine and Reardon, 1991; Gopalakrishnan et al., 1992). X-Ray crystallographic analysis of HIV-1 RT, complexed with an 18-nucleotide DNA template, showed the distance between the RNase H and polymerase active sites to be 18 to 19 nucleotides (Jacobo-Molina et al., 1993). The predicted base distance was based on a linear B-conformation helix (Arnold et al., 1992), whereas the high-resolution crystal structure revealed a dsDNA duplex, which changes conformations in the primer-template binding pocket of RT (Jacobo-Molina et al., 1993).The l-base overhang from the DNA template is aligned just over the polymerase active site in the palm subdomain. Residues in the fingers and palm subdomain of p66 of heterodimeric RT may play a role in positioning the template in the palm subdomain of p66. The HIV-1 Ft” mutants LlOOS, L109S, A114S, and V118S are significantly less processive than wild-type RT (Boyer et al., 1994). All of the latter mutations are in close proximity to the polymerase active site and form part of the “floor”in the palm subdomain, onto which the template is bound (Fig. 2). The p51 subunit may also play a role in primer-template binding and positioning. Increasing the extent of deletions at the amino terminus of p51 in heterodimeric HIV-1 w decreased both processivity in DDDP reactions and binding to tRNAL@ (Jacques et al., 1994b). Hydroxyl-radical footprinting of HIV-1 RT on a DNA primer-DNA template complex revealed that HIV-1 RT protected from + 3 to -15 nucleotides from the primer terminus on DNA template (Metzger et al., 1993). A “window” of accessibility for hydroxyl-radical-dependent cleavage existed between -8 and -11 nucleotides in the region of template protected by HIV-1 RT (Metzger et al., 1993). This “window” is found just upstream of the RNase H active site and was later shown

112

ERIC J. ARTS AND MARK A. WAINBERG

to be the location of an A- to B-DNA transition in the primer-template complexed with HIV-1 RT (Jacobo-Molina et al., 1993). The crystal structure of rat DNA polymerase (3 with DNA has revealed that the bulk of the enzyme interacts in front of the primer terminus (Pelletier et al., 1994 Sawaya et al., 1994). The fingers, palm, and thumb subdomains of rat DNA polymerase (3 are quite similar in structure to that of HIV-1 RT and Klenow fragment (Ollis et al., 1985; Kohlstaedt et al., 1992; Jacobo-Molina et al., 1993). As a result, it was suggested that HIV-1 RT may interact with primer-template in a n orientation opposite to what was observed with crystals of HIV-1 RT and a relatively short fragment of dsDNA (Jacobo-Molina et al., 1993). With the crystal structure of rat DNA polymerase p as a model, the majority of retroviral RT, including the RNase H domain, would be positioned over the single-stranded template rather than the primer-template complex. This hypothesis is, however, at odds with the hydroxyl radical footprinting results described above (Metzger et al., 1993) and recent nuclease footprinting studies with HIV-1 RT and MLV RT (Wohrl et al., 1995a,b). Nuclease footprinting (DNase I and S1 nuclease) of HIV-1 RT during DDDP shows a protected complex that extends beyond the hydroxyl footprint, namely, a protected region of +7 to -22 nucleotides from the primer terminus (Wohrl et al., 1995b). The MLV RT enzyme protects a region of + 5 / + 6 to -27 nucleotides in a nuclease footprint (Wohrl et al., 1995a). Unlike the case for HIV-1 RT, deletion of the RNase H domain of MLV RT still results in primer-template binding and DNA polymerization (Telesnitsky and Goff, 1993b). Deletion of the RNase H domain of MLV RT reduced the protected template region in the nuclease footprint by 12 nucleotides but downstream of the primer terminus (Wohrl et al., 1995a). This provides strong evidence that retroviral RT enzymes likely align over a primer-template with the polymerase active site over the primer terminus and the RNase H domain over the nucleic acid duplex. The DNA template-DNA primer duplex, found 5' of the template overhang, is in a n A-conformation (A-DNA). In the connection subdomain of RT and approximately 10 nucleotides into the dsDNA, a n A- to B-DNA switch in the dsDNA is accompanied by a 40" to 45" bend in the helical axis (Jacobo-Molina et al., 1993; Metzger et al., 1993). Free dsDNA is predominantly found as B-DNA, which differs from A-DNA in that it resembles a stretched A-conformation with a wider major groove, 11 bases per turn, and a pitch of 34 A (Selsing et al., 1979). Sequence-dependent transition from an A- to a B-conformation in dsDNA is accompanied by a 26" bend in the helical axis (Selsing et al., 1979). However, HIV-1 RT showed no sequence specificity in binding of

HIV-1 REVERSE TRANSCRIPTASE

113

DNA (Bakhanashvili and Hizi, 1994);thus, the bend in the helical axis of dsDNA was not sequence-dependent. Torsion-induced bends in DNA can also be the result of proteinnucleic acid interactions (reviewed by Crother, 1992). For example, crystallographic analysis of the catabolite gene activator protein (CAP) of E . coli complexed to dsDNA revealed a bend of about 90" in the DNA (Schultz et al., 1991).Therefore, specific interactions of HIV-1 RT with dsDNA may induce torsional and conformational changes on the dsDNA (Jacobo-Molina et al., 1993). In the polymerase region of RT, the a H and a1 helices of the thumb subdomain make contact with the sugar-phosphate backbone of the primer and template, respectively (Fig. 2A) (Jacobo-Molina et al., 1993).In the RNase H subdomain, the aA, aB, and p1-p2 hairpin loop make contact with the sugar-phosphate backbone of the primer strand (Davies et al., 1991. Single residues (253 to 271) of HIV-1 RT, found in the H a helix of the polymerase domain, were mutated to alanine residues (alanine scanning mutagenesis) (Beard et al., 1994). Several of the mutant p66/p66 RTs (i.e., Q258A, G262A, and W266A) showed reduced binding to primer-template when compared to wild-type p66/p66 RT (Beard et al., 1994). It should be noted that the crystal structure for HIV-1 RT complexed with dsDNA can only be compared with the HIV-1 RT-nucleic acid complex involved in DDDP, the last step in HIV-1 reverse transcription. During initiation of reverse transcription, RDDP is primed by tRNALys3 and utilizes an RNA template. An RNA duplex, unlike the DNA duplex, can only assume a linear A-conformation, owing to steric hindrance from oxygens on the 2' position of the ribose sugars (reviewed by Chastain and Tinoco, 1991). After initial HIV-1 DNA synthesis by HIV-1 RT, the DNA-RNA hybrid becomes covalently attached and positioned adjacent to the RNA-tRNALyss duplex. The DNA-RNA hybrid assumes an A-like conformation with the RNA strand remaining in an A-conformation, but the DNA strand assumes neither an A nor a B form, but rather one that ensures an optimum phosphate group separation between bases for nucleophilic attack by the RNase H active site of HIV-1 RT on the RNA strand (Federoff et al., 1993; Yu et al., 1993). The structure of HIV-1 RT complexed with these various forms of nucleic acid duplexes has not been determined. The specific binding of HIV-1 RT to a primer-template complex positions the 3'-hydroxyl of the primer over the polymerase active site in the palm subdomain of RT (Jacobo-Molina et al., 1993). The RT sequences involved in primer binding have been roughly mapped using photoaffinity labeling (Andreola et al., 1993; Cheng et al., 1991;

114

ERIC J. ARTS AND MARK A. WAINBERG

Kumar et al., 1993; Sobol et al., 1991). After UV-irradiation, proteolytic digestion, and amino acid sequencing, a fragment spanning residues 195 and 300 was described as the sequence binding to the photoreactive primer (Table I) (Sobol et al., 1991). Residues 195 to 300 are localized to the thumb and palm subdomains of HIV-1 RT (Kohlstaedt et al., 1992). The p12-pl3 hairpin (residues 227-2351, termed the “primer grip,” appears to be in close association with the sugarphosphate backbone of the primer (Jacobo-Molina et al., 1993). Alanine scanning mutagenesis of residues 224 to 229 revealed that W229 in the proposed primer grip region, and not carboxyl-terminal residues, was essential for template binding but not tRNALys3 primer binding (Jacques et al., 1994a). A W229A mutant RT was nonprocessive in DDDP assays and failed to provide a DNase I footprint, but it was still capable of binding primer tRNALyd in an RT-tRNALysa band shift assay (Jacques et al., 1994a). Interestingly, this residue forms part of the proposed binding pocket for nevirapine (Kohlstaedt et al., 1992; Smerdon et al., 1994) but is never mutated in nevirapineresistant viruses, suggesting its essential role in reverse transcription.

C . Polymerase Active Site and Deoxynucleoside 5’-Triphosphate Binding Site of HIV-1 Reverse Transcriptase As mentioned above, the catalytic site for polymerization was found in the palm subdomain (Larder et al., 198713). The YXDD sequence, the proposed polymerase active site, is semiconserved in nearly all viral and cellular RNA-dependent and DNA-dependent polymerases (Argos, 1988; Xiong and Eickbush, 1990). This sequence is found as 183YMDD186 in HIV-1 and HIV-2 RT (Hizi et al., 1991; Lowe et al., 19911, YVDD in MLV (Schinnich et al., 1981) and feline leukemia virus RT (Donahue et al., 1988), YADD in the RT of Myxococcus xanthus (Inouye et al., 19891, and YGDTDS in most DNA-dependent DNA polymerases (Wong et al., 1988). In HIV-1, substitution mutations of Y183, D185, and D186 or linker insertions in this region resulted in uninfectious virus and the complete loss of polymerization activity of RT (Table I) (Boyer et al., 1992a; Larder et al., 198713, 1989b; Prasad and Goff, 1989b; Wakefield et al., 1992).Mutations M184 to S, G, P, or L also resulted in both polymerase-deficient RT and noninfectious virus, whereas M184V or M184A mutations resulted in wild-type RT activity and replication-competent virus that was resistant to 2‘ ,3’dideoxyinosine (ddI), 2‘,3’-dideoxycytidine (ddC), and (- )-2‘,3’-dideoxy-3’4hiacytidine (3TC) (Gao et al., 1993; Gu et al., 1992; Schinazi et al., 1993; Wakefield et al., 1992). None of the latter mutations affected the RNase H activity of HIV-1 RT (Boyer et al., 1992a).

HIV-1 REVERSE TRANSCRIPTASE

115

It is thought that D185 and D186 with DllO bind to a divalent cation (Mg2+ or Mn2+), which is necessary for a nucleophilic attack by the 3’-hydroxyl of the primer on the a-phosphate of the incoming dNTP (Table I) (Kohlstaedt et al., 1992). The D185 and D186 residues of the YMDD sequence are found in a short loop between pl0 and p l l and in close proximity to a DlOO residue found in the p6 sheet (Fig. 2A) (Kohlstaedt et al., 1992; Jacobo-Molina et al., 1993). The resultant divalent cation binding pocket of HIV-1 RT is analogous to the polymerase active site of E . coli DNA polymerase I (Polesky et al., 1990, 1992). Residues D113, A114, Y115, Q155, K154, Y181, Y183, and M184 of HIV-1 RT, not directly implicated in Mgz+ binding, may be important for maintaining the structural integrity of the active site (Boyer et al., 1992a,b; Kohlstaedt et al., 1992; Larder et al., 1987b, 1989b; Wakefield et al., 1992). In p66, all of the latter residues are folded out from the floor of the palm for contact with the primer, whereas in p51, these residues are buried in the collapsed palm and connection subdomains (Jacobo-Molina et al., 1993; Kohlstaedt et al., 1992). Interestingly, selective substitution of part of the HIV-1 RT polymerization active site (amino acids 176-190) with the same region of HIV-2 or vice versa resulted in mutated HIV-1 and HIV-2 RT enzymes that were active in polymerization (Shih et al., 1991). Although these active sites are similar in sequence and function, there are slight variations in structure between the active sites of HIV-1 and HIV-2 RT, exemplified by TIBO (tetrahydroimidazo[4,5,l-jklbenzodiazepin-2[1K]-one) sensitivity of HIV-1 RT and TIBO resistance of HIV-2 RT. The selective substitution of HIV-1 RT sequences into HIV-2 RT rendered mutated HIV-2 RT sensitive to TIBO (Shih et al., 1991). Chimeric HIV-1IHIV-2 RT, containing changes around the polymerization active site, provided details on the binding affinity of nonnucleoside inhibitors such as nevirapine and TIBO and gave rise to models for new drug design based on structural characteristics of HIV RT (Bacolla et al., 1993; Hizi et al., 1993; reviewed by Nanni et al., 1993; Shaharabany and Hizi, 1992). The dNTP binding site on HIV-1 RT has been roughly mapped by monoclonal antibodies, mutational analysis, and photoaffinity labeling (Boyer et al., 1992b; Cheng et al., 1993; Wu et al., 1993; Painter et al., 1993). Inhibition of polymerization was achieved with a monoclonal antibody that bound to HIV-1 RT;this could be reversed by addition of increasing concentrations of dNTP but not primer-template. This antibody (1E8) recognized residues 65 to 73 in HIV-1 RT (Table I and Fig. 2) (Wu et al., 19931, the site of several mutations that confer resistance to nucleoside analogs, namely, D67N and K70R for AZT

116

ERIC J. ARTS AND MARK A. WAINBERG

resistance (Larder and Kemp, 1989), T69D for ddC resistance (Fitzgibbon et al., 1992), K65R for ddC resistance (Gu et al., 1994a,b; D. Zhang et al., 1994), L74V for ddI resistance (St. Clair et al., 1991), and V75T for 2’,3’-didehydro-2’,3’-dideoxythymidine (d4T) resistance (Lacey and Larder, 1994). The L74V mutation resulted in a moderate increase in the Ki value (five-fold) for 2’,3’-dideoxyadenosine 5’-triphosphate (ddATP) with no change in the K , for the native dNTP (Martin et al., 199313). Other substitution mutations (I63S, K64R, K65R, T691, R72K, and D76E), resulted in selective inhibition of polymerization without affecting RNase H activity (Boyer et al., 1992b). Finally, UV-induced cross-linking of [rnethyl-3H]dTTP on HIV-1 RT, followed by proteinase digestion and amino acid sequencing of the radiolabeled peptides, identified residue 73 as the binding site for dTTP (Table I) (Cheng et al., 1993). Residues 65-73 are found flanking or in the loop connecting the (33 and p4 sheets in the fingers subdomain of RT (Fig. 2) (JacoboMolina et al., 1993; Kohlstaedt et al., 1992). Hence, this proposed dNTP binding site is found opposing the primer-template binding site on the thumb subdomain (Section II1,A) and in a position to drop a bound dNTP into the polymerization active site in the palm (Jacobo-Molina et al., 1993; Kohlstaedt et al., 1992). Another proposed dNTP binding site was located a t K263 in the a H helix of the thumb subdomain (Basu et al., 1989). The effects of residues in the a H helix on dNTP binding and deoxynucleoside 5’-mOnOphosphate (dNMP) incorporation are still subject to debate and further research. Two groups failed to show differences in dNTP binding for the wild-type and K263 mutated RTs but demonstrated a slight increase in the polymerization rate constant (kcat) with K263 mutated enzyme (Martin et al., 1993a; Beard et al., 1994). Another HIV-1 RT, mutated in the a H helix a t residue Y271, displayed the same phenotype as the K263 mutated RT (Beard et al., 1994). It is conceivable that these residues are not involved in initial dNTP binding but, rather, are required for a second dNTP binding event during enzyme isomerization, pyrophosphorolysis, and dNMP incorporation (see below). As mentioned previously, the homodimeric (p66/p66 and p51/p51) and heterodimeric forms of HIV-1 RT are enzymatically active for polymerization from a n RNA or DNA template. Several studies have tried to establish if only one or if both subunits are needed for polymerization activity in the heterodimer (Basu et al., 1993; Bavand et al., 1993; LeGrice et al., 1991a). Although both subunits are capable of polymerization, only p66 in the p66/p51 heterodimer is responsible for the full polymerization activity of RT. The heterodimer-associated p51 likely contributes to polymerization activity by maintaining and/or establishing the active conformation of p66 (LeGrice et al., 1991a).In

HIV-1 REVERSE TRANSCRIPTASE

117

this regard, mutations in the coding region of RT may not directly affect the catalytic center of p66 but rather disrupt the p51 structure that interacts with p66 to ensure proper polymerase function. The p51 subunit, mutated in the 183YMDD186 sequence and reconstituted with a wild-type p66 subunit, had no effect on either RDDP or RNase H activity of the RT. However, the same mutation in p66 dimerized with a wild-type p51 and resulted in complete abolition of polymerization (LeGrice et al., 1991a). The contribution of the p51 subunit in polymerase activity was also examined using subunit chimeras of HIV-1 and HIV-2 RT (Howard et al., 1991). The wild-type HIV-1 RT (p66/p51) and a chimera of p66 (HIV-1) and p51 (HIV-2) were sensitive to TIBO. The chimeric heterodimer of p66 (HIV-2) and p51 (HIV-1) was resistant to TIBO, indicating that the TIBO-sensitive subunit, p51, was not involved in polymerization (Howard et al., 1991). The polymerase active site of p66 is obviously inactive in p51 of the heterodimer. However, residues that play a structural role in p66 of the heterodimer may actually be positioned in the p51 subunit to play a functional role in polymerization or RNase H digestion enacted for the most part by p66. The E138K mutation in HIV-1 RT encodes resistance to TSAO [2',5'-bis-O-(tert-butyldimethylsilyl)-3'-spir0-5~ '44''amino-l",2"-oxathiole 2",2"-dioxide)l but only when present in the p51 subunit and not in p66 (Jonckheere et al., 1994). In p66, E l 3 8 is found in the p7 sheet which is distal from the polymerase active site (Fig. 2). In p51, E l 3 8 is found adjacent to the polymerase active site of p66 (Kohlstaedt et al., 1992; Smeardon et al., 1994) and is likely to be expendable for polymerization.

Iv. HUMANIMMUNODEFICIENCY VIRUS TYPE 1 REVERSETRANSCRIPTION A . Overview of Reverse Transcription Scheme of Retroviruses The following description of reverse transcription is illustrated in schematic form in Fig. 3. The process begins with a transcription complex, consisting of tRNA primer, RT, and possibly nucleocapsid protein and cellular factors, that interact with viral genomic RNA (Barat et al., 1989; reviewed by Leis et al., 1993; Meric and Goff, 1989; Panet et al., 1975). The latter complex can initiate RDDP from the primer binding sequence (PBS) in an immature virus during assembly, in the virion after budding, or in the host cell on core entry (Biswal et al., 1971; Levinson et al., 1970; Lori et al., 1992; Trono, 1992; Zhang et al.,

118

ERIC J. ARTS AND MARK A. WAINBERG

I.lnitiatlon of RDDP env (A)

I 1

atrong*slop

U3 I R

HIV RNA Genome

I

reverse transcriptasa

(Interstrand awitching)

R 1 U5 I PBS

1

gag

Po'

env

1

2, First template switch

3. RNase H digestion and (-) DNA synthesis

4. ppt priming of DDDP

I-)

I

LEGEND

.I.*.I II -

DNASynthesized due to interstrand switching single-stranded RNA

D

, . 1 , , , , .

5. Second template

ANase H digested RNA

RDDP RNA-dependent DNA polymerizaiion

DDDP DNA-dependent DNA polymerization

ppt

pdypurine tract

6. Campletton of ds proviral DNA (+) DNA (-) DNA

FIG.3. Reverse transcription scheme of retroviruses.

HIV-1 REVERSE TRANSCRIPTASE

119

1993). Minus (-1 strong-stop DNA, the first transcribed DNA fragment, is a complementary copy of the 5’ unique (U5) sequence and R region of the LTR of viral genomic RNA and is covalently linked to tRNA primer. In HIV, synthesis of (-1 strong-stop DNA is accompanied by exonucleolytic-like endonucleolytic digestion of the RNA template by the RNase H activity of RT (DeStefano et al., 1991; Furfine and Reardon, 1991; Schatz et al., 1990). Digestion of the RNA in the R region permits the annealing of the R region of (-1 strong-stop DNA to the complementary R region (Oyama et al., 1989) at the 3’ end of the same genomic RNA used as template for synthesis of (-) strong-stop DNA (intrastrand) or the 3’ end of an R region on a new genomic RNA template (interstrand) (Hu and Temin, 1990; Panganiban and Fiore, 1988).The (-1 strong-stop DNA, once annealed to a new R region, now primes the continuation of (-) DNA synthesis and RNase H digestion of the RNA template (Luo and Taylor, 1990; Peliska and Benkovic, 1992).This event is termed the first template switch or strand transfer event and is described in greater detail in Section IV,G. The tRNA primer, after initiation of RDDP, likely remains annealed to the PBS on the RNA template. RNase H* cleavage of the PBS may be required for its removal from the annealed tRNA sequences (BenArtzi et al., 1992a; Fbth et al., 1989; Swanstrom et al., 1982).The tRNA sequence can then be used as template for the synthesis of (+) DNA. An RNase H* activity of HIV-1 RT was shown to make two cuts in the PBS but only in the presence of the divalent cation Mn2+ (Ben-Artzi et al., 1992a). In a reverse transcription scheme employing an intrastrand first template switch, RNase H* cleavage prior to completion of (-) DNA would result in transcription of a deleted PBS at the 3’ end of (-) DNA. The full PBS could only be transcribed if full (-1 DNA synthesis preceded RNase H* cleavage. In contrast, (-1 HIV DNA can be extended through the U3 region to the end of R after an interstrand template switch. A stretch of purine residues, found 5’ of the U3 region in the genomic RNA, is resistant to RNase H digestion during (-) DNA synthesis (Sorge and Hughes, 1982).Lack of RNase H digestion in this region may be due t o both RNA secondary structure and sequence (Champoux et al., 1984). The polypurine tract primes DDDP and (+) strong-stop DNA synthesis (Omer et al., 1984; Resnick et al., 1984; Smith et al., 1984a,b), but specific RNase H cleavage around the polypurine tract may be necessary for this priming event to occur (Huber and Richardson, 1990; Luo et al., 1990; Pullen et al., 1993). The initial product of DDDP is the (+) strong-stop DNA comprising the U3/R/U5 of the LTR (Resnick et al., 1984). The template for the PBS in (+) DNA is the 3’ terminal sequence of the tRNA primer, originally annealed to the PBS in the genomic RNA (Swanstrom et al., 1982).

120

ERIC J. ARTS A N D MARK A. WAINBERG

The use of tRNALys3 as a template requires dissociation of the PBS sequence and t R N A b 3 which remain associated after initiation of (-1 strong-stop DNA. Dissociation of the PBS from tRNALys3 may be promoted by RNase H* digestion of PBS RNA by HIV-1 RT (Ben-Artzi et al., 199213) or, in the absence of RNase H” cleavage, by a strand displacement activity of HIV-1 RT. The MLV RT is capable of displacing DNA strands of up t o 1000 nucleotides from a DNA template during polymerization (Whiting and Champoux, 1994). This activity has not been characterized for HIV-1 RT. However, regions of nucleic acid secondary structure result in pausing of HIV-1 RT during polymerization (Abbotts et al., 1993; Bebenek et al., 1989,1993; Klarmann et al., 1993; J i et al., 1994). Continued polymerization into the PBS a t the 3’ end of (+) strongstop DNA requires a switch in RT enzymatic activity from DDDP to RDDP. A precise stop position must exist in the tRNA template for the future extension of (+) DNA after the second template switch (Roth et al., 1989). This stop site is likely the first modified base on the tRNA template. The tRNA template is likely removed by RNase H digestion (Murphy and Goff, 1989; Smith and Fbth, 1992)but is initiated between the terminal adenosine and cytidine bases of tRNALys3 (Whitcomb et al., 1990). This remaining terminal ribonucleotide could be the single base found in preintegrated, circular proviral DNA (Whitcomb et al., 1990) but is absent in integrated proviral DNA (Ratner et al., 1985). If the first template switch were a n intrastrand event, the PBS of the (+ strong-stop DNA should anneal to the 3’ end of PBS or the full PBS in (-1 DNA (see above). If the first template switch were an interstrand event, the PBS from (+) strong-stop DNA might also anneal to the complementary PBS of (-) DNA. In addition, strand displacement of the U5 and R regions of (+) strong-stop DNA from the 5’ end to the 3’ end of (-1 DNA could further stabilize the annealing of the two complementary PBS in (-1 and (+) DNA (reviewed by Telesnitsky and Goff, 1993a). The binding of (+) strong-stop DNA to the PBS of (-) DNA can prime DDDP for the completion of (+) DNA. This process is referred to as the second template switch or transfer and can only be an intrastrand event (Panganiban and Fiore, 1988). The 3’ terminus of (- ) DNA, ending in either the R region (interstrand first template switch) or the PBS (intrastrand first template switch), can prime DDDP for the completion of (-) DNA synthesis but only after the second template switch (Gilboa et al., 1979).

B . Origin of HIV-1 Reverse Transcription There has been a resurgence in interest in the significance of viral DNA in retroviral particles (Arts et al., 1994a; Lori et al., 1992; Trono,

HIV-1 REVERSE TRANSCRIPTASE

121

1992; H. Zhang et al., 1993, 1994). Qualitative detection of viral DNA in such viruses was first demonstrated by Levinson et al. (1970) and Biswal et al. (1971). The advent of new sensitive and quantitative nucleic acid detection procedures, including the polymerase chain reaction (PCR), demonstrated that retrovirions contained viral DNA of heterogeneous size (Lori et al., 1992; Trono, 1992). The heterogeneity in viral DNA appeared to result from random stops during both (-) and (+) strand synthesis during HIV-1 reverse transcription (Lori et al., 1992; Trono, 1992). More recently, studies have focused on the comparative presence of DNA versus RNA in HIV particles and the significance of such DNA in infectiousness (Arts et al., 1994a; H. Zhang et al., 1993, 1994). These studies employed near-identical techniques to isolate and lyse viruses and to purify nucleic acids. Different viral isolates or clones (i.e., HIV-lHXBZ and HIV-1A,") were isolated from chronically infected H9 cells as well as from cos-7 and Jurkat cells that had been transfected with proviral HXB2-containing vectors. Quantitative PCR amplifications of DNA or in uitro reverse-transcribed RNA, using different primer pairs that detected different segments of the HIV genome, were performed on nucleic acids isolated from viral lysates. Both groups obtained a ratio of approximately 1:lOOO of HIV DNA to RNA in the latter HIV isolates. In addition, there was approximately a 10- to 100-fold decrease in HIV gag DNA as compared with (-1 strong-stop DNA. Viral DNA has also been identified and quantitated in various patient isolates, all of which had a viral DNA t o viral RNA ratio of approximately 1:lOOO (H. Zhang et al., 1994). Previous reports (Lori et al., 1992; Trono, 1992) had suggested that the viral DNA found in retroviral particles may contribute to the latent viral DNA transcripts found in quiescent peripheral blood lymphocytes during HIV-1 infection (Zack et al., 1990). Treatment of phytohemagglutinin (PHA)-stimulatedperipheral blood lymphocytes with AZT and 2',3'-didehydro-2',3'-dideoxythymidine(d4T), prior to HIV-1 infection, did not inhibit synthesis of (-1 strong-stop DNA (Zack et al., 1990), suggesting that this DNA was carried into the host cell by the virion core (Trono, 1992). However, lack of AZT-mediated chain termination during synthesis of (-) strong-stop DNA may be due to preferential chain termination by nucleoside analogs observed after the first template switch in activated CD4+ lymphocytes and in uitro reverse transcription/template switching reactions (Arts and Wainberg, 1994). Near complete inhibition of synthesis of (-1 strong-stop DNA by AZT and other nucleoside analogs was observed in HIV-exposed quiescent brain macrophages and quiescent peripheral blood mononuclear cells (Geleziunas et al., 1993; Zack et al., 1992).These results suggest that viral DNA found in retroviral particles may not be the main requirement

122

ERIC J. ARTS AND MARK A. WAINBERG

for reverse transcription during host cell infection. However, the small virion-associated DNA may contribute a small fraction of the partial HIV DNA transcripts found in HIV-exposed,quiescent peripheral blood lymphocytes (Zack et al., 1990; 1992; H. Zhang et al., 1993, 1994). Earlier studies have suggested that the protein precursor of reverse transcriptase, pl6Ogag-~ol,might be responsible for the synthesis of viral DNA in immature cores prior to or during virus budding (Lori et al., 1992; Trono, 1992). Thus, they likened the retrovirus genome to that of hepadnaviruses, which carry an asymmetric DNA genome that is reverse-transcribed in host cells during viral assembly but prior to virus release (Ganem and Varmus, 1987). Presence of viral DNA and RNA was examined in viruses isolated from Cos cells transfected with a protease-defective HIV-1 expression vector (Arts et al., 1994a). The protease-defective virus contains only the 160gq-pol precursor and lacks mature p66/p51 reverse transcriptase (Gottlinger et al., 1989). The viral DNA:RNA ratio was only 1:500,000 in protease-defective virus (Arts et al., 1994a); in other words, there was approximately 500fold less DNA than that found in wild-type HIV-1 with processed precursors (Fig. 1).The significant decrease in HIV-1 DNA found in protease-defective virus suggests that mature RT is responsible for the reverse-transcribed DNA found in wild-type viruses. In support of this hypothesis, it was shown that incubation of wild-type HIV-1 in increasing concentrations of dNTPs results in an increased presence and length of HIV-1 DNA in virions (H. Zhang et al., 1993). Deoxynucleoside5’4riphosphates may represent the only ingredient needed for full reverse transcription and synthesis of proviral DNA not found in mature virions (Fig. 1). For example, tRNALys3, the cognate primer of HIV-1 reverse transcription, is preferentially packaged into HIV-1 and is tightly associated with the genome (Jiang et al., 1993).The other factor limiting reverse transcription in the virion may be space and hydrostatic interference. Dissolution of the core on viral entry may be necessary for complete synthesis of proviral DNA. However, studies suggest that the core remains somewhat intact even after completion of reverse transcription due to the interaction of core elements, namely, nucleocapsid and matrix proteins, with integrase and proviral DNA (Bukrinsky et al., 1993a,b). What is the significance of viral DNA in HIV-1 infection? A dNTPmediated increase in viral DNA was also correlated with a slight increase in viral infectivity (Zhang et al., 1993), suggesting that the latter was not affected by the site of initiation of reverse transcription (i.e., in free virus or in the host cell cytoplasm). Interestingly, the ratio of partial viral DNA transcripts to genomic RNA in the genome (1:lOOO) correlates with the ratio of infectious retrovirions in a pool of

HIV-1 REVERSE TRANSCRIPTASE

123

retrovirus particles. Therefore, it may be that only retrovirus particles, containing extendable viral DNA, are infectious on entry into a host cell. To test this hypothesis, cos-7 cells were treated with AZT (2 pit0 prior to transfection with proviral DNA, generating HIV-1 particles with viral DNA that had been chain-terminated immediately after the first template switch. Viral DNA that contains a 3’4erminal AZT 5’-monophosphate cannot support further elongation due to lack of a 3 ’-hydroxyl necessary for incorporation of deoxynucleoside 5’-mOnOphosphates. Equal quantities of viruses from untreated and AZTtreated cos-7 cell transfections were then used to infect H9 cells and other CD4+ cell lines (Arts et al., 1994a). Similar p24 antigen production levels were observed in culture fluids of H9 cells infected with the wild-type and AZT-treated viruses. If viral DNA in HIV particles were required for infection, then viruses derived from the AZT-treated, transfected cells should have yielded little or no de novo p24 production. A similar, independent study using AZT-treated and wild-type HIV yielded identical results (J. A. Zack, personal communication, 1994). Thus, viral DNA is likely not essential for HIV infection, although such DNA, when present, may presumably serve as a primer for proviral DNA synthesis during infection of host cells.

C . Host tRNALys3 Primer in HIV-1 Reverse Transcription Shortly after the discovery of an RNA-dependent DNA polymerase in RNA tumor viruses (Baltimore, 1970; Temin and Mizutani, 1970),it became evident that polymerization from the retroviral RNA genome was initiated by a host-derived RNA molecule (Leis and Hurwitz, 1972; Verma et al., 1971).This primer molecule was shown to be transfer RNA, namely, tRNA*rp for RSV (Dahlberg et al., 1974; Harada et al., 1975). The 3‘ end of the specific tRNA isoacceptor species was shown to anneal to the 5’ end of viral genomic RNA at a site known as the primer binding sequence (PBS) (Taylor and Illmensee, 1975). The binding of a retrovirus-specific tRNA to viral genomic RNA and the specific interactions of viral RT with tRNA (Barat et al., 1989; Panet et al., 1975) were thought to initiate or “prime” RNA-dependent DNA polymerization (see Fig. 5). No single tRNA isoacceptor species is employed as primer for all retroviruses. Indeed, there is no apparent relationship between selection of a tRNA species for priming and retroviral sequence homology. For example, HIV-1 and the mouse mammary tumor virus both utilize tRNAW3 to prime RDDP, but these viruses share little homology in the RT coding region and are members of the subfamilies Lentivirinae and Oncovirinue, respectively (Leis et al., 1993). However, retroviruses

124

ERIC J. ARTS AND MARK A. WAINBERG

of the same subfamily or type may employ the same tRNA as primer; for instance, HIV-1, HIV-2, simian immunodeficiency virus type 1 (SIV-11, and SIV-2 of the subfamily Lentiuirinue all employ tRNALys3 as primer. All members of the avian sarcoma and leukosis virus group use tRNATrp as primer (Harada et al., 1975; reviewed by Leis et al., 1993, Waters and Mullin 1977). The tRNA isoacceptors are incorporated into virus particles during assembly (Fig. 1)(Erikson and Erikson, 1971; Faras et al., 1973; Peters and Hu, 1980). In HIV-1, tRNALys3 is found tightly associated with the viral RNA genome and is preferentially packaged into virions (Jiang et al., 1992, 1993). Reduction or the absence of viral genomic RNA did not affect tRNA packaging, suggesting that the PBS and other RNA sequences do not control tRNA incorporation into virus particles (Levin and Seidman, 1979; Mak et al., 1994; Peters and Hu, 1980).Instead, the pl60gag-pol precursor protein of HIV-1 may be responsible for the preferential tRNALys incorporation (Mak et al., 1994). A proteinasedefective HIV-1 virus had wild-type amounts of tRNALys in its core but contained only the pl60g"g-~land p 5 5 g ~precursors of the core proteins (Mak et al., 1994). It was shown that overexpressing wild-type or mutant suppressor tRNALys3 from a tRNALys3 expression vector, cotransfected into a host cell with an HIV-1 expression vector, resulted in viruses containing 50fold increases of wild-type or mutant tRNALys3 over wild-type tRNALysI3 (Huang et al., 1994). However, there was no change in viral infectivity when comparing viruses containing excess wild-type tRNAb3 or mutant suppressor tRNALys3 with wild-type HIV-1 (Huang et al., 1994). These results suggest that there is a select tRNALys3 isoacceptor species which acts as primer for HIV-1 reverse transcription. Interestingly, analysis of [3Hl1 or [Wllysine-charged tRNALys isoacceptor species in HIV-1 and host cells by RPC-5 HPLC revealed multiple tRNALys3 species in HIV-1 but not in the infected or uninfected CD4+ lymphocyte cell lines (X. Li et al., 1994).In cells, only two peaks were present, namely, a peak containing tRNALysl2 and one containing tRNALys3. tRNALys analyzed from HIV-1 produced by monocytic U937 cells, lymphocytic H9 cells, and peripheral blood lymphocytes had six peaks. The first and third tRNALys isoacceptor peaks, containing mostly tRNALysI3 and tRNALyd, respectively, were the most abundant and coresponded to the two peaks found in cells. The first peak, which also contained trace amounts of tRNALys3, and second and fifth peaks, which contained only tRNALys3, were the most efficient at initiating RDDP in an in uitro reverse transcription assay employing HIV RNA templates (Arts et al., 1994b; X. Li et al., 1994). Although tRNALys isoacceptor species were preferentially incorpo-

HIV-1 REVERSE TRANSCRIF’TASE

125

rated into HIV-1 particles, the status of the tRNALys3 has not been determined (i.e., free in the core or tightly associated with viral genomic RNA). There are several hypotheses as to how the 3’ end of free tRNALys3 anneals with the PBS and interacts with HIV-1 RT to initiate RDDP (Fig. 5). First, pl60gag-pol and/or p5&’*, the possible carrier molecules of tRNALys3 and viral genomic RNA (Berkowitz et al., 1993; Mak et al., 19941, may also place tRNAL@ onto the PBS. This placement may be facilitated by the two RNA-binding zinc finger motifs found in the nucleocapsid protein domain of the precursors (Barat et al., 1993; De Rocquigny et al., 1992; South et al., 1990). Second, the proteolytic products NCpl5 and NCp7 may catalyze the annealing of tRNALys3 to the PBS. The nucleocapsid proteins possess strong renaturation activities for tRNA (Dib-Hajj et al., 1993; Khan and Giedroc, 1992) as well as binding specificity to the T W loop of tRNALys3 (Barat et al., 1993) and to the uncoding region found 3’ of the PBS in viral genomic RNA (Darlix et al., 1990; Sakaguchi et al., 1993). Interactions of tRNALys3 with viral genomic RNA may extend to sequences found 3’ and/or 5’ from the PBS (Isel et al., 1993; Kohlstaedt and Steitz, 1992). Extended interactions between tRNALys3 and viral genomic RNA may suffice for annealing of these two molecules in the virion. Interactions between the U5 RNA of avian leukosis virus (ALV) genomic RNA and the T W loop of tRNATrp were shown to be required for efficient initiation of reverse transcription (Aiyar et al., 1992). The U5/PBS/leader sequence of ALV forms a stem-loop in which the PBS is looped out on annealing with the 3‘ end of tRNATrp. Another 8-nucleotide segment, found downstream of the 18-nucleotide anti-PBS sequence on the 3‘ end of tRNATrP, may anneal to the U5 sequence opposite the PBS in the latter stem-loop. Disruption of this stem-loop results in reduced reverse transcription and viral infectivity (Aiyar et al., 1992,1994; Cobrinik et al., 1987,1988,1991). A similar stem-loop containing the PBS exists in both HIV-1 and HIV-2 genomic RNA (Fig. 5 ) (Baudin et al., 1993; Berkhout and Schoneveld, 1993). Interactions between synthetic (tRNALys3and HIV-1 genomic RNA (nucleotides + 1 to +311) reduced the chemical reactivity of bases opposite the PBS in the PBS stem-loop (Isel et al., 1993). This protected region was extended using a dethiolated synthetic tRNALys3, suggesting the importance of the 2’-thio-5‘-carboxymethyluridine in the anticodon loop of natural human tRNALys3 for efficient interaction with viral genomic RNA (Fig. 4) (Isel et al., 1993). Therefore, in addition to annealing of tRNAL@-PBS, interactions and base pairing between the tRNALys3 and the viral genomic RNA, outside the PBS, may promote association of free tRNALyd with the viral genomic RNA in the virus particle. Other modifications in the tRNALys3 isoacceptor species may further

-W C stem

acceptor stem

LEGEND squared base lnteracls with the HIV RNA template

0 U

circled base protected by HIV RT in RNA fwtprinting bolded base bound to the primer bindlng sequence

7mG methylaled base, i.e. 7-methyl-guanosine oligo~bonudeotideused in compnitin inhibition studies

D

dihydmayridine

y

pseudouridine

S

2'-lh10-5'-carb0xymethyluridine

T,

methyl tibothymidine

FIG.4. tRNALy* and sites for interaction with HIV-1 reverse transcriptase and with genomic RNA template. The bases on tRNALyd known to interact with HIV-1 RT are circled in the secondary and tertiary structures of tRNALyd (Sarih-Cottin et al., 1992; Wohrl et al., 1993). A base thought to interact with the RNA template is squared (Isel et al., 1993). The lines outlining the D and anticodon loops represent oligoribonucleotides used in competitive inhibition studies (Sarih-Cottin et al., 1992).

HIV-1 REVERSE TRANSCRIFTASE

127

strengthen interactions with genomic RNA. This may be an underlying reason for the increased priming efficiency of certain unique tRNALys3 species found in HIV-1 (X. Li et al., 1994). In uztro, human placental tRNALysl3 and tRNATrp could bind to complementary primer binding sequences, substituted for the wild-type PBS in an HIV RNA template (X. Li et al., 1994). However, only tRNALys3 could efficiently prime RDDP with HIV-1 RT from its respective PBS (X. Li et al.,1994). Interestingly, a deoxyoligonucleotide complementary to the wild-type PBS was more proficient at initiating RDDP than even tRNALys3 (Arts et al., 199413). This was due to increased annealing of the deoxyoligonucleotide to PBS. However, (-) DNA polymerization primed by a deoxyoligonucleotideprimer was less processive than that observed with tRNALy.3 as a primer (E. J. Arts and M. A. Wainberg, 1994, unpublished data. Two HIV-1 RNA templates, one containing R, U5, PBS, and uncoding (between gag start codon and PBS) sequences (HIV RNA PBS) and the other containing R and U3 sequences (HIV RNA R/U3), were utilized in HIV-1 reverse transcription/template switching reactions primed by human placental tRNALys3 (Arts et al., 1994b). Increasing the extent of 5' cis-acting deletions of the R region in HIV RNA PBS template resulted in decrased synthesis of (-1 strong-stop DNA primed by tRNALys3 from the PBS. This decrease in synthesis of (-) strong-stop DNA was not observed with a trans-acting deletion in HIV RNA R/U3 template (Arts et al., 1994b). Therefore, sequences upstream of the PBS stem-loop, such as the tat response element (TAR) stem-loop, may be necessary for efficient priming of RDDP by tRNALys3 and HIV-1 RT (Fig. 4). The HIV-1 RT may directly associate with tRNALys3, promoting its annealing to the PBS and initiation of RDDP. In 1989, Barat et al. showed in a band shift assay that the association of HIV-1 and AMV RT with cellular tRNA was specifically inhibited by addition of the respective cognate primers, tRNALys3 and tRNATrp. The HIV-1 RT could bind specifically to synthetic tRNALw3, but increased binding affinity to bovine t R N A b 3 suggested a role for modified bases in this interaction (Barat et al., 1991; Sarih-Cottin et al., 1992; Wohrl et al., 1993). RNase A footprinting analysis and oligoribonucleotide competition experiments indicated that HIV-1 RT associated with the D (or variable) and anticodon loop of tRNALys3, and also promoted the unwinding of bases 72 to 61 (Fig. 4) (Sarih-Cottin et al., 1992). A more detailed nuclease footprinting study found that bases in the T W , D, and anticodon loops, inaccessible to RNase T1, RNase A, and RNase S1 digestion, were all found on one side of the L-shaped tertiary structure of tRNALW, suggesting that HIV-1 RT interacts with only one face of

128

ERIC J. ARTS AND MARK A. WAINBERG

tRNALys3 (Fig. 4) (Wohrl et al,., 1993). Destabilization of the T W loop of tRNALys3 through interactions with HIV-1 RT may promote unwinding of the acceptor stem and annealing of the 18 nucleotides at the 3’ end of tRNALys3 to the PBS (Wohrl et al., 1993). The sites on HIV-1 RT that are important for interactions with tRNALys3 have not been well characterized. It is known that all three dimers of RT (i.e., p66/p66, p51/p51, and p66/p51) can specifically interact with tRNALrs3 (Richter-Cook et al., 1992). However, free tRNALys3 has an inhibitory effect on the polymerase activity of p66/p51 but stimulatory effects on that of p66/p66 HIV-1 RT. This difference could be due to different tRNALys3 binding properties of the p66 and p51 subunits in both the heterodimers and homodimers of RT (Andreola et al., 1992; Bordier et al., 1990). In addition, interaction of heterodimeric RT with tRNALys3 resulted in conformational changes in HIV-1 RT (Robert et al., 1990).As discussed in Section III,A, the site of primer binding in HIV-1 RT has been roughly mapped. The RT residues important for primer binding are found near the polymerase active site and were characterized with a short oligoribonucleotide primer and not tRNALyd (Andreola et al., 1993; Kumar et al., 1993; Sobol et al., 1991). The tRNALys3 molecule, being 80 A in length (Susman and Kim, 19761, is almost as large as HIV-1 RT (100 A in length) (Kohlstaedt et al., 1992; Lederer et al., 1992). Because the nucleotides implicated in interactions with HIV-1 RT are located in all three loops of tRNALys3 (Barat et al., 1989; Sarih-Cottin et al., 1992; Wohrl et al., 19931, it is likely that several RT subdomains are required for these interactions. However, they may not all be necessary for initiation of reverse transcription from the 3’-hydroxyl of tRNALys3. Heterodimeric RT with a selective deletion of 13 amino acids from the carboxyl terminus of p51 was capable of DDDP from a DNA primer but did not bind tRNALyd in band shift assays (Jacques et al., 1994b). These 13 amino acids could be crucial for HIV-1 RT-tRNALys3 interactions. In conclusion, complex mechanisms, involving several viral proteinRNA template interactions with primer tRNALys3, are required for incorporation of tRNALys3 into virions, annealing of tRNALys3 to the PBS, and initiation of RDDP from tRNALys3 on the PBS. The complexities of these mechanisms are illustrated through experiments that studied the replication kinetics of mutated viruses containing altered PBS in their viral RNA genomes. Substitution mutations in the PBS, which changed its sequence complementarity to that of other tRNA isoacceptor species such as tRNAPhe or tRNALysl.2 (X. Li et al., 1994; Wakefield et al., 19941, or deletions at the 3’ end (Nagashunmugam et al., 1992; Rhim et al., 1991), resulted in viruses with slower initial rates of replication than the wild type. However, continued replication

HIV-1 REVERSE TRANSCRIPTASE

129

of the mutated viruses yielded progeny with wild-type replication kinetics (X. Li et al., 1994; Rhim et al., 1991; Wakefield et al., 19941, corresponding to reversion to a wild-type PBS form complementary to tRNALys3 (X.Li et al., 1994; Rhim et al., 1991;Wakefield et al., 1994). In reverse transcription, the primer used to initiate RDDP is also used as a template during (+) strand synthesis (see Section 111,C). Therefore, viruses that reverted from mutated to wild-type PBS must have employed tRNAL@ as a primer for RDDP. As explained above, this utilization of tRNALys3 might involve selective sequences outside of the PBS, preferential incorporation of the tRNALys3 isoacceptor into the virion, and/or specific interactions between HIV-1 RT and tRNALys3.

D . R N A - and DNA-Dependent D N A Polymerization The mechanisms and kinetics of RDDP and DDDP of HIV-1 RT have been derived from pre-steady-state or steady-state kinetic assays using short primer-templates (Cheng et al., 1987; Majumdar et al., 1988). These reactions do not always reflect the true nature of HIV-1 reverse transcription but are necessary in determining the kinetics and affinity of substrate binding to HIV-1 RT. In addition, the simplicity of these assays has been invaluable for rapid drug screening of anti-IM' drugs. However, the complexity of these polymerization reactions requires a fully endogenous RT reaction, in order to study a single enzymatic stage during a multistage reaction (Painter et al., 1991; Reardon, 1993). Active RDDP and DDDP are multistep reactions, requiring the binding of primer-template (p/t) and dNTPs to HIV-1 RT prior to deoxynucleotide 5'-monophosphate addition to the 3' end of the primer (Majumdar et ul., 1988): Rt

k, + p/t e RT-p/t + dNTPs k- 1

KC,

-+

RT

+ extended p/t

The association constants (Kd)for HIV-1 RT binding to DNA primerDNA template and to DNA primer-RNA template were similar but dependent on length of both primer and template (Reardon, 1993). The dissociation constant (k- was slightly increased for the DNA primerDNA template. Using analytical ultracentrifugation for the separation of different HIV RT-primer-template complexes, there appeared to be only one primer-template binding site on heterodimeric RT (Kruhoffer et al., 1993). However, there may be a two-step mechanism for HIV-1 RT binding to primer-template that involves an initial binding event followed by an Mg2+-independent conformational rearrangement of the RT-primer-template complex (Hsieh et al., 1993; Kruhoffer et al., 1993).

130

ERIC J. ARTS AND MARK A. WAINBERG

The binding of dNTP to an enzyme-primer-template complex may also be a two-step reaction (Hsieh et al., 1993; Kati et al., 1992; Painter et al., 1991; Reardon, 1993). First, dNTP must bind nonspecifically, which is followed by enzyme isomerization in the presence of cognate dNTP. This step likely catalyzes dNTP pyrophosphorolysis, which is required for the addition of dNMP to the 3’-hydroxyl of the primer and release of the pyrophosphate (PPJ (Kati et al., 1992; Reardon, 1993).A more complete reaction scheme for RDDP is as follows: RT

+

(1) primer/template,, (pit,,)

dNTP

(3) RT-p/t,, S RT*-p/t, (2)

k, k-

I

h2

k-2

k3

G=

(4)

RT-p/t,,-dNTP

k-3

dNTP

(4)

RT-p/t,,-dNTP

k4

C k-4

(5) RT*-p/t,-dNTP

(6)

kC.91 +

RT

+ p/t,,+, + PP,

In addition to being a substrate for polymerization, dNTP may also act as a competitive inhibitor of RDDP (Furman et al., 1991; West et al., 1992). However, competitive inhibition by dNTP has been demonstrated only with homopolymeric primer-templates and may not be significant in an endogenous reaction. The polymerization rate constant (kcat) was 20-fold greater for a RNA primer-DNA template than a DNA primer-DNA template (Reardon, 1993). This may be related to the dissociation rate constant of the two different primer/templates, but the rate-determining step of polymerization is likely initial primer-template binding to RT. These assays utilized DNA primer annealed to DNA or RNA templates to study the DDDP or RDDP activities of HIV-1 RT. However, studies on RNA primer-RNA template complexes are needed to investigate the steps required for initiation of RDDP primed by an RNA or tRNALys3 primer. The process of primer extension is likely a cycling event between complexes (3)and ( 6 )in the above RDDP reaction scheme. Therefore, it is difficult to ascertain the true kcat and V,, values for an extension greater than one nucleotide (Kati et al., 1992). On a heteropolymeric template, the HIV-1 RT enzyme does not maintain a fluent and invariable rate of nucleotide addition (Huber et al., 1989; Yu and Goodman, 1992). The rate of RDDP, with substrates in excess but below competitive inhibitory concentrations, was 5 to 15 nucleotides/sec using a homopolymeric primer-template (Huber et al., 1989) but only 0.3 to 1.5 nucleotides/sec with a heteropolymeric primer-template (Yu and Goodman, 1992). The difference in rates (V,,,) is attributable to pausing of HIV-1 RT during polymerization through a heteropolymeric template (Huber et al., 1989).

HIV-1 REVERSE TRANSCRIPTASE

131

The HIV-1 RT, unlike other retroviral RTs or cellular DNA polymerases, is not a highly processive enzyme (Bakhanashvili and Hizi, 1992, 1993; Yu and Goodman, 1992). Processivity measures the average number of nucleotides incorporated prior to dissociation of the enzyme and template. Increased pausing by HIV-1 RT during polymerization is directly related to the processivity of the enzyme because the addition of an enzyme trap such as heparin sulfate or poly(rA)/oligo(dT) showed that HIV-1 RT generally dissociates from the template at these pause sites (Klarmann et al., 1993). Processivity and pausing of HIV-1 RT were observed during both RDDP and DDDP (Bebenek et al., 1989; Huber et al., 1989).A decrease in RT concentration or an increase in potassium chloride concentration resulted in decreased RDDP and DDDP processivity and an increase in the number and intensity of pauses (Huber et al., 1989). Templatedirected pausing was sequence-specific as well as structure-specific. Homopolymeric sequences in a heteropolymeric template or a hairpin RNA secondary structure were efficient at disrupting polymerization and, often, at dissociating RT from the template (Abbotts et al., 1993; Bebenek et al., 1989, 1993; Klarmann et al., 1993; J i et al., 1994). Pausing was also more pronounced during polymerization from a DNA template (DDDP) than from an RNA template (RDDP) (Klarmann et al., 1993; Yu and Goodman, 1992). This may be related to an increased dissociation constant for HIV-1 RT-DNA primer-DNA template complexes than for HIV-1 RT-DNA primer-RNA template complexes (Reardon, 1993). Furthermore, there appears to be a sequence specificity in regard to pausing at homopopolymeric sequences in RNA and DNA heteropolymeric templates (Abbotts et al., 1993; Klarmann et al., 1993). In RNA templates, runs of greater than three cytidine and guanosine bases resulted in efficient pausing, whereas, in DNA templates, runs of four deoxythymidine and deoxyadenosine bases were required for efficient pausing (Klarmann et al., 1993).

E . Fidelity of Polymerization by HIV-1 Reverse Transcriptase The HIV-1 RT shows a high mutation rate ( l / l O 4 nucleotides/replication event) when compared to other viral RTs such as that of AMV ( - l / l O 5 nucleotides/replication event) (Preston et al., 1988; Roberts et al., 1988). In large part, this is due to decreased fidelity or high error frequency during HIV-1 reverse transcription (Preston et al., 1988; Roberts et al., 1988; Takeuchi et al., 1988; reviewed by Williams and Loeb, 1992). A 3’ + 5’-exonuclease activity found in cellular DNA polymerases, but not in HIV-1 RT, prevents most substitution, deletion, and insertion errors during polymerization by removing misaligned bases. However, the lack of such activity in HIV-1 RT cannot explain

132

ERIC J. ARTS AND MARK A. WAINBERG

the higher mutation frequency associated with this enzyme as compared with other viral or cellular DNA polymerases (reviewed by Williams and Loeb, 1992). A relationship exists between the sites of hypermutability and the sites of pausing during reverse transcription (Bakhanashvili and Hizi, 1993; Bebenek et al., 1993; Ricchetti and Buc, 1990). Substitution errors were generally found immediately preceding a pause site, mostly in homopolymeric stretches in the template (Bebenek et aZ., 1993). These substitution errors may be due to preferential dNTP binding by HIV-1 RT, independent of the complementary base on the template (Cai et al., 1993). The enzyme may be predisposed to this binding when stalled at a homopolymeric stretch in the template. Mutations in the homopolymeric stretch may be the result of primer slippage on the template followed by incorrect alignment of complementary bases in this region (Bebenek et aZ., 1993; Streisinger et aZ., 1966). This could result in single base mispairings and loops in the primer or template. Not surprisingly, RDDP is 10 times more faithful than DDDP (Yu and Goodman, 1992). This correlates with decreased pausing and increased kcat observed during polymerization from RNA as compared to DNA templates (Klarmann et al., 1993; Reardon, 1993). In addition, increased mutagenesis was observed near homopolymeric runs in both DNA and RNA templates that had high potential for pausing during reverse transcription (Abbotts et al., 1993; Bebenek et al., 1993). Both sequence-specific hypermutability and pausing may be the result of differential dNTP insertion kinetics by HIV-1 RT (Cai et al., 1993). Insertion of purines was favored over pyrimidines during arrested polymerization on a DNA template (Cai et al., 1993). Finally, imbalances in the dNTP pool of cells has been shown to increase the frequency of insertion, deletion, and substitution mutations in both replicating cellular DNA and infecting viral DNA (Bebenek et al,, 1992; Kunz, 1988; Phear et al., 1987; Vartanian et al., 1994). Further studies into dNTP incorporation by HIV-1 RT may provide understanding for the incorrect base substitutions that occur prior to pause sites.

F. Ribonucleases of HIV-1 Reverse Transcriptase Shortly after the discovery of an RDDP activity in retroviruses, it was predicted that synthesis of the second or (+) DNA strand in double-stranded proviral DNA would have to use the reversetranscribed (-) DNA as template (Molling et al., 1971). This event would require the removal of the RNA template from the (+) RNA(-) DNA hybrid. On this basis, a DNAiRNA-dependent ribonuclease (RNase HI activity was identified that copurified with AMV RT (Molling et al., 1971).

HIV-1 REVERSE TRANSCRIPTASE

133

An RNase H activity was also copurified with HIV-1 RT heterodimer (Hansen et al., 1987; Starnes and Cheng, 1989). In addition, a 15-kDa polypeptide (amino acids 441 to 5601, arising from the cleavage of p66 to pi51 (Mizrahi et al., 1989; Schatz et al., 1989) and identified by monoclonal antibodies to the carboxyl terminus of p66 (Hansen et al., 19881, showed a high sequence similarity to E. coli RNase H as well as to RNase H domains of other retroviral RTs (Johnson et al., 1986). Both the 95- and 63-kDa subunits to the AMV RT heterodimer had active RNase H domains at carboxyl termini (Soltis and Skalka, 1988), whereas only the p66 subunit in heterodimer HIV-1 RT possessed RNase H activity (Hansen et al., 1988; Starnes and Cheng, 1989). The HIV-1 p15 protein was enzymatically active when purified from virions (Hansen et al., 1988). However, recombinant p15, cleaved by HIV-1 proteinase between sites F440 and Y441 or purified with short amino-terminal extensions, was devoid of RNase H activity (Becerra et al., 1990; Hostomsky et al., 1991).Surprisingly, RNase H activity could be restored in an inactive, recombinant p15 when combined with the RNase H-deficient p51 subunit of HIV-1 RT (Hostomsky et al., 1991). Therefore, the RNase H activity observed in p15 purified from HIV-1 virions (Hansen et al., 1988) may have been the result of p51 contamination. Recombinant p15, containing residues 426 to 560 of RT and a polyhistidine tag on the C terminus, showed RNase H activity after metal chelate affinity purification (Evans et al., 1991; Smith and Roth, 1993). The covalent addition of amino-terminal sequences, derived from the polymerase domain, to p15 resulted in a progressive increase in Mnz+-dependent RNase H activity (Smith et al., 1994).This increase in RNase H activity can also be modulated by the addition of aminoterminal peptides of p51 (Smith et al., 1994). The crystal structure for the RNase H domain (p15) of HIV-1 RT was determined to 2.4 resolution (Davies et al., 1991; Hostomska et al., 1991). The RNase H domain is composed of a five-stranded mixed p sheet flanked by four a helices. The nomenclature for the secondary structure of p15 and p51 are the same in name but not sequence; that is, the aB sheet in p15 (amino acids 500-508) and p51 (amino acids 7883) are not the same and do not consist of the same amino acid sequence. Only one a helix (aB) was found perpendicular to all of the other parallel a helices and p sheets in p15 (Davies et al., 1991).Except for the active site pocket, the structures of the RNase H domain of HIV-1 RT and of E.coli RNase H are very similar (Davies et al., 1991; Yang et al., 1990).The flexible active site of the HIV-1 RNase H domain is found folded into the protein in p15 and acquires the proper, rigid conformation for RNase H digestion only when covalently linked to p51 to form p66 or noncovalently associated with the p51 subunit (Hostomsky et al., 1991; Powers et al., 1992). An active recombinant

134

ERIC J. ARTS AND MARK A. WAINBERG

p15 likely assumes a different conformation than the RNase H domain found in p66 of the heterodimer, since both showed different substrate inhibition constants and binding preferences to primer-template (Evans et al., 1994). It was also shown that p15, with the addition of four residues at the amino terminus, could bind primer-template but was still enzymatically inactive (Cirino et al., 1993). The sequence 534AWVPAHKGIGGN545, found in the RNase H domain of HIV-1 RT, is conserved in nearly all enzymes capable of RNase H digestion (Johnson et al., 1986). This conserved sequence as well as residues D443, E478, D494, D498, and A549 (Mizrahi et al., 1994; Schatz et al., 1989) are thought to form the divalent cation binding pocket and active site of the RNase H domain of HIV-1 RT (Table I) (Davies et al., 1991). Mutations D443N, E478Q, and D443N/D498N significantly reduced RNase H activity, whereas a N494D mutation had no effect on RNase H activity (Mizrahi et al., 1994). Stable enzyme could not be obtained when residue D498 was mutated (Mizrahi et al., 1994). Interestingly, E. coli DNA polymerase I, which binds two Mn2+ ions at its exonucleolytic active site (Beese and Steitz, 19911, has a similar divalent cation binding pocket to that found in the RNase H domain of HIV-1 RT (Davies et al., 1991).The possible binding of a pair of divalent cations may be responsible for the 3’ + 5’-exonuclease activity of HIV-1 RNase H, not found in nonlentiviral RTs (Schatz et al., 1990). This notion is further supported by results showing the separation of endonucleolytic and exonucleolytic or 5’-directed RNase H activity, using mutant HIV-1 RTs with carboxyl-terminal deletions in p66 (Ghosh et al., 1995).Destabilization of D549, through deletion of a residue conserved among lentiviruses (i.e., S5531, may impair the binding of one divalent cation and result in the loss of 5’-directed RNase H activity of HIV-1 RT but not its endonucleolytic activity. Orientation of the polymerase active site over the 3’-hydroxyl of the primer in an HIV-1 RT-primer-template complex results in an endonucleolytic cut in the RNA template, approximately 15 to 19 nucleotides behind the polymerization initiation site (DeStefano et al., 1991; Furfine and Reardon, 1991; Gopalakrishnan et al., 1992; Oyama et al., 1989; Schatz et al., 1990).In the case of most retroviral RTs, this initial endonucleolytic cut is Mg2+-dependent and is independent of polymerization (Krug and Berger, 1989). As mentioned above, HIV-1 Rl’ and MLV RT are capable of a 3’ +. 5’-exonucleolytic RNase H activity (Schatz et al., 1990; Wohrl et al., 1991; Post et al., 1993). In the absence of dNTP, a reduction of the initial cut at positions -15 (-17) to -7 (9) may be due to an exonucleolytic activity of RT displaced on the primer-template complex (Schatz et al., 1990). Deletion of 8 residues from the carboxyl terminus of p66 in heterodimeric RT resulted in the

HIV-1 REVERSE TRANSCRIPTASE

135

loss of exonucleolytic or 5’-directed RNase H activity (-17 to -9 cleavage) but the retention of endonucleoytic RNase H activity (-17 cut). Deletions of 16 and 23 amino acids resulted in the complete loss of both activities (Ghosh et al., 1995). All three deleted enzymes maintained wild-type polymerase properties. The addition of an HIV-1 RT trap prevented this polymerizationindependent exonuclease or 5’-directed nuclease cleavage event on the RNA template (Gopalakrishnan et al., 1992).In addition, the use of an HIV-1 RT trap during polymerization suggested that RNase H digestion and polymerization by HIV-1, AMV, and MLV RTs were functionally uncoupled (DeStefano et al., 1991). However, programmed extensions of +1 to + 5 nucleotides from the polymerization active site resulted in a corresponding exonuclease digestion from the initial endonuclease cut at -15 or -16 (Furfine and Reardon, 1991). This supports the notion that an exonucleoytic-like RNase H digestion is a polymerization-dependent event. As described in Section IV,A, RNase H digestion is necessary for the specific cleavages around the polypurine tract used to prime (+) strong-stop DNA (Pullen et al., 1993), and for the removal of tRNA sequences used as a template for PBS synthesis in (+I strong-stop DNA (Smith and Roth, 1992). An RNA/RNA-dependent RNase activity (RNase D) has been characterized for HIV-1 RT. The HIV-1 enzyme was capable of cutting the PBS RNA, annealed to the tRNALys3 primer, at two sites but only in the presence of Mn2+ (Ben-Artzi et al., 1992a). An RNase H* activity in the PBS would aid the (+1 strong-stop DNA to switch templates via the PBS and complete (+) DNA synthesis. However, the recombinant HIV-1 RT used in these studies may have been contaminated with E . coli RNase I11 (Hostomsky et al., 1992), which can also enact the Mnz+-dependent cleavage of one strand in a RNA duplex (Hostomsky et al., 1992). Nevertheless, RNase H* activity can be copurified with HIV-1 RT (Ben-Artzi et al., 1992b). In addition, the E478Q mutation in HIV-1 RT, which rendered RT deficient in RNase H activity (Schatz et al., 19891, also resulted in a lack of Mnz+-dependent RNase H* cleavage (Ben-Artzi et al., 1992b).These results suggest that RNase H and RNase H* activities, dependent on Mg2+ and Mn2+, respectively, may be controlled by the same enzymatic domain and/or active site in HIV-1 RT.

G . First Template Switch The synthesis of (-1 DNA from retroviral RNA templates is not a direct and continuous polymerization process. The initiation of RDDP by a tRNA primer near the 5’ end of the viral RNA genome results in a

5' HIW RNA GENOME

c

w

m

B +1

LEGEND

-

(-)HWDNA HCV gemmic RNA IFINA'W

+187 base position in RNA genomr R re ion * .

%paration 01 regions in lhe RNA aenorne or

RNase H acttve sle of RT

3

polymerkationactive site of R'

DISLOCATED RT

of

138

ERIC J. ARTS AND MARK A. WAINBERG

short (-) DNA fragment. This fragment, consisting of the R and U5 regions of the LTR covalently linked to the tRNA primer, is termed (-1 strong-stop DNA. For the continuation of HIV-1 (-1 DNA polymerization from (-1 strong-stop DNA, two events must occur in succession: (1) active RNase H digestion of the RNA template annealed to (-> strong-stop DNA and (2) annealing of the complementary R regions in the (-) strong-stop DNA and in the acceptor viral RNA template (Gilboa et al., 1979; reviewed by Telesnitsky and Goff, 1993b). The cognate primer of HIV-1 RT, tRNAL@, anneals to the PBS in conjunction with HIV-1 RT, secondary RNA sequences surrounding the PBS, and possibly the nucleocapsid protein (see Section IV,E). Structure 1 in Fig. 5 schematically depicts tRNALrs3p bound to the 5’ end of the HIV RNA, in its predicted secondary structure. The specific binding of RT to the tRNAL@-viral RNA complex initiates RDDP in the presence of dNTPs. Polymerization of at least 15 nucleotides may precede the initiation of RNase H digestion by HIV-1 RT (Furfine and Reardon, 1991; DeStefano et al., 1991; Schatz et al., 1990). The tRNAPBS RNA duplex, found in a linear A-conformation (Chastain and Tinoco et al., 19911, may be resistant t o MgZ+-dependentRNase digestion in a translocating HIV-1 RT. The RNase D activity of HIV-1 RT is a Mnz+-dependent, endonuclease cleavage (Ben-Artzi et al., 1992a),but HIV-1 RT preferentially binds Mg2+ over Mn2+ (Hostomska et al., 1991). The initial endonucleolytic RNase H cleavage, near or at the first base in the RNA-DNA duplex, may relax the RNA secondary structure of the template for more processive RDDP and exonucleolytic RNase H digestion (structure 2 in Fig. 5). RNA-dependent DNA polymerization and the lagging polymerization-dependent exonucleolytic RNase H activity proceed to transcribe and degrade, in a 3’ to 5’ direction, nucleotides + 181 to + 1of the HIV RNA genome (structure 2 in Fig. 5). Eventually, the 5’ end of the viral RNA genome is met by the polymerization active site of HIV-1 RT, resulting in the last deoxynucleotide addition and dissociation of FIG.5. First template switch during HIV-1 reverse transcription. Structure 1: Depiction of tRNALyd annealing to the PBS prior to the initiation of RNA-dependent DNA polymerization. The RNA secondary structure for the RIU5IPBSluncoding region of HIV-1 genomic RNA is schematically represented as predicted by the Zucher formula (Zucher and Steigler, 1981). Structure 2 Initiation of RDDP and RNase H activities of HIV-1 RT from tRNALys3. The polymerization of (-) strong-stop DNA is preceded by unwinding of the RNA secondary structure. Structure 3 Annealing (or zippering) of the R region of the acceptor RNA template (3’HIV RNA genome) with the complementary R region of the (-) strongstop DNA. The latter occurs during the completion of (-) strongstop DNA and RNase H digestion of the R region. Structure 4 After the complete annealing of the two R regions, (-) strong-stop DNA can prime continued RDDP of (-1 HIV DNA as described.

HIV-1 REVERSE TRANSCRIPTASE

139

HIV-1 RT from the primer-template complex. Dissociation of HIV-1 RT prior to completion of RNase H digestion of the RNA template results in a short RNA fragment annealed to the 3’ end of (-) strongstop DNA (Fu and Taylor, 1992; Oyama et al., 1989). The size of this RNA fragment was 14 to 18 nucleotides when RDDP was primed with AMV RT, and was reduced to 8 nucleotides from the 5‘ end in the absence of potassium chloride (Ben-Artzi et al., 1993; Fu and Taylor, 1992; Oyama et al., 1989). There appeared to be an inherent inability of retroviral RTs (Le., HIV-1, MLV, and AMV enzymes) to RNase H digest the RNA of DNA-RNA hybrids near the 5’ end of an LTR (BenArtzi et al., 1993). However, in HIV-1, exonuclease activity in nonpolymerizing RT may reduce this distance to 7 nucleotides from the 5’ end (Schatz et al., 1990). Interestingly, a mutant HIV-1 RT, deficient in exonucleolytic RNase H digestion (5’-directed nuclease activity) but still capable of degrading RNA by an endonucleolytic activity, was incapable of template switching (Ghosh et al., 1995). This is likely due to a lack of exonucleolytic digestion at the 5’end of the RNA template. Full RNase H digestion of the entire RNA template at the 5’ end may not be required for the first template switch. Immediately following initial RNase H degradation of the R region during polymerization of (-) strong-stop DNA, the 3’ end of the acceptor RNA template can begin to anneal to the 5’end of the R region of (-) strong-stop DNA (Peliska and Benkovic, 1992). This event could be described schematically as a “zippering up” of the two R regions (structure 3, Fig. 5). At one instant during this zippering process, a quaternary complex may exist, consisting of RT bound to the 3‘ end of the acceptor RNA template, the 5’ end of the oligoribonucleotide of the donor RNA template, and the 3‘ end of the elongating (-1 strong-stop DNA (structure 3,Fig. 5) (Peliska and Benkovic, 1992). The oligoribonucleotide at the 5‘ end of the R region is likely displaced by the zippering of the R region from the acceptor template with the R region of (-1 strong-stop DNA (structure 4, Fig. 5 ) (Ben-Artzi et al., 1993; Fu and Taylor, 1992). The size of the oligoribonucleotide may be associated with the length of the R region. For example, the R region of HIV-1 (96 nucleotides) is considerably longer than that of MLV (68 nucleotides), but the oligoribonucleotide associated with the 5’ end of the genome in HIV-1 is also longer than that of MLV (Fu and Taylor, 1992). Therefore, complexes with a longer R or region of complementarity may provide the extra energy required to displace a loner oligoribonucleotide. Absence of the 5’ directed nuclease activity of HIV-1 RT results in a longer oligoribonucleotide fragment (Ghosh et al., 1995). The “zippering” of the two R regions likely did not provide the energy required to displace a longer oligoribonucleotide. This could have caused inhibition of a strand transfer event.

140

ERIC J. ARTS AND MARK A. WAINBERG

Displacement of the oligoribonucleotide is required for the annealing of the 3' end of the acceptor template and the initiation of continued (- 1 DNA synthesis starting in the U3 region (nucleotides + 9098 to +8630 in an interstrand template switch; see structures 3 and 4, Fig. 5). Although the majority of (-) DNA transcripts switched templates after complete synthesis of (-1 strong-stop DNA, it was shown that partial (-1 strong-stop DNA transcripts can switch from the donor to acceptor templates in the R region homologs, prior to complete reverse transcription of the R from the donor template (Klaver and Berkhout, 1994). As stated above, the complementary R region overlap is required for template switching. In an in uitro template switching reaction with MLV and HIV-1 RT, an increase in R region overlap between an acceptor DNA or RNA template and the RNA donor template resulted in an augmented efficiency of template switching (Luo and Taylor, 1990; Peliska and Benkovic, 1992).In addition, a polymerized DNA product, generated from a DNA donor template by MLV RT, could efficiently switch to an RNA but not DNA template (Luo and Taylor, 1990). This increase is likely due to increased annealing energy of a DNA-RNA hybrid duplex as compared to a DNA-DNA duplex. Template switching efficiency was also augmented by an increase in incubation temperature (i.e., from 37" to 50°C) in reactions catalyzed by the AMV RT, which is fully competent for both polymerization and RNase H digestion at these temperatures (Ouhammouch and Brody, 1992). An increase in temperature could augment the dissociation of the DNA product from the shorter RNase H-digested donor RNA more than the dissociation of the DNA product from the acceptor RNA template with a longer complementary overlap. With HIV-1, AMV, and MLV RTs enzymes, the increase of acceptor template relative to a donor template augmented the efficiency of template switching (Arts et al., 1994b; Luo and Taylor, 1990; Ouhammouch and Brody, 1992; Peliska and Benkovic, 1992). Finally, a helicase activity of RT may serve to unwind and dissociate the oligoribonucleotide/(- strong-stop DNA hybrid duplex during the annealing of acceptor template (Collett et al., 1978). The genomic RNA of HIV-1 is greater than 9 kilobases in length. The reverse transcription scheme predicts that elongating (-1 DNA must switch from the 5' end of one genomic RNA template to the 3' end of the same or another genomic RNA template. The efficiency of an in uitro template switching reaction, mimicking the first template switch of many retroviruses, is less than 100% even with a ratio of acceptor:donor template as high as 1O:l (Arts et al., 199413; Luo and Taylor, 1990; Ouhammouch and Brody, 1992; Peliska and Benkovic,

HIV-1 REVERSE TRANSCRIFTASE

141

1992). However, in activated CD4+ cells infected with HIV-1, where the ratio of acceptor to donor genomic template is estimated at 2:1, the efficiency of the first template switch was nearly 100%(Arts and Wainberg, 199413; Zack et al., 1990). Thus, other factors, aside from reverse transcriptase and tRNA primer, must be required for the first template switch. Dimerization of genomic RNA would likely increase the frequency of interaction of the two ends. Currently, no studies have investigated the effects of genomic dimerization on the first template switch. A dimer linkage structure has been mapped in HIV-1 to 100 nucleotides 5' of the major splice donor site (Darlix et al., 1990; Paillart et al., 1994). Using the HIV RNA PBS template (Arts et al., 1994b), a potential dimerization sequence was further refined to 19 nucleotides (+ 233 to +251) in the dimer linkage structure (Laughrea and Jettk, 1994). In Fig. 5 , the HIV-1 RNA templates involved in the first template switch are schematically represented in the most energy efficient secondary structure calculated by the Zucher formula (Zucher and Steigler, 1981).However, these structures fail to represent interactions between the two ends of genomic RNA induced by dimerization. Retroviral nucleocapsid was shown to bind, denature, and renature retroviral genomic RNA but was not necessary for dimerization (De Rocquigny et al., 1992; Fu and Rein, 1993; South et al., 1990). Nucleocapsid protein, added to reverse transcription reactions in uitro, was shown to increase the efficiency and rate of the first template switch and nonspecific strand transfer events (Allain et al., 1994; Peliska et al., 1994; Tsuchihashi and Brown, 1994).The binding of RNA by nucleocapsid protein likely causes relaxing of the RNA secondary structure, thus promoting both displacement and annealing of nucleic acids during nonspecific strand transfer events and the first template switch. The majority of studies to date on the first template switch in uitro have employed synthetic RNA templates with nonretroviral R regions and synthetic primers to initiate RDDP (Luo and Taylor, 1990; Ouhammouch and Brody, 1992; Peliska and Benkovic, 1992). A synthetic deoxyoligonucleotide, complementary to the PBS, and human placental tRNALys3 were compared as primers in an in uitro reverse transcription/template switch assay (Arts et al., 199413). This assay employed the actual HIV RNA templates utilized for the first template switch and recombinant HIV-1 IET (Fig. 5). Use of a deoxyoligonucleotide primer resulted in three times less template switching than was obtained with tRNALys3 (Arts et al., 1994b).The increase in template switching efficiency with tRNALys3 as primer, as compared to a deoxyoligonucleotide, may be attributable to several factors including an increase in RNase H activity in reactions primed by tRNALys3and

142

ERIC J. ARTS AND MARK A. WAINBERG

specific interactions of tRNALys3 with HIV-1 RT and RNA template. The addition of AZT 5'-triphosphate (AZT-TP)to the reverse transcriptionhemplate switching reaction resulted in preferential chain termination immediately following the template switch (Arts and Wainberg, 1994). This preferential inhibition by AZT-TP was observed in reactions primed by tRNALys3 but not in those primed by a deoxyoligonucleotide primer (Arts et al., 199413). This difference with tRNALys3, as compared with deoxyoligonucleotide as primer, further highlights the specificity of RDDP initiation and the first template switch during HIV-1 reverse transcription. The MLV RT, unlike HIV-1 RT, has relatively weak RNase H activity relative to polymerase activity (DeStefano et al., 1991; Krug and Berger, 1989). Therefore, it is not surprising that mutant MLV RT, deficient in RNase H activity, could still switch templates, albeit at a lower efficiency than the wild type (Luo and Taylor, 1990). On the other hand, an active RNase H domain in HIV-1 RT was required for template switching (Peliska and Benkovic, 1992). Finally, AMV has a shorter R region length (21 nucleotides) than MLV (68 nucleotides) or HIV-1 (97 nucleotides) (Weiss et al., 1985).The RNase H activity relative to polymerase activity was compared for both HIV-1 RT and AMV RT (DeStefano et al., 1994). The HIV-1 RT enacted on average 1 RNA cleavage per every 10 to 15 nucleotides incorporated, whereas AMV RT enacted RNA cleavage at a much lower rate (DeStefano et al., 1994). Similar to the correlation between R region length and the length of the 5' end oligoribonucleotide, there appears to be a correlation between the length of the R region and the relative strength of the RTRNase H activity of a given retrovirus. MLV has the shorter R region and a weaker RT-RNase H activity, relative to RT-polymerase activity, than is the case for HIV-1. RNase H cleavage, during AMV reverse transcription, may not be necessary for the first template switch but would still be required for synthesis of (+) DNA. The first template switch in AMV may be more dependent on displacement of the R region of the donor template from (-1 strong-stop DNA and transfer of (-1 strong-stop to the R region on the acceptor RNA template. Therefore, although the general scheme of reverse transcription is similar for all retroviruses, significant differences exist among different enzymes with regard to individual steps. The first template switch for MLV and AMV is likely more dependent on the R region than the RNase H activity of these RTs. Thus, the short R region and nucleocapsid could increase displacement of the annealed donor template and increase annealing of (-1 strong-stop DNA for acceptor template, respectively (Allain et al., 1994; Luo and Taylor, 1990). The first template switch during HIV-1 reverse transcription is likely dependent on

HIV-1 REVERSE TRANSCRIPTASE

143

RNase H activity for complete degradation of donor RNA to occur during polymerization of (-1 strong-stop DNA (DeStefano et al., 1992; Peliska and Benkovic, 1992).This would promote annealing of free (-1 strong-stop DNA to acceptor template in the R region. Only an interstrand first template switch has been demonstrated in uitro, yet both interstrand and intrastrand first template switching events may occur during reverse transcription in uiuo. Two groups have studied the specificity of template switching in tissue culture infections with recombinant retroviral-like particles. In one study, a helper cell line that produced empty spleen necrosis virus particles was transfected with two vectors, termed constructs A and B, and expressed pseudogenomes with genes encoding hygromycin B resistance and neomycin resistance, respectively (Panganiban and Fiore, 1988). The LTRs flanking the resistance genes of both constructs differed by only two restriction enzyme sites, HindIII found in a U5 region of construct A and SacI found in a U3 region of construct B. Retroviral particles, heterozygous for the two encapsidated RNA genomes, were used to infect chicken embryo fibroblast cells which were then selected for resistance to hygromycin B and neomycin. Through differential HindIII and SacI restriction endonuclease cleavage of the proviral DNA in the latter infected cells, it was shown that the first template switch was generally an interstrand event, whereas the second template switch was always an intrastrand event. Hu and Temin (1990) used a similar approach with the same packaging cell line but with different LTR constructs to ensure equal packaging of the two heterozygous pseudogenomes. This study suggested that there was no preference for either an interstrand or intrastrand event during the first template switch. The discrepancy between the two observations for the mode of the first template switch may be the result of differences in experimental protocols (reviewed by Telesnitsky and Goff, 1993b). First, neither study could assume that retroviral particles packaged only two RNA pseudogenomes. Preferential packaging of one of the two RNA pseudogenomes in the first study may have caused more than two encapsidated RNA pseudogenomes, thus skewing the results so that the interstrand template switch appeared more frequent during (-1 strong-stop DNA transfer. This could be due to increases in trans over cis templates. In the second study, recombination between the two heterozygous templates must have occurred during reverse transcription in order for double toxin resistance t o have been achieved. Such recombination likely resulted from strand transfer events between the two LTRs, as well as the first and second template switches. An interstrand template switch could only be assayed in cells infected with

144

ERIC J. ARTS AND MARK A. WAINBERG

virus undergoing such rare recombination. Thus, the system may have been biased in favor of the intrastrand first template switch. ACKNOWLEDGMENTS Research performed in our laboratory was supported by the Medical Research Council of Canada and by Health Canada. Eric J. Arts is the recipient of a fellowship from Health Canada and Mark A. Wainberg holds a National AIDS Scientist award from Health Canada.

REFERENCES Abbotts, J., Bebenek, K., Kunkel, T. A., and Wilson, S. H., (1993). Mechanism of HIV-1 reverse transcriptase: Termination of processive synthesis on a natural DNA template is influenced by the sequence of the template-primer stem. J. Biol. Chem. 268, 10312-10323. Aiyar, A., Cobrinik, D., Ge, Z., Kung, H.J., and Leis, J. (1992). Interaction between retroviral U5 RNA and the TVrC loop of the tRNATp primer is required for efficient initiation of reverse transcription. J. Virol. 66, 2464-2472. Aiyar, A., Ge, Z., and Leis, J. (1994). A specific orientation of RNA secondary structures is required for initiation of reverse transcription. J. Virol. 68, 611-168. Allain, B., Lapadat-Tapolsky, M., Berlioz, C., and Darlix, J.-L. (1994). Transactivation of the minus-strand DNA transfer by nucleocapsid protein during reverse transcription of the retroviral genome. EMBO J. 13, 973-981. Andreola, M.-L., Nevinsky, G. A., Barr, P. J., Sarih-Cottin, L., Bordier, B., et al. (1992). Interaction of tRNA1YS with the p66/p66 form of HIV-1 reverse transcriptase stimulates DNA polymerase and ribonuclease H activities. J. Biol. Chem. 267, 1935619362. Andreola, M.-L., Tarrago-Litvak, L., Levina, A. S., Kolocheva, T. I., El Dirani-Diab, R., et al. (1993). Affinity labeling and functional analysis of the primer binding domain of HIV-1 reverse transcriptase. Biochemistry 32, 3629-3637. Argos, P. (1988). A sequence motif in many polymerases. Nucleic Acids Res. 16, 99099916. Arnold, E., Jacobo-Molina, A., Nanni, R. G., Williams, R. L., Lu, X., et al. (1992). Structure of HIV-1 reverse transcriptase/DNA complex at 7 A resolution showing active site locations. Nature (London) 357, 85-89. Arts, E. J., and Wainberg, M. A. (1994). Preferential incorporation of nucleoside analogs after the first template switch during human immunodeficiency virus reverse transcription. Antimicrob. Agents Chemother. 38, 1008-1016. Arts, E. J., Mak, J., Kleiman, L., and Wainberg, M. A., (1994a). DNA found in human immunodeficiency virus type 1 particles may not be required for infectivity. J. Gen. Virol. 75, 1605-1613. Arts, E. J., Li, X., Gu, Z., Kleiman, L., Parniak, M. A., and Wainberg, M. A. (1994b). Comparison of deoxyoligonucleotide and tRNAL@ as primers in an endogenous human immunodeficiency virus-1 in uitro reverse transcription/template switching assay. J. Biol. Chem. 269,14672-14680. Arya, S. K.,Guo, C., Josephs, S. F., and Wong-Staal, F. (1985). Trans-activator gene of human T-lymphotropic virus type I11 (HTLV-111). Science 229, 69-73. Bacolla, A., Shih, C. K., Rose, J. M., Piras, G., Warren, T. C., et al. (1993). Amino acid substitutions in HIV-1 reverse transcriptase with corresponding residues from HIV-2.

HIV-1 REVERSE TRANSCRIPTASE

145

Effect on kinetic constants and inhibition by non-nucleoside analogs. J.Biol. Chem. 268,16571-16577. Bakhanashvili, M., and Hizi, A. (1992). Fidelity of the RNA-dependent DNA synthesis exhibited by the reverse transcriptases of human immunodeficiency virus types 1and 2 and of murine leukemia virus: Mispair extension frequencies. Biochemistry 31, 9393-9398. Bakhanashvili, M., and Hizi, A. (1993). The fidelity of the reverse transcriptases of human immunodeficiency viruses and murine leukemia virus, exhibited by the mispair extension frequencies is sequence dependent and enzyme related. FEBS Lett. 319,201-205. Bakhanashvili, M., and Hizi, A. (1994). Interaction of the reverse transcriptase of human immunodeficiency virus type 1 with DNA. Biochemistry 33, 12222-12228. Baltimore, D. (1970). RNA-dependent DNA polymerase in virions of RNA tumor viruses. Nature (London) 226, 1209-1211. Barat, C., Lullien, V., Schatz, O., Keith, G., Nugeyre, M. T., Gruninger-Leitch, F., BarreSinoussi, F., LeGrice, S. F. J., and Darlix, J. L. (1989). HIV-1 reverse transcriptase specifically interacts with the anticodon domain of its cognate primer tRNA. EMBO J. 8,3279-3285. Barat, C., LeGrice, S. F. J., and Darlix, J. L. (1991). Interaction of HIV-1 reverse transcriptase with a synthetic form of its replication primer. Nucleic Acids Res. 19, 751757. Barat, C., Schatz, O., LeGrice, S., and Darlix, J. L. (1993). Analysis of the interactions of HIV-1 replication primer tRNA(Lys3) with nucleocapsid protein and reverse transcriptase. J. Mol. Biol. 231, 185-190. Barr, P. J., Power, M. D., Lee-Ng, C. T., Gibson, H. L., and Luciw, P. A. (1987). Expression of active human immunodeficiency virus reverse transcriptase in Sacchuromyces cereuisiae. BiolTechnology 5, 486-489. Barre-Sinoussi, F., Chermann, J. C., Rey, F., Nugeybe, M. T., Chamaret, S., et al. (1983). Isolation of a T-lymphocyte retrovirus from a patient at risk of acquired immune deficiency syndrome (AIDS). Science 220,868-870. Basu, A., Tirumalai, R. S., and Modak, M. J. (1989). Substrate binding in human immunodeficiency virus reverse transcriptase. An analysis of pyridoxal5’-phosphate sensitivity and identification of lysine 263 in the substrate-binding domain. J.Biol. Chem. 264,8746-8752. Basu, A., Basu, S., and Modak, M. J. (1993). Structure-activity analyses of HIV-1 reverse transcriptase. Biochem. Biophys. Res. Commun. 183, 1131-1138. Baudin, F., Marquet, R., Isel, C., Darlix, J. L., Ehresmann, B., and Ehresmann, C. (1993). Functional sites in the 5’ region of human immunodeficiency virus type1 RNA form defined structural domains. J. Mol. Biol. 229, 382-397. Bavand, M. R., Wagner, R., and Richmond, M. J. (1993). HIV-1 reverse transcriptase: Polymerization properties of the p51 homodimer compared to the p66/p51 heterodimer. Biochemistry 32,10543-10552. Beard, W . A., Stahl, S. J., Kim, H.-R., Bebenek, K., Kumar, A., Strub, M.-P., Becerra, S. P., Kunkel, T. A., and Wilson, S. H. (1994). Structure/function studies of human immunodeficiency virus type 1 reverse transcriptase. Alanine scanning mutagenesis of a n a-helix in the thumb subdomain. J.Biol. Chem. 269, 28091-28097. Bebenek, K., Abbotts, J., Roberts, J. D., Wilson, S. H., and Kunkel, T. A. (1989). Specificity and mechanism of error-prone replication by human immunodeficiency virus-1 reverse transcriptase. J.Biol. Chem. 264, 16948-16956. Bebenek, K., Roberts, J. D., and Kunkel, T. A. (1992). The effects of dNTP pool imbalances on frameshift fidelity during DNA replication. J. Biol. Chem. 267, 3589-3596. Bebenek, K., Abbotts, J., Wilson, S. H., and Kunkel, T. A. (1993). Error-prone polymeriz-

146

ERIC J. ARTS AND

MARK A. WAINBERG

ation by HIV-1 reverse transcriptase: Contribution of template-primer misalignment, miscoding and termination probability to mutational hot spots. J.Biol. Chem. 268, 10324-10334. Becerra, S. P., Clore, G. M., Gronenborn, A. M., Karlstrom, A. R., Stahl, S. J., et al., (1990). Purification and characterization of the RNase H domain of HIV-1 reverse transcriptase expressed in recombinant Escherichia coli. FEBS Lett. 270, 76-80. Becerra, S. P., Kumar, A., Lewis, M. S., Widen, S. G., Abbotts, J., Karawya, E. M., Hughes, S. H., Shiloach, J., and Wilson, S. H. (1991). Protein-protein interactions of HIV-1 reverse transcriptase: Implications of central and C-terminal regions in subunit binding. Biochemistry 30, 11707-11719. Beese, L. S., and Steitz, T. A. (1991). Structural basis for the 3’-5’ exonuclease activity of Escherichia coli DNA polymerase I: A metal ion mechanism. EMBO J. 10, 25-33. Ben-Artzi, H., Zeelon, E., Gorecki, M., and Panet, A. (1992a). Double-stranded RNAdependent RNase activity associated with human immunodeficiency virus type 1 reverse transcriptase. Proc. Natl. Acad. Sci. U.S.A. 89, 927-931. Ben-Artzi, H., Zeelon, E., LeGrice, S. F. J., Gorecki, M., and Panet, A. (199213).Characterization of the double stranded RNA dependent RNase activity associated with recombinant reverse transcriptases. Nucleic Acids Res. 20, 5115-5118. Ben-Artzi, H., Zeelon, E., Amit, B., Wortzel, A., Gorecki, M., and Panet, A. (1993). RNase H activity of reverse transcriptases on substrates derived from the 5’ end of retroviral genome. J.Biol. Chem. 268, 16465-16471. Berkhout, B., and Schoneveld, I. (1993). Secondary structure of the HIV-2 leader RNA comprising the tRNA-primer binding site. Nucleic Acids Res. 21, 1171-1178. Berkowitz, R. D., Luban, J., and Goff, S. P. (1993). Specific binding of human immunodeficiency virus type 1 gag polyprotein and nucleocapsid protein to viral RNAs detected by RNA mobility shift assays. J. Virol. 67, 7190-7200. Biswal, N., McCain, B., and Bensyesh-Melnick, M. (1971).The DNA of murine sarcomaleukemia virus. Virology 45, 697-706. Boeke, J. D., Garfinkel, D. J., Styles, C. A,, and Fink, G. R. (1985). Ty elements transpose through a n RNA intermediate. Cell (Cambridge, Mass.)40, 491-500. Bordier, B., Tarrago-Litvak, L., Sallafranque-Andreola, M.-L., Robert, D., Tharaud, D., et al. (1990). Inhibition of the p66/p51 form of human immunodeficiency virus reverse transcriptase by tRNALy8. Nucleic Acids Res. 18, 429-436. Borroto-Esoda, K., and Boone, L. R. (1991). Equine infectious anemia virus and human immunodeficiency virus DNA synthesis in uitro: Characterization of the endogenous reverse transcriptase reaction. J. Virol. 65, 1952-1959. Boshl, M. L., Earl, P. L., Fargnoli, K., Picciafuoco, S., Giombini, F., et al. (1989). Identification of the fusion peptide of primate immunodeficiency virus. Science 244, 694697. Boyer, P. L., Ferris, A. L., and Hughes, S. H. (1992a). Cassette mutagenesis of the reverse transcriptase of human immunodeficiency virus type 1. J. Virol. 66, 1031-1039. Boyer, P. L., Ferris, A. L., and Hughes, S. H. (1992b). Mutational analysis of the fingers domain of human immunodeficiency virus type 1 reverse transcriptase. J. Virol. 66, 7533-7537. Boyer, P. L., Ferris, A. L., Clark, P., Whitmer, J., Frank, P., Tantillo, C., Arnold, E., and Hughes, S. H. (1994). Mutational analysis of the fingers and palm subdomains of human immunodeficiency virus type-1 (HIV-1) reverse transcriptase. J. Mol. Biol. 243,472-483. Bukrinsky, M. I., Haggerty, S., Dempsey, M. P., Sharova, N., Adzhubel, A., et al. (1993a). A nuclear localization signal within HIV-1 matrix protein that governs infection of non-dividing cells. Nature (London) 365, 666-669. Bukrinsky, M., Sharova, N., and Stevenson, M. (1993b). Human immunodeficiency virus

HIV-1 REVERSE TRANSCRIPTASE

147

type 1 2-LTR circles reside in a nucleoprotein complex which is different from the preintegration complex. J . Virol. 67, 6863-6865. Bushman, F. D., Fujiwara, T., and Craigie, R. (1990). Retroviral DNA integration directed by HIV integration protein in vitro. Science 249, 1555-1558. Cai, H., Bloom, L. B., Eritja, R., and Goodman, M. F. (1993). Kinetic of deoxyribonucleotide insertion and extension at abasic template lesions in different sequence contexts using HIV-1 reverse transcriptase. J. Biol. Chem. 268, 23567-23572. Champoux, J. J., Gilboa, E., and Baltimore, D. (1984).Mechanism of RNA primer removal by RNase H activity of avian myeloblastosis virus reverse transcriptase. J. Virol. 49,686-691. Chang, D. D., and Sharp, P. A. (1989). Regulation by HIV rev depends upon recognition of splice sites. Cell (Cambridge, Mass.) 59, 789-795. Chastain, M., and Tinoco, I., Jr. (1991). Structural elements in RNA. Prog. Nucleic Acid Res. Mol. Biol. 41, 131-177. Chattopadhyay, D., Evans, D. B., Deibel, M.R., Jr., Voster, A. F., Eckenrode, F. M., et al. (1992). Purification and characterization of heterodimeric human immunodeficiency virus type 1 (HIV-1)reverse transcriptase produced by in vitro processing of p66 with recombinant HIV-1 protease. J. Biol. Chem. 267, 14227-14232. Cheevers, W. P., and McGuire, T. C. (1985). Equine infectious anemia virus: Immunopathogenesis and persistence. Rev. Infect. Dis. 7, 83-88. Cheng, N., Painter, G. R., and Furman, P. A. (1991). Cross-linking of substrates occurs exclusively to the p66 subunit of heterodimeric HIV-1 reverse transcriptase. Biochem. Biophys. Res. Commun. 174, 785-789. Cheng, N., Merril, B. M., Painter, G . R., Frick, L. W., and Furman, P. A. (1993). Identification of the nucleotide binding site of HIV-1 reverse transcriptase using d’M’P as a photoaffinity label. Biochemistry 32, 7630-7634. Cheng, Y.-C., Dutchman, G. E., Bastow, M. K., Sarngadharan, M. G., and Ting, R. Y.C. (1987). Human immunodeficiency virus reverse transcriptase: General properties and its interaction with nucleoside triphosphate analogs. J . Biol. Chem. 262,2187-2189. Cirino, N. M., Kalayjian, R. C., Jentoff, J . E., and LeGrice, S. F. J . (1993). Fluorimetric analysis of recombinant p15 HIV-1 ribonuclease H. J.Biol. Chem. 268,14743-14749. Cobrinik, D., Katz, R., Terry, R., Skalka, A. M., and Leis, J . (1987). Avian sarcoma and leukosis virus pol-endonuclease recognition of the tandem long terminal repeat junction: Minimum site required for cleavage is also required for viral growth. J . Viro.61, 1999-2008. Cobrinik, D., Saskey, L., and Leis, J . (1988). A retroviral secondary structure required for efficient initiation for reverse transcription. J. Viro. 62, 3622-3630. Cobrinik, D., Aiyar, A., Ge, Z., Katzman, M., Huang, H., and Leis, J . (1991).Overlapping U5 sequence elements are required for efficient integration and initiation of reverse transcription. J . Virol. 65, 3864-3872. Collett, M. S., Leis, J . P., Smith, M. S., and Faras, A. J. (1978). Unwinding-like activity associated with retrovirus RNA-directed DNA polymerase. J . Virol. 65, 69616968. Craven, R. C., Burnet, R. P., and Wills, J . W. (1992).Role of the avian retroviral protease in the activation of reverse transcriptase during virion assembly. J. Virol. 65, 62056217. Crother, D. (1992). DNA bends and bendability. In “Molecular Structure and Life” (Y.Kyogoku and Y. Nishimura, eds.), pp. 67-77. Japan Science Society Press, Tokyo/CRC Press, Boca Raton, Florida. Cullen, B. R. (1991). Regulation of humay immunodeficiency virus replication. Annu. Rev. Microb. 45, 219-250. Dahlberg, J. E., Saywer, R. C., Taylor, J . M., Faras, A. J., and Levinson, W. E. (1974).

148

ERIC J. ARTS AND MARK A. WAINBERG

Transcription of DNA from the 70s RNA of Rous sarcoma virus. I. Identification of a specific 4 s RNA which serves as primer. J. Virol. 13, 1126-1133. Darlix, J. L., Gabus, C., Nugeyre, M. T., Clavel, F., and Barre-Sinoussi, F. (1990). Cis elements and trans-acting factors involved in the RNA dimerization of the human immunodeficiency virus HIV-1. J. Mol. Biol. 216, 689-699. Davies, 11,J. F., Hostomska, Z., Hostomsky, Z., Jordan, S. R., and Matthews, D. A. (1991). Crystal structure of the ribonuclease H domain of HIV-1 reverse transcriptase. Science 252,88-95. Davis, J. L., Molineaux, S., and Clements, J. E. (1987). Visna virus exhibits a complex transcriptional pattern: One aspect of gene expression shared with acquired immunodeficiency syndrome retrovirus. J. Virol. 61, 1325-1331. De Rocquigny, H., Gabus, C., Vincent, A., Fournie-Zaluski, M. C., Roques, B., and Darlix, J. L. (1992). Viral RNA annealing activities of human immunodeficiency virus type 1 nucleocapsid protein require only peptide domains outside the zinc fingers. Proc. Nutl. Acud. Sci. U S A . 89, 6472-6476. DeStefano, J. J., Buiser, R. G., Mallaber, L. M., Myers, T. W., Bambara, R. A., and Fay, P. J. (1991). Polymerization and RNase H activities of the reverse transcriptases from avian myeloblastosis, human immunodeficiency, and Moloney murine leukemia viruses are functionally uncoupled. J. Biol. Chem. 266, 7423-7431. DeStefano, J. J.,Mallaber, L. M., Rodriguez-Rodriguez, L., Fay, P. J., and Bambara, R. A. (1992). Requirements for strand transfer between internal regions of heteropolymer templates by human immunodeficiency virus reverse transcriptase. J. Virol. 66, 6370-6378. DeStefano, J. J., Mallaber, L. M., Fay, P. J., and Bambara, R. A. (1994). Quantitative analysis of RNA cleavage during DNA synthesis by human immunodeficiency and avian myeloblastosis virus reverse transcriptases. Nucleic Acids Res. 22,3793-3800. Dhundale, A., Furuichi, T., Inouye, S., and Inouye, M. (1987). Structure of msDNA from Myxococcus runthus: Evidence for a long, self-annealing RNA precursor for the covalently linked, branched RNA. Cell (Cambridge, Muss.)51, 1105-1112. Dib-Hajj, F., Khan, R., and Giedroc, D. P. (1993). Retroviral nucleocapsid proteins possess potent nucleic acid strand renaturation activity. Protein Sci. 2, 231-243. Di Marzo Veronese, F., Copeland, T. D., DeVico, A. L., Rahman, R., Oroszlan, S., et ul. (1986). Characterization of highly immunogenic ~ 6 6 1 ~ as 5 1the reverse transcriptase of HTLV-III/LAV. Science 231, 1289-1291. Divita, G., Restle, T., Goody, R. S., Chermann, J.-C., and Baillon, J. G. (1994). Inhibition of human immunodeficiency virus type 1 reverse transcriptase dimerization using synthetic peptides derived from the connection domain. J. Biol. Chem. 269, 1308013083. Donahue, P. R., Hoover, E. A., Beltz, G. A., Riedel, N., Hirsh, V. M., etal. (1988). Strong sequence conservation among horizontally transmissible, minimally pathogenic feline leukemia viruses. J. Virol. 62, 722-731. Earl, P. L., Moss, B., and Doms, R. W. (1991). Folding, interaction with GRP-BiP, assembly, and transport of the human immunodeficiency virus type 1 envelope protein. J. Virol. 65, 2047-2055. Engelman, A., Mizuuchi, K., and Craigie, R. (1991).HIV-1 DNA integration: Mechanism of viral DNA cleavage and DNA strand transfer. Cell (Cambridge, Muss.) 67, 12111221. Erikson, E., and Erikson, R. L. (1971). Association of4S RNA with oncornavirus RNA. J. Virol. 8, 254-256. Evans, D. B., Brawn, K., Deibel M. R., Jr., Tarpley, W. G., and Sharma, S. K. (1991). A recombinant ribonuclease H domain of HIV-1 reverse transcriptase that is enzymatically active. J. Biol. Chem. 266, 20583-20585.

HIV-1 REVERSE TRANSCRIPTASE

149

Evans, D. B., Fan, N., Swaney, S. M., Tarpley, W. G., and Sharma, S. K. (1994). An active recombinant p15 RNase H domain is functionally distinct from the RNase H domain associated with human immunodeficiency virus type 1 reverse transcriptase. J.Biol. Chem. 269, 21741-21747. Fache, M., Janetzko, A,, Shoeman, R. L., and Krausslich, H. G. (1993).A large deletion in the matrix domain of the human immunodeficiency virus gag gene redirects virus particle assembly from the plasma membrane to the endoplasmic reticulum. J. Virol. 67,4972-4980. Faras, A. J., Garapin, A. C., Levinson, W. E., Bishop, J. M., and Goodman, J. M. (1973). Characterization of the low-molecular-weight RNAs associated with the 70s RNA of b u s sarcoma virus. J. Virol. 12, 334-342. Farmerie, W. G., Loeb, D. D., Casavant, N. C., Hutchison 111, C. A., Edgell, M. A,, and Swanstrom, R. (1987). Expression and processing of the AIDS virus reverse transcriptase in Escherichia coli. Science 236, 305-308. Federoff, 0. Y., Salazar, M., and Reid, B. R. (1993). Structure of a DNA:RNA hybrid duplex. Why RNase H does not cleave pure RNA. J. Mol. Biol. 233, 509-523. Fitzgerald, P. M., McKeever, B. M., van Middlesworth, J. F., Singer, J. P., Heimbach, J . C., et al. (1990). Crystallographic analysis of a complex between human immunodeficiency virus type 1 protease and acetyl-pepstatin a t 2.0-A resolution. J . Biol. Chem. 265, 14209-14219. Fitzgibbon, J. E., Howell, R. M., Haberzettl, C. A., Sperber, S. J., Gocke, D. J., and Dubin, D. T. (1992). Human immunodeficiency virus type 1pol gene mutations which cause decreased susceptibility to 2’,3’-dideoxycytidine. Antimicrob. Agents Chernother. 36, 153-157. Fu, T.-B., and Taylor, J. (1992). When retroviral reverse transcriptase reach the end of their RNA templates. J. Virol. 66, 4271-4278. Fu, W., and Rein, A. (1993). Maturation of dimeric viral RNA of Moloney murine leukemia virus. J. Virol. 67, 5443-5449. Furfine, E. S., and Reardon, J. E. (1991). Reverse transcriptase-RNase H from the human immunodeficiency virus: Relationship of the DNA polymerase and RNA hydrolysis activities. J.Biol. Chem. 266, 406-412. Furman, P. A., Painter, G., Wilson, J. E., Cheng, N., and Hopkins, S. (1991). Substrate inhibition of the human immunodeficiency virus type 1 reverse transcriptase. Proc. Natl. Acad. Sci. U.S.A.88, 6013-6017. Ganem, D., and Varmus, H. E. (1987). The molecular biology of the hepatitis B virus. Annu. Rev. Biochem. 56,. 651-693. Gao, Q., Gu, X., Cameron, J. M., Cammack, N., Boucher, C., and Wainberg, M. A. (1993). The same mutation that encodes low-level HIV-1 resistance to 2’,3’-dideoxycytidine confers high level resistance against the (-1 enantiomer of 2’,3’-dideoxy-3’thiacytidine (3TC).Antimicrob. Agents Chemother. 37, 1390-1392. Geleziunas, R., Arts, E. J., Boulerice, F., Goldman, H., and Wainberg, M. A. (1993). Effect of 3’-azido-3’-deoxythymidine on human immunodeficiency virus type 1 replication in human fetal brain macrophages. Antimicrob. Agents Chemother. 37, 13051319. Ghosh, M., Howard, K. J., Cameron, C. E., Benkovic, S. J., Hughes, S. H., and LeGrice, S. F. J. (1995). Truncating a-helix E’ of p66 human immunodeficiency virus reverse transcriptase modulates RNase H function and impairs DNA strand transfer. J. Biol. Chem. 270, 7068-7076. Gilboa, E., Mitra, S. W., Goff, S. P., and Baltimore, D. (1979). A detailed model of reverse transcription and tests of crucial aspects. Cell (Cambridge, Mass.) 18,93-100. Gopalakrishnan, V., Peliska, J. A,, and Benkovic, S. J. (1992). Human immunodeficiency virus type 1 reverse transcriptase: Spatial and temporal relationship between

150

ERIC J. ARTS AND MARK A. WAINBERG

the polymerase and RNase H activities. Proc. Natl. Acad. Sci. U.S.A. 89, 1076310767. Gottlieb, M. S., Schroff, R., Schanler, H. M., Weisman, J. D., Fan, P. T. et al. (1981). Pneumocystis carinii pneumonia and mucosal candidiasis in previously healthy homosexual men: Evidence of a new acquired cellular immuno-deficiency. N Engl. J. Med. 305, 1426-1431. Gottlinger, H. G., Sodroski, J. G., and Hasseltine, W. A. (1989). Role of capsid precursor processing and myristoylation in morphogenesis and infectivity of human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. U.S.A.86,5781-5785. Gottlinger, H. G., Dorfman, T., Cohen, E. A., and Haseltine, W. A. (1993). upu protein of human immunodeficiency virus type 1 enhances the release of capsids produced by gag gene constructs of widely divergent retroviruses. Proc. Natl. Acad. Sci. U.S.A.90, 7381-7385. Greider, C. W., and Blackburn, E. H. (1985). Identification of a specific telomere terminal transferase activity in Tetrahymena extracts. Cell (Cambridge, Mass.) 43, 405-413. Gu, Z., Gao, Q., Li, X., Parniak, M. A., and Wainberg, M. A. (1992). Novel mutation in the human immunodeficiency virus type 1 reverse transcriptase gene encodes crossresistance to 2',3'-dideoxyinosine and 2',3'-dideoxycytidine. J. Virol. 66, 7128-7135. Gu, Z., Fletcher, R. S., Arts, E. J., Wainberg, M. A., and Parniak, M. A. (1994a). The K65R mutant reverse transcriptase of HIV-1 cross-resistant to 2',3'-dideoxycytidine, 2',3'-dideoxy-3'-thiacytidine,and 2',3'-dideoxyinosine shows reduced sensitivity to dideoxynucleoside triphosphate inhibitors in uitro. J. Biol. Chem. 262, 28118-28122. Gu, Z., Gao, Q., Fang, H., Solomon, H., Parniak, M. A., et al. (1994b). Identification of a mutation a t codon 65 in the IKKK motif of reverse transcriptase that encodes human immunodeficiency virus resistance to 2',3'-dideoxycytidine and 2',3'-dideoxy-3'thiacytidine. Antimicrob. Agents Chemother. 38, 275-281. Guilley, H., Richards, K. E., and Jonard, G. (1983). Observations concerning the discontinuous DNAs of cauliflower mosaic virus. EMBO J. 7, 277-282. Hadzopoulos-Cladaras, M., Felber, B., Cladaras, C., Athanassopoulos, A., Tse, A., et al. (1989). The rev (trslart) protein of human immunodeficiency virus type 1 affects viral mRNA and protein expression via a cis-acting sequence in the enu region. J. Virol. 63, 1265-1274. Hallenberger, S., Bosch, V., Angliker, H., Shaw, E., Klenk, H.-D., and Garten, W.(1992). Inhibition of furin-mediated cleavage activation of HIV-1 glycoprotein gp 160. Nature (London) 360,358-361. Hansen, J., Schulze, T. and Moelling, K. (1987). RNase H activity associated with bacterially expressed reverse transcriptase of human T-cell lymphotropic virus III/lymphadenopathy-associatedvirus. J . Biol. Chem. 262, 12393-12396. Hansen, J., Schulze, T., Muller, W., and Moelling, K. (1988). Identification and characterization of HIV-specific RNase H by monoclonal antibody. EMBO J . 7 , 239-243. Harada, F., Sawyer, R. C., and Dahlberg, J. E. (1975). A primer ribonucleic acid for initiation in in uitro b u s sarcoma virus deoxyribonucleic acid synthesis. Nucleotide sequence and amino acid acceptor activity. J. Biol. Chem. 250, 3487-3497. Haseltine, W. A. (1991) Molecular biology of the human immunodeficiency virus type 1. FASEB J . 5,2349-2360. Henderson, L. E., Bowers, M. A., Sowder 11, R. C., Serabyn, S. A., Johnson, D. G., et al. (1992). Gag proteins of the highly replicative MN strain of human immunodeficiency virus type 1: Postranslational modifications, proteolytic processing, and complete amino acid sequences. J. Virol. 66, 1856-1865. Hizi, A., Tal, R., and Hughes, S. H. (1991). Mutational analysis of the DNA polymerase and ribonuclease H activities of human immunodeficiency type 2 reverse transcriptase expressed in Escherichiu coli. Virology 180, 339-346.

HIV-1 REVERSE TRANSCRIPTASE

151

Hizi, A., Tal, R., Shaharbany, M., Currens, M. J., Boyd, M. R., et al. (1993). Specific inhibition of reverse transcriptase of human immunodeficiency virus type 1 and the chimeric enzymes of human immunodeficiency virus type 1 and type 2 by nonnucleoside inhibitors. Antimicrob. Agents Chemother. 37, 1037-1042. Hostomska, Z., Mathews, D. A,, Davies 11, J. F., Nodes, B. F., and Hostomsky, Z. (1991). Proteolytic release and crystallization of the RNase H domain of human immunodeficiency virus type 1 reverse transcriptase. J. Biol. Chem. 266, 14697-14702. Hostomsky, Z., Hostomska, Z., Hudson, G. O., Moomaw, E. W., and Nodes, B. R. (1991). Reconstitution in vitro of RNase H activity by using purified N-terminal and C-terminal domains of human immunodeficiency virus type 1 reverse transcriptase. Proc. Natl. Acad. Sci. U.S.A. 88, 1148-1152. Hostomsky, Z., Hudson, G. O., Rahmati, S., and Hostomska, Z. (1992). RNase D, a reported new activity associated with HIV-1 reverse transcriptase, display the same cleavage specificity as Escherichia coli RNase 111. Nucleic Acids Res. 20, 5819-5824. Hotta, J., and Loh, P. C. (1987). Enhanced production of a human spumavirus (Retrouiridae) in semi-permissive cell cultures after treatment with 5’-azacytidine. J. Gen. Virol. 68, 1183-1186. Howard, K. J., Frank, D. B., Sim, I. S., and LeGrice S. F. J . (1991). Reconstitution and properties of homologous and chimeric HIV-1/HIV-2 p66/p51 reverse transcriptase. J . Biol. Chem. 266, 23003-23009. Hsieh, J.-C., Zinnen, S., and Modrich, P. (1993). Kinetic mechanism of DNA-dependent DNA polymerase activity of human immunodeficiency virus reverse transcriptase. J. Biol. Chem. 268,24607-24613. Hu, W.-S., and Temin, H. M. (1990). Retroviral recombination and reverse transcription. Science 250, 1227-1233. Huang, Y., Mak, J., Cao, Q., Li, Z., Wainberg, M. A., and Kleiman, L. (1994). Incorporation of excess wild-type and mutant tRNALyd into human immunodeficiency virus type 1. J. Virol. 68, 7676-7683. Huber, H. E., and Richardson, C. C. (1990). Processing of the primer for plus strand DNA synthesis by human immunodeficiency virus type 1 reverse transcriptase. J. Biol. Chem. 265,10565-10573. Huber, H. E., McCoy, J. M., Seehra, J . S., and Richardson, C. C. (1989). Human immunodeficiency virus 1reverse transcriptase: Template binding, processivity, strand displacement synthesis, and template switching. J. Biol. Chem. 264, 4669-4678. Hwang, S. S., Boyle, T. J., Lyerly, H. K., and Cullen, B. R. (1991). Identification of the envelope V3 loop as the primary determinant of cell tropism in HIV-1. Science 253, 71-72. Inouye, S., Hsu, M.-Y., Angle, S., and Inouye, M. (1989). Reverse transcriptase associated with the biosynthesis of the branched RNA-linked msDNA in Myzococcus zanthus. Cell (Cambridge, Mass.) 56, 709-717. Isel, C., Marquet, R., Keith, G., Ehresmann, C., and Ehresmann, B. (1993). Modified nucleotides of tRNALyJ3 modulate primerhemplate loop-loop interaction in the initiation complex of HIV-1 reverse transcription. J . Biol. Chem. 268, 25269-25272. Jacks, T., Townsley, K., Varmus, H. E., and Majors, J. (1987). Two efficient ribosomal frameshifting events are required for synthesis of mouse mammary tumor virus gagrelated polyproteins. Proc. Natl. Acad. Sci. U S A . 84, 4298-4302. Jacks, T., Power, M. D., Masiarz, F. R., Luciw, P. A., Barr, P. J., and Varmus, H. E. (1988). Characterization of ribosomal frameshifting in HIV-1 gag-pol expression. Nature (London)331,280-283. Jacobo-Molina, A., and Arnold, E. (1991). HIV reverse transcriptase-function relationships. Biochemistry 30, 6351-6361. Jacobo-Molina, A., Ding, J., Nanni, R. G., Clark, A. D., Jr., Lu, S. et al. (1993). Crystal

152

ERIC J. ARTS AND MARK A. WAINBERG

structure of human immunodeficiency virus type 1 reverse transcriptase complexed with double-stranded DNA a t 3.0 A resolution shows bent DNA. Proc. Natl. Acad. Sci. U.S.A.90,6320-6324. Jacques, P. S., Wohrl, B. M., Ottman, M., Darlix, J.-L., and LeGrice, F. J. (1994a). Mutating the “primer grip” of p66 HIV-1 reverse transcriptase implicates tryptophan 229 in template-primer utilization. J. Biol. Chem. 269, 2672-26478. Jacques, P. S., Wohrl, B. M., Howard, K. J., and LeGrice, S. F. J. (1994b). Modulation of HIV-1 reverse transcriptase function in “selectively deleted p66/p51 heterodimers. J. Biol. Chem. 269, 1388-1393. Ji, J., Hoffman, J.-S., and Loeb, L. (1994). Mutagenicity and pausing of HIV reverse transcriptase during HIV plus-strand DNA synthesis. Nucleic Acids Res. 22, 47-52. Jiang, M., Mak, J., Wainberg, M. A,, Parniak, M. A., Cohen, E., and Kleiman, L. (1992). Variable tRNA content in HIV-lIIIB,Biochem. Biophys. Res. Commun. 185, 10051015. Jiang, M., Mak, J.,Ladha, A., Cohen, E., Klein, M., Rovinski, B., and Kleiman, L. (1993). Identification of tRNAs incorporated into wild-type and mutant human immunodeficiency virus type 1.J. Virol. 67, 3246-3253. Johnson, M. S., McClure, M. A., Feng, D.-F., Gray, J., and Doolittle, R. F. (1986). Computer analysis of retroviral pol genes: Assignment of enzymatic functions to specific sequences and homologies with nonviral enzymes. Proc. Natl. Acad. Sci. U.S.A. 83, 7646-7652. Jonckheere, H., Taymans, J.-M., Balzarini, J., Velazquez, S., Camarasa, M.J., Desmyter, J., De Clerq, E., and Anne, J. (1994). Resistance of HIV-1 reverse transcriptase against [2’,5‘-bis-0-~tert-butyldimethylsilyl~-3’-spiro-5”-~4‘-amino-1’,2”-oxathiole-2”,2”-dioxide)] (TSAO) derivatives by the muatation Glu-138 to Lys on the p51 subunit. J. Biol. Chem. 269, 26255-26258. Kao, S.-Y., Calman, A. F., Luciw, P. A,, and Peterlin, B. M. (1987). Anti-termination of transcription within the long terminal repeat of HIV-1 by tat gene production. Nature (London) 330,489-493. Kashanchi, F., Piras, G., Radonovich, M. F., Duvall, J. F., Fattacy, A., et al. (1994). Direct interactions of huamn TFIID with HIV-1 transactivator tat. Nature (London) 367, 295-299. Kati, W. M., Johnson, K. A., Jerva, L. F., and Anderson, K. S. (1992). Mechanism and fidelity of HIV reverse transcriptase. J. Biol. Chem. 267, 25988-25997. Kawa, S., Kumar, A., Smith, J . S., Becerra, S. P., and Beard, W. A. (1993). Expression and purification of the HIV-1 reverse transcriptase using the baculovirus expression vector system. Protein Expression Purif. 4, 298-303. Khan, R., and Giedroc, D. P. (1992). Recombinant human immunodeficiency virus type 1 nucleocapsid (NCp7) protein unwinds tRNA. J. Biol. Chem. 267, 6689-6695. Klarmann, G. J., Schauber, C. A,, and Preston, B. D. (1993). Template-directed pausing of DNA synthesis by HIV-1 reverse transcriptase during polymerization of HIV-1 sequences in vitro. J. Biol. Chem. 268, 9793-9802. Klaver, B., and Berkhout, B. (1994). Premature strand transfer by the HIV-1 reverse transcriptase during strong-stop DNA synthesis. Nucleic Acids Res. 22, 137-144. Kohl, N. E., Emini, E. A., Schleif, W. A., Davis, L. J., and Heimbach, J. C. (1988). Active human immunodeficiency virus protease is required for viral infectivity. Proc. Natl. Acad. Sci. U.S.A.85, 4686-4690. Kohlstaedt, L. A., and Steitz, T. A. (1992). Reverse transcriptase of human immunodeficiency virus can use either human tRNALyS3or Escherichia coli tRNAGLa as a primer in a n in vitro primer-utilization assay. Proc. Natl. Acad. Sci. U.S.A.89,96529656.

HIV-1 REVERSE TRANSCRIPTASE

153

Kohlstaedt, L. A., Wang, J., Friedman, J. M., Rice, P. A., and Steitz, T. A. (1992). Crystal structure at 3.5 A resolution of HIV-1 reverse transcriptase complexed with a n inhibitor. Science 256, 1783-1790. Kroup, R. A., Merluzzi, V. J., Hargrave, K. D., Adams, J., and Grozinger, K. (1991). Inhibition of human immunodeficiency virus type 1 (HIV-1) replication by the dipyridodiazepinone BI-RG-587. J. Infect. Dis. 163, 966-970. Krug, M. S., and Berger, S. L. (1989). Ribonuclease H activities associated with reverse transcriptases are endonucleases. Proc. Natl. Acad. Sci. U.S.A.86, 3539-3543. Kruhoffer, M., Urbanke, C., and Grosse, F. (1993). l k o step binding of HIV-1 reverse transcriptase to nucleic acid substrates. Nucleic Acids Res. 21, 3943-3949. Kumar, A., Kim, H.-R., Sobol, R. W., Becerra, S. P., Lee, B.J., et al. (1993). Mapping of nucleic acid binding in proteolytic domains of HIV-1 reverse transcriptase. Biochemistry 32,7466-7474. Kunz, B. A. (1988). Mutagenesis and deoxyribonucleotide pool imbalances. Mufat. Res. 200, 133-147. Lacey, S. F., and Larder, B. A. (1994). Novel mutation (V75T) in human immunodeficiency virus type 1 reverse transcriptase confers resistance to 2’,3’-didehydro-2’,3’dideoxythymidine. Antimicrob. Agents Chemother. 38, 1428-1432. Larder, B. A., and Kemp, S. D. (1989). Multiple mutations in HIV-1 reverse transcriptase confer high-level resistance to zidovudine (AZT). Science 246, 1155-1158. Larder, B., Purifoy, D., Powell, K., and Darby, G. (1987a). AIDS virus reverse transcriptase defined by high level expression in Escherichia coli. EMBO J . 6, 3133-3137. Larder, B., Purifoy, D. J. M., Powell, K., and Darby, G. (198713). Site-specific mutagenesis of AIDS virus reverse transcriptase. Nature (London) 327,716-717. Larder, B. A., Darby, G., and Richman, D. D. (1989a). HIV with reduced sensitivity to zidovudine (AZT) isolated during prolonged therapy. Science 243, 1731-1734. Larder, B. A., Kemp, S. D., and Purifoy, D. J . M. (198913). Infectious potential of human immunodeficiency virus type 1 reverse transcriptase mutants with altered inhibitor sensitivity. Proc. Natl. Acad. Sci. U S A . 86, 4803-4807. Laspia, M. F., Rice, A. P., and Mathews, M. B. (1989). HIV-1 tat protein increases transcriptional initiation and stabilizes elongation. Cell (Cambridge, Mass.) 59,283292. Laughrea, M., and Jette, L. (1994). A 19-nucleotide sequence upstream of the 5’major splice donor is part of the dimerization domain of human immunodeficiency virus 1 genomic RNA. Biochemistry 33, 13464-13474. Lavallee, C., Yao, X. J., Ladha, A., Gottlinger, H., Haseltine, W. A., and Cohen, E. (1994). Requirements of the Pr55ew precursor for incorporation of the upr product into human immunodeficiency virus type 1 viral particles. J. Virol. 68, 1926-1934. Lederer, H., Schatz, O., May, R., Crespi, H., Darlix, J.-L., et al. (1992). Domain structure of the human immunodeficiency virus reverse transcriptase. EMBO J . 11, 11311139. LeGrice, S. F. J., and Gruninger-Leitch, F. (1990). Rapid purification of homodimer and heterodimer HIV-1 reverse transcriptase by metal chelate affinity chromatography. Eur. J . Biochem. 187,307-314. LeGrice, S . F. J., Beuck, V., and Mous, J. (1987). Expression of biologically active human T-cell lymphotropic virus type I11 reverse transcriptase in Bacillus subtilis. Gene 55, 95-103. LeGrice, S. F. J., Mills, J., and Mous, J. (1988). Active site mutagenesis of the AIDS virus proteaase and its alleviation by trans complementation. EMBO J . 7, 46864690. LeGrice, S. F. J., Mills, J., Ette, R., and Mous, J. (1989). Comparison of the human

154

ERIC J. ARTS AND MARK A. WAINBERG

immunodeficiency virus type 1and 2 proteases by hybrid gene construction and transcomplementation. J.Biol. Chem. 264, 14902-14908. LeGrice, S. F. J., Naas, T., Wohlgensinger, B., and Schatz, 0. (1991a). Subunit-selective mutagenesis indicates minimal polymerase activity in heterodimer-associated p51 HIV-1 reverse transcriptase. EMBO J. 10,3905-3911. LeGrice, S. F. J., Panin, M., Kalayjian, R. C., Richter, N. J., Keith, G., Darlix, J. L., and Payne, S. L. (1991b). Purification and characterization of recombinant equine infectious anemia virus reverse transcriptase. J. Virol. 65, 7004-7007. LeGrice, S. F. J., Cameron, C. E., and Benkovic, S. (1994). Purification of retroviral transcriptase. In Methods in Enzymology (J.L. Campbell, ed.), in press. Academic Press, New York. Leis, J., and Hurwitz, J. (1972). RNA-dependent DNA polymerase activity of RNA tumor viruses. 11. Directing influence of RNA in the reaction. J. Virol. 9, 130-142. Leis, J., Aiyar, A., and Cobrinik, D. (1993). Regulation of initiation of reverse transcription of retroviruses. In “Reverse Transcriptase” (A. M. Skalka and S. P. Goff, eds.), pp. 33-47. Cold Spring Harbor Lboratory, Cold Spring Harbor, New York. Levin, J. G., and Seidman, J. G. (1979). Selective packaging of host tRNAs by murine leukemia virus particles does not require genomic RNA. J. Virol. 29, 328-335. Levin, J. G., Hatfield, D. L., Oroszlan, S., and Rein, A. (1993). Mechanism of translational suppression used in the biosynthesis of reverse transcriptase. In “Reverse Transcriptase” (A. M. Skalka and S. P. Goff, eds.), pp. 5-31. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Levinson, W., Bishop, J. M., Quintrell, N., and Jackson, J. (1970). Presence of DNA in b u s sarcoma virus. Nature (London) 227, 1023-1025. Lewis, P., Hensel, M., and Emerman, M. (1992). Human immunodeficiency virus infection of cells arrested in the cell cycle. EMBO J. 8, 3053-3058. Li, X., Mak, J., Arts, E. J., Kleiman, L., Wainberg, M. A., and Parniak, M. A. (1994). Effects of alterations of primer binding site sequences on human immunodeficiency virus type 1replication. J. Virol. 68,6198-6206. Li, X., Gu, Z., Geleziunas, R., Kleiman, L., Wainberg, M. A., and Parniak, M. A. (1993). Expression, purification, and RNA-binding properties of HIV-1 p l S nucleocapsid ~ protein. Protein Expression Purif. 4, 304-311. Lightfoote, M. M., Coligan, J. E., Folks, T. M., Fauci, A. S., Martin, M. A., and Venkatesan, S. (1986). Structural characterization of reverse transcriptase and endonuclease polypeptides of the acquired immunodeficiency syndrome virus. J. Virol. 60, 771-775. Lori, F., Di Marzo Veronese, F., DeVico, A. L., Lusso, P., Reitz, M. S., and Gallo, R. C. (1992). Viral DNA carried by human immunodeficiency virus type 1virions. J. Virol. 66,5067-5074. Lowe, D. M., Aitkin, A., Bradley, C., Darby, G. K., and Larder, B. A. (1988). HIV-1 reverse transcriptase: Crystallization and analysis of domain structure by limited proteolysis. Biochemistry 27, 8884-8889. Lowe, D. M., Parmor, V., Kemp, S. D., and Larder, B. A. (1991). Mutational analysis of two conserved sequence motifs in HIV-1 reverse transcriptase. FEBS Lett. 282,231234. Loya, S., Tal, R., Hughes, S. H., and Hizi, A. (1992). The effects of cysteine mutations on the catalytic activities of the reverse transcriptase of human immunodeficiency virus type-1. J. Biol. Chem. 267, 13879-13883. Luo, G. X., and Taylor, J. (1990). Template switching by reverse transcriptase during DNA synthesis. J. Virol. 64, 4321-4328. Luo, G., Sharmeen, L., and Taylor, J. (1990). Specificities involved in the initiation of retroviral plus-strand DNA. J. Virol. 64, 592-597.

HIV-1 REVERSE TRANSCRIPTASE

155

McDougal, J. J., Kennedy, M. S., Sligh, J. M., Cort, S. P., Mawle, A., and Nicholson, J. K. A. (1986). Binding of HTLV-III/LAV to T4+ cells by a complex of a llOK viral protein and the T4 molecule. Science 231, 382-385. Maddon, P. J., Dalgleish, A. G., McDougal, J. S., Clapham, P. R., Weiss, R. A., and Axel, R. (1986). The T4 gene encodes the AIDS virus receptor and is expressed in the immune system and the brain. Cell (Cambridge, Mass.) 47,333-348. Majumdar, C., Abbotts, J., Broder, S., and Wilson, S. H. (1988). Studies on the mechanism of human immunodeficiency virus reverse transcriptase: Steady-state kinetics, processivity, and polynucleotide inhibition. J. Biol. Chem. 263, 15657-15665. Mak, J., Jiang, M., Wainberg, M. A., Hammarskjold, M.-L., Rekosh, D., and Kleiman, L. (1994). Role of P r l 6 o B ~ - p oin~ mediating the selective incorporation of tRNALys into human immunodeficiency virus type 1 particles. J. Virol. 68, 2065-2072. Malim, M. H., Hauber, J.,Le, S.-Y., Maizel, J. V., and Cullen, B. R. (1989). The HIV-1 reu trans-activator acts through a structured target sequence to activate nuclear export of unspliced viral mRNA. Nature (London) 338, 254-257. Martin, J. L., Wilson, J. E., Furfine, E. S., Hopkins, S. E., and Furman, P. A. (1993a). Biochemical analysis of human immunodeficiency virus-1 reverse transcriptase containing a mutation a t position lysine 263. J . Biol. Chem. 268, 2565-2570. Martin, J. L., Wilson, J. E., Haynes, R. L., and Furman, P. A. (199313). Mechanism of resistance of human immunodeficiency virus type 1 to 2',3'-dideoxyinosine. Proc. Natl. Acad. Sci. U.S.A. 90, 6135-6139. Masur, H., Michelis, M. A., Greene, J. B., Onovato, I., Van de Stowe, R. A., et al. (1981). An outbreak of community-acquired pneumocystis carinii pneumonia: Initial manifestation of cellular immune dysfunction. N . Engl. J. Med. 305, 1431-1438. Meric, C., and Goff, S. (1989). Characterization of Moloney murine leukemia virus mutants with single amino acid substitutions in the Cys-His box of the NC protein. J. Virol. 63, 1558-1568. Metzger, W., Herman, T., Schatz, O., LeGrice, S. F. J., and Hermann, H. (1993). Hydroxyl radical footprint analysis of human immunodeficiency virus reverse transcriptasetemplate/primer complexes. Proc. Natl. Acad. Sci. U.S.A. 90,5909-5913. Mitsuya, H., Weinhold, K. J., Furman, P. A., St. Clair, M. H., Lehrman, S. N., et al. (1985). 3'-Azido-3'-deoxythymidine(BWA509V):An antiviral agent that inhibits the infectivity and cytopathic effects of human T-lymphotropic virus type III/lymphadenopathy-associated virus in vitro. Proc. Natl. Acad. Sci.U S A . 82,70967100. Mizrahi, V., Lazarus, G. M., Miles, L. M., Myers, C. A., and Debouck, C. (1989). Recombinant HIV-1 reverse transcriptase: Purification, primary structure, and polymerase/ribonuclease H activities. Arch. Biochem. Biophys. 273, 347-358. Mizrahi, V., Brooksbank, R. L., and Nkabinde, N. C. (1994). Mutagenesis of the conserved aspartic acid 443, glutamic acid 478, asparagine 494, and aspartic acid 498 residues in the ribonuclease H domain of p66/p51 human immunodeficiency virus type 1 reverse transcriptase. J.Biol. Chem. 269, 19245-19249. Molling, K., Bolognesi, D. P., Bauser, H., Bussen, W., Plassmann, H. W., and Hausen, P. (1971). Association of viral reverse transcriptase with a n enzyme degrading the RNA moiety of RNA-DNA hybrids. Nature (London) 234, 240-243. Mous, J., Heimer, E. P., and LeGrice, S. F. J. (1988). Processing protease and reverse transcriptase from human immunodeficiency type 1polyprotein in Escherichia coli. J. Virol. 62,1433-1436. Murphy, J. E., and Goff, S. P. (1989). Construction and analysis of deletion mutations in the U5 region of Moloney murine leukemia virus: Effects on RNA packaging and reverse transcription. J. Virol. 63, 319-327. Nagashunmugam, T., Velpandi, A., Goldsmith, C. S., Zaki, S. R., Kalyanaraman, V. S.,

156

ERIC J. ARTS AND MARK A. WAINBERG

and Srinivasan, A. (1992). Mutation in the primer binding site of the type 1 human immunodeficiency virus genome affects virus production and infectivity. Proc. Nutl. Acud. Sci. U.S.A.89, 4114-4118. Nanni, R. G., Ding, J., Jacobo-Molina, A,, Hughes, S. H., and Arnold, E. (1993). Review of HIV-1 reverse transcriptase three-dimensional structure: Implications for drug design. Perspect. Drug Discovery Des. 1, 129-150. Ollis, D., Brick, P., Hamlin, R., Xuong, N. G., and Steitz, T. A. (1985). Structure of large fragment of Escherichiu coli DNA polymerase I complexed with dTMP. Nature (London 313, 762-766. Omer, C. A., Resnick, R., and Faras, A. J. (1984). Evidence for involvement of a n RNA primer in initiation of strong-stop plus DNA synthesis during reverse transcription in vitro. J . Virol. 50, 465-470. Ouhammouch, M., and Brody, E. N. (1992). Temperature-dependent template switching during in vitro cDNA synthesis by the AMV-reverse transcriptase. Nucleic Acids Res. 20,5443-5450. Oyama, F. R., Kikuchi, R., Crouch, R. J., and Uchida, T. (1989). Intrinsic properties of reverse transcriptase in reverse transcription. Associated RNase H is essentially regarded as a n endonuclease. J. Biol. Chem. 264, 18808-18817. Paillart, J.-C, Marquet, R., Skripkin, E., Ehresmann, B., and Ehresmann, C. (1994). Mutational analysis of the bipartite dimer linkage structure of human immunodeficiency virus type 1 genomic RNA. J . Biol. Chem. 269, 27486-27493. Painter, G. R., Wright, L. L., Hopkins, S., and Furman, P. A. (1991). Initial binding of 2‘-deoxynucleoside 5’4riphosphates to human immunodeficiency virus type 1reverse transcriptase. J. Biol. Chem. 266, 19362-19368. Painter, G. R., Aulabaugh, A. E., and Andrews, C. W. (1993). A comparison of the conformations of the 5’-triphosphates of zidovudine (AZT) and thymidine bound to HIV-1 reverse transcriptase. Biochem. Biophys. Res. Cornmun. 191, 1166-1171. Panet, A., Haseltine, W. A., Baltimore, D., Peters, G., Harada, F., and Dahlberg, J. E. (1975). Specific binding of tRNATv to avian myeloblastosis virus RNA-dependent DNA polymerase. Proc. Nutl. Acud. Sci. U.S.A. 72, 2535-2539. Paganiban, A. T., and Fiore, D. (1988). Ordered interstrand and intrastrand DNA transfer during reverse transcription. Science 241, 1064-1069. Park, J., and Morrow, C. D. (1992). The nonmyristylated P r 1 6 0 8 ~ - polyprotein ~0~ of human immunodeficiency virus type 1interacts with Pr55e~gand is incorporated into viruslike particles. J. Virol. 66, 6304-6314. Parkin, N. T., Chamorro, M., and Varmus, H. E. (1992). Human immunodeficiency virus type 1gag-pol frameshifting is dependent on downstream messenger RNA secondary structure: Demonstration by expression in uiuo. J. Virol. 66, 5147-5151. Peliska, J. A., and Benkovic, S. J. (1992). Mechanism of DNA strand transfer reactions catalyzed by HIV-1 reverse transcriptase. Science 258, 1112-1118. Peliska, J. A., Balasubranmanian, S., Giedroc, D. P., and Benkovic, S. J. (1994). Recombinant HIV-1 nucleocapsid protein accelerates HIV-1 reverse transcriptase catalyzed DNA strand transfer reactions and modulates RNase H activity. Biochemistry 33, 13817-13823. Pelletier, H., Sawaya, M. R., Kumar, A., Wilson, S. H., and Kraut, J. (1994).Structures of ternary complexes of rat DNA polymerase p, a DNA template-primer, and ddCTP. Science 264, 1891-1903. Peng, C., Ho, B. K., Chang, T. W., and Chang, N. T. (1989). Role of human immunodeficiency virus type 1-specificprotease in core protein maturation and viral infectivity. J. Virol. 63, 2550-2556. Peng, C., Cheng, N. T., and Chang, T. W. (1991). Identification and characterization of

HIV-1 REVERSE TRANSCRIPTASE

157

human immunodeficiency virus type 1 gag-pol fusion protein in transfected mammalian cells. J. Virol. 63, 2550-2556. Peters, G. G., and Hu, J. (1980). Reverse transcriptase as the major determinant for selective packaging of tRNAs into avian sarcoma virus particles. J. Virol. 26, 962700. Peters, G., Harada, F., Dahlberg, J . E., Panet, A., Haseltine, W. A., and Baltimore, D. (1977). Low-molecular-weight RNAs of Moloney murine leukemia virus: Identification of the primer for RNA-directed DNA synthesis. J. Virol. 21, 1031-1041. Pfeiffer, P., and Hohn, T.(1983). Involvement of reverse transcription in the replication of cauliflower mosaic virus: A detailed model and test of some aspects. Cell 33 (Cambridge, Mass.) 33, 781-789. Phear, G., Nalbantoglu, J., and Meuth, M. (1987). Next nucleotide effects in mutation driven by DNA precursor pool imbalances a t the aprt locus of Chinese hamster ovary cells. Proc. Natl. Acad. Sci.U.S.A.84, 3465-3469. Polesky, A. H., Steitz, T. A., Grindle, N. D. F., and Joyce, C. W. (1990). Identification of residues critical for the polymerase activity of the Klenow fragment of DNA polymerase I from Escherichia coli. J. Biol Chem. 265, 14579-14591. Polesky, A. H., Dahlberg, M., Benkovic, S. J., Grindley, N. D., and Joyce, C. W. (1992). Side chains involved in catalysis of the polymerase reaction of DNA polymerase I from Escherichia coli. J. Biol Chem. 267, 8417-8428. Pomerantz, R. J., Trono, D., Feinberg, M. B., and Baltimore, D. (1990). Cells nonproductively infected with HIV-1 exhibit a n aberrant pattern of viral RNA expression: A molecular model for latency. Cell (Cambridge, Mass.) 61, 1271-1276. Popovic, M., Sarngadharan, M. G., Read, E., and Gallo, R. C. (1984). Detection, isolation, and continuous production of cytopathic retroviruses (HTLV-111) from patients with AIDS and pre-AIDS. Science 224,497-500. Post, K., Guo, J., Kalman, E., Uchida, T., Crouch, R. J., and Levin, J. G. (1993). A large deletion in the connection subdomain of murine leukemia virus reverse transcriptase or replacement of the RNase H domain with Escherichia coli RNase H results in alterred polymerase and RNase H activities. Biochemistry 32, 5508-5517. Powers, R., Clore, G. M., Stahl, S. J., Wingtield, P. T., and Gronenborn, A. (1992). Analysis of the backbone dynamics of the ribonuclease H domain of the human immunodeficiency virus reverse transcriptase using 16N relaxation measurements. Biochemistry 31,9150-9157. Prasad, V. R., and Goff, S. P. (1989a). A novel in situ colony screening method to detect immunodeficiency virus reverse transcriptase activity in bacteria. J. Biol. Chem. 264, 16689-16693. Prasad, V. R., and Goff, S. P. (1989b). Linker insertion mutagensis of the human immunodeficiency virus reverse transcriptase expressed in bacteria: Definition of the minimal polymerase domain. Proc. Natl. Acad. Sci.U S A . 86, 3104-3108. Preston, B. D., Poiesz, B. J., and Loeb, L. A. (1988). Fidelity of HIV-1 reverse transcriptase. Science 242, 1168-1171. Pullen, K. A,, Rattray, A. J., and Champoux, J . J. (1993). The sequence features important for plus strand priming by human immunodeficiency virus type 1 reverse transcriptase. J. Biol. Chem. 268, 6221-6227. Ratner, L., Haseltine, W., Patarca, R., Litvak, K. J., Starcich, B., et al. (1985). Complete nucleotide sequence of the AIDS virus, HTLV-111. Nature (London) 313, 277-284. Reardon, J. E. (1993). Human immunodeficiency virus reverse transcriptase: A kinetic analysis of RNA-dependent and DNA-dependent DNA polymerization. J.Biol. Chem. 268, 8743-8751. Resnick, R., Omer, C. A., and Faras, A. J. (1984). Involvement of retrovirus reverse

158

ERIC J. ARTS AND MARK A. WAINBERG

transcriptase-associated RNase H in the initiation of strongstop (+) DNA synthesis and the generation of the long terminal repeat. J . Virol. 61, 813-821. Restle, T., Muller, B., and Goody, R. S. (1990). Dimerization of human immunodeficiency virus type 1 reverse transcriptase. J. Biol. Chem. 266, 8986-8988. Rhim, H., Park, J.,and Morrow, C. (1991). Deletions in the tRNALyd primer-binding site of human immunodeficiency virus type 1 identified essential regions for reverse transcription. J. Virol. 66,4555-4564. Ricchetti, M.; and Buc, H. (1990). Reverse transcriptase and genomic variability: The accuracy of DNA replication is enzyme specific and sequence dependent. EMBO J. 9, 1583-1593. Richter-Cook, N. J., Howard, K. J., Cirino, N. J., Wohrl, B. M., and LeGrice, S. F. J. (1992). Interaction of tRNALyd with multiple forms of human immunodeficiency virus reverse transcriptase. J. Biol. Chem. 267, 15952-15957. Robert, D., Sallafranque-Andreola, M.-L., Bordier, B., Sarih-Cottin, L., Tarrago-Litvak, L., Graves, P. V., Barr, P. J., Fournier, M., and Litvak, S. (1990). Interactions with tRNALys induce important structural changes in human immunodeficiency virus reverse transcription. FEBS Lett. 277, 239-242. Roberts, J. D., Bebenek, K., and Kunkel, T. A. (1988).The accuracy of reverse transcriptase from HIV-1. Science 242, 1171-1173. Roe, T.-Y., Reynolds, T. C., Yu, G., and Brown, P. 0. (1993). Integration of murine leukemia virus DNA depends on mitosis. EMBO J. 12, 2099-2108. Roth, M., Tanese, N., and Goff, S. P. (1985). Purification and characterization of murine retroviral reverse transcriptase expressed in E . coli. J. Biol. Chem. 290, 93269335. Roth, M. J., Schwartzberg, P. L., and Goff, S. P. (1989). Structure of the termini of DNA intermediates in the integration of retroviral DNA Dependence on IN function and terminal DNA sequence. Cell (Cambridge, Mass.) 58,47-54. St. Clair, M. H., Martin, J. L., Tudor-Williams, G., Bach, M. L., Vavro, C. L., King, D. M., Kellam, P., Kemp, S. D., and Larder, B. A. (1991). Resistance to ddI and sensitivity to AZT induced by a mutation in HIV-1 reverse transcriptase. Science 263, 1557-1559. Sakaguchi, K., Zambrano, N., Baldwin, E. T., Shapiro, B. A., Erickson, J. W., et al. (1993). Identification of a binding site for the human immunodeficiency virus type 1 nucleocapsid protein. Proc. Natl. Acad. Sci.U.S.A.90, 5219-5223. Sakai, H., Shibata, R., Sakuragi, J., Sakuragi, S.,Kawamura, M., and Adachi, A. (1993). Cell-dependent requirement of human immunodeficiency virus type 1 uif protein for maturation of virus particles. J. Virol. 67, 1663-1666. Sakai, K., Ma, X. Y., Gordienko, I., and Volsky, D. J. (1991). Recombinational analysis of a natural noncytopathic human immunodeficiency virus type 1 (HIV-1) isolate: Role of the uif gene in HIV-1 infection kinetics and cytopathicity. J. Virol. 66, 5765-5773. Sarih-Cottin, L., Bordier, B., Musier-Forsyth, D., Andreola, A.-L., Barr, P. J., and Litvak, S. (1992). Preferential interaction of human immunodeficiency virus reverse transcriptase with two regions of [ro,er tRMALY6 as evidenced by footprinting studies and inhibition with synthetic oligoribonucleotides. J. Mol. Biol. 226, 1-6. Sawaya, M. R., Pelletier, H., Kumar, A,, Wilson, S. H., and Kraut, J. (1994). Crystal structure of rat DNA polymerase p: Evidence for a common polymerase mechanism. Science 284,1930-1935. Schatz, O., Cromme, F. V., Gruninger-Leitch, F., and LeGrice, S. F. J. (1989). Point mutations in conserved amino acid residues within the C-terminal domain of HIV-1 reverse transcriptase specifically repress RNase H function. FEBS Lett. 267, 311314. Schatz, O., Mous, J., and LeGrice, S. F. J. (1990). HIV-1 RT-associated ribonuclease H displays both endonuclease and 3’-5’ exonuclease activity. EMBO J. 9, 1171-1176.

HIV-1 REVERSE TRANSCRIF’TASE

159

Schinazi, R. F., Lloyd, R. M., Nguyen, M.-H., Cannon, D. L., McMillan, A., et al. (1993). Characterization of human immunodeficiency viruses resistant to oxathiolanecytosine nucleosides. Antimicrob. Agents Chemother. 37, 875-881. Schinnich, T. M., Lerner, R. A., and Sutcliffe, J. G. (1981). Nucleotide sequence of Moloney murine leukemia virus. Nature (London) 293, 543-548. Schneider, J., Kaaden, O., Copeland, T. D., Oroszlan, S., and Hunsmann, G. (1986). Shedding and interspecies type seroreactivity of the envelope gp 120 of the human immunodeficiency virus. J. Gen. Virol. 67, 25331-2538. Schultz, S. C., Schields, G. C., and Steitz, T. A. (1991). Crystal structure of a CAP-DNA complex: The DNA is bent by 90 degrees. Science 253,1001-1007. Selsing, E., Wells, R. D., Alden, C. J., and Arnott, S. (1979). Bent DNA: Visualization of a base-paired and stacked A-B conformation junction. J.Biol. Chem. 254, 5417-5422. Shaharabany, M., and Hizi, A. (1992). The catalytic functions of chimeric reverse transcriptase of human immunodeficiency viruses type 1and 2. J.Biol. Chem. 267,36743678. Sharma, S. K., Fan, N., and Evans, D. B. (1994). Human immunodeficiency virus type 1 (HIV-1) recombinant reverse transcriptase. Asymmetry in p66 subunits of the ~ 6 6 1 ~ homodimer. 66 FEBS Lett. 343, 125-130. Sharp, P. A., and Mariniak, R. A. (1989). HIV TAR: An RNA enhancer? Cell (Cambridge, Mass.) 59, 229-230. Shih, C. K., Rose, J. M., Hansen, G. L., Wu, J. C., Bacolla, A., and Griffin, J. A. (1991). Chimeric human immunodeficiency virus type l/type 2 reverse transcriptase displays altered sensitivity to nonnucleoside analog inhibitors. Proc. Natl. Acad. Sci. U S A . 88.9878-9882. Skalka, A. M., and Goff, S. P., eds. (1993). “Reverse Transcriptase.” Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Smerdon, S. J., Jager, J., Wang, J., Kohlstaedt, L. A., Chirino, A. J., Friedman, J. M., Rice, P. A., and Steitz, T. A. (1994). Structure of the binding site for nonnucleoside inhibitors of the reverse transcriptase of human immunodeficiency virus type 1.Proc. Natl. Acad. Sci. U S A . 91, 3911-3915. Smith, A. J., Srinivasakumar, N., Hammarskjold, M.-L., and Rekosh, D. (1993). Requirements for incorporation of Prl6Oew-po’ from human immunodeficiency virus type 1 into virus-like particles. J. Virol. 67, 2266-2275. Smith, J. K., Cywinski, A,, and Taylor, J. M. (1984a). Initiation of plus-strand DNA synthesis during reverse transcription of an avian retrovirus genome. J. Virol. 49, 200-204. Smith, J. K., Cywinski, A., and Taylor, J. M. (198413). Specificity of initiation of plusstrand DNA by Rous sarcoma virus. J. Virol. 52, 314-319. Smith, J. S., and Roth, M. J. (1992). Specificity of human immunodeficiency virus-1 reverse transcriptase-associated ribonuclease H in removal of the minus-strand primer, tRNALyd. J. Bwl. Chem. 267,15071-15079. Smith, J. S., and Roth, M. J. (1993). Purification and characterization of an active human immunodeficiency virus type 1RNase H domain. J. Virol. 67,4037-4049. Smith, J. S., Gritsman, K., and Roth, M. J. (1994). Contributions of DNA polymerase subdomains to the RNase H activity of human immunodeficiency virus type 1reverse transcriptase. J.Biol. Chem. 68, 5721-5729. Sobol, R. W., Suhadolnik, R. J., Kumar, A., Lee, B. J., Hatfield, D. L., and Wilson, S. H. (1991). Localization of a polynucleotide binding region in the HIV-1 reverse transcriptase: Implications for primer binding. Biochemistry 30, 10623-10631. Sodroski, J., Rosen, C. A., and Haseltine, W. A. (1984). Trans-acting transcriptional activation of the long terminal repeat of human T lymphotropic viruses in infected cells. Science 225, 381-421.

160

ERIC J. ARTS AND MARK A. WAINBERG

Sodroski, J., Goh, W. C., Rosen, C., Dayton, A., Terwilliger, E., and Haseltine. (1986). A second post-transcriptional trans-activator gene required for HTLV-I11 replication. Nature (London) 321, 412-417. Soltis, D. A., and Skalka, A. M. (1988). The (I and p chains of avian retrovirus reverse transcriptase independently expressed in Escherzchia coli: Characterization of enzymatic activities. Proc. Natl. Acad. Sci. U S A . 85, 3372-3376. Sorge, J., and Hughes, S. H. (1982). Polypurine tract adjacent to the U3 region of the Rous sarcoma virus genome provides a cis-acting function. J. Virol. 43, 482-488. South, T. L., Blake, P. R., Sowder 111, R. C., Arthur, L. O., Henderson, L. E., and Summers, M. F. (1990). The nucleocapsid protein isolated from HIV-1 particles binds zinc and forms retroviral-type zinc fingers. Biochemistry 29, 7786-7789. Starnes, M. C., and Cheng, Y.-C. (1989). Human immunodeficiency virus reverse transcriptase-associated RNase H activity. J. Biol. Chem. 264, 7073-7077. Stein, B. S., Gouda, S. D., Lifson, J. D., Penhallow, R. C., Bensch, K. G., and Engleman, E. G. (1987). pH- Independent HIV entry into CD4-positive T cells via virus envelope fusion to the plasma membrane. Cell (Cambridge, Mass.) 49, 659-668. Stevenson, M., Stanwick, T. L., Dempsey, M. P., and Lamonica, C. A. (1990). HIV-1 replication is controlled a t the level of T cell activation and proviral integration. EMBO J. 9, 1551-1560. Streisinger, G., Okada, Y., Emrich, J., Newton, J., Tsugita, A., et al. (1966). Frameshift mutations and the genetic code. Cold Spring Harbor Symp. Quant. Biol. 31, 77-84. Summers, J., and Mason, W. S. (1982). Replication of the genome of an hepatitis-B like virus by reverse transcription of a n RNA intermediate. Cell (Cambridge, Mass.) 29, 403-415. Susman, J. L., and Kim, S.-H. (1976).Three-dimensional structure of a transfer RNA in two crystal forms. Science 192, 853-858. Swanstrom, R., Bishop, J. M., and Varmus, H. E. (1982). Structure of a replication intermediate in the synthesis of Rous sarcoma virus DNA in uiuo. J. Virol. 42, 337341. Takeuchi, Y., Nagumo, T., and Hoshino, H. (1988). Low fidelity of cell-free DNA synthesis by reverse transcriptase of human immunodeficiency virus. J. Virol. 62, 39003902. Tanese, N., Sodroski, J., Hazeltine, W. A., and Goff, S. P. (1986). Expression of reverse transcriptase activity of human T-lymphotropic virus type I11 (HTLV-III/LAV)in Escherichia coli. J. Virol. 59, 743-745. Taylor, J. M., and Illmensee, R. (1975). Site on the RNA of a n avian sarcoma virus at which primer is bound. J. Virol. 16, 553-558. Telesnitsky, A., and Goff, S. P. (1993a). Strong-stop strand transfer during reverse transcription. In “Reverse Transcriptase” (A. M. Skalka and S. P. Goff, eds.), pp. 49-83. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Telesnitsky, A., and Goff, S. P. (1993b). RNase H domain mutations affect the interaction between Moloney murine leukemia virus reverse transcriptase and its primertemplate. Proc. Natl. Acad. Sci. U S A . 90, 1276-1280. Temin, H. M. (1985). Reverse transcription in the eukaryotic genome: Retroviruses, pararetroviruses, retrotransposons, and retrotranscripts. Mol. Biol. Euol. 2,455-468. Temin, H., and Mizutani, S. (1970). RNA-directed DNA polymerase in virions of Rous sarcoma virus. Nature (London) 226, 1211-1213. Threadgill, D. S., Steagall, W. K., Flaherty, N. T., Fuller, F. J., Perry, S. T., et al. (1993). Characterization of equine infectious anemia virus UTPase: Growth properties of a dUTPase-deficient mutant. J. Virol. 67, 2592-2600. Tisdale, M., Erlt, P., Larder, B. A., Purifoy, D. J. M., Darby, G., and Powell, K. (1988).

HIV-1 REVERSE TRANSCRIPTASE

161

Characterization of human immunodeficiency virus type 1 reverse transcriptase by using monoclonal antibodies: Role of the C-terminus in antibody reactivity and enzyme function. J. Virol. 62, 3662-3667. Tisdale, M., Schulze, T., Larder, B. A., and Moelling, K. (1991). Mutations within the RNase H domain of human immunodeficiency virus type 1 reverse transcriptase abolish virus infectivity. J. Gen, Virol. 72, 59-66. Trono, P. (1992). Partial reverse transcription in virions from human immunodeficiency and murine leukemia virus. J. Virol. 66, 4893-4900. Tsuchihashi, Z., and Brown, P. 0. (1994). DNA strand exchange and selective DNA annealing promoted by the human immunodeficiency virus type 1 nucleocapsid protein. J. Virol. 68, 5863-5870. Vartanian, J . P., Meyerans, A., Sala, M., Wain-Hobson, S. (1994).G to A hypermutation of the human immunodeficiency virus type 1 genome: Evidence for dCTP pool imbalance during reverse transcription. Proc. Natl. Acad. Sci. U.S.A.91, 3092-3096. Verma, I. M., Meuth, N. L., Bromfeld, E., Manly, K. F., and Baltimore, D. (1971). Covalently linked RNA-DNA molecule as initial product of RNA tumor virus DNA polymerase. Nature (London) 233, 131-134. von Poblotzki, A., Wagner, R., Niedrig, M., Wanner, G., Wolf, H., and Modrow, S. (1993). Identification of a region in the Pr55e~g-polyproteinessential for HIV-1 particle formation. Virology 193, 981-985. Wakefield, J. K., Jablonski, S. A., and Morrow, C. D. (1992).Zn uitro enzymatic activity of human immunodeficiency virus type 1 reverse transcriptase mutants in the highly conserved YMDD amino acid motif correlates with the infectious potential of the proviral genome. J. Virol. 66, 6806-6812. Wakefield, J. K., Rhim, H., and Morrow, C. D. (1994).Minimal sequence requirements of a functional human immunodeficiency virus type 1 primer binding site. J. Virol. 68, 1605-1614. Wang, C. T., Zhang, Y., McDermott, J., and Barklis, E. (1993).Conditional infectivity of a human immunodeficiency virus matrix domain deletion mutant. J. Virol. 67, 70677076. Wang, J., Smerdon, S. J., Jager, J., Kohlstaedt, L. A., Rice, P. A., Friedman, J. M., and Steitz, T. A. (1994). Structural basis of asymmetry in the human immunodeficiency virus type 1 reverse transcriptase heterodimer. Proc. Natl. Acad. Sci. U.S.A. 91, 7242-7246. Waters, L. C., and Mullin, B. C. (1977). Transfer RNA in RNA tumor viruses. Prog. Nucleic Acid Res. Mol. Biol. 20, 131-160. Weiss, R., Teich, N., Varmus, H., and Coffin, J., eds. (1985). “RNA Tumor Viruses.” Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. West, A. B., Roberts, T. M., and Kolodner, R. D. (1992). Regulation of the reverse transcriptase of human immunodeficiency virus type 1 by dNTPs. Proc. Natl. Acad. Sci. U.S.A.89, 9720-9724. Whitcomb, J. M., Kumar, R., and Hughes, S. H. (1990). Sequence of the circle junction of human immunodeficiency virus type 1: Implications for reverse transcription and integration. J. Virol. 64, 4903-4906. Whiting, S. H., and Champoux, J. J. (1994). Strand displacement synthesis capability of Moloney murine leukemia virus reverse transcriptase. J. Virol. 68.4747-4758. Williams, K. J., and Loeb, L. A. (1992). Retroviral reverse transcriptases error frequencies and mutagenesis. Curr. Top. Microb. Zmmunol. 176, 165-180. Winslow, B. J., Pomerantz, R. J., Bagasra, O., and Trono, D. (1993). HIV-1 latency due to the site of proviral integration. Virology 196, 849-854. Wohrl, B. M., Volkmann, S., and Moelling, K. (1991). Mutations of a conserved residue

ERIC J. A W S AND MARK A. WAINBERG within HIV-1 ribonuclease H affect its exo- and endonuclease activities. J. Mol. Biol. 220,801-818. Wohrl, B. M., Ehresmann, B., Keith, G., and LeGrice, S. F. J. (1993). Nuclease footprinting of human immunodeficiency virus reverse transcriptaseltRNALyd complexes. J. Biol. Chem. 268,13617-13624. Wohrl, B. M., Howard, K. J., Jacques, P. S., and LeGrice, S. F. J. (1994). Alternative modes of polymerization distinguish the subunits of equine infectious anemia virus reverse transcriptase. J. Biol. Chem. 269, 8541-8548. Wohrl, B. M., Georgiadis, M. M., Telesnitsky, A., Hendrickson, W. A., and LeGrice, S. F. J. (1995a). Footprinting analysis of replicating murine leukemia virus reverse transcriptase. Science 267, 96-99. Wohrl, B. M., Tantillo, C., Arnold, E., and LeGrice, S. F. J. (1995b). An expanded model of replicating human immunodeficiency virus reverse transcriptase. Biochemistry 34, in press. Wong, S. F., Wahl, A. F., Yuan, P.-M., Arai, N., Pearson, B. E., et al. (1988). Human polymerase a gene expression is cell proliferation dependent and its primary structure is similar to both prokaryotic and eukaryotic replicative DNA polymerases. EMBO J . 7,37-44. WongStaal, F., and Gallo, R. C. (1985). Human T-lymphotropic retroviruses. Nature (London) 317,395-403. Wu, J., Amandoron, E., Li, X., Wainberg, M. A., and Parniak, M. A. (1993). Monoclonal antibody-mediated inhibition of HIV-1 reverse transcriptase polymerase activity. J . Biol. Chem. 268,9980-9985. Xiong, Y., and Eickbush, T. J. (1990). Origin and evolution of retroelements based upon their reverse transcriptase sequences. EMBO J . 9, 3353-3362. Yang, W., Hendrickson, W. A., Crouch, R. J., and Satow, Y. (1990). Structure of ribonuclease H phased at 2 A resolution by MAD analysis of the selenomethionyl protein. Science 249, 1398-1405. Yu, G., Salazar, M., and Reid, B. R. (1993). Structure of a DNA:RNA hybrid duplex. J. Mol. Biol. 233, 509-523. Yu, H., and Goodman, M. F. (1992). Comparison of HIV-1 and avian myeloblastosis virus reverse transcriptase fidelity on RNA and DNA templates. J . Biol. Chem. 267,1088810896. Yu, X., Yuan, X., Matsuda, Z., Lee, T. H., and Essex, M. (1992). The matrix protein of human immunodeficiency virus type 1 is required for incorporation of viral envelope protein into mature virions. J. Virol. 66, 4966-4971. Yuan, X., Yu, X., Lee, T. H., and Essex, M. (1993). Mutations in the N-terminal region of human immunodeficiency virus type 1matrix protein block intracellular transport of the gag precusor. J. Virol. 67, 6387-6394. Zack, J. A., Arrigo, S. J., Weitman, S. R., Go, A. S., Haislip, A., and Chen, I. S. Y. (1990). HIV-1 entry into quiescent primary lymphocytes: Molecular analysis reveals a labile, latent viral structure. Cell (Cambridge, Mass.) 61, 213-222. Zack, J. A., Haislip, A. M., Krogstad, P., and Chen, I. S. (1992). Incompletely reversetranscribed human immunodeficiency virus type 1 genomes in quiescent cells can function as intermediates in the retroviral life cycle. J. Virol. 66, 1717-1725. Zhang, D., Caliendo, A. M., Eron, J. J., DeVore, K. M., Kaplan, J. C., Hirsch, M. S., and D'Aquila, R. T. (1994). Resistance to 2',3'-dideoxycytidine conferred by a mutation in codon 65 of the human immunodeficiency virus type 1 reverse transcriptase. Antimicrob. Agents Chemother. 38, 282-287. Zhang, H., Zhang, Y., Spicer, T. P., Abbott, L. Z., and Poiesz, B. J. (1993). Reverse

HIV-1 REVERSE TRANSCRIPTASE

163

transcription takes place within extracelluiar HIV-1 virions: Potential biological significance. AIDS Res. Hum. Retroviruses 9, 1287-1296. Zhang, H., Bagasra, O., Niikura, M., Poiesz, B. J., and Pomerantz, R. J. (1994). Intravirion reverse transcripts in the peripheral blood plasma of human immunodeficiency virus type 1-infected individuals. J. Virol. 68, 7591-7597. Zucher, H., and Steigler, B. (1981). Optimal computer folding of large RNA sequences using thermodynamics and auxillary information. Nucleic Acids Res. 18,1839-1846. Zybarth, G., Krausslich, H. G., Partin, K., and Carter, C. (1994). Proteolytic activity of novel human immunodeficiency virus type 1 proteinase proteins from a precursor with a blocking mutation at the N terminus of the PR domain. J. Virol. 68,240-250.

This Page Intentionally Left Blank

ADVANCES IN VIRUS RESEARCH, VOL. 46

HEPADNAVIRUSES: CURRENT MODELS OF RNA ENCAPSIDATION AND REVERSE TRANSCRIPTION Dorothy A. Fallows and Stephen P. Goff Howard Hughes Medical Institute and Deportment of Biochemistry and Molecular Biophysics Columbia University College of Physicians and Surgeons New York, New York 10032

I. Introduction A. Historical Background B. Overview of the Hepadnaviral Life Cycle 11. Transcription and Translation A. Major Transcripts B. Protein Products 111. RNA Encapsidation A. Core Particle Assembly B. Cis-Acting Signals on Pregenomic RNA IV. Hepadnaviral Polymerase A. Experimental Approaches B. Sequence Similarities C. Mutational Analyses D. Minus Strand Priming V. Reverse Transcription A. Organization of Pregenomic RNA B. Minus Strand DNA Priming and Synthesis C. Plus Strand DNA Priming and Synthesis D. Formation of Covalently Closed Circular DNA VI. Concluding Remarks References

I. INTRODUCTION Human hepatitis B virus, originally known as serum hepatitis, is a major worldwide health threat and is considered responsible for most of the 1to 2 million deaths estimated to occur in the world each year from hepatitis. Although most infections are cleared by immune surveillance, some 5 to 10% of infections progress to a chronic state in which the likelihood is great that complications such as cirrhosis and hepatocellular carcinoma will arise in later life. In children, the chances of progressing from acute to chronic hepatitis B virus (HBV) infection are 20 to 50%. In the United States, about 300,000 HBV 165

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

166

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

infections occur each year, leading to an estimated 18,000 to 30,000 new chronic cases (Storch, 1993). In large parts of the world where HBV infection is endemic and vertical transmission from mother to infant is common, primary hepatocellular carcinoma is a major cause of death. The only treatments currently available for HBV infection, interferon therapy or liver transplant, are expensive, have serious consequences, and are, at best, only partially successful. Because of its role as a major human pathogen, HBV has long been the subject of intense study which, owing to the early lack of animal models for the disease and difficulties in culturing the virus in the laboratory, is only recently beginning to yield answers.

A. Historical Background The earliest demonstration of the infectious agent responsible for transmitting serum hepatitis occurred in 1965 with Baruch Blumberg’s identification of the Australia antigen in the sera of multiply transfused hemophilia patients (Blumberg et al., 1965). Electron microscopy studies reported by Dane and co-workers in 1970 revealed the presence of several species of particles in the serum of an infected individual (Dane et al., 1970). These included mostly spheres and filaments roughly 16-25 nm in diameter, as well as less abundant 42-nm spheres, all of which aggregated in the presence of antibodies against the Australia antigen. On the basis of morphology, the 42-nm spheres were proposed to represent complete virion structures, whereas the smaller spheres and filaments were thought to consist of excess viral coat material. Examination of the structure of the 42-nm spheres, or “Dane particles,” revealed a detergent-soluble coat surrounding a 28-nm electrondense core (Almeida et al., 1971). Further studies confirmed the presence of DNA within the 28-nm cores and, moreover, demonstrated an associated DNA polymerase activity (Kaplan et al., 1973; Robinson et al., 1974). Extraction and characterization of viral DNA revealed an unusual genome structure consisting of a small nicked circular molecule of double-stranded DNA with a single-stranded gap variably extending over 15 to 50% of the circle. The genome comprised one DNA strand about 3 kb in length annealed to a shorter DNA strand of variable length, held together in a circle by overlapping cohesive ends (Sattler and Robinson, 1979; Summers et al., 1975). Endogenous DNA polymerase activity was shown to involve filling in this gapped region by chain elongation on the 3’ end of the shorter strand. Sequence analysis of cloned DNA identified two open reading frames (ORFs) overlapping a third ORF on the long strand, termed the minus strand, which were proposed to encode the viral core and surface antigens and

HEPADNAVIRUSES

167

the polymerase (Galibert et al., 1979; Pasek et aZ.,1979; Valenzuela et al., 1980).A further clue to the mechanism of hepadnavirus replication came with the discovery of a protein covalently attached to the 5' end of minus strand DNA (Gerlich and Robinson, 1980; Molnar-Kimber et al., 1983). The source of this terminal protein was unknown, but the possibility of its functioning as the primer for DNA replication was proposed. Although a means of producing virions in tissue culture systems was not initially available, research on human hepatitis B virus (HBV) was greatly assisted by the discovery of related viruses in several animal species that were amenable to more detailed study. Related members of the hepadnavirus family include woodchuck hepatitis virus (WHV) (Summers et al., 19781, ground squirrel hepatitis virus (GSHV) (Marion et al., 1980), duck hepatitis B virus (DHBV) (Mason et al., 1980), and heron hepatitis B virus (HHBV) (Sprengel et al., 1988). Mason et al. (1982) were the first to describe replicative intermediates of DHBV isolated from infected duck livers. They found, in addition to relaxed circular DNA molecules, a heterogeneous population of single-stranded viral DNA molecules ranging in size up to the full length of the DHBV genome, which they suggested might be nascent DNA chains in early stages of replication. Southern analysis with strand-specific probes showed these DNA species to be exclusively minus strand, indicating an asymmetric mechanism of DNA synthesis in DHBV. Prior to the reported findings of Mason and Summers, it had been assumed that hepadnaviruses replicate by a mechanism involving DNA-directed DNA synthesis. On the basis of observations of asymmetric replication in DHBV, Mason and Summers suggested an alternative model. They proposed that minus strand DNA synthesis first occurs on an RNA template followed by plus strand synthesis using the minus strand DNA as a template, in a process resembling retroviral reverse transcription (Summers and Mason, 1982). In support of this model, they demonstrated that minus strand synthesis in DHBV was sensitive to RNase treatment but uninhibited by actinomycin D, whereas plus strand synthesis was unaffected by RNase but inhibited by actinomycin D. Subsequent studies on other hepadnaviral species revealed a similar RNA-dependent, asymmetric mechanism of replication, consistent with reverse transcription (Miller and Robinson, 1984; Miller et al., 1984; Weiser et al., 1983). Moreover, comparative sequence analyses have identified sequence similarities within the coding sequences of the putative polymerases of several hepadnaviruses and the reverse transcriptases of classic retroviruses (Doolittle et al., 1989; Miller, 1988; Miller and Robinson, 1986; Toh et al., 1983). Further elucidation of hepadnaviral reverse transcription awaited

168

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

the development of systems for propagating the virus in tissue culture. Efforts to identify cellular receptors of the various hepadnaviruses have, thus far, proved fruitless. This lacuna in our knowledge has precluded the possibility of constructing stable infectable cell lines for viral expression. Primary hepatocytes of ducks, humans, and woodchucks have been successfully infected in culture, but these cells do not maintain their infectable state for long (Aldrich et al., 1989; Gripon et al., 1988; Tuttleman et al., 198613). To date, the most useful expression systems have been provided by hepatocyte cell lines which are readily maintained in culture and transfectable either stably or transiently (Chang et al., 1987; Condreay et al., 1990; Sells et al., 1987; Shih et al., 1989; Sureau et al., 1988; Yaginuma et al., 1987). Evidence that such systems are good approximations of in vivo viral replication has been provided by the observation of all previously described replicative intermediates in transfected cells, and by the demonstration that virions produced in such systems are infectious in vivo (Acs et al., 1987; Pugh et al., 1988; Sells et al., 1988; Sureau et al., 1988).Heterologous promoter constructs have been used to improve viral expression in transfections and to allow expression in nonhepatocyte lines (JunkerNiepmann et al., 1990; Seeger et al., 1989). With cell culture systems such as these, it has been possible to examine in detail the highly unusual life cycle of the hepadnaviruses.

B . Overview of Hepadnaviral Life Cycle Hepadnaviruses are small enveloped viruses that replicate by reverse transcription within hepatocytes of an infected individual (Loeb and Ganem, 1993; Nassal and Schaller, 1993). Infectious particles released into the serum display a typical structure that is represented in Fig. 1A. Mature virions are surrounded by an outer membrane layer, the envelope, into which are inserted three related forms of a virally encoded transmembrane protein, the surface antigen. The envelope encloses the nucleocapsid, an icosahedral structure formed by the viral core protein. The nucleocapsid contains the relaxed circular DNA genome, including the terminal protein covalently bound to the 5’ end of the minus strand DNA, and the viral polymerase. A schematic of the viral life cycle is presented in Fig. 1B. On infection, the nucleocapsid is released into the host cell, and the viral genome is delivered into the nucleus by means that are not currently well understood. Within the nucleus, the single-stranded region of the viral DNA is completed, the nicks in each strand are repaired, and the genome is converted into a supercoiled DNA molecule, referred to as the covalently closed circular (CCC) DNA. Unlike the case of retro-

HEPADNAVIRUSES

169

FIG.1. (A) Structure of the virion. (B) Hepadnaviral life cycle.

viruses and retrotransposons, the hepadnaviral genome is not integrated into the host cell chromosome as part of its normal life cycle. Instead, unintegrated molecules of CCC DNA serve as the template for transcription by host RNA polymerase and nuclear transcription factors (Mason et al., 1983; Tagawa et al., 1986). After transport to the cytoplasm, viral transcripts are translated by the host cell machinery to produce the viral proteins. The surface anti-

170

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

gen proteins are cotranslationally inserted into the membrane of the endoplasmic reticulum (Eble et al., 1986). The template for reverse transcription, termed the pregenomic RNA (pgRNA), and the polymerase are assembled into nucleocapsids, or core particles, where reverse transcription begins. In the final stages of replication, core particles associate with the surface antigen proteins; mature enveloped virions bud into the lumen of the endoplasmic reticulum and are transported to the cell surface for release (Patzer et al., 1986). In contrast with retroviral reverse transcription, most hepadnaviral reverse transcription takes place within intracellular core particles prior to release from infected cells. This leads to a possibility for amplification of the CCC DNA via intracellular replication, without the necessity for new rounds of infection. Several studies have indicated that a pool of amplified CCC DNA accumulates within the nuclei of hepatocytes in infected livers, as well as in cells transfected or infected in culture (Condreay et al., 1990; Miller and Robinson, 1984; RuizOpazo et al., 1982; Sells et al., 1988; Weiser et al., 1983). Moreover, analyses of infected primary duck hepatocytes have revealed up to 50 copies of CCC DNA per infected cell nucleus, and confirmed that this DNA is amplified via an intracellular asymmetric process of DNA synthesis resembling reverse transcription, rather than semiconservative replication (Tuttleman et al., 1986a; Wu et al., 1990). Interestingly, surface antigen mutations which interfere with viral release have been found to result in hyperaccumulation of CCC DNA (Summers et al., 1990, 1991). This finding suggests that the level of nuclear CCC DNA may be determined by the rate at which core particles are removed from the pool of intracellular replicative intermediates by association with the viral surface antigens and transport out of the cell. In the absence of a mechanism for genomic integration such as that employed by retroviruses, the pathway of intracellular CCC DNA amplification likely plays an important role in maintaining viral persistence.

11.

TRANSCRIPTION

AND

TRANSLATION

The full-length hepadnaviral genomes range from about 3000 to 3300 nucleotides. In spite of their small genome size, hepadnaviruses are able to produce a remarkable variety of proteins. This is achieved through a combination of overlapping open reading frames (ORFs) and differential start sites for translation t o produce multiple related products from a single ORF with varying amino-terminal additions. The genomic map of the human hepatitis B virus (HBV) is presented in

HEPADNAVIRUSES

171

Fig. 2. The general features of this map are conserved among the mammalian and avian hepadnaviruses, with a few important exceptions described in the text.

A . Major Transcripts Hepadnaviral transcripts have been divided into two classes on the basis of size: the genomic and the subgenomic mRNAs. The genomic transcripts include terminally redundant sequences of about 150 to 250 nucleotides and range in size from about 3.0 to 3.4 kb, depending on the viral species. In HBV, there are three start sites for transcription of the genomic (preC/C) mRNAs within a distance of 31 nucleotides from one another. The two major species of subgenomic mRNAs

FIG. 2. Organization of the HBV genome. The covalently closed circular (CCC) DNA is represented in the center; direct repeat sequences (DR1 and DR2) are shown as boxes on the molecule. The four open reading frames (ORFs) are indicated, as follows: preC/C encodes the e antigen and core protein, P encodes the polymerase, preSl/preS2/S encodes the three forms of the surface antigen, and X encodes the X protein. The preCIC, preS1, preS2/S, and X mRNAs are shown as lines on the outside.

172

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

that have been described in mammalian and avian hepadnaviruses are the 2.1-kb preS2IS transcripts and the 2.4-kb preSl transcript. Mammalian hepadnaviruses produce an additional subgenomic mRNA, the X mRNA, that is roughly 0.7 kb in length and present in very low abundance. The avian hepadnaviruses do not produce an X mRNA. All of the viral transcripts described above remain unspliced. A few spliced transcripts have been found in transfected cells and in the livers of infected individuals (Chen et al., 1989; Su et al., 1989; Suzuki et al., 1989). However, these spliced mRNAs do not seem to be consistently present in infected or transfected cells, and there has been no evidence to indicate that they are required for replication. Thus, the relevance of these spliced mRNAs in the hepadnaviral life cycle remains unclear. All hepadnaviral transcripts are capped, and all are polyadenylated by means of a common polyadenylation signal. The poly(A) signal is present only once in the subgenomic transcripts (see Fig. 2). However, in the terminally redundant genomic transcripts, the poly(A) signal is represented twice and must be read through without termination on the first pass of the transcription machinery. Studies on the mechanism of read-through have shown that recognition of the hepadnaviral poly(A) signal requires the presence of upstream activating signals on the nascent mRNA; these sequences are not present at the 5' end of the genomic transcripts (Russnak and Ganem, 1990; Russnak, 1991).

B . Protein Products The HBV genome contains four ORFs which encode the core protein (0,the polymerase (€9,the surface antigen proteins, and the X protein (see Fig. 2). The C ORF additionally encodes a protein referred to as the e antigen, which is produced from an upstream translational start site in the same reading frame. The longer genomic mRNAs include the initiation site for translation of the e antigen, whereas the shortest genomic mRNA does not. Thus, the template for translation of the e antigen is provided by the longer genomic transcripts. The additional sequences encoding the amino terminus of the e antigen, termed the preC sequences, comprise a signal peptide that targets this protein to the endoplasmic reticulum for post-translational processing and secretion (Jean-Jean et al., 1989a; Ou et al., 1986).The full-length peptide is processed, most likely by a host cell protease, to yield a final product of 17 kDa which is released into the bloodstream of infected individuals. Although a role for the e antigen in mediating host immune responses has been suggested (Ganem, 1982; Milich et al., 19901, the nature of this role is unclear. That e antigen production is unnecessary

HEPADNAVIRUSES

173

for viral replication in tissue culture systems has been shown by the expression of fully replication-competent viral genomes from heterologous promoter constructs that do not include the e antigen AUG (Huang and Summers, 1991; Junker-Niepmann et al., 1990; Seeger et al., 1989). The 21-kDa core protein is translated from the shortest of the three genomic transcripts. The use of alternative transcriptional start sites determines which of the two forms of the core protein will be synthesized. In addition to encoding the core protein, the shortest genomic mRNA serves as the template for translation of the polymerase, as well as providing the substrate for reverse transcription (Huang and Summers, 1991). The P ORF spans roughly three-quarters of the fulllength hepadnaviral genome and encodes a protein of about 90 kDa in mammalian variants and 83 kDa in the avian hepadnaviruses. The P ORF overlaps the C ORF and is in the +1 frame relative to C . According to conventional models of eukaryotic translation, the position of the P AUG is unfavorable for de nouo translation of P protein (Kozak, 1989). This observation initially led investigators to suggest that P may be translated as part of a core-polymerase polyprotein, possibly by a mechanism of ribosomal frameshifting similar to that used by retroviruses (Levin et al., 1993). However, analysis of a variety of nonsense mutations introduced in the vicinity of the P AUG has demonstrated that the hepadnaviral C and P proteins are synthesized as separate translation products (Chang et al., 1990a; Schlicht et al., 198913). The mechanism for translation of the P protein is not clear, but two alternative explanations have been proposed: a “leaky” scanning model and a model based on direct internal initiation of P synthesis (Chang et al., 198913, 1990a; Jean-Jean et al., 198913; Roychoudhury and Shih, 1990; Schlicht et al., 198913). The surface antigen ORF entirely overlaps the P ORF and is in the + 1 frame with respect to P. Mammalian hepadnaviruses synthesize three different forms of the surface antigen, termed the small ($3, medium (MI, and large (L) surface proteins. These are produced from a single ORF with three distinct translational start sites. The shortest unit of open sequences is referred to as the S ORF; the open sequences upstream of S are referred to as the preSl and preS2 sequences. Translation of the L protein initiates at the first AUG in the series and proceeds through the preS1, preS2, and S sequences. The M protein is translated from the second AUG and is encoded by the preS2 and S sequences, whereas S protein translation initiates at the S AUG. Thus, the M and L proteins share a common carboxyl terminus identical in sequence to the full-length S protein and vary by the addition of amino-terminal sequences. The relative proportions of S , M, and L are

174

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

regulated by the differential use of transcriptional start sites that determine the position of the first available site for translation initiation. Avian hepadnaviruses employ a similar scheme to produce only two forms of surface proteins. The X ORF overlaps the 3' end of the P ORF and the 5' end of the C ORF. The X protein is unique to the mammalian hepadnaviruses; avian viral genomes do not include an X ORF. Although the X protein has been shown to trans-activate a variety of promoters in uitro (Scheck et al., 1991),and to play a role in the establishment of infections in uiuo (Chen et al., 1993);Zoulim et al., 1994),the mechanism of its action remains to be elucidated. Deletions and frameshift mutations in the X gene show no adverse effects on viral replication in cultured cells, indicating that the protein is unnecessary for viral expression in these systems (Blum et al., 1992;Yaginuma et al., 1987). 111. RNA ENCAPSIDATION

A . Core Particle Assembly Core particle assembly involves the interactions of the structural proteins, core (C) and polymerase (P),with the pregenomic RNA (pgRNA) which provides the template for reverse transcription. In the absence of pgRNA and P protein, the C protein is capable of selfassembling into particles that resemble the native icosahedral structures (Nassal, 1988;Onodera et al., 1983;Zhou and Standring, 1992). Deletion analyses have identified a minimal amino-terminal portion of the core protein that is required for self-assembly (Birnbaum and Nassal, 1990;Nassal, 1992).The carboxyl terminus of C protein contains a highly arginine-rich region capable of acting as a nonspecific RNAand DNA-binding domain (Hatton et al., 1992). Although frameshift mutations that completely eliminate this domain do not interfere with C protein self-assembly, these mutations completely abrogate packaging of viral RNA (Nassal, 1992;Schlicht et al., 1989a).Limited deletions of portions of the arginine-rich region allow RNA encapsidation but are deleterious to the production of plus-strand DNA, implying that core protein also provides an essential function in reverse transcription (Nassal, 1992;Yu and Summers, 1991). In classic retroviruses, incorporation of the viral polymerase into assembling nucleocapsids is accomplished by translation of the polymerase as a gag-pol polyprotein (Levin et al., 1993).The capsid, or gag, portion of the gag-pol protein contains assembly domains which enable both gag and gag-pol to be incorporated via the same mechanism.

HEPADNAVIRUSES

175

Unlike the case in retroviruses, the hepadnaviral C and P proteins are synthesized as two separate translation products. This raises the interesting question of how the hepadnaviral polymerase is incorporated into assembling core particles. According to one reported estimate, only one or two molecules of P protein are incorporated per virion (Bartenschlager and Schaller, 1992). P protein has been shown to play a critical role in the process of RNA encapsidation. Several mutations in P have been identified which either reduce the efficiency of pgRNA packaging (Chen et al., 1992,1994; Roychoudhury et al., 1991) or completely abrogate RNA packaging (Bartenschlager et al., 1990; Hirsch et al., 1990).The results of these analyses indicate that all regions of the viral polymerase, including the DNA polymerase and RNase H domains, are involved in RNA packaging. Missense mutations within the catalytic site of the reverse transcriptase do not interfere with RNA encapsidation, indicating that the enzymatic function of P protein is not necessary for its packaging function (Bartenschlager et al., 1990; Hirsch et al., 1990).Interestingly, in cotransfections of either wild-type or core-defective genomes with polymerase-defective genomes, pgRNAs expressing functional P protein were preferentially encapsidated over pgRNAs carrying mutations in the P ORF. This cis preference may reflect a requirement for cotranslational RNA packaging or may simply be the result of the extremely limiting quantities of P protein synthesized by hepadnaviruses. Additional studies have demonstrated that the incorporation of P protein and pgRNA into assembling core particles is mutually dependent (Bartenschlager and Schaller, 1992; Pollack and Ganem, 1994). These results suggest a model for viral assembly in which the P protein first interacts with the viral pregenome, and then the P protein-pgRNA complex is recognized by the core protein and incorporated into assembling particles.

B . cis-Acting Signals on Pregenomic RNA Despite the similarities between the three genomic mRNAs, only the shortest of the transcripts is encapsidated and reverse transcribed (Enders et al., 1987; Lien et al., 1986; Seeger et al., 1986). In retroviruses, the packaging signal is located within a region that is spliced out of the subgenomic mRNAs, thus ensuring selective packaging of full-length replication-competent genomes. In contrast, the longer preC transcripts of hepadnaviruses that are not packaged contain all of the sequences present in the shorter pregenomic RNA that is packaged. Therefore, linear sequences alone cannot specify the signal for differential packaging of the pregenomic RNA into assembling core parti-

176

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

cles. Disruption of translation by the introduction of point mutations into the preC coding sequences allowed the preC transcripts to be packaged nearly as efficiently as pgRNA (Nassal et al., 1990).These results suggest that there is a competition between the processes of translation and packaging of the genomic mRNAs, and that preC mRNAs are excluded from core particles by translational commitment. Packaging of the pregenomic RNA probably occurs following translation of the polymerase and is likely assisted by the strong affinity of this protein for packaging its own mRNA in cis. In HBV, the cis-acting signal for encapsidation has been localized by deletion analyses to sequences that overlap the preC region near the 5’ end of pgRNA (Chiang et al., 1992;Junker-Niepmann et al., 1990).The packaging signal, termed E, includes 85 to 94 nucleotides of sequence from the 5’ end of pgRNA that are sufficient to direct encapsidation of foreign RNA sequences into core particles (Junker-Niepmann et al., 1990; Pollack and Ganem, 1993). Deletion analyses in DHBV have identified a similar cis-acting packaging domain at the 5’ end of pgRNA, but showed that additional sequences near the middle of pgRNA are required for RNA encapsidation (Calvert and Summers, 1994; Hirsch et al., 1991; Lavine et al., 1989). The E sequences are located within the terminal redundancy on pgRNA and are present at both the 5’ and 3’ ends of the RNA (see Fig. 3). However, mutations introduced into the 3’ copy of E had no effect on RNA packaging, indicating that some aspect of the position of the 3’ sequences excluded them from being recognized as an appropriate packaging signal (Hirsch et al., 1991). This finding is not surprising, as subgenomic mRNAs include the 3‘ E sequences and yet are not packaged. It is possible that the packaging signal must be located in close proximity to a 5‘ capped mRNA end in order t o be recognized. However, the ability of the longer preC mRNAs to be packaged, if translation is disrupted, implies some flexibility in the required distance between E and the cap (Nassal et al., 1990). The 5’ packaging regions of both HBV and DHBV contain several nested inverted repeats predicted by computer analyses to form a bulged stem-loop structure (Junker-Niepmann et al., 1990) (see Fig. 3). Although RNA packaging has not been studied in other hepadnaviral variants, computer-simulated foldings of 5’ pgRNA sequences have predicted the presence of similar bulged stem-loops in the RNAs of WHV, GSHV, and HHBV (Junker-Niepmann et al., 1990; Pollack and Ganem, 1993). Phylogenetic conservation of this predicted RNA structure may indicate conservation of a functional requirement. RNA encapsidation in HBV has been examined by the use of cotransfection systems in which E-Z~CZ chimeric mRNA was packaged and C and P

177

HEPADNAVIRUSES

A

C Gu G AA*.'' cG u CC,' cu,', GGu G~ \ G A \'GG U UUCA c G 'Uu ucA-U U -G C-G C --G U -G 0-C u --A 4 C ORF A-U C -0

'

u -G

U --A G-C 5'ACUUU~UCACCUCUG@XJAAUCAUCUCUU A2CGACCCUUAU

DR 1

I

0 /

DR1

-3'

0

E

0

0

0

/ I

n

"

DR2

DR1

m An3 ' E

FIG.3. The E encapsidation signal. (A) Predicted stem-loop of the E sequences. The position of the C AUG is indicated by an arrow. (B) Position of the E sequences on the pregenomic RNA. DR1 and DR2 are direct repeat sequences.

proteins were provided in trans from a helper plasmid that did not include the E sequences (Knaus and Nassal, 1993;Pollack and Ganem, 1993). Structural probing with single strand- and double strandspecific RNases has confirmed the presence of this folded structure in chimeric HBV E-ZUCZRNA isolated from virions produced in such cotransfections (Pollack and Ganem, 19931,as well as in in uitro transcribed RNAs containing the HBV E sequences (Knaus and Nassal, 1993). Genetic studies in HBV and in DHBV have shown that major disruptions of the stem-loop interfere with encapsidation, thereby confirming a functional role for this folded structure in uivo (Fallows and Goff, 1995;Knaus and Nassal, 1993;Pollack and Ganem, 1993,1994).Moreover, in uitro studies in DHBV have demonstrated a direct interaction between P protein and the E RNA sequences (Pollack and Ganem, 1994;Wang et al., 1994).In this system, E mutations that interfered with formation of the RNA-P protein complex in uitro were found to abrogate packaging in uiuo, implying a correlation between the two capabilities (Pollack and Ganem, 1994). However, some of the packaging-defective E mutants were still capable of forming the RNAP complex. The effects of P protein mutations on formation of the

178

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

RNA-P complex have also been examined (Wang et al., 1994). The results of this study indicated that, although the amino-terminal and the DNA polymerase domains of P were necessary for the interaction with E, the RNase H domain was dispensable in this assay. This finding contrasts with previous data on RNA packaging requirements, which demonstrated that all of the P protein domains were necessary for packaging. Taken together, the results of the in vitro binding studies indicate that RNA-P protein complex formation is necessary but not sufficient for RNA packaging. An alternative explanation for exclusion of the subgenomic mRNAs from virions could be that the 3‘ E sequences are incapable of forming a proper stem-loop within the context of the surrounding RNA sequences. However, structural probing of an in vitro transcribed RNA, designed to mimic the major subgenomic transcript, has indicated that the 3’ E sequences form a folded structure identical to that of the 5’ E domain (Pollack and Ganem, 1993). Thus, it remains unclear what distinguishes the 5’ E sequences as the signal for encapsidation.

IV. HEPADNAVIRAL POLYMERASE During reverse transcription, the hepadnaviral P protein functions as an RNA-dependent DNA polymerase, a DNA-dependent DNA polymerase, and an RNase H nuclease, in much the same way as traditional reverse transcriptases. In addition to these functions, P protein also mediates RNA encapsidation, as previously discussed, and serves as a protein primer for minus strand synthesis in a process described below.

A . Experimental Approaches Our ability to characterize the hepadnaviral polymerase has been severely hampered by the lack of experimental systems for examining its activity outside of the virion. In spite of a wide variety of eukaryotic and prokaryotic expression systems tested for the purpose, all but the most recent efforts to express an enzymatically active P in vitro have been unsuccessful. Attempts to demonstrate the activity of virally derived P protein on exogenously provided templates were also unsuccessful (Radziwill et al., 1988). Low pH treatment was found to permeabilize core particles, rendering viral DNA accessible to restriction endonucleases without destroying the endogenous activity of the viral polymerase. However, even under these conditions, the polymerase remained tightly associated with the viral DNA and was unable to switch to a variety of exogenous templates (Radziwill et al., 1988).

HEPADNAVIRUSES

179

Problems in examining the structure and function of the native P protein have been compounded by the lack of high-titer antisera to the protein (Bosch et al., 1988; Chang et al., 1989a; Mack et al., 1988). Without such reagents, it has been especially difficult to resolve continuing debates about the size and structure of the protein in virions (Bavand and Laub, 1988; Oberhaus and Neubold, 1993; Wu et al., 1991). The introduction of target phosphorylation sites into P protein followed by 32P-labeling with protein kinase A has been used to improve the sensitivity of P protein detection. Immunoprecipitations of virions carrying the tagged polymerase identified a single product of the expected size of full-length P protein, indicating that the polymerase most likely remains unprocessed in virions (Bartenschlager et al., 1991).

B . Sequence Similarities Amino acid sequence comparisons have identified several regions of similarity in the coding sequences of hepadnaviral polymerases and the reverse transcriptases of retroviruses and retrotransposons (Doolittle et al., 1989; McClure, 1993).In particular, a highly conserved TyrX-Asp-Asp motif associated with the catalytic site of the retroviral DNA polymerase domain is conserved in all variants of hepadnaviral polymerases (Miller, 1988; Miller and Robinson, 1986; Toh et al., 1983). Comparative studies of retroviral and E. coli RNase H have allowed identification of specific residues associated with the RNase H catalytic site that are similarly conserved in all hepadnaviral polymerases (Schodel et al., 1988). The conserved DNA polymerase and RNase H domains together comprise the carboxyl-terminal two-thirds of the hepadnaviral polymerase. Alignments of the P protein sequences from different hepadnaviruses have identified a region of unconserved sequences between the DNA polymerase domain and a highly conserved domain that forms the amino terminus of P (Radziwill et al., 1990; Sprengel et al., 1985). The amino-terminal domain contains no homologies to any proteins of other retroid family members, although some very limited homologies to the genome-linked proteins of the picornaviruses have been suggested (Khudyakov and Makhov, 1989; Weber et al., 1994).

C . Mutational Analyses Analysis of the effects of mutations within the P gene on viral replication in transfected cells has, for the most part, confirmed the predictions of comparative sequence analyses. In HBV, point mutations of

180

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

the conserved Tyr-X-Asp-Asp residues abolished all DNA synthesis, whereas mutations in the conserved residues of the predicted RNase H domain allowed minus-strand but not plus-strand DNA synthesis (Radziwill et al., 1990). Point mutations in the conserved residues of the predicted DNA polymerase domain of DHBV similarly disrupted DNA synthesis (Chang et al., 1990b). However, the DNA polymerase and RNase H domains of DHBV P protein were not functionally separable; mutations in the RNase H domain were deleterious to synthesis of both minus and plus strand DNA. Presumably, the overall structure of the DHBV polymerase is more sensitive than HBV P to the presence of mutations in the carboxyl terminus. In contrast, the variable domain has been quite tolerant to a variety of in-frame substitutions, deletions, and insertions (Li et aZ., 1989; Radziwill et al., 1990). These observations and the lack of sequence conservation imply that the region is not functionally essential, but rather serves as a spacer or tether between the amino-terminal domain and the DNA polymerase and RNase H domains.

D . Minus Strand Priming The presence of a protein attached to the 5’ end of the minus strand DNA was noted in early reports on the hepadnaviral genome, although the identity and source of the terminal protein were unknown (Ganem et aZ., 1982; Gerlich and Robinson, 1980; Molnar-Kimber et al., 1983). The resistance of the protein-DNA association to treatment with high pH or boiling in SDS suggested that a covalent linkage held the two species together (Gerlich and Robinson, 1980). The identity of the protein was revealed when antisera directed against sequences of the conserved amino terminus of P were found to immunoprecipitate selectively minus strand DNA, demonstrating that the terminal protein is derived from the viral polymerase (Bartenschlager and Schaller, 1988; Bosch et al., 1988). On the basis of these observations, the aminoterminal domain of P was predicted to function as a primer for initiation of minus strand DNA similar to the terminal proteins that prime DNA synthesis in Adenouirus and bacteriophage $29 (Salas, 1991). Direct evidence in support of this theory has been provided only relatively recently by the expression of enzymatically active P protein in several different in uitro systems (Howe et al., 1992; Seifer and Standring, 1993; Tavis and Ganem, 1993; Wang and Seeger, 1992). The P protein of DHBV synthesized in a cell-free rabbit reticulocyte system was capable of incorporating radiolabeled deoxynucleotides into extended chains up to 500 nucleotides in length and covalently at-

HEPADNAVIRUSES

181

tached to P (Wang and Seeger, 1992).The DNA products were shown to be templated on the DHBV RNA, and the sequence of the first four nucleotides incorporated corresponded exactly to the sequences of the 5' end of the authentic DHBV minus strand. No labeled DNA products were synthesized by a P protein carrying a mutation in the conserved Tyr-X-Asp-Aspmotif, implying that terminal protein alone is not sufficient to form a covalent linkage with a deoxynucleoside 5'-monophosphate (dNMP), but requires the activity of the DNA polymerase domain. Expression of P in the presence of [ c Y - ~ ~ P I ~ Gthe T P first , nucleotide incorporated into DHBV minus strand, produced a radiolabeled protein the same size as the predicted full-length polymerase, confirming that the terminal protein functions as a true primer rather than being added to the DNA after synthesis (Wang and Seeger, 1992). Protease digestion of the labeled polymerase followed by amino acid analysis of the peptide fragments localized the 32P-dGMP to a specific tyrosine residue at position 96 in the amino-terminal domain, thus identifying Qr-96 as the site of minus strand priming (Zoulim and Seeger, 1994). Genetic analysis of DHBV in transfected cells has demonstrated that the Tyr-96 residue is essential for minus strand synthesis in uiuo,consistent with a role for this residue as the substrate for minus strand priming (Weber et al., 1994). The DHBV P ORF has also been inserted into the genome of the Q retrotransposon and expressed in yeast cells within viruslike particles (VLPs) formed by the Ty A gene product (Tavis and Ganem, 1993). Within this context, DHBV P was able to prime and elongate DNA chains up t o 2500 nucleotides in length. Further analysis demonstrated that the DNA products were derived from DHBV sequences initiated at the appropriate position expected for minus strand, and that the products were covalently bound to protein. Thus, not only was the polymerase capable of using its own template to reverse transcribe authentic minus strand DNA, but the products of the reaction could be extended to much greater lengths than was apparently possible in the cell-free rabbit reticulocyte expression system. Perhaps the Ty A protein was able to supply some specific function required for elongation normally provided by the hepadnaviral core protein, or maybe the spatial arrangements imposed by enclosure within VLPs may have been more amenable for elongation. A third report describes the expression of HBV polymerase from RNA injected into Xenopus oocytes that was capable of synthesizing protein-linked DNA products of an undetermined sequence (Seifer and Standring, 1993). These expression systems have enabled investigators to begin detailed analyses of the mechanism of reverse transcription in hepadnaviruses.

182

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

V. REVERSE TRANSCRIPTION

A . Organization of Pregenomic RNA A model of hepadnaviral reverse transcription is presented in Fig. 4, along with the structure of the pregenomic RNA. In addition to the terminally redundant sequences (R), pgRNAs contain identical direct repeat sequences of 11 to 12 nucleotides, designated DR1 and DR2 (Buscher et al., 1985; Enders et al., 1985; Horwich et al., 1990). The DR1 sequences are represented twice on pgRNA within the terminal redundancy, and thus are located at both the 5’ and 3‘ ends of the RNA. DR2 is located at the 3’ end of pgRNA upstream of the 3‘ DR1 and separated from it by a distance of about 50 nucleotides in avian hepadnaviruses and about 200 nucleotides in the mammalian variants. The direct repeat sequences are involved in the priming of plus and minus strand DNA, and the terminal redundancy is important for plus strand transfer and formation of the circular genome. Details of these processes are discussed below.

B. Minus Strand DNA Priming and Synthesis The primer for synthesis of minus strand DNA is provided by a tyrosine residue in the amino-terminal domain of the viral polymerase. Current knowledge indicates that the terminal protein remains connected to the rest of the polymerase throughout reverse transcription; so, in Fig. 4, the polymerase is shown as a full-length protein on the end of the minus strand (Bartenschlager et al., 1991).The 5’ end of the minus strand has been mapped to sequences within DR1 (Lien et al., 1987; Molnar-Kimber et al., 1984; Seeger et al., 1986; Will et al., 1987). However these experiments did not address which of the two copies of DR1 is employed. Minus strand could initiate within the 5’ DR1 close to the capped end of pgRNA and then transfer to the 3‘ end of the template for elongation; alternatively, initiation could occur within the 3’ DR1 where synthesis of full-length minus strand would require no strand transfer. Mapping of the 3‘ end of the minus strand revealed a short terminal redundancy of 9 to 10 nucleotides (r) on the molecule and showed that the 3’ end ofthe DNA coincided exactly with the 5‘ end of pgRNA (see Fig. 4) (Lien et al., 1987; Seeger et al., 1986). This observation implied that the 5’ end of pgRNA remained intact throughout synthesis of the minus strand, rather than being degraded by RNase H to release the nascent DNA for strand transfer or “jumping,” as occurs in retroviral reverse transcription. Thus, the 3’ DR1 was proposed as the site of minus strand initiation. Consistent with

183

HEPADNAVIRUSES

R

h h DR 1

2.

n A

n

J P

3.

kza

l u P

4.

Relaxed Circular (RC) DNA

FIG.4. Reverse transcription in hepadnaviruses: Model I. The pregenomic RNA is represented at the top, where R symbolizes terminal redundancy on pgRNA and DR1 and DR2 are direct repeats. Step 1:Minus strand DNA synthesis initiates within the 3' DRl sequences. The polymerase (P) provides the primer for minus strand synthesis through a covalent linkage between the first nucleotide on the DNA and a tyrosine residue in P. Step 2: Elongation of the minus strand DNA proceeds with concomitant degradation of the pgRNA by RNase H activity of the viral polymerase. Step 3: Minus strand synthesis is completed and the 5' terminal sequences of pgRNA are released by RNase H to provide the primer for plus strand DNA synthesis. Step 4: The plus strand RNA primer is translocated from DR1 to DR2, and plus strand DNA synthesis is initiated. Step 5: Plus strand jumping occurs, mediated by the terminally redundant sequences (r) on minus strand DNA, to form the relaxed circular genome present in mature virions. Step 6: Following infection of the host cell, plus strand DNA synthesis proceeds to completion, the primers and redundant sequences are removed, and the free ends of the plus and minus strand DNAs are ligated with supercoiling to form the covalent closed circular molecule (CCC)that provides the template for transcription.

184

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

this model, genetically marked sequences in the 3’ DR1 but not in the 5’ DR1 were transferred into the 5’ end of the minus-strand DNA in WHV (Seeger and Maragos, 1990). Furthermore, in DHBV, deletion of the 3’ DR1 was found to impair minus-strand synthesis, whereas deletion of the 5’ DR1 allowed minus strand synthesis to proceed (Condreay et al., 1992). The significance of the short terminal redundancy on the minus strand in mediating the plus strand jump is discussed below. It is interesting to note that RNA packaging depends on interactions between P and the E encapsidation signal situated at the 5’ end of pgRNA. Thus, to carry out the early steps of viral assembly and DNA replication, the polymerase must recognize specific sequences located on opposite ends of the viral genome. The mechanism by which these steps occur and the cis-acting signals responsible for positioning the P protein at the 3’ end of pgRNA are not well understood. Deletion of the 3’ DR1 in WHV has revealed two cryptic initiation sites for minus strand synthesis upstream of DR2 (Seeger and Maragos, 1990, 1991). Alignments of these cryptic sites and other novel initiation sites produced by the introduction of point mutations into DR1 have identified a common sequence motif that seems to specify the position of the 5’ end of minus strand in each case. This short motif consists of the sequences UUUC. Although over 30 of these motifs are present in the WHV genome, only a few are employed by the mutants as sites for minus strand initiation, suggesting that additional sequences play a role in positioning the polymerase on the RNA. Analysis of deletions in the WHV genome has indicated that such signals may be located in a region between the surface antigen ORF and the 3’ DR1 (Seeger and Maragos, 1990, 1991). Similar experiments have been carried out in DHBV transfections to determine the cis-acting signals responsible for directing initiation of minus strand DNA synthesis (Condreay et al., 1992). Deletion of the 3’ DR1 sequences in DHBV revealed two cryptic sites of minus strand initiation located between the 3’ DR1 and the poly(A) site. A short motif of the sequences UUA was common to both cryptic initiation sites and the wild-type site at DR1. As in the case of WHV, this motif is almost certainly too short to constitute the full signal for minus strand initiation, and additional unidentified sequences are likely to participate in specifying the correct origin. Preliminary reports on DHBV have indicated that deletions and mutations of the sequences immediately upstream of the 3’ DR1 significantly reduced the efficiency of correct minus strand initiation (Loeb and Ganem, 1992,1993).Thus, in both WHV and DHBV, very short sequence motifs were found to specify the site for initiation of minus-strand synthesis. Moreover, in both

HEPADNAVIRUSES

185

systems, sequences located just outside of the terminal redundancy at the 3’ end of pgRNA have been implicated as part of a signal necessary for positioning the polymerase at the correct origin. The successful expression of DHBV P protein in functional form in uitro has enabled investigators to study the priming reaction in more detail (Tavis et al., 1994; Wang and Seeger, 1993). The results of the analyses have lead to the proposal of a modified scheme of hepadnaviral reverse transcription, which is summarized in Fig. 5 . Prior to these experiments, it had been assumed that priming of the minus strand occurs at the position where the 5’ end of minus strand is located in the mature genome, that is, within the 3’ DR1 sequences. Contrary to this prediction, in uitro experiments with DHBV have indicated that minus strand synthesis initiates within sequences in the E encapsidation signal; the nascent DNA is then dissociated from its template and translocated to complementary sequences in DR1 (Tavis et al., 1994; Wang and Seeger, 1993).Primer extensions revealed two DNA products synthesized by the in uitro expressed DHBV polymerase, corresponding to minus strand DNAs initiating at the expected DR1 site and at an additional site located within the bulge sequences of the E stem-loop. Both sites contained a 4-nucleotide sequence that is complementary to the first 4 nucleotides of the DHBV minus strand DNA. To investigate which of the two sites provided the template for priming, mutations were introduced at each of the sites. Surprisingly, mutations at the E bulge were found to direct the synthesis of DNA carrying altered sequences, whereas mutations at DR1 had no effect on the nucleotides incorporated into the DNA (Tavis et al., 1994; Wang and Seeger, 1993). Moreover, the presence of mutations at either DR1 or E selectively abolished synthesis of the DNA product initiated at DR1, without affecting initiation at the E site. When the same mutation was introduced into both DR1 and E, a partial rescue of the DR1 initiated product resulted, indicating that DR1 initiation was dependent on homology with the sequences at E (Tavis et al., 1994; Wang and Seeger, 1993). Mutational analysis in the cell-free expression system has been used to examine RNA and P protein requirements in the priming reaction (Wang et al., 1994). The results of the analysis indicate that the E RNA sequences are both necessary and sufficient for protein-mediated priming. Unexpectedly, these experiments also identified a DNA polymerase activity in P protein that does not depend on the presence of E or the Qr-96 primer. This activity results in the synthesis of DNA that is not covalently bound to protein and displays heterogeneous 5’ ends; this DNA is possibly primed by random RNA primers that fortuitously hybridize with the template.

186

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

1.

An

FIG. 5. Reverse transcription in hepadnaviruses: Model 11. The pregenomic RNA, including the bulged stem-loop formed by the E encapsidation sequences, is shown at top; R, terminal redundancy on pgRNA; DR1 and DR2, direct repeat sequences. Step 1: Minus strand DNA synthesis initiates within the bulge sequences of the 5' E stem-loop structure. The primer for the minus strand is provided by the polymerase (P) that forms a covalent linkage with the first nucleotide in the chain via a tyrosine residue. DNA synthesis arrests after the incorporation of a few nucleotides. Step 2: Translocation of the minus strand DNA oligomer to the 3' DR1 site. Steps 3-6: As in Fig. 4.

Although mutations in the E bulge region were previously shown to have no effect on RNA packaging in transfections, the impact of these mutations on reverse transcription was not examined in the earlier studies (Knaus and Nassal, 1993; Pollack and Ganem, 1993).To inves-

HEPADNAVIRUSES

187

tigate the relevance of €-mediated priming in uiuo, mutations were introduced into the DHBV genome for further analysis in transfected cells. The results of these studies have confirmed the model of e-templated minus strand priming (Tavis et al., 1994; Wang and Seeger, 1993). Consistent with the model, point mutations introduced into the bulge of the E stem-loop were transferred into the DR1 sequences, whereas a point mutation in the 3' DR1 was not maintained in culture but readily reverted to wild type, These results indicated that the minus strand sequences were not derived from the DR1 site, but were determined by the nucleotides present in the E bulge. More extensive mutations in the 5' E signal completely abrogated DNA synthesis, which was partially rescued by the presence of compensatory mutations in the 3' DR1. Thus, reverse transcription in DHBV depends on the presence of homologous sequences in the 5' E and 3' DR1, as expected if the nascent DNA must be translocated from one site to the other. A similar mutational analysis in HBV has confirmed that the 5' E sequences play a critical role in reverse transcription in uiuo, consistent with minus strand priming at the E site (Fallows and Goff, 1995). However, in this study, viruses carrying compensatory mutations at both the 5' E and 3' DR1 sites were also unable to synthesize DNA. These results and the incomplete rescue in compensatory mutants of DHBV indicate that complementary sequences at the two sites are necessary but not sufficient to allow minus strand synthesis. One possible explanation is that specific primary sequences in either DR1 or E serve an additional function as a recognition signal necessary for translocation of the primer. An attractive feature of the €-mediated priming model is that it suggests that the processes of RNA packaging and DNA synthesis are closely connected, thus providing a possible mechanism for the coordinated regulation of viral assembly and reverse transcription.

C. Plus Strand DNA Priming and Synthesis During the synthesis of minus strand DNA, the newly copied pgRNA is degraded by the RNase H domain of P (Radziwill et al., 1990; Summers and Mason, 1982). This action of the RNase H releases a short oligoribonucleotide from the 5' end of the pgRNA which serves as the primer for plus strand synthesis (Lien et al., 1986, 1987; Seeger et al., 1986). The plus strand primer is 15 to 18 nucleotides in length and contains the 5' cap structure and all of the sequences at the 5' end of pgRNA through DR1. Mapping studies have shown that plus strand

188

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

synthesis initiates at the 3’ end of the DR2 sequences (Lien et al., 1986; Molnar-Kimber et al., 1984; Seeger et al., 1986; Will et al., 1987).Thus, for proper plus strand synthesis to proceed, the primer must be translocated on the minus strand from its original position at DR1 to the homologous sequences at DR2. Two alternative scenarios could account for generation of the plus strand primer. The cleavage reaction that determines the position of the 3’ end of the primer could occur at the DR1 site prior to translocation, or a short piece of RNA could be translocated to the DR2 site where the unduplexed bases at the 3‘ end could be trimmed off. However, extension of the sequence similarity between DR1 and DR2 did not increase the length of the primer, indicating that the final cleavage site is determined prior to translocation (Seeger and Maragos, 1989). The possibility that sequences at the DR1 site are responsible for determining the specificity of the cleavage has also been examined by mutational analyses. Although nucleotide substitutions introduced in the vicinity of the cleavage site had no impact on the position of cleavage (Seeger and Maragos, 1989; Staprans et al., 1991), insertions and deletions were found to alter the cleavage specificity (Loeb et al., 1991). In the latter mutants, regardless of the sequences present, the position of the cleavage was maintained at a constant distance of 15 to 18 nucleotides from the end of the pgRNA. These results demonstrated that the cleavage reaction is sequence-independent, instead relying on measurement of the distance from the 5’ end of the pregenome. The processes of primer cleavage and primer translocation were found to be separable by mutations in the sequences at the 5’ end of pgRNA (Staprans et al., 1991). Even though mutations in the DR sequences had no effect on cleavage, the mutants were unable to translocate the primer to DR2. Not surprisingly, the translocation step depended on the presence of similar sequences at DR1 and DR2. Primer translocation was also inhibited by the presence of mutations in the sequences immediately 3‘ of DR1, thus implicating this region as part of a signal required for translocation. Although DHBV mutants that were defective in translocation were unable to synthesize relaxed circular genomes, they produced a linear form of duplex DNA by elongation of the untranslocated primer (Condreay et al., 1992; Loeb et al., 1991; Staprans et al., 1991). In contrast, a mutation of WHV that abrogated primer translocation was able to prime plus strand DNA from an alternative site located between DR2 and the 5’ end of the minus strand (Seeger and Maragos, 1989). The sequences at the new priming site comprised a purine-rich motif that was apparently resistant to RNase H digestion, similar to the polypurine tracts used for

HEPADNAVIRUSES

189

plus strand priming in retroviral reverse transcription. In situ priming of plus strand from the DR1 site was not observed in the WHV translocation mutants. The DHBV genome does not contain a purine-rich region, but the motif is present in genomes of GSHV and HBV; this alternative mechanism for plus strand priming may be common to all the mammalian hepadnaviruses. After primer translocation, plus strand synthesis proceeds to the 5’ end of the minus strand. A t this point, the plus strand must jump from the 5’ end to the 3’ end of the minus strand template in order for elongation to continue. As a result of the terminal redundancy on the minus strand, the nascent plus strand carries sequences that are complementary to both ends of the minus strand DNA. To carry out the jump, the plus strand dissociates from the 5‘ end of the minus strand and reanneals to the complementary sequences located on the 3’ end of the template. Preliminary evidence suggests that jumping occurs before the plus strand reaches the end of the template (Condreay et al., 1992; Loeb and Ganem, 1993). This finding indicates that less than 10 nucleotides of complementarity on the plus strand are sufficient to effect the jump. The proteins responsible for directing the plus strand primer translocation and strand transfer remain to be identified. The core protein is probably involved in at least some of these steps, as mutations in C that interfered with either primer translocation or plus strand elongation have been described (Nassal, 1992; Yu and Summers, 1991). As the major protein responsible for the enzymatic reactions in reverse transcription, the polymerase seems likely to play a central role in these processes. However, there have been no reports to indicate specifically how the P protein is involved.

D . Formation of Covalently Closed Circular DNA In mature virions of mammalian hepadnaviruses, the plus strands are considerably shorter than full length, and synthesis must be completed in order to form CCC DNA (Ganem et al., 1982; Summers et al., 1975,1978). In contrast, most of the plus strands in DHBV virions are complete, but the RNA primer is not displaced by the polymerase and remains hybridized to the DR2 sequences (Lien et al., 1987; Mason et al., 1980). These later steps in viral replication have not been studied extensively. Consequently, many questions remain to be answered about the process of CCC DNA formation. How are the terminal protein on minus strand and the RNA primer on plus strand removed? Is it the viral P protein or a host cell polymerase that copies the DR2 se-

190

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

quences to complete plus strand DNA? Finally, what proteins are responsible for ligating the DNA ends together and supercoiling the resulting circular molecule?

VI. CONCLUDING REMARKS Progress in the field of hepadnaviral research has advanced considerably through the use of tissue culture systems for viral expression and the development of systems for the in uitro expression of the polymerases. Although the essential similarities between hepadnaviruses and the classically studied retroviruses have been confirmed, differences between the two groups of viruses have become strikingly apparent. Synthesis of the hepadnaviral polymerase as a distinct translation product and the central role played by this protein in directing RNA encapsidation may be contrasted with the retroviral model in which polymerase is expressed as a gag-pol fusion protein and the major determinants of RNA packaging specificity reside on the gag protein. Whereas retroviruses typically carry out both proteolytic processing and RNA splicing, there is no evidence for either protein processing or splicing in hepadnaviruses. The use of a protein primer for the synthesis of minus strand DNA is without precedent among the retroviruses, which employ a host cell tRNA for this function. In both cases, strand transfer is mediated through the use of terminal sequence similarities. However, the plus and minus strand primer translocations intrinsic to hepadnaviral reverse transcription are very different from the strongstop DNA translocations that occur in retroviruses. A characteristic feature of hepadnaviral replication is transcription from the unintegrated CCC DNA and the maintenance of this DNA through a pathway of intracellular amplification, whereas retroviral expression and persistence depend on a mechanism of genomic integration. Although the processes of hepadnaviral assembly and reverse transcription are now understood in some detail, research in the field continues to reveal surprising discoveries about the unique strategies employed in the replication of these fascinating viruses. The early steps of viral infection and uncoating remain a phase of the viral life cycle about which little is known because of the lack of infectable cell lines. Success in the on-going efforts to identify host cell receptors would allow the construction of such lines and would open the field to research on this important topic. As our knowledge of hepadnaviral replication grows, the possibilities should increase for the development of urgently needed drug therapies to treat the millions of chronically infected individuals.

HEPADNAVIRUSES

191

REFERENCES Acs, G., Sells, M. A., Purcell, R. H., Price, P., Engle, R., Shapiro, M., and Popper, H. (1987).Proc. Natl. Acad. Sci. U S A . 84, 4641-4644. Aldrich, C. E., Coates, L., Wu, T.-T., Newbold, J., Tennant, B. C., Summers, J., Seeger, C., and Mason, W. S. (1989).Virology 172,247-252. Almeida, J. D., Rubenstein, D., and Stott, E.J. (1971).Lancet 2, 1225-1227. Bartenschlager, R., and Schaller, H. (1988).EMBO J . 7,4185-4192. Bartenschlager, R., and Schaller, H. (1992).EMBO J . 11, 3413-3420. Bartenschlager, R.,Junker-Niepmann, M., and Schaller, H. (1990).J . Virol. 64, 53245332. Bartenschlager, R., Kuhn, C., and Schaller, H. (1991).Nucleic Acids Res. 20, 195-202. Bavand, M. R., and Laub, 0. (1988).J. Virol. 62,626-628. Birnbaum, F., and Nassal, M. (1990).J . Virol. 64,3319-3330. Blum, H. E., Zhang, Z.-S., Galun, E., von Weizsacker, F., Garner, B., Liang, T. J., and Wands, J. R. (1992).J . Virol. 66, 1223-1227. Blumberg, B. S., Alter, H. J., and Visnich, S. (1965).JAMA, J . Am. Med. Assoc. 191, 541-546. Bosch, V., Bartenschlager, R., Radziwill, G., and Schaller, H. (1988).Virology 166,475485. Buscher, M.,Reiser, W., Will, H., and Schaller, H. (1985).Cell (Cambridge, Mass.) 40, 717-724. Calvert, J., and Summers, J. (1994).J. Virol. 68,2084-2090. Chang, C., Jeng, K., Hu, C., Lo, S. J., Su, T., Ting, L.-P., Chou, C.-K., Han, S., Pfaff, E., Salfeld, J., and Schaller, H. (1987).EMBO J. 6, 675-680. Chang, L.J., Dienstag, J., Ganem, D., and Varmus, H. E. (1989a).Hepatology (Baltimore) 10,332-335. Chang, L.-J., Pryciak, P., Ganem, D., and Varmus, H. (1989b).Nature (London) 337, 364-368. Chang, LA., Ganem, D., and Varmus, H. E. (1990a).Proc. Nutl. Acad. Sci. U.S.A.87, 5158-5162. Chang, L A , Hirsch, R., Ganem, D., and Varmus, H. E. (1990b).J . Virol. 64,5553-5558. Chen, H.-S., Kaneko, S., Girones, R., Anderson, R. W., Hornbuckle, W. E., Tennant, B. C., Cote, P. J., Gerin, J. L., Purcell, R. H., and Miller, R. H. (1993).J. Virol. 67, 12181226. Chen, P.J., Chen, C.-R., Sun, J.-L., and Chen, D.-S. (1989).J. Virol. 63,4165-4171. Chen, Y.,Robinson, W. S., and Marion, P. (1992).J . Virol. 66,1282-1287. Chen, Y.,Robinson, W. S., and Marion, P. L. (1994).J. Virol. 68,5232-5238. Chiang, P.-W., Jeng, K.-S., Hu, C.-P., and Chang, C. (1992).Virology 186, 701-711. Condreay, L. D., Aldrich, C. E., Coates, L., Mason, W. S., and Wu, T.-T. (1990).J . Virol. 64, 3249-3258. Condreay, L. D., Wu, T.-T., Aldrich, C. E., Delaney, M. A., Summers, J., Seeger, C., and Mason, W. S. (1992).Virology 188, 208-216. Dane, D. S., Cameron, C. H., and Briggs, M. (1970).Lancet 1, 695-698. Doolittle, R. F., Feng, D.-F., Johnson, M. S., and McClure, M. A. (1989).Q.Rev. Biol. 64, 1-30. Eble, B. E.,Lingappa, V. R., and Ganem, D. (1986).Mol. Cell. Biol. 6, 1454-1463. Enders, G. H., Ganem, D., and Varmus, H. (1985).Cell (Cambridge, Mass.) 42, 297308. Enders, G. H., Ganem, D., and Varmus, H. E. (1987).J . Virol. 61,35-41. Fallows, D. A., and Goff, S. P. (1995).J. Virol. 69,3067-3073.

192

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

Galibert, F., Mandart, E., Fitoussi, F., Tiollais, P., and Charnay, P. (1979).Nature (London) 281,646-650. Ganem, D. (1982).Rev. Infect. Dis. 4, 1026-1047. Ganem, D., Greenbaum, L., and Varmus, H. E. (1982).J. Virol. 44,374-383. Gerlich, W. H., and Robinson, W. S. (1980).Cell (Cambridge, Mass.) 21,801-809. Gripon, P.,Diot, C., Theze, N., Fourel, I., Loreal, O., Brechot, C., and Guguen-Guillouzo, C. (1988).J. Virol. 62,4136-4143. Hatton, T., Zhou, S., and Standring, D. N. (1992).J. Virol. 66, 5232-5241. Hirsch, R. C., Lavine, J. E., Chang, L., Varmus, H. E., and Ganem, D. (1990).Nature (London) 344,552-555. Hirsch, R. C., Loeb, D. D., Pollack, J. R., and Ganem, D. (1991).J. Virol. 65,3309-3316. Horwich, A. L., Furtak, K., Pugh, J., and Summers, J. (1990).J. Virol. 64,642-650. Howe, A,, Elliott, J. F., and Tyrrell, D. L. J. (1992).Biochem. Biophys. Res. Commun. 189, 1170-1176. Huang, M., and Summers, J. (1991).J. Virol. 65, 5435-5439. Jean-Jean, O., Levrero, M., Will, H., Perricaudet, M., and Rossignol, J. M. (1989a). Virology 170,99-106. Jean-Jean, O., Weimer, T., de Recondo, A.-M., Will, A. H., and Rossignol, J.-M. (1989b).J. Virol. 63, 5451-5454. Junker-Niepmann, M., Bartenschlager, R., and Schaller, H. (1990).EMBO J. 9, 33893396. Kaplan, P. M., Greenman, R. L., Gerin, J. L., Purcell, R. H., and Robinson, W. S. (1973).J. Virol. 12, 995-1005. Khudyakov, Y. E., and Makhov, A. M. (1989).FEES Lett. 243, 115-118. Knaus, T., and Nassal, M. (1993).Nucleic Acids Res. 21, 3967-3975. Kozak, M. (1989).J. Cell Biol. 108,229-241. Lavine, J., Hirsch, R., and Ganem, D. (1989).J. Virol. 63, 4257-4263. Levin, J. G., Hatfield, D. L., Oroszlan, S., and Rein, A. (1993).In “Reverse Transcriptase” (A. M. Skalka and S. P. Goff, eds.), pp. 5-31. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Li, J.-S., Cova, L., Buckland, R., Lambert, V., Deleage, G., and Trepo, C. (1989).J. Virol. 63,4965-4968. Lien, J.-M., Aldrich, C. E., and Mason, W. S. (1986).J. Virol. 57, 229-236. Lien, J.-M., Petcu, D. J., Aldrich, C. E., and Mason, W. S. (1987).J. Virol. 61,3832-3840. Loeb, D., and Ganem, D. (1992).Meeting abstract from Molecular Biology of Hepatitis B Viruses. UCSD, San Diego, California, 1992. Loeb, D. D., and Ganem, D. (1993).In “Reverse Transcriptase” (A. M. Skalka and S. P. Goff, eds.), pp. 329-355. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Loeb, D. D., Hirsch, R., and Ganem, D. (1991).EMBO J. 10,3533-3540. McClure, M. A. (1993).I n “Reverse Transcriptase” (A. M. Skalka and S. P. Goff, eds.), pp. 425-444. Cold Spring Harbor Laboratories, Cold Spring Harbor, New York. Mack, D. H., Block, W., Nath, N., and Sninsky, J. J. (1988).J. Virol. 62,4786-4790. Marion, P. L., Oshiro, L., Regnery, D. C., Scullard, G. H., and Robinson, W. S. (1980). Pmc. Natl. Acad. Sci. U S A . 77, 2941-2945. Mason, W. S., Seal, G., and Summers, J. (1980).J. Virol. 36, 829-836. Mason, W.S., Aldrich, C., Summers, J., and Taylor, J. M. (1982).Proc. Natl. Acad. Sci. U S A . 79,3997-4001. Mason, W. S., Halpern, M. S., England, J. M., Seq, G., Egan, J., Coates, L., Aldrich, C., and Summers, J. (1983).Virology 131,375-384. Milich, D. R., Jones, J. E., Hughes, J. L., Price, J., Raney, A. K., and McLachlan, A. (1990).Proc. Natl. Acad. Sci. U.S.A. 87,6599-6603.

HEPADN AVIRUSES

193

Miller, R. H. (1988).Virology 164, 147-155. Miller, R. H., and Robinson, W. S. (1984).Virology 137, 390-399. Miller, R. H.,and Robinson, W. S. (1986).Proc. Natl. Acad. Sci.U S A . 83,2531-2535. Miller, R. H., Tran, C.-T., and Robinson, W. S. (1984).Virology 139,53-63. Molnar-Kimber, K. L.,Summers, J., Taylor, J. M., and Mason, W. S. (1983).J. Virol. 45, 165-172. Molnar-Kimber, K. L., Summers, J. W., and Mason, W. S. (1984).J. Virol. 51, 181-191. Nassal, M. (1988).Gene 66,279-294. Nassal, M. (1992).J . Virol. 66,4107-4116. Nassal, M., and Schaller, H. (1993).Trends Microbiol. 1, 221-228. Nassal, M., Junker-Niepmann, M., and Schaller, H. (1990).Cell (Cambridge,Mass.) 63, 1357-1363. Oberhaus, S. M., and Neubold, J. E. (1993).J. Virol. 67, 6558-6566. Onodera, S.,Onda, H., Sasada, R., Igarishi, K., Sugino, Y., and Nishioka, K. (1983).J . Med. Virol. 10, 147-155. Ou, J.-H., Laub, O., and Rutter, W.(1986).Proc. Natl. Acad. Sci.U S A . 83, 1578-1582. Pasek, M., Goto, T., Gilbert, W., Zink, B., Schaller, H., MacKay, P., Leadbetter, G., and Murray, K. (1979).Nature (London) 282,575-579. Patzer, E. J., Nakamura, G. R., Simonsen, C. C., Levinson, A. D., and Brands, R. (1986). J . Virol. 58, 884-892. Pollack, J. R.,and Ganem, D. (1993).J. Virol. 67, 3254-3263. Pollack, J. R., and Ganem, D. (1994).J. Virol. 68, 5579-5587. Pugh, J. C., Yaginuma, K., Koike, K., and Summers, J. (1988).J. Virol. 62,3513-3516. Radziwill, G., Zentgraf, H., Schaller, H., and Bosch, V. (1988).Virology 163, 123132. Radziwill, G., Tucker, W., and Schaller, H. (1990).J. Virol. 64, 613-620. Robinson, W. S.,Clayton, D. A., and Greenman, R. L. (1974).J. Virol. 14,384-391. Roychoudhury, S.,and Shih, C. (1990).J. Virol. 64, 1063-1069. Roychoudhury, S.,Faruqui, A. F., and Shih, C. (1991).J. Virol. 65,3617-3624. Ruiz-Opazo, N., Chakraborty, P. R., and Shafritz, D. A. (1982).Cell (Cambridge,Mass.) 29, 129-138. Russnak, R. H. (1991).Nucleic Acids Res. 19, 6449-6456. Russnak, R.,and Ganem, D. (1990).Genes Deu. 4,764-776. Salas, M. (1991).Annu. Reu. Biochem. 60,39-71. Sattler, F., and Robinson, W. S. (1979).J. Virol. 32, 226-233. Scheck, N., Fischer, M., and Schaller, H. (1991).In “Molecular Biology of the Hepatitis B Virus” (A. McLachlan, ed.), pp. 181-192. CRC Press, Boca Raton, Florida. Schlicht, H.J., Bartenschlager, R., and Schaller, H. (1989a).J . Virol. 63,2995-3000. Schlicht, H A , Radziwill, G., and Schaller, H. (1989b).Cell (Cambridge,Mass.) 56,8592. Schodel, F., Weimer, T., Will, H., and Sprengel, R. (1988).AIDS Res. Hum. Retroviruses 4, 9. Seeger, C., and Maragos, J. (1989).J . Virol. 63, 1907-1915. Seeger, C., and Maragos, J. (1990).J . Virol. 64, 16-23. Seeger, C., and Maragos, J. (1991).J. Virol. 65, 5190-5195. Seeger, C., Ganem, D., and Varmus, H. E. (1986).Science 232,477-484. Seeger, C., Baldwin, B., and Tennant, B. C. (1989).J . Virol. 63,4665-4669. Seifer, M., and Standring, D. N. (1993).J . Virol. 67, 4513-4520. Sells, M. A,, Chen, M.-L., and Acs, G. (1987).Proc. Natl. Acad. Sci. U S A . 84, 10051009. Sells, M. A., Zelent, A. Z., Shvartsman, M., and Acs, G. (1988).J. Virol. 62,2836-2844.

194

DOROTHY A. FALLOWS AND STEPHEN P. GOFF

Shih, C., Li, L.-S., Roychoudhury, S., and Ho, M.-H. (1989).Proc. Nutl. Acud. Sci. U.S.A. 86,6323-6327. Sprengel, R., Kuhn, C., Will, H., and Schaller, H. (1985). J . Med. Virol. 15, 323-333. Sprengel, R., Kaleta, E. F., and Will, H. (1988).J. Virol. 62, 3832-3839. Staprans, S., Loeb, D. D., and Ganem, D. (1991).J. Virol. 65, 1255-1262. Storch, G. A. (1993).Pediatr. Infect. Dis. J. 12, 427-453. Su, T.-S., Lai, C.J., Huang, J.-L., Lin, L.-H., Yauk, Y.-K., Chang, C. M., Lo, S. J., and Han, S.-H. (1989).J . Virol. 63, 4011-4018. Summers, J., and Mason, W. S. (1982). Cell (Cambridge,Muss.) 29, 403-415. Summers, J., O'Connell, A,, and Millman, I. (1975). Proc. Nutl. Acud. Sci. U S A . 72, 4597-4601. Summers, J., Smolec, J. M., and Snyder, R. (1978). Proc. Nutl. Acud. Sci. U S A . 75, 4533-4537. Summers, J., Smith, P. M., and Horwich, A. L. (1990). J . Virol. 64,2819-2824. Summers, J., Smith, P., Huang, M., and Yu, M. (1991).J . Virol. 65, 1310-1317. Sureau, C., Eichberg, J. W., Hubbard, G. B., Romet-Lemonne, J. L., and Essex, M. (1988). J . Virol. 62, 3064-3067. Suzuki, T., Masui, N., Kajino, K., Saito, I., and Miyamura, T. (1989).J. Virol. 86,84228426. Tagawa, M., Omata, M., and Okuda, K. (1986). Virology 152,477-482. Tavis, J. E., and Ganem, D. (1993).Proc. Nutl. Acud. Sci. U.S.A.90,4107-4111. Tavis, J. E., Perri, S., and Ganem, D. (1994). J . Virol. 68, 3536-3543. Toh, H,. Hayashida, H., and Miyata, T. (1983). Nature (London) 305,827-829. Tuttleman, J. S,. Pourcel, C., and Summers, J. (1986a).Cell (Cambridge,Muss.)47,451460. Tuttleman, J. S., Pugh, J. C., and Summers, J. W. (1986b).J . Virol. 58, 17-25. Valenzuela, P., Quiroga, M., Zaldivar, J., Gray, P., and Rutter, W. J. (1980).Anim. Virus Genet. (ICN-UCLA Symp. Mol. Cell. Biol. 18), 57-70. Wang, G.-H., and Seeger, C. (1992). Cell (Cambridge,Muss.) 71, 1-20. Wang, G.-H., and Seeger, C. (1993).J. Virol. 67, 6507-6512. Wang, G.-H., Zoulim, F., Leber, E. H., Kitson, J., and Seeger, C. (1994). J . Virol. 68, 8437-8442. Weber, M., Bronsema, V., Bartos, H., Bosserhoff, A., Bartenschlager, R., and Schaller, H. (1994). J . Virol. 68, 2994-2999. Weiser, B., Ganem, D., Seeger, C., and Varmus, H. E. (1983).J. Virol. 48, 1-9. Will, H., Reiser, W., Weimer, T., Pfaff, E., Bucher, M., Sprengel, R., Cattaneo, R., and Schaller, H. (1987). J . Virol. 61, 904-911. Wu, T.-T., Coates, L., Aldrich, C . E., Summers, J., and Mason, W. S. (1990).Virology 175, 255-261. Wu, T.-T., Condreay, L. D., Coates, L., Aldrich, C., and Mason, W. (1991). J. Virol. 65, 2155-2163. Yaginuma, K., Shirakata, Y., Kobayashi, M., and Koike, K. (1987).Proc. Nutl. Acud. Sci. U.S.A. 84, 2678-2682. Yu, M., and Summers, J. (1991). J . Virol. 65, 2511-2517. Zhou, S., and Standring, D. N. (1992). Proc. Nutl. Acud. Sci. U S A . 89, 10046-10050. Zoulim, F., and Seeger, C. (1994). J . Virol. 68, 6-13. Zoulim, F., Saputelli, J., and Seeger, C. (1994).J. Virol. 68, 2026-2030.

ADVANCES IN VIRUS RESEARCH, VOL. 46

CELL TYPES INVOLVED IN REPLICATION AND DISTRIBUTION OF HUMAN CYTOMEGALOVIRUS Bod0 Plachter,’ Christian Sinzger,t and Gerhard Ja hnt ‘Institut fur Klinische und Molekulare Viralogie Univenitat Erlangen-Nurnberg 0-91054 Erlangen, Germany tHygiene-lnstitut der Univenitat Tiibingen Abteilung Medizinische Virologie und Epidemiologie der Viruskrankheiten D-72076 Tiibingen, Germany

I. Introduction 11. Determinants of Human Cytomegalovirus Infection A. Viral Gene Expression and Replication in Permissive Culture Cells B. Nonpermissive Human Cytomegalovirus Infection C. Initial Events in Human Cytomegalovirus Infection D. Role of Virion Proteins in Initiating Infection E. Strain Variabilities 111. Organ Tropism of Human Cytomegalovirus IV. Cell Types Involved in Acute Human Cytomegalovirus Disease A. Human Cytomegalovirus Infection in Tissue Cells B. Hematopoietic System and Circulating Cells V. Viral Spread and Pathogenesis A. Modes of Transmission B. Cell Q p e s Involved in Spread and Pathogenesis VI. Latent Cytomegalovirus Infection A. Latent Murine Cytomegalovirus Infection as Model B. Site of Human Cytomegalovirus Latency C. Cell Culture Models for Human Cytomegalovirus Latency VII. Summary Fkferences

1. INTRODUCTION Human cytomegalovirus (HCMV) was first noted more than a century ago by typical morphological alterations of infected cells. In 1881, Ribbert found large “protozoan-like” cells in kidney sections of a stillborn infant, an observation published over 20 years later (Ribbert, 1904). Similar findings were reported by several authors during the early 19OOs, and pathological changes were attributed to either syphilitic or protozoan infection (Jesionek and Kiolemenoglou, 1904; Lowenstein, 1907; Pisano, 1910). Later, Cole and Kuttner (1926) substantiated that cytopathic changes of cells caused by viral infection were responsible for these histological findings; they could show that a virus 195

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

196

BOD0 PLACHTER et al.

was present in submaxillary glands of guinea pigs and that the presence of this virus was associated with cells containing inclusion bodies (Huang and Kowalik, 1993). After that, “cytomegalic inclusion disease” (CID) became widely accepted as a frequent viral cause of morbidity and mortality in newborns. A first systematic study on the prevalence of this infection in children, who had died for various reasons, was carried out by Farber and Wolbach (1932). These authors found typical cytomegalic cells (owl eye cells) in 14% of the salivary glands of those infants (for a more detailed coverage of the history of HCMV, see Ho, 1991). With the introduction of cell culture techniques, HCMV was independently isolated by three different laboratories in the early 1950s (Rowe et al., 1956; Smith, 1956; Weller et al., 1957). Since then, the virus has been well recognized as an ubiquitous pathogenic agent, which causes a wide array of clinical symptoms not only in the neonate but also in adults. Especially under conditions of immunosuppression, as in patients with acquired immunodeficiency syndrome (AIDS) or in transplant recipients, HCMV infection may lead to severe and lifethreatening disease (reviewed by Alford and Britt, 1990; Ho, 1991). The molecular biology of this virus has therefore been the subject of intensive studies during the past years and has been reviewed (Stinski, 1991; Gibson, 1993; Mocarski, 1993).Investigations of virus-cell interactions and of host factors that influence multiplication and distribution of HCMV, however, have been hampered by the strict species specificity of the virus, a characteristic of HCMV already noted by Smith (19561, and by the lack of an animal model. Through the invention of new molecular techniques and the availability of complete sequence data of the HCMV genome (Chee et al., 1990), more information is being gathered about virus-cell interactions. This review primarily focuses on recent findings concerning the interaction of HCMV with different host cells in uitro and in uivo. The role of different cell populations in replication, latency, and distribution of HCMV is discussed.

11. DETERMINANTS OF HUMAN CYTOMEGALOVIRUS INFECTION Even during the first attempts to isolate the “salivary gland virus,” it became evident that HCMV has a very narrow host cell range for efficient productive infection in culture (&we et al., 1956; Smith, 1956; Weller et al., 1957). Only primary and, more recently, transformed human fibroblasts have been found to support HCMV growth to high titers (Compton, 1993). No other culture system, may it consist of

HUMAN CYTOMEGALOVIRUS

197

primary or immortalized cells, has been found that supports HCMV replication in a way comparable to human fibroblasts (Table I). In addition, replication of HCMV and production of progeny virions is highly species specific; other than human cells, replication and subsequent production of progeny virions have only been demonstrated in chimpanzee fibroblasts (Perot et al., 1992). The molecular mechanisms that determine the permissivity of cells for HCMV replication are not understood. In other viral systems, adsorption and penetration is one level at which the permissivity of cells is determined. However, in a number of studies it could be demonstrated that HCMV can penetrate a variety of cells of human and nonhuman origin without being able to replicate (Fioretti et al., 1973; Waner and Weller, 1974; Albrecht et al., 1976; Rosenthal et al., 1981; Einhorn et al., 1982; DeMarchi, 1983; Lafemina and Hayward, 1983, 1986; Rice et al., 1984; Smith, 1986; Nelson et al., 1987; Taylor and Cooper, 1989; Wright et al., 1994). In line with this are results of studies which identified surface receptor molecules that bind HCMV; some of these receptors have been found on a wide variety of cells (Keay et al., 1989; Taylor and Cooper, 1989, 1990; Adlish et al., 1990; Nowlin et al., 1991; Compton et al., 1993; Soderberg et al., 1993).Therefore, it appears reasonable to postulate that cellular factors determine the outcome of HCMV infection after viral entry.

A . Viral Gene Expression and Replication in Permissive Culture Cells After adsorption of HCMV to the cell surface, penetration and subsequent delivery of the viral DNA genome of 230 kb to the cell nucleus are rapidly performed (Smith and de-Harven, 1973). In addition, constituents of the viral tegument are detectable early within infected cells and are translocated to the nucleus prior to viral gene expression (Geballe et al., 1986; Britt and Vugler, 1987; Grefte et al., 1992a). The events that take place in a permissively infected cell are schematically depicted in Fig. 1. Viral gene expression and part of the maturation of virus particles occur in the nucleus of infected cells. Viral genes are expressed in a cascade fashion comparable to the replicative cycle of other herpesviruses (Roizman and Sears, 1990; Stinski, 1990; Mocarski, 1993). In analogy, three phases of viral gene expression have been operationally defined in the human fibroblast system (Honess and Roizman, 1974), Immediate early (IE) genes are transcribed in the absence of de nouo synthesis of viral proteins. In HCMV, these genes are assumed to carry out key regulatory functions in permissive as well as in latent infection. Proteins necessary for the replication of the

TABLE I SYSTEMS FOR HUMAN CYTOMEGAL~VIRUS INFECTION AND DETECTION OF VIRALANTIGENS, CELLCULTURE IN INFECTED CELLCULTURESQ DNA, OR VIRALPROGENY Cell type

Cell identification

HCMV strains used for infection

Viral gene products

CPE

Progeny virus

Reference

Bone marrow leukocytes, monocytes, PMNL, lymphocytes B lymphocytes, T lymphocytes, NK cells, monocytes T lymphocytes, stimulated by mixed lymphocyte culture Monocytes Macrophages

Morphology

Isolates

2.3% EA (Pol)

Einhorn and Ost (1984)

Immunostaining

Isolates

IEA ( ~ ~ 7 2 )

Rice et al. (1984)

Immunostaining

AD169

1.7% EA (?), LA

Immunostaining Immunostaining

AD169, isolates Isolate

Macrophages

Immunostaining

AD169

CD13+ cells, CD8 lymphocytes, CD14 monocytes

FACS

Bone marrow fibroblasts, bone marrow stem cells Bone marrow fibroblasts Bone marrow adipocytes Bone marrow mononuclear cells Bone marrow CD34 cells Bone marrow-derived macrophages

Plastic adherence, immunorosetting Morphology

Immunomagnetic beads

AD169, isolate

AD169 AD169, isolates

Isolate

0.5% IEA (pp72) 40% IEA (pp72), LA 1.7% IEA ( ~ ~ 7 2LA % 21.6% IEA (pp72), 17% LA (pp65) on cell surface IEA, pp150 RNA Undefined

+ -

+ (cell associated)

+ (supernatant)

+

+

Braun and Reiser (1986) Scott et al. (1989) Ibanez et al. (1991) Lathey and Spector (1991) Soderberg et a1 (1993)

Reiser et a1 (1986) Apperley et al. (1989)

Minton et al. (1994)

Human arterial smooth muscle H W E C , HAEC HASMC HUVEC

Immunostaining

AD169

EA

?

?

Factor VIII Actin Immunostaining

AD169 Isolate


4

+

Brain endothelial Astrocyes Glioblastoma Neuroblasotoma Macrophagal-microglial

Factor VIII GFAP

Towne

IEA (pp72), LA

+ + neg

Immunostaining

Towne

IEA (pp72) IEA (PP72), LA EA

Kidney mesangial cells

Immunostaining

AD169 Isolate AD169 Isolate Towne, isolate

IEA, EA, LA

Glomerular epithelium, tubular epithelium Endometrial stromal cell

Immunostaining

Towne, isolate

IEA, EA, LA

Towne

IEA, EA, LA

Towne

0.01% IEA

?

45% IEA; EA; LA

Brain aggregate

Morphology

Tumilowicz (1990) Hosenpud et al. (1991) Waldman et al. (1991) Poland et al. (1990)

+

EM ?

cytomegaly neg

EM EM EM EM

+

+ +

+

TPC-1 (epithelial cell line) TPC-1 + sodium butyrate Immortalized fibroblasts

Morphology

AD169

+

+ +

Chimpanzee fibroblasts

Morphology

Towne

+

+

Pulliam (1991)

McCarthy et a1 (1991) Heieren et a1 (1988a) Heieren et a1 (1988b) Kowalik et a1 (1994) Tanaka et a1 (1991)

Compton (1993) Perot et a1 (1992)

PMNL, Polymorphonuclear cells; NK cells, natural killer cells; IEA, HCMV immediate early antigen; EA, HCMV early antigen; LA, HCMV late antigen; CPE, cytopathogenic effect; FACS, fluorescence-activated cell sorting; neg, negative; HUVEC, human umbilical vein endothelial cells; HAEC, human aortic endothelial cells; HASMC, human aortic smooth muscle cells; EM, electron microscopy. Q

200

B OD0 PLACHTER et al.

FIG.1. Schematic representation of the replication cycle of HCMV in permissive cells. Virions (V),noninfectious enveloped particles (NIEPs; N) and dense bodies (D) adsorb to the cell membrane via specific receptor interaction (left). After adsorption and penetration, the viral DNA and constituents of the viral tegument (pp65) are transported to the cell nucleus (central part). Viral gene expression is ordered in a sequential fashion. During the immediate early phase, viral regulatory proteins are expressed. Proteins necessary for replication of viral DNA and additional regulatory proteins are synthesized during the early phase. After DNA replication, the late genes are expressed and capsids are assembled within the nucleus (right). Subsequently, the capsids are tegumented and enveloped, then released from the cell.

viral DNA are expressed in the early phase, and their synthesis depends on the prior expression of immediate early proteins. After DNA replication, late genes are expressed, most of which encode proteins necessary for the generation of progeny virions. The molecular mechanisms that lead to this sequentially ordered replication have been reviewed in detail (Mocarski et al., 1990;Nelson et al., 1990;Stamminger and Fleckenstein, 1990; Stinski, 1991; Stinski et al., 1991; Mocarski, 1993). Earlier publications have shown that transcription during the immediate early phase of viral infection is restricted to discrete regions on the viral genome (DeMarchi et al., 1980; Wathen et al., 1981; Wathen and Stinski, 1982; McDonough and Spector, 1983). More detailed analyses have characterized four genomic regions of IE gene expression on the genome of HCMV (Stinski et al., 1983; Jahn et al., 1984; Stenberg et al., 1984,1985; Wilkinson et al., 1984; Weston, 1988; Tenney and Colberg-Poley, 1990, 1991a,b; Stasiak and Mocarski, 1992) (Fig. 2).

20 1

HUMAN CYTOMEGALOVIRUS

One of these regions is heavily transcribed immediately after HCMV infection of fibroblasts and has been described as the major immediate early region (Stinski et al., 1983;Jahn et al., 1984).Two loci of IE gene expression were found in the 20-kb Hind111 E DNA fragment (Jahn et al., 1984; Stenberg et al., 1984,1985;Akrigg et al., 1985; Plachter et al., 1988). Sequence analyses revealed that one of these two regions, although heavily transcribed into two RNA species of 5 kb throughout the infectious cycle, did not encode a larger polypeptide (Nelson et al., 1984; Plachter et al., 1988; Martignetti and Barrell, 1991). The role of this transcription unit (IE4) in HCMV replication is unclear. Some studies indicate, however, that this genomic region might be involved in transformation of cells to permanent growth (Nelson et al., 1984; el-Beik et al., 1986; Buonaguro et al., 1987).

-

I

I

4

2.30kb 1.7Okb

IE2.pp86 ( 8 2 4 6 D a ) . IE2-pp55 (52.55kDa)

1.50kb

IEZ.pp40 (40.45kDa)

1

++

+ect

sc

t

200

225 229

I

0

i

25

60

7s

100

125

150

115

kbp

FIG.2. Location of immediate early (IE) genes on the HCMV genome. The structure of the HCMV genome with unique (thin line) and repeated sequences (black boxes) is given at bottom. The location of the IE genes is given on top of that (nomenclature according to Chee et al., 1990). In the upper part, a blowup of the structure of the major IE gene region, coding for UL122 and UL123, is shown, including the splicing pattern of the different mRNAs. Numbers indicate the different exons; solid lines, exon sequences; thin lines, introns; ATG, translational start codon; TAA, translational stop codons; AATAAAICAATAA, polyadenylation signals. Sizes of the corresponding proteins are given on the right-hand side.

202

BOD0 PLACHTER et al.

The second IE region within the Hind111 E fragment of HCMV (AD169 strain) encodes two different immediate early transcription units. One genomic segment, termed IE3 (UL119-115), is transcribed a t IE, early (E), and late (L) times and potentially codes for viral glycoproteins (Leatham et al., 1991). However, the functions of these proteins have not been described. Most attention has focused on the UL123 and UL122 gene region, also termed IE1 and IE2, respectively (Chee et al., 1990). Both open reading frames are transcribed from a single strong promoter-enhancer element (major IE promoter; MIEP) under immediate early conditions (Thomsen et al., 1984; Boshart et al., 1985). The major gene products that are translated from differentially spliced mRNAs are two phosphorylated polypeptides of 72 and 86 kDa. These polypeptides have been given different names and are termed here IE1-pp72 and IE2-pp86. The IE 1-pp72 polypeptide was originally described as a phosphorylated polypeptide of 68-72 kDa, depending on the virus strain used for analysis (Michelson-Fiske et al., 1977; Stinski, 1978; Michelson et al., 1979; Blanton and Tevethia, 1981; Gibson, 1981). The protein is the most abundant viral polypeptide expressed immediately after infection. It is present in the nuclei of infected cells throughout the replicative cycle of HCMV. Although no DNA-binding activity could be demonstrated, IE1-pp72 associates with chromatin (Lafemina et al., 1989). In transient transfection analyses, IE1-pp72 has been shown to be involved in autoregulation of its own promoter and, alone or in concert with other viral or cellular proteins, in trans-activation of viral and cellular promoters (Stenberg and Stinski, 1985; Cherrington and Mocarski, 1989; Hunninghake et al., 1989; Malone et al., 1990; Biegalke and Geballe, 1991; Ghazal et al., 1991; Crump et al., 1992; Hagemeier et al., 1992a,b; Monick et al., 1992; Walker et al., 1992; Geist et al., 1994; Lukac et al., 1994). It could be shown that IE1-pp72 can activate TATA-less promoters (Caswell et al., 1993a); in addition, NF-KBhas been suggested to be involved in the autoregulatory function of IE1-pp72 (Sambucetti et al., 1989). However, the molecular mechanisms by which IE1-pp72 regulates gene expression are not completely understood. Considerable attention has focused on elucidating the functions of the IE2 protein of 82-86 kDa (IE2-pp86). This protein is present in abundance in permissively infected cells (Hermiston et al., 1987; Stenberg et al., 1989; Plachter et al., 1993). It is encoded by a differentially spliced mRNA and thus shares amino-terminal sequence with IE l pp72 (Stenberg et al., 1984, 1985) (Fig. 2). The IE2-pp86 polypeptide has been shown to be a promiscuous transactivator. Alone or in cooperation with IE1-pp72, IE2-pp86 can stimulate transcription from a

HUMAN CYTOMEGALOVIRUS

203

number of cellular as well as viral genes (Hermiston et al., 1987; Tevethia et al., 1987; Pizzorno et al., 1988, 1991; Chang et al., 1989; Arya and Sethi, 1990; Iwamoto et al., 1990; Malone et al., 1990; Gartenhaus et al., 1991; Geist et al., 1991; Paya et al., 1991; Colberg-Poley et al., 1992; Hagemeier et al., 1992a,b; Monick et al., 1992; Stasiak and Mocarski, 1992; Klucher et al., 1993; Kline et al., 1994; Lukac et al., 1994).In addition, IE2-pp86 negatively autoregulates its own promoter (Hermiston et al., 1987; Pizzorno et al., 1988; Stenberg et al., 1990). This repression has been shown t o be mediated by a short nucleotide motif located at the transcriptional start site (Pizzorno and Hayward, 1990; Cherrington et al., 1991; Liu et al., 1991);binding of IE2-pp86 to that element has been demonstrated (Lang and Stamminger, 1993; Macias and Stinski, 19931, and site-specific inhibition of the RNA polymerase I1 preinitiation complex by IE2-pp86 has been suggested as one possible mechanism (Wuet al., 1993). Interaction with upstream sequence motifs has also been shown to be responsible for the IE2mediated trans-activation of early viral promoters (Arlt et al., 1994; Schwartz et al., 1994). In addition, IE2 interacts with itself and a number of viral and cellular proteins (Chiou et al., 1993; Furnari et al., 1993). Association of IE2-pp86 with the gene product of UL84 has been demonstrated, and it has been suggested that this interaction may have some effect on the replication of the viral DNA from the lytic origin of HCMV (Pari and Anders, 1993; Spector and Tevethia, 1994). The TATAbinding protein (TBP) associates with IE2-pp86 both in solution and when complexed with DNA, suggesting that the interaction of IE2 with the basal transcription machinery regulates activation functions of this viral protein (Caswell et al., 1993b; Jupp et al., 1993a,b). Furthermore, it has been shown that IE2-pp86 binds to the retinoblastoma protein Rb (Hagemeier et al., 1994) and interacts with the tumor suppressor protein p53. The latter interaction may be functionally relevant in the proliferation of artery smooth muscle cells in restenosis after coronary angioplasty (Speir et al., 1994). These findings suggest that IE2-pp86 may also be involved in cell cycle regulation and stimulation of cell growth. Therefore, IE2-pp86 appears to be a multifunctional protein that influences cellular pathways in different ways in order to provide optimal conditions for HCMV infection. Additional forms of proteins have been suggested to be synthesized from alternatively spliced RNAs encoded by the major immediate early gene region (Stinski et al., 1983; Hermiston et al., 1987; Stenberg et al., 1990; Pizzorno et al., 1991). A protein of 52-55 kDa has been described that shares functional domains with IE2-pp86 but lacks amino acids 366-520 (Stenberg et al., 1985, 1990; Pizzorno et al., 1988,

204

BOD0 PLACHTER et al.

1991; Hermiston et al., 1990). This protein has also been suggested to be involved in trans-activation and autoregulation, although it is not clear yet whether it is synthesized during natural infection in fibroblasts (Hermiston et al., 1987; Baracchini et al., 1992; Lukac et al., 1994).Another protein of 40-45 kDa (IE2-p40) has been reported that most likely contains amino acids 242-580 of IE2-pp86 (Stenberg et al., 1989; Plachter et al., 1993). The RNA encoding this protein is transcribed from a late promoter, located within exon 5 (Fig. 2) (Stenberg et al., 1989; Puchtler and Stamminger, 1991).The IE2-p40 protein carries sequences necessary for autorepression (Hermiston et al., 1990; Stenberg et al., 1990; Pizzorno et al., 1991). In a study using cotransfection analyses it was found that IE2-p40 represses the IE promoter in a fashion comparable t o IE2-pp86 and that, in concert with IE1-pp72, this protein trans-activates cellular but not viral early promoters (Jenkins et al., 1994). Still, the roles of both IE2-pp55 and IE2-p40 in the replication of HCMV require further investigation. Using a panel of monoclonal antibodies (MAbs) generated against bacterially expressed portions of the IE-proteins (Plachter et al., 1993), we investigated the different isoforms of proteins synthesized from the major immediate early gene region of HCMV. Using antibodies directed against either exons 2 or 3, common to IE1-pp72 and IE2-pp86, a time course of expression concordant with what has been described for the two proteins was observed (Fig. 3). The IE1-pp72 was detectable very early and accumulated to increasing amounts in the course of infection. In contrast, the kinetics of IE2-pp86 accumulation were biphasic; after an initial increase, the amount of detectable protein decreased at early times and increased at late times. Besides these two proteins, two minor polypeptides in a size range of 36-40 kDa were found. The 40 kDa protein showed a time course of accumulation comparable to that of IE2-pp86, whereas the 36-kDa protein accumulated similarly to IE1-pp72. It is not clear whether these proteins are degradation products of the larger proteins or whether they are synthesized from alternatively spliced RNAs. When immunoblots were carried out with MAbs directed against exon 4 encoded sequences, only the IE1-pp72 protein was found. In contrast, MAbs directed against exon 5 encoded epitopes detected a number of different proteins, especially at late times after infection. This is concordant with previously published results (Stenberg et al., 1989; Pizzorno et al., 1991). As expected, the IE2-pp86 appeared with the same kinetics as seen with antibodies directed against exon2/3 encoded sequences. A minor protein of approximately 66 kDa with the same kinetics as IE2-pp86 was detected with all the exon 5 specific MAbs (Fig. 3). As evidenced by a parallel experiment, the kinetics of

HUMAN CYTOMEGALOVIRUS

205

appearance of this protein were clearly distinct from those of the abundant tegument protein pp65 (UL83), which was previously found to bind unrelated antibodies in immunoblot assays (Plachter et al., 1990). It is not clear whether the 66-kDa protein represents a degradation product of IE2-pp86 or is synthesized from an alternatively processed RNA. As exon213 specific antibodies did not detect this protein, it may be hypothesized that an RNA containing exons 1 and 5 may encode this protein. To clarify this, cloning and sequencing of cDNAs, made from late infected cell RNA, is required. Two less abundant proteins of around 30 kDa were detectable only with MAb 2-9-5, but not with the other two MAbs (Fig. 3). One of the MAbs, “XS,”is directed against sequences that are spliced out in some of the IE2 RNAs. Therefore, these two proteins presumably lack the intron encoded portions of IE2-pp86. Their expression was weak at IE and E times but the kinetics appeared to be comparable to that of IE2-pp86; therefore the RNAs for these proteins are presumably initiated at the MIEP. Again, the structure of these RNAs awaits further analysis. All other proteins detected with exon 5 specific antibodies showed clear late kinetics of accumulation. One major band of approximately 40 kDa corresponded to the previously described IE2-p40 protein (Stenberg et al., 1989; Plachter et al., 1993). A minor protein with a slightly slower mobility and the same kinetics probably represents a modified derivative of IE2-p40 (Fig. 3). In addition, another dominant protein in the size range 60-63 kDa was also detected with all IE2specific MAbs. As described before (Plachter et al., 1993), this protein most likely corresponds to a protein of 55 kDa seen with another laboratory strain of HCMV (Pizzorno et al., 1991). Again, the origin of this protein is unclear. Because of its late kinetics, it can be assumed that it is encoded by an RNA transcribed from the late promoter within UL122 (Stenberg et al., 1989; Puchtler and Stamminger, 1991). An AUG codon is located immediately at the initiation site for this late RNA. However, it is not known whether translation of IE2-p63 is initiated at this site. Alternatively, the IE2-p63 protein may represent a modified form of IE2-p40. Another band of 72 kDa was inconsistently seen with MAb 2-9-5 (Fig. 3). This band showed the same kinetics as IE1-pp72. Therefore, it is likely that this is the result of unspecific detection of the IE1 protein by this MAb. Finally, it should be noted that during our kinetic analyses, which were done without blocking cells with substances such as cycloheximide, we did not detect a protein of 52-55 kDa corresponding to the trans-activator described before. However, low amounts of this protein may have been undetectable in Western blots even when highly reactive MAbs were used.

206

BOD0 PLACHTER et al.

FIG.3. Immunoblot analyses of polypeptides encoded by the major immediate early region of HCMV using a panel of MAbs. (A1-C) Immunoblot analyses of lysates of infected foreskin fibroblasts from different times after infection. Cells were infected at a n MOI of approximately 1. Infection was terminated by lysis of the cells at the times indicated at top. The molecular masses of the proteins detected are given on the righthand side. Immunoblots in A-C were performed with MAbs E l 3 (Mazeron et al., 1992), SMX, and 2-9-5,respectively (Plachter, et al., 1993). (D)Schematic representation summarizing results of the immunoblot analyses with different MAbs. The splicing pattern of different mRNAs generated from the major IE region of HCMV is given at top, with

207

HUMAN CYTOMEGALOVIRUS

D

IE E L 63kDa

43kDa 40kDa 3lkDa 29kDa

'1 i

IE E L 86kDa 72kDa

40kDa 36kDa

solid lines representing exon sequences and dashed lines representing intron sequences. Question marks indicate transcripts that have not been mapped by cDNA cloning and sequencing. The arrow indicates splicing from exon 5 sequences of IE region 2 to the downstream IE region 3 (UL118) as proposed by Rawlinson and Barrel1 (1993), a region also shown to be transcribed a t different times after infection (Leatham et al., 1991). Vertical lines indicate the location of initiation codons on the different RNAs. The approximate location of the epitopes for the different MAbs used is given at top. In the lower half, results of the immunoblot experiments are schematically depicted. Solid horizontal bars represent dominant proteins detected with the different MAbs indicated at top. Stippled bars represent proteins that were found in low abundance, and open bars stand for minor protein bands. On the left- and right-hand sides, the relative abundances of these bands at different time points after infection and the respective molecular masses of the proteins are indicated.

Taken together, these analyses indicate that multiple forms of proteins encoded by IE regions 1 and 2 are synthesized in the course of infection. The relative abundance of these proteins at different phases of the replicative cycle may depend on the strain of HCMV, the multiplicity of infection (MOI), and the type of infected cells. For instance, we noted that the 1.7-kb RNA abundantly synthesized under IE conditions in permissively infected fibroblasts was barely detectable when the permissive glioblastoma cell line U138-MG was infected under the same conditions (Wolff et al., 1994). It is tempting to speculate that

208

BOD0 PLACHTER et al.

differential expression from the major IE region of HCMV may be one way to regulate the initial events of viral replication in different cell types and therefore may be one mechanism of the virus t o meet different requirements while replicating in its natural host. In addition to the proteins encoded by UL122/123, open reading frames TRSl/IRSl, US3, and UL36/37 (Fig. 2) are transcribed at immediate early times, and their protein products have been shown to be involved in regulating early and late viral gene activity. Proteins encoded by open reading frames TRSl and IRSl have been shown to be important in the trans-activation of the promoter of UL44, encoding the DNA-binding protein p52. This gene is required for replication of the viral genome (Mocarski et al., 1985; Stasiak and Mocarski, 1992; Pari et al., 1993). It was suggested that the TRSl/IRSl regions express regulatory functions that influence a subset of viral genes important for viral replication (Mocarski, 1993); at least one copy of the homologous genes was required for HCMV replication (Pari et al., 1993). The proteins encoded by UL36/UL37 and US3 have been shown to act synergistically in the trans-activation of the cellular heat-shock promoter (hsp),as well as of viral early promoters (Colberg-Poley et al., 1992; Tenney et al., 1993). Trans-activation appeared to be cell type specific and was in some instances augmented by IE1 and IE2. Thus, different viral immediate early trans-activators appear to be active during HCMV infection and may be required for the ordered induction of the early functions necessary for viral DNA replication. The early phase of viral gene expression has been long known to be prolonged, lasting at least 24 hr. Genes have been grouped into the early phase by virtue of the insensitivity of their transcription to substances that block viral DNA replication. According to these criteria, a number of viral genes have been identified as being early (for review, see Spector et al., 1990; Mocarski, 1993).The most intensively studied early proteins have been those involved in DNA replication from the lytic origin of HCMV (Hamzeh et al., 1990; Anders and Punturieri 1991; Anders et al., 1992). Among those were the DNA polymerase (Heilbronn et al., 1987; Kouzarides et al., 1987; D’Aquila et al., 1989) and two different DNA-binding proteins (Gibson et al., 1981; Mocarski et al., 1985; Anders and Gibson, 19881, one of which appears to be a polymerase processivity factor (Ertl and Powell, 1992). Two additional open reading frames have been identified on the HCMV genome that show sequence similarity with the helicase primase complex described for HSV (Chee et al., 1990).Using transient complementation studies, all of these genes have been proved to be essential for replication the lytic origin; in total, eleven loci of the HCMV genome were required (Pari and Anders, 1993).

HUMAN CYTOMEGALOVIRUS

209

In addition to these replicative functions, two early-late viral proteins, encoded by UL82 (pp71) and UL69, have been described which act as transcriptional transactivators. Both gene products were able to induce expression from homologous and heterologous promotors (Liu and Stinski, 1992; Winkler et al., 1994; Winkler, M., Schmolke, S., Plachter, B., and Stamminger, T., 1995, unpublished observations). Transcription of these genes is initiated prior to DNA-replication and peaks at late times. The precise role of these proteins for the replication of HCMV, however, remains to be determined. After DNA replication, late gene expression ensues. Most of the late genes characterized to date encode proteins that are contained within intra- or extracellular particles (Jahn and Mach, 1990; Gibson, 1993; Mocarski, 1993). Some of these proteins have been suggested to carry functions other than structural (Roby and Gibson, 1986; Britt and Vugler, 1987; Liu and Stinski, 1992). Concomitant with the synthesis of virion components and DNA replication, assembly of particles is initiated within the cell nucleus. Three forms of intracellular particles have been described. A protein termed the assembly protein has been shown to be involved in the maturation of virions. The reader is referred to reviews detailing current knowledge about virion assembly and maturation (Gibson, 1993; Rixon, 1993). Immature viral particles are thought to bud through the inner nuclear membrane into the perinuclear space, thereby being enveloped and tegumented. Data suggest that the final envelopment and subsequent egress of HCMV from the cell is mediated by early endosomal compartments (Tooze et al., 1993). In culture supernatant from late-stage infected fibroblasts, three forms of extracellular particles can be recovered (Irmiere and Gibson, 1983). Besides typical herpes virions, two forms of defective particles are generated. Noninfectious enveloped particles (NIEPs) have a protein composition comparable to virions but lack genomic DNA (Gibson and Irmiere, 1984). In contrast, dense bodies (DB) are enveloped particles containing an electron-dense mass. These particles carry a complete set of viral glycoproteins but lack DNA and most of the tegument and capsid proteins (Gibson and Irmiere, 1984; %by and Gibson, 1986). Dense bodies have been shown to consist to more than 90% of the tegument protein pp65 (UL83). Large amounts of DB are produced in fibroblast cultures. The number of DB synthesized appears to correlate with MOI and the number of passages of HCMV in culture (Klages et al., 1989). In electron microscopy studies, DB have also been found in uiuo in circulating cytomegalic cells from patients with acute HCMV infection, indicating that the synthesis of these particles may be of relevance for natural infection (Grefte et al., 1993a). Although their function in the replication of HCMV is unclear, both DB and

210

BOD0 PLACHTER et al.

NIEPs are able to deliver at least some of their protein constituents into cells (Topilko and Michelson, 1994;Schmolke et al., 1995).This is exemplified by the presence of the tegument protein pp65 in the nucleus immediately after exposure of cells in the absence of viral gene expression (Fig. 4). Therefore, all three forms of particles are apparently able to bind to cells, and it has been suggested that they can enter the cell via the same pathway (Topilko and Michelson, 1994).

B . Nonpermissive Human Cytomegalovirus Infection As discussed above, permissive HCMV infection in culture is restricted to human cells or cells from closely related species. However, a variety of different host cells can be infected by HCMV in culture, and

FIG.4. Indirect immunofluorescence analysis of human fibroblasts exposed to purified virions and DB. Particles were separated by centrifugation in glycerol tartrate gradients (Irmiere and Gibson, 1983). Cells were incubated with gradient-purified particles in the presence of 100 pg/ml cycloheximide for 30 min. After that, cells were stained with MAbs directed against IE1-pp72 or pp65 (Plachter et al., 1990), generously provided by Dr. W.Britt, Birmingham, Alabama.

HUMAN CYTOMEGALOVIRUS

211

they express a limited repertoire of viral genes (Table I). Substantial effort has focused on the elucidation of the molecular mechanisms that restrict viral replication in different host cells. In early studies, it was shown that HCMV could replicate in differentiated but not in nondifferentiated teratocarcinoma cells (Gonczol et al., 1984). Expression from the major IE gene region has been found to be downregulated in nondifferentiated cells but can be induced when cells are treated with agents promoting differentiation (Gonczol et al., 1984; Lafemina and Hayward, 1986; Shelbourn et al., 1989; Sinclair et al., 1992). A regulatory element termed the modulator sequence, could be identified upstream of the major immediate early enhancer promoter; this element regulated the activity of the MIEP in different host cells. Cellular proteins have been found to bind to this region (Nelson et al., 1987; Lubon et al., 1989).A protein complex binding to the modulator region and to a 21-bp repeat element within the enhancer in undifferentiated but not in differentiated cells was described, and one component of this complex was identified as the cellular protein YY1 (Shelbourn et al., 1989; Sinclair et al., 1992; Hagemeier et al., 1994). Thus, as the state of cell differentiation is considered crucial for HCMV infection, it is hypothesized that undifferentiated cells may provide an environment in which the virus can persist in a latent state. Differentiation of the same cells may lead to derepression of the IE genes and subsequent permissive replication. Although such a process has not formally been proved, studies done on hematopoietic cells isolated from seropositive individuals seem to support this hypothesis (Taylor-Wiedeman et al., 1994). In addition, the susceptibility of bone marrow precursor cells to HCMV infection appears to lead along the same lines; it could be shown that permissiveness of these cells to HCMV gene expression correlated with their state of differentiation (Minton et al., 1994). Nevertheless, induction of expression from the major IE gene locus cannot be the sole determinant of permissive versus nonpermissive infection. The MIEP is known to be active in a wide variety of human and nonhuman cell lines, most of which are nonpermissive to HCMV replication. Therefore, additional factors are supposed t o control the permissiveness for HCMV replication. As discussed, different isoforms of proteins may be expressed by the major IE region in different cell types. As most investigations on nonpermissive cells were done with immunologic reagents that would not distinguish between these isoforms of proteins, the pattern of expression of the polypeptides in different cells is not clear yet. In addition, no detailed data are available on the expression from other IE loci in nonpermissive cells. Thus,

212

BOD0 PLACHTER et al.

the differential expression of IE proteins in different cells may be critical for their permissiveness for HCMV replication.

C. Initial Events in Human Cytomegalovirus Infection Although there is evidence that HCMV can enter a wide variety of cells and that the MIEP is active in these cells, most permanently growing human cell lines tested are nonpermissive to HCMV replication (Nowlin et al., 1991). As opposed to most other herpesviruses, HCMV is able to stimulate host cell macromolecule synthesis very early after infection (Furukawa et al., 1975a, 1976; Tanaka et al., 1975; Albrecht et al., 1976; Stinski, 1978). Stimulation of host cells appears to be a prerequisite for efficient replication in HCMV and murine cytomegalovirus (Furukawa et al., 1975b; Muller and Hudson, 1977; Landini et al., 1979). On the molecular level, many different cellular proteins have been shown to be induced by HCMV infection, and some of these effects were attributable to the prior expression of IE1-pp72 and IE2-pp86 (Geist et al., 1991, 1993; Grundy and Downes, 1993; Mocarski, 1993; Kline et al., 1994). Yet, other genes such as the cellular oncogenes c-myc, c-fos, and c-jun have been shown to be transcriptionally transactivated by interaction of cells with inactivated virus particles (Boldogh et al., 1990). It was suggested that interaction of HCMV with its membrane-bound receptors leads to activation of intracellular signaling pathways including induction of protein kinases A (PKA) and C (PKC) (Boldogh et al., 1993) and transcription factors such as AP1, NF-KB,and members of the CREB family (Sambucetti et al., 1989; Stamminger and Fleckenstein, 1990; Lang et al., 1992). The MIEP of HCMV has been shown to be responsive to several cellular transcription factors including NF-KB,AP1, and CREB. In addition, stimulation of cells with CAMPor phorbol esters leads to activation of transcription from the MIEP, and these effects are mediated via PKA- or PKC-dependent pathways (Hunninghake et al., 1989; Chang et al., 1990; Stamminger et al., 1990). Therefore, it appears that physiological cellular signaling pathways are used by the virus to induce immediate early and, possibly, early gene expression. This would imply that the way in which cells can react to activation signals set by HCMV may be critically linked to viral replication and yield of infectious virus. It might be hypothesized that a controlled and sequentially ordered induction of viral and cellular trans-inducing factors is required t o initiate viral DNA replication and late gene expression; most established cell lines may not be able to

HUMAN CYTOMEGALOVIRUS

213

respond to viral infection in a fashion which allows the expression of all the viral genes necessary to initiate DNA replication.

D. Role of Virion Proteins in Initiating Infection

A virion component, the tegument protein VP16, is required for the induction of immediate early gene expression in herpes simplex virus (HSV) (Roizman and Sears, 1993). In contrast, the MIEP of HCMV does not require any viral trans-activating protein to be transcriptionally active. Nevertheless, early studies indicated that a virion component could enhance expression from the MIEP (Spaete and Mocarski, 1985; Stinski and Roehr, 1985). More recent analyses have identified the tegument protein pp71 (UL82) as one trans-activating protein, which stimulates transcription from the MIEP via upstream ATF/CREB sites (Liu and Stinski, 1992). This phosphorylated protein is a major constituent of the extracellular virion and is presumed to be translocated into cells via infection (Roby and Gibson, 1986). However, its role in natural infection remains to be elucidated. Another tegument protein, which by virtue of limited sequence similarity is grouped into one family together with pp71, is the phosphorylated tegument protein pp65 (UL83) (Chee et al., 1990).As mentioned above, this protein is a major constituent of extracellular DB and virions. It appears to be exceptional in several respects. During active infection in uiuo, this protein is abundantly found in peripheral polymorphonuclear granulocytes (PMNL) (Grefte et al., 1992a,b). After infection, pp65 is readily detectable in the nuclei of infected cells before onset of viral gene expression (Geballe et al., 1986; Britt and Vugler, 1987; Grefte et al., 1992a),and protein kinase activity has been associated with this polypeptide (Somogyi et al., 1990).In more recent analyses, it could be shown that the rapid nuclear transport of this protein is mediated by at least two independent nuclear localization signals, one of which displayed exceptional functional features as compared with other known targeting sequences (Schmolke et al., 1995). Although its nuclear accumulation would suggest some role in the induction of viral gene expression, no trans-activator function could be assigned to pp65 until now (Liu and Stinski, 1992; S. Schmolke and B. Plachter, 1995, unpublished). In this respect, the kinase activity associated with pp65 appears to be interesting because it may be able to regulate viral or cellular proteins via phosphorylation. Kinase activity has been associated with extracellular particles, but the association of that activity with pp65 has been questioned (Roby and Gibson, 1986). Therefore, additional constituents of the virion may be found in the future that influence activation of cellular processes after HCMV infection.

214

BOD0 PLACHTER et al.

E . Strain Variabilities Although it is well acknowledged that HCMV strains vary in genome size, restriction pattern, and nucleotide sequences, only limited information is available about strain variabilities and resulting differences in the interaction of such strains with host cells. Early studies showed that although the genomes of different HCMV strains are homologous, they contain unique sequences (Pritchett, 1980). However, these unique regions have not been accurately mapped. Reduced infectivity of laboratory strains versus fresh clinical isolates has been reported for hematopoietic and endothelial cells (Reiser et al., 1986; Waldman et al., 1989; Simmons et al., 1990; Mocarski et al., 1993; Minton et al., 1994). In addition, impairment of the host cell range of HCMV strains passaged for longer periods on fibroblasts versus more recent isolates has also been reported (Waldman et al., 1991; Mocarski et al., 1993). It is unclear whether these differences result from mutations within the HCMV genome during fibroblast passage or alterations in gene expression during culturing. On the grounds of sequence analyses in viral structural proteins, it was suggested that HCMV strains can be functionally grouped into different subtypes (Chou and Dennison, 1991; Lehner et al., 1991; Chou, 1992; Meyer et al., 1992; Urban et al., 1992; Fries et al., 1994).In addition, differences in the susceptibility of different viral strains to ganciclovir have been attributed to sequence variations within the DNA polymerase and UL97 genes (Sullivan et al., 1992,1993). On the other hand, only minor modifications have been found in the coding regions of regulatory proteins (Lehner et al., 1991). Alterations in the infectability of endothelial cells after a limited number of passages of fibroblasts have been reported, suggesting that a mechanism associated with the adaptation of HCMV to fibroblasts may alter its host range and biological properties (Waldman et al., 1991). However, some of the alterations seen during culture in fibroblasts may be reversible and may possibly be linked to the high-level production of virions and defective particles in such culture systems (Klages et al., 1989).Future studies will focus on the question of whether viral genes can be identified that define altered biological properties of different variant strains of HCMV.

111. ORGANTROPISM OF HUMAN CYTOMEGALOVIRUS One of the most intriguing aspects of HCMV biology is the highly efficient spread of the virus throughout the population while only infrequently causing severe symptoms. In adults, 40-100% of individu-

HUMAN CYTOMEGALOVIRUS

215

als are infected, indicated by the presence of antibodies against HCMV. Although the site of latency of HCMV has not been defined, the different modes of transmission during the neonatal period and adolescence suggest that various organs may harbor the virus during latent infection in immunocompetent individuals (Alford and Britt, 1990). In addition, during HCMV disease in immunocompromised patients, the virus can be detected in solid organs as well as in peripheral blood. Understanding organ tropism and spread is one of the prerequisites for the elucidation of pathogenic mechanisms in HCMV infection. Involvement of the central nervous (CNS) and sensory systems leads to the most prominent sequelae seen during HCMV infection. In congenital cytomegalic inclusion disease, damage of the developing CNS is a frequent feature. Symptoms include mental retardation, seizures, hypotonia, and hearing loss, which are reflected in various pathological alterations in the CNS (Weller and Henshaw, 1962; McCracken et al., 1969; Pass et al., 1980; Perlman and Argyle, 1992). The HCMVinfected cells were detected in histological sections of affected brains by demonstration of characteristic inclusion body cells, viral antigens, or DNA (Haymaker et al., 1954; Benirschke et al., 1974; Schmidbauer et al., 1989). Moreover, the presence of major cerebral migrational defects were reported in severely handicapped children with congenital CID (Hayward et al., 1991). Therefore, it seems likely that CNS damage is caused, at least in part, by direct or indirect cytopathic effects of HCMV (Davis et al., 1987; Strauss, 1990; Perlman and Argyle, 1992). Central nervous tissues are also major target organs of HCMV in patients suffering from AIDS (Snider et al., 1983; Helweg-Larsen et al., 1986; Behar et al., 1987; Morgello et al., 1987;Jensen and Klinken, 1989; Anders et al., 1986; Matthiessen et al., 1992; Weber et al., 1994). In these individuals, HCMV is able to replicate efficiently in various brain structures and the retina, but the pathogenic role of HCMV in brain disease in AIDS patients is sometimes obscured by coinfections with other pathogens (Wiley and Nelson, 1988;Grafe and Wiley, 1989; Vinters et al., 1989; Horn et al., 1992; Rummelt et al., 1994). The HCMV infection seems to spread independently from human immunodeficiency virus (HIV) as, in general, no colocalization was detectable (Schmidbauer et al., 1989). However, other authors reported colocalization of HCMV and HIV antigens in some infected cells (Nelson et al., 1988; Skolnik et al., 1989). The role of in uiuo interaction of both viruses in the pathogenesis of brain disease in AIDS patients, however, remains to be determined. Another prominent target of HCMV in AIDS is the retina. Acute retinal infection frequently leads to blind-

216

BOD0 PLACHTER et al.

ness in these patients. In cell culture, permissive infection of retinal pigment epithelial cells could be demonstrated, rendering a direct cytopathic effect set by HCMV infection likely to be the pathogenic mechanism (Miceli et al., 1989; L. Pereira, 1994, personal communication). In adults other than AIDS patients, infection of the CNS or the retina is a rare event, although there have been reports of HCMV encephalitis and myelitis in organ transplant recipients or even in immunocompetent individuals (Qler et al., 1986; Pantoni et al., 1991; Bamborschke et al., 1992; Studahl et al., 1994). In contrast, elevated liver enzymes, indicative of a mostly subclinical hepatitis, are frequently associated with HCMV infection in immunocompetent individuals (Klemola et al., 1970; Sterner et al., 1970; Horwitz et al., 1979). Yet, signs of hepatitis can also be seen in connatally infected newborns and in HCMV infection of the immunocompromised host (McCracken et al., 1969;Luby et al., 1974; Aldrete et al., 1975; Sopko and Anuras, 1978; Ware et al., 1979; Barkholt et al., 1994). In adults with HCMV related hepatitis, cytomegalic inclusion body cells and/or round cell infiltrates are a dominant feature in histological sections (Fig. 5 ) (Bonkowsky et al., 1975; Macasaet et al., 1975). This suggests that liver dysfunction might directly result from HCMV cytopathogenicity or indirectly from the inflammatory reaction to HCMV infection. In HCMV-infected newborns, however, hepatomegaly is most likely the result of extramedullary hematopoiesis with round cell infiltration (McCracken et al., 1969; Benirschke et al., 1974). Infection with HCMV in immunocompromised hosts often affects other organs of the gastrointestinal tract (Myerson et al., 1984; Francis et al., 1989; Roberts et al., 1989; Aqel et al., 1991; Shintaku et al., 1991; Escudero-Fabre et al., 1992). The whole digestive system can be a site of viral replication (Fig. 5), including the esophagus, stomach, small intestine, and colon. Direct cytopathogenicity leading to erosions, nodules, ulcerations, and even perforations dominates the pathological aspect, but necrosis due to vasculitic processes has also been suggested (Goodman and Porter, 1973; Frank and Raicht, 1984; Fernandes et al., 1986;Francis et al., 1989;Aqel et al., 1991).In contrast, the pathogenetic mechanisms operative in HCMV pneumonia are not completely clear. Pneumonia occurs infrequently during HCMV disease, but symptoms are often severe. Successful prophylaxis of early onset HCMV intersitial pneumonitis in bone marrow recipients has been established; however, late onset pneumonitis is still a major clinical problem in these patients (Ljungman et al., 1994). Extensive cytopathogenicity due t o viral replication may be present (Fig. 5 ) (Klatt and Shibata, 1988;Aukrust et al., 1992; Humbert et al., 1992);in many cases, however, the subtle histological alterations do not correlate with

HUMAN CYTOMEGALOVIRUS

217

the severity of disease; immunopathological factors may contribute to the development of symptoms in these patients (Grundy et al., 1987; Grundy, 1990; Humbert et al., 1992). In summary, organ dysfunction in most cases coincides with HCMV replication in the affected organs and with an inflammatory reaction to the replicating virus. In some instances, however, symptoms cannot be exclusively attributed to direct effects of the virus, and other pathogenetic mechanisms appear to be operative to result in HCMV disease (Ho, 1990; Alford and Britt, 1990). IV. CELLTYPES INVOLVED IN ACUTE HUMAN CYTOMEGALOVIRUS DISEASE In cell culture, a variety of different cell types besides fibroblasts have been found to be permissive to HCMV infection, though to moderate extents; these included epithelial cells, endothelial cells, stromal cells, hematopoietic cells, and smooth muscle cells (Table I). In the following section we summarize what is known about the cells involved in HCMV infection in its natural host.

A . Human Cytomegalouirus Infection in Tissue Cells The cytomegalic alterations characteristic for HCMV-infected cells in tissue sections have hindered an unequivocal definition of the target cells for the virus. This problem was, to some extent, resolved when immunohistochemical techniques were applied to HCMV-infected tissues. Monoclonal antibodies against HCMV immediate early antigens allowed the detection of morphologically unaltered, infected cells (Sinzger et al., 1993a). A broad variety of infected cell types were identified, including smooth muscle cells, inflammatory cells, and endothelial cells (Table 11). However, the origin of the cytomegalic cells, presumably representing late stage permissive infection, remained obscure. Furthermore, there was the striking paradox that, although fibroblasts were the standard culture system for HCMV replication, they were not found to be a major target population in uiuo. Immunohistochemical double-labeling techniques or in situ cytohybridization in combination with antigen detection led to the identification of infected cells being endothelial cells and macrophages (Myerson et al., 1984; Roberts et al., 1989; Schwartz et al., 1992). Using a severe combined immunodeficiency (SCID) mouse model, thymic epithelial cells were found to be a prominent target of HCMV infection in uiuo (Mocarski et al., 1993). In more recent studies, we employed

218

BOD0 PLACHTER et al.

FIG.5. Detection of HCMV-infected cells in various tissues by immunoperoxidase staining using MAbs with defined specificity against HCMV immediate early, early, and late antigens. Peroxidase staining with diaminobenzidine yielded dark staining of HCMV antigens. (A) Massive HCMV infection in a colon ulceration in an AIDS patient; (B) Massive infection of duodenal gland epithelium in a n AIDS patient; (C) HCMV infection of the liver during a rejection period after liver transplantation; (D) Vesselassociated HCMV infection of duodenal tissue in a renal transplant recipient; (E) HCMV-associated pneumonitis in an AIDS patient; (F) HCMV retinitis in a n AIDS patient; (G) HCMV infection of the pituitary gland in an AIDS patient; HCMV-specific MAbs were directed against (C,D, F, GI, the immediate early antigen 1and 2 (MAb E l 3 Biosoft, Paris, France), the immediate early antigen 1 (MAb BS 500 Biotest, Dreieich, Germany), (E) the early-late antigen pp65, and (G) the early antigen p52 (BS 510, Biotest). Bar in each panel = 100 pm.

immunohistochemical double stainings, simultaneously labeling HCMV antigens and cell marker proteins, to identify unequivocally the different cell types involved in active HCMV infection in various organs (Sinzger et al., 199313, 1995a,b). We found that HCMV can infect a broad spectrum of cells in uiuo. Fibroblasts, epithelial cells, endothelial cells, and smooth muscle cells were the major targets of permissive HCMV infection, whereas infected granulocytes and macrophages were present to a minor extent (Fig. 6). In contrast to what was thought before, fibroblasts comprised one dominant cell population infected in uiuo. This finding has major implications, as much of the knowledge on the molecular biology of HCMV has been gained through studies on cultured human fibroblasts.

HUMAN CYTOMEGALOVIRUS

219

220

BOD0 PLACHTER et al.

HUMAN CYTOMEGALOVIRUS

221

TABLE 11 DETECTION OF HUMAN CYTOMECALOVIRUS-INFECTED CELLS IN HISTOPATHOLOGICAL SECTIONS OF INFECTED ORGANS AND IN BLOOD CELLPREPARATIONS~ Organ Brain (AIDS)

Brain (connatal, AIDS)

tu

Peripheral nerve Myelon

Cell type Endothelial Glial Neurons

Cell identification Factor VII GFAP NSE

Neurons, Morphology astrocytes, oligodendrocytes, ependyma, perineural, endoneural, leptomeningal Schwann cells SlOO Glial Neurons Ependymal Mesenchymal Endothelial Macrophages Various cells

GFAP, SlOO NSE, synaptophysin Morphology Vimentin Factor VIII MAC 387,EBM 11 Morphology Morphology

Placenta

Stromal, endothelial

Placenta

Stromal Syncytiotrophoblast Endothelial

Morphology Morphology Factor VIII

Macrophages Stromal

HAM 56 Morphology

Placenta

HCMV antigen

HCMV DNA

+ (undefined) + (undefined)

+ (undefined)

-

EA (p52/UL44)

+ (undefined) -

-

IEA, EA, LA

+ (undefined)

Cytomegaly

Reference Wiley and Nelson (1988)

+

+ + + +

-

+

Grafe and Wiley (1989)

Schmidbauer et al. (1989)

Horn et al. (1992)

-

-

+

Garcia et al. (1989)

+ + + +

Muhlemann et al. (1992)

neg

Schwartz et al. (1992)

Placenta

+ + +

Fibroblasts Endothelial Macrophages Trophoblast Smooth muscle

Vimentin Factor VIII CD68 P-HCG, placental AP Actin

Lung

Epithelial

Morphology

Lung

Epithelial Endothelial Fibroblasts Smooth muscle Macrophages Granulocytes Lymphocytes

Keratin Factor VIII Vimentin Actin CD68 Neurophil elastase LCA

Gastrointestinal tract

Endothelial Epithelial Stromal

Factor VIII Morphology Morphology

Esophagus

Endothelial Stromal Epithelial, endothelial Stromal, epithelial, endothelial, muscle

Morphology

EA lp52/UIA4)

Stomach

Stromal, endothelial, smooth muscle, epithelial

Morphology

-

-

+

Aqel et al. (1991)

Jejunum

Endothelial

Morphology

-

-

+

Nabeshima et al. (1992)

Colon

Endothelial

Morphology

-

-

+

Escudero-Fabre (1992)

N

IEA

-

Sinzger et a1 (1993b)

neg

+

-

Aukrust et al. (1992)

IEA IEA IEA IEA IJ3A IEA neg -

Sinzger et al. (1995a,b)

Roberts et al. (1989)

N

Small intestine Colorectum

Francis et al. (1989)

et al.

TABLE I1 (Continued) Organ

Lu

Cell type

Cell identification

HCMV antigen

+

HCMV DNA

Cytomegaly

Reference

Gastrointestinal tract

Stromal, endothelial, smooth muscle, epithelial

Morph o1ogy

Gastrointestinal tract

Epithelial Endothelial Fibroblasts Smooth muscle Macrophages Granulocytes Lymphocytes

Keratin Factor VIII Vimentin Actin CD68 Neurophil Elastase LCA

Liver

Hepatocyte

Morphology

-

Liver

Sinusoidal, hepatocytes, mesenchymal, bile ductal

Morphology

IEA, EA

Theise et a1 (1993)

Endomyocardium

Endothelial cells Fibroblasts

Morphology

-

Millett et al. (1991)

Heart

Mwcyte, endothelial cells

Morphology

EA, LA

Arbustini et al. (1992)

Heart

Endothelial cells, lymphocytes, smooth muscle

Morphology

-

Wu et al. (1992)

Kidney

Inflammatory cells, tubular cells, glomerular cells

Morphology

IEA IEA IEA IEA IEA IEA neg

Escudero-Fabre et a1 (1992)

Sinzger et al. (1995a,b)

Arnold et al. (1992)

Gnann et al. (1988)

Uvea Cornea

Endothelial Endothelial, smooth muscle

Morphology Morphology

Cornea

Epithelial

Morphology

-

Yee et al. (1991)

Lymph node

T lymphocytes

UCHLl

-

Younes et al. (1991)

Various

Endothelial Epithelial, stromal

Factor VIII Morphology

-

Myerson et al. (1984)

Peripheral blood

Lymphocytes

FACS

-

Schrier et al. (1985)

Peripheral blood

Monocytes

FACS

-

Taylor-Wiedeman et al. (1991)

Peripheral blood

PBMC, PMN

Morphology

IEA, LA (pp65)

Grefte et al. (1992a)

Peripheral blood

PMNL Monocytes T cells B cells

FACS

IEA, LA (pp65) IEA, LA (pp65) LA (pp65) neg

Gerna et al. (1992b)

Peripheral blood

PMNL

Morphology

IEA, LA (pp65), EA (p52/UL44), LA?

Gerna et al. (1992a)

Peripheral blood

Endothelial cells

PAL-E, UEAL, Factor VnI, CD51

IEA, EA (p52/UL44) LA (gB, gh), LA (MCP)

Grefte et al. (1993a,b)

Peripheral blood

Endothelial cells

Tight junctions

E3

N

cn

Daiker (1988)

-

-

a IEA, HCMV immediate early antigen; EA, HCMV early antigen; LA, HCMV, late antigen. Identification of cells was done using morphological criteria, by FACS analysis, or with antisera against the antigens listed. neg, Negative; P-HCG, beta-human chorionic gonadotrophin; F, factor; EM, electron microscopy; GFAP, glial fibrillary acidic protein; NSE, neuron-specific enolase; S100, SlOO protein; MAC 387, monocyte marker MAC 387; EBM 11, macrophage marker EBM 11; HAM 56, macrophage marker HAM 56; HCG, human chorionic gonadotrophin; AP, alkaline phosphatase; LCA, leukocyte common antigen; PAL-E, endothelial cell marker PAL-E; UCHL-1, T-cell linage antibody UCHL-1; UEA-1, Ulex europaeus lectin.

226

BOD0 PLACHTER et al.

FIG.6. Identification of cell types in HCMV-infected tissues by immunohistochemical double staining. Detection of HCMV immediate early antigen (IEA) by the immunoperoxidase technique yielded dark nuclear staining. Detection of specific cell markers by the immunoalkaline phosphatase technique resulted in cytoplasmic staining. Counterstaining was performed with hematoxylin. (A) HCMV placentitis, showing detection of the mesenchymal marker vimentin in HCMV-infected fibroblasts; (B) HCMV gastritis, showing detection of the endothelial marker factor. VIII in an infected gastric venule; (C) HCMV duodenitis, showing detection of the epithelial cell marker keratin in HCMVinfected cells of duodenal glands; (D) HCMV gastritis, showing detection of the smooth muscle cell marker actin in infected cells of the muscularis mucosae; (E) HCMV colitis, showing detection of the polymorphonuclear cell marker neutrophil elastase in a colonic stromal cell. Bar (A-E): 30 pm. Arrowheads indicate the locations of HCMV-infected cells.

Although the well-known cytopathic effect of cell enlargement and the formation of nuclear inclusions could be demonstrated in most cell types, granulocytes and cells within the trophoblastic layer of the placental villi consistently did not display typical morphological alterations. This agrees well with our observations that early and late viral gene products were not detectable in these cells (Sinzger et al., 1993b, 1995a,b; Grefte et al., 1994).Therefore, they may be abortively infectable by HCMV, whereas the other cell types seem to support the complete viral replication cycle.

B . Hematopoietic System and Circulating Cells Human cytomegalovirus can be isolated from peripheral blood cells during acute infection. In addition, HCMV can be transmitted by blood

HUMAN CYTOMEGALOVIRUS

227

228

BOD0 PLACHTER et al.

HUMAN CYTOMEGALOVIRUS

229

transfusions obtained from seropositive donors. Removal of the white blood cell fraction decreases the infectivity of blood products considerably (Gilbert et al., 1989; Smith et al., 1993a). Consequently, much attention focused on circulating white blood cells as being targets of HCMV infection. Viral components have been detected by several techniques in peripheral blood lymphocytes (PBL), peripheral blood monocytes (PBMC),and polymorphonuclear leukocytes (PMNL). Lymphocytes were among the first blood cells where HCMV DNA, RNA, and antigen were detected (Rice et al., 1984). Furthermore T lymphocytes were reported to be permissive for HCMV infection in uitro (Braun and Reiser, 1986); however, studies have demonstrated that, in uiuo, lymphocytes are at best sporadically infected by HCMV (TaylorWiedeman et al., 1991; Gerna et al., 1992a). During acute infection, viral antigen has been most frequently found in granulocytes; these “antigen positive” cells, which can make up 1% of buffy coat preparations in acute infection, were primarily thought to express viral immediate early antigen as a result of active infection (Schirm et al., 1987). The detection of such cells has proved to be very useful in monitoring patients at risk for severe HCMV disease (Boland et al., 1990; The et al., 1990). Later, it was found that only a fraction of these cells expressed the nonstructural IE1-pp72 protein, thus being actively infected by HCMV. However, the major viral antigen detectable in granulocytes was the tegument protein pp65 (UL83), which is a structural component of extracellular particles (Grefte et al., 1992a,b; Revello et al., 1992).I n situ cytohybridization analyses of such cells showed that the delayed early gene coding for pp65 was not transcriptionally active in granulocytes (Grefte et al., 1994). During acute HCMV infection, PMNLs may take up pp65 through adhesion to permissively infected endothelial cells (Grefte et al., 1993a, 1994). Therefore, granulocytes may be susceptible to HCMV infection, but viral gene expression in such cells appears t o be abortive and restricted to IE genes. This is consistent with the findings of Taylor-Wiedeman and collegues, who could show that, during latent infection in uiuo, HCMV DNA cannot be detected in PMNL using sensitive nested PCR (polymerase chain reaction) technology (Taylor-Wiedeman et al., 1993). In contrast, peripheral monocytes have been found to carry viral DNA during acute and latent HCMV infection (Dankner et al., 1990; Taylor-Wiedeman et al., 1991). In addition, viral antigen could be detected in these cells during the acute phase of HCMV infection. The presence of viral proteins in the cells could be attributed in part to the expression of viral genes in the cells (Grefte et al., 1994). However, no late viral gene expression was found in PBMCs isolated from actively infected patients. Concordant with this, we found that cytomegalic

230

BOD0 PLACH'IER et al.

monocytes, characteristic for permissive infection, were absent in peripheral blood but were found in tissue sections. However, tissue macrophages, which are the differentiated equivalent of circulating monocytes, showed expression of viral antigens from all stages of viral replication in sections from various organs (Sinzger et al., 1995a,b). Nevertheless, it remains unclear where monocytes/macrophages become infected with HCMV. Finally, HCMV-infected cytomegalic inclusion body cells (CCIC) have been reported to be present in peripheral blood during acute infection (Grefte et al., 1993a,b). Immunocytochemical analyses demonstrated that the cytomegalic cells were of nonhematopoietic origin. Using a panel of MAbs directed against different cell marker proteins, the CCIC were identified as endothelial cells (Grefte et al., 1993b). Viral proteins from all stages of the replicative cycle were found, consistent with the notion that these circulating endothelial cells were fully permissive for HCMV. AND PATHOGENESIS V. VIRALSPREAD

The role of distinct cell types in the spread of HCMV in the organism and throughout the population has not yet been determined unequivocally. The same is true for the role that HCMV-infected cells play in the pathogenesis of symptomatic HCMV disease. In general, the ability of HCMV to infect productively a broad spectrum of cell types in different tissues seems to enable the virus to spread under various circumstances (Fig. 7). The occurrence of asymptomatic chronic and latent infection results in prolonged initial shedding of HCMV and in repeated periods of excretion of infectious virus during reactivation from the latent state (Kumar et al., 1973; Reynolds et al., 1973; Stagno et al., 1975; Huang et al., 1980). Both conditions facilitate the silent distribution of the virus among individuals.

A . Modes of Transmission Although a number of morphological studies focused on the sites of viral replication in symptomatic disease, the definition of organs and cell types involved in the spread of the virus among asymptomatically infected individuals is still mainly restricted to epidemiological research. Again, as no animal model for HCMV is available, only the investigation of the shedding of HCMV in body fluids can supply indirect evidence for the sites of viral replication and the modes of transmission. Human cytomegaloviral antigen, DNA, or infectious virus

HUMAN CYTOMEGALOVIRUS

23 1

FIG.7. Schematic representation of hypothetical routes of HCMV spreading in acute infection. Endothelial cells, macrophages, and polymorphonuclear cells seem to be involved in the hematogenous spread of the virus, whereas fibroblasts and smooth muscle cells may be important for the expansion of infection in an infected organ. Additionally, HCMV may spread via the lumen of alveoli and the gastrointestinal tract by disconnected epithelial cells.

has been detected in various body fluids including blood, breast milk, saliva, urine, genital secretions, and throat swabs. The role of body fluids in the spread of HCMV throughout the population varies with different periods of life. Although in congenital inclusion disease the virus is assumed to be transplacentally transmitted by blood cells, genital HCMV infection of the mother seems to cause perinatal infection. In addition, breast feeding is an important source of the virus in perinatal infection (Stagno et al., 1980; Dworsky et al., 1983; Stagno and Cloud, 1994). Postnatal infections in general appear to be mediated by infectious body secretions rather than by the transmission of infected cells. For instance, infectious virus was recovered more frequently from the cellfree fraction of bronchoalveolar lavage specimens as compared t o the cellular fraction (Clarke et al., 1992). This would suggest the upper respiratory tract as a primary portal of entry after transmission via droplets. In support of this, studies of unselected autopsy material from children revealed a very high prevalence of silent HCMV infection of up to 32% in salivary glands (Alford and Britt, 1990). Silent infection of salivary glands is also a possible source for viral spread among adolescents and young adults, thus resembling the transmission of Epstein-Barr virus in infectious mononucleosis (Stagno and Cloud, 1994). However, no formal proof for such a route of transmission has been provided. Also ill-defined are the cells involved in primar y replication. Esophageal squamous mucosal epithelium appears un-

232

BOD0 PLACHTER et al.

susceptible to HCMV infection, whereas the epithelium of the lower gastrointestinal tract has been shown to be highly susceptible to HCMV infection (Francis et al., 1989) (Table 11). Another site of entry of HCMV appears to be the genital tract. It has been shown that HCMV is present in up to 9% of cases in the genital tract of pregnant or nonpregnant women (Montgomery et al., 1972; Knox et al., 1979). In addition, HCMV is detectable in the semen of young adults, declining after the age of 30 years (Spector et al., 1984; Lang and Kummer, 1975; Leach et al., 1994). Therefore, some authors emphasize the role of sexual activity in adolescents for the transmission of HCMV (Handsfield et al., 1985; Chandler et al., 1985; Ho, 1990).

B . Cell Types Involved in Spread and Pathogenesis The multiorgan involvement of HCMV infection is suggestive of effective mechanisms of viral spread in the body. During acute infection, virus can be isolated from buffy coat preparations of leukocytes, thus implicating one or more of the leukocyte populations as likely to be the carrier cells of the virus. Viral antigens have been predominantly found in neutrophils and monocytes in acute infection (Grefte et al., 1992a; Gerna et al., 1992b). Although evidence is lacking that granulocytes are permissive to HCMV replication, monocytes have been demonstrated to synthesize progeny virus on differentiation (Weinshenker et al., 1988; Ibanez et al., 1991; Lathey and Spector, 1991; Grefte et al., 1994; Taylor-Wiedeman et al., 1994).This is consistent with our findings in immunohistochemical double-labeling analyses, where infiltrating macrophages in organ tissues were found to be permissively infected, whereas granulocytes showed no signs of viral late gene expression (Sinzger et al., 199513).It has been suggested that infected blood monocytes can adhere to activated vascular endothelial cells and subsequently infiltrate solid tissues (Nelson et al., 1990). On differentiation, such cells then would be able to transmit infectious virus to highly permissive cell populations in solid organs. Supportive of this hypothesis are results of immunohistochemical analyses of biopsies from heart allografts, where infection with HCMV correlated with elevated levels of expression of the adhesion molecule VCAM-1 (Koskinen, 1993). In addition, in vitro studies on cultured endothelial cells suggest that the process of adherence of leukocytes to the vascular wall is enhanced by HCMV infection in such cells (Span et al., 1989, 1991; Waldman et al., 1993). Another population of cells that have been suggested to be involved in hematogenous spread of the virus are circulating endothelial cells, which have detached from the basal membrane as a result of HCMV

HUMAN CYTOMEGALOVIRUS

233

infection (Grefte et al., 1993b; Sinzger et al., 1995a). The size of such cells prevents their passage through capillaries. They are presumed to get trapped in small vessels and thus may be able to transmit HCMV to adjacent tissues. However, the molecular mechanisms of interaction of circulating monocytes or endothelial cells with the microvascular system are still poorly understood. Blood cells are also thought to be involved in prenatal transmission of HCMV via the placenta (Benirschke et al., 1974). Infection of the placenta is a constant finding in connatally infected newborns (Garcia et al., 1989).However, the routes of transmission past the trophoblast cells separating fetal from maternal blood is still unclear (Rosenthal et al., 1981; Sinzger et al., 1993b). In particular, it is not known whether maternal leukocytes, infected with HCMV, can infiltrate the villous stroma of the placenta and thus gain entry to the fetal circulation. Nevertheless, it has been shown that macrophages, fibroblasts, and endothelial cells are infected in the villous stroma after prenatal transmission (Schwartz et al., 1992; Muhlemann et al., 1992; Sinzger et al., 1993b).Therefore, a mechanism comparable to infiltration in solid organs may also be operative in prenatal transmission. The multiplication and expansion of HCMV after initial infection of solid organs are thought t o determine the development of symptomatic disease in many instances. The virus may spread within infected organs by direct cell-cell contact or by distribution of infected cells via the lumen of preformed cavities. A number of different studies supported the role of stromal cells in the expansion of HCMV organ infection, although the nature of these cells was not clear (Table 11). Fibroblasts, smooth muscle cells, and endothelial cells could be identified to be infected in tissue sections from lung, gastrointestinal tract, and placenta (Sinzger et al., 1993a,b; Sinzger et al., 1995a).In addition, late stage infected epithelial cells have been found in the lumen of gastrointestinal glands as well as in alveolar spaces (Sinzger et al., 1995a). Therefore, once the virus has gained entry to a particular organ, it may replicate in various cell populations and thus may cause organ malfunction. In certain tissues the virus seems to be present without affecting cell or organ function and without inducing an inflammatory response (Toorkey and Carrigan, 1989) (Table 11).There is ongoing discussion about the pathogenic relevance of these findings. In this context it appears interesting that mechanisms have been reported by which HCMV is able to evade immune defense mechanisms of the host (Barnes and Grundy, 1992; Del-Val et al., 1992; Gilbert et al., 1993; Warren et al., 1994; for reviews on the immune response to cytomegalovirus infection, see Koszinowski, 1991; Koszinowski et al., 1992). On the other hand, cyto-

234

BOD0 PLACHTER et al.

pathogenicity is clearly correlated with symptoms in some clinical settings (Kennedy et al., 1986; Jacobson and Mills, 1988a; Francis et al., 1989; Aqel et al., 1991; Aukrust et al., 1992; Rummelt et al., 1994; Sinzger et al., 1995a). Direct effects of HCMV on cell function is further supported by the beneficial effect of gancyclovir therapy in patients suffering from HCMV-associated retinitis and gastroenteritis and by successful prophylaxis using gancyclovir in the prevention of pneumonitis (Jacobson and Mills, 1988b; Francis et al., 1989; Li et al., 1994). Virus-induced inflammatory reactions may add to the impairment of parenchymal cell function. Yet, based on the discrepancy of massive immunoreaction in contrast to very few HCMV-infected cells in certain cases, immunopathological mechanisms induced by an otherwise irrelevant HCMV have been suggested (Grundy et al., 1987; Grundy, 1990; Humbert et al. 1992). Therefore, different pathogenic mechanisms appear to be operative which lead to the impairment of organ cell function in HCMV infection, and part of this may be related to immune functions inadequate to eliminate the virus. VI. LATENT CYTOMEGALOVIRUS INFECTION Primary infection with HCMV leads to a state of lifelong persistence, from which the virus can be reactivated. Latency in this clinical context is defined as the inability to detect infectious virus despite evidence that the virus is still present in the organism. Although no direct proof of HCMV latency has been provided thus far, epidemiological data strongly suggest that, as with other herpesviruses, there is HCMV latency and, resulting from it, recurrence of the original virus (Huang et al., 1980). Reactivation of HCMV and subsequent development of disease are usually associated with conditions of immunosuppression (for review, see Ho, 1991). Considerable scientific effort has therefore focused on elucidating the site of viral latency and the molecular and immunologic mechanisms that lead to latency and reactivation. However, these have met with only limited success, mainly because of the lack of an animal model system for HCMV. No firm evidence has been presented as to whether HCMV infects its host in a latent or chronic form. Chronic persistent infection is commonly looked on as permanent shedding of virus from a limited number of cells, a process that is controlled by a competent immune system. This contrasts with latent persistent infection, which is characterized by the lack of production of infectious virus particles. As these two alternative mechanisms are difficult to distinguish during infection

HUMAN CYTOMEGALOVIRUS

235

of HCMV in its natural host, we focus on what is known about viral latency/persistence on the cellular level. Here, latency is defined as the reversibly nonproductive infection of a cell by replication-competent virus; this has been distinguished from irreversibly nonproductive (abortive) infection and chronic persistent infection, the latter of which is associated with the continuous shedding of virus (Garcia-Blanco and Cullen, 1991). In this section we summarize the knowledge that has been gathered from studies on HCMV and on the closely related murine cytomegalovirus (MCMV). We use the term “latent infection,” but we acknowledge that future studies may show that cytomegalovirus causes chronic persistent rather than latent persistent infection on the cellular level.

A . Latent Murine Cytomegalovirus Infection as Model Murine cytomegalovirus (MCMV) has been frequently used as a model system for HCMV (Hudson et al., 1979; Hudson, 1979; Jordan, 1983). Several aspects of MCMV concerning transmission, shedding, pathogenesis, immune defense, and molecular biology appear to be sufficiently comparable in order to aid to the understanding of the pathogenetic mechanisms underlying HCMV infection (Jordan et al., 1978; Hudson et al., 1979; Hudson, 1979; Shanley et al., 1979; Jordan, 1983; Loh and Hudson, 1980, 1982; Keil et al., 1984, 1985, 1987a,b; Dorsch-Hasler et al., 1985; Koszinowski et al., 1986, 1990; Schickedanz et al., 1988; Buhler et al., 1990; Messerle et al., 1991, 1992). Latent MCMV infection has been one issue of preferential interest in this respect (Jordan et al., 1984; Gonczol et al., 1985). Reactivation of MCMV to replicate in permissive culture from tissue devoid of detectable virus has been used as a criterion t o define MCMV latent infection. By this definition, several organs of mice have been found to be latently infected by MCMV (Olding et al., 1975; Mayo et al., 1978; Wise et al., 1979; Jordan and Mar, 1982; Jordan et al., 1982; Porter et al., 1985; Wilson et al., 1985; Schmader et al., 1991). Using the polymerase chain reaction, an even broader spectrum of organs including spleen, salivary glands, kidneys, liver, lungs, heart, and brain have been found to be latently infected (Collins et al., 1993).Consistent with this, both MCMV and HCMV are well known to infect multiple organs during acute infection. Most attention focused on the spleen as being the major site of MCMV latency. Using in situ cytohybridization, immunohistochemistry, electron microscopy, or PCR, sinusoidal lining cells and stromal cells within the spleen have been favored as being latently infected (Mercer et al., 1988; Pomeroy et al., 1991). In a study using PCR an-

236

BOD0 PLACHTER et al.

alyses, RNA encoded by the viral IE1 gene could be detected in latently infected mouse spleens, indicating that, during latent infection, the MCMV genome may be transcriptionally active (Henry and Hamilton, 1993). Alternatively, low-level productive infections, undetectable by other methods, might have generated positive results. Using quantitative PCR technology, Balthesen et al., (1993) found several organs latently infected by MCMV, the lungs being the major site of latency. Interestingly, MCMV DNA could be detected in the peripheral blood of mice for prolonged times over 6 months after acute infection; however, the virus was eventually cleared from peripheral blood in many animals, but DNA remained detectable in solid organs. In another study, it could be shown that the overall burden of viral DNA in latently infected organs correlated with the organ-specificand overall risk for reactivation of MCMV (Fkddehase et al., 1994). The establishment of latency in a particular organ, however, did not correlate with local virus production (Balthesen et al., 1994). These results, together with the data from other studies, indicate that (1) MCMV can latently infect multiple organs; (2) different cell types may harbor latent MCMV in vivo; and (3) the MCMV genome may be transcriptionally active during latent infection.

B . Site

of

Human Cytomegalovirus Latency

Multiorgan latency as suggested for MCMV is an attractive model that would explain several clinical findings concerning HCMV infection. It has long been known that HCMV can be transmitted from latently infected donors via blood and blood products (Ho, 1991). Furthermore, transmission of HCMV via a graft from a seropositive donor is well acknowledged as a major source of posttransplantation infection with HCMV (Chou, 1986, 1987; Ho, 1990). Therefore, HCMV is supposed to be located at different sites in the body of asymptomatic seropositive subjects and may thus be reactivated from latency in different organs. One mechanism to explain this would be that HCMV can latently infect peripheral blood cells, which would distribute the virus through different organs. As a matter of fact, early studies claimed that HCMV was contained in a latent state in peripheral blood leukocytes (Schrier et al., 1985). Analyses using sensitive PCR methods have confirmed the prolonged presence of HCMV DNA in peripheral leukocytes after acute HCMV infection (Taylor-Wiedeman et al., 1991; Delgado et al., 1992; Stanier et al., 1992; Smith et al., 1993b; Bitsch et al., 1993). Monocytes have been suggested in this respect to be the latently infected population, whereas granulocytes were found to be negative by

HUMAN CYTOMEGALOVIRUS

237

PCR (Taylor-Wiedeman et al., 1993). In addition, using in situ cytohybridization on renal grafts, infiltrating monocytic cells were suggested to be the main reservoir of latent infection (Nelson et al., 1990). Therefore, it is reasonable to assume that monocytes can distribute the complete genetic formation of HCMV throughout the body. However, doubt has been raised about short-lived cells such as peripheral blood cells being an optimal site of lifelong latency (Balthesen et al., 1993). Bone marrow cells, including hematogenic precursors, have been discussed as being a major latently infected cell population. In vitro, bone marrow stromal cells, fibroblasts, and progenitor cells have been reported to be infectable by HCMV (Reiser et al., 1986; Simmons et al., 1990; Maciejewski et al., 1992).Replication of HCMV dependent on the state of differentiation of progenitor cells could be shown (Minton et al., 1994). However, it remains to be determined whether bone marrow cells of normal seropositive persons contain HCMV. Besides peripheral blood, few studies have been carried out addressing the question of HCMV DNA being present in tissues from seropositive persons using methods sensitive enough to detect low levels of viral nucleic acids. Using PCR, HCMV has been detected in arterial walls and synovial membranes (Hendrix et al., 1990; Einsele et al., 1992). Also by PCR and in situ cytohybridization, HCMV DNA has been found in the vascular tree of patients with disease not related to HCMV (Hendrix et al., 1990, 1991; Wu et al., 1992). One study found HCMV antigen in multiple organs from seropositive subjects employing immunohistochemical analyses (Toorkey and Carrigan, 1989). Future studies using well-defined MAbs and improved PCR protocols will resolve the question of whether, as in murine cytomegalovirus, multiple organ involvement is also characteristic of HCMV latency.

C . Cell Culture Models for Human Cytomegalovirus Latency Given the strict host specificity of HCMV, little information is available on the molecular mechanisms that regulate latency and reactivation. In other herpesviruses, considerable information about viral gene expression during latency has been gathered from culture models (reviewed in Liebowitz and Kieff, 1993; Fbizman and Sears, 1993). Persistent infection of cultured cells by HCMV has been reported (Tocci and St-Jeor, 1979; Furukawa, 1979a,b; Mocarski and Stinski, 1979; Ogura et al., 1986; Tanaka et al., 1987, 1991). Most of these cells, however, produced low levels of infectious particles, consistent with chronic persistent rather than with latent infection. One early report focused on the infection of fibroblasts with HCMV at high MOI, thereby establishing persistent productive infection; the synthesis of infectious par-

238

BOD0 PLACHTER et al.

ticles, however, became undetectable while treating the cells with antisera against HCMV (Mocarski and Stinski, 1979). In one study, latent nonproductive infection in thyroid papillary carcinoma cells when kept at elevated temperatures was reported (Tanaka et al., 1987). Under these conditions, expression from the IE locus but no viral replication was found. However, conflicting data were presented in a subsequent publication (Tanaka et al., 1991). We have developed a strategy to select those culture cells that are latently infected by HCMV using selectable mutants of HCMV. These viruses carried the genetic information for bacterial neomycin phosphotransferase (neo)inserted in a genomic region dispensable for growth in culture (Wolff et al., 1993). Viral recombinants were used to infect different cell lines of human and nonhuman origin. Treatment of infected cultures with G418 was used successfully to isolate human glioblastoma cells latently infected with HCMV (Wolff et aZ.,1994). These cells, termed U13WRVAneo2, expressed the viral IE1-pp72 in a frequency of over 90%. The IE2-pp86 was reduced in its expression as compared with permissive infection in the same cell line, but it could be induced by mitogenic stimulation with phorbol esters, frequent passaging of cells, or treatment with retinoic acid. This subsequently led to the induction of early gene expression as exemplified by the expression of pp65 (UL83) and p52 (UL44). However, no late protein expression could be detected under such conditions. Therefore U138/RVAneo2 provides a culture system where the regulatory mechanisms leading to the induction of early and late viral function from a state of latency of the genome can be experimentally addressed. Such studies will have to focus on IE1-pp72. This protein is present in high amounts in latently infected U138/RVAneo2. As expected from transient analyses, this protein alone did not significantly induce early gene expression. It did not influence cell growth significantly as compared to uninfected glioblastoma cells. Besides its trans-activating functions, association of this protein with metaphase chromosomes has been demonstrated (Lafemina et al., 1989). It therefore might be reasonable to assume that IE1-pp72 is also expressed during latency in uiuo. Another indication that the major immediate early region of HCMV is active for prolonged periods during latency would be that this genomic region is deficient in CpG dinucleotides when compared to the rest of the genome (Honess et al., 1989). Deficiency of CpG dinucleotides in DNA of higher eukaryotes is diagnostic for the accessability of this region to methylation. Methylation of DNA is a slow process and is preferentially found in DNA regions that are transcriptionally active in contrast to regions covered by histones. Future investigations will have to focus on the expression of IE1-pp72 in latently

HUMAN CYTOMEGALOVIRUS

239

infected tissues of human origin. As exemplified by this, culture models for HCMV latency are important in the understanding of the molecular mechanisms that regulate HCMV latency; results obtained from culture latency will provide the basis for molecular analyses of the role of particular viral and cellular genes in latency and reactivation in uiuo. VII. SUMMARY

As the number of patients suffering from severe HCMV infections has steadily increased, there is a growing need to understand the molecular mechanisms by which the virus causes disease. The factors that control infection at one time and the events leading t o virus multiplication at another time are only beginning to be understood. The interaction of HCMV with different host cells is one key for elucidating these processes. Through modern techniques, much has been learned about the biology of HCMV infections in culture systems. In addition to endothelial cells, epithelial cells, and smooth muscle cells, fibroblasts are one cell population preferentially infected in solid tissues in uiuo. From these sites of multiplication, the virus may be carried by peripheral monocytes and circulating endothelial cells to reach distant sites of the body. This would explain the multiorgan involvement in acute HCMV infection and the modes of viral transmission. From what has been learned mainly from human fibroblast culture systems, future studies will focus on how HCMV regulates the expression of its putative 200 genes in different host cells at different stages of cell differentiation and activation to result in viral latency and pathogenesis. ACKNOWLEDGMENTS We thank Michael Mach, Susi Schmolke, and Thomas Stamminger for critical comments regarding the manuscript. Part of the work presented here was supported by the German Research Foundation (DFG, Grant F191/10-4) and the German Department for Research and Technology (BMFT, Grant 01 KI 9301).

REFERENCES Adlish, J. D., Lahijani, R. S., and St.Jeor, S. C. (1990). Identification of a putative cell receptor for human cytomegalovirus. Virology 176, 337-345. Akrigg, A., Wilkinson, G. W., and Oram, J. D. (1985). The structure ofthe major immediate early gene of human cytomegalovirus strain AD169. Virus Res. 2, 107-121. Albrecht, T., Nachtigal, M., StJeor, S. C., and Rapp, F. (1976). Induction of cellular DNA

240

BOD0 PLACHTER et al.

synthesis and increased mitotic activity in Syrian hamster embryo cells abortively infected with human cytomegalovirus. J. Gen. Virol. 30, 167-177. Aldrete, J. S., Sterling, W. A., Hathaway, B. M., Morgan, J. M., and Diethelm, A. G. (1975). Gastrointestinal and hepatic complications affecting patients with renal allografts. A m . J. Surg. 129, 115-124. Alford, C. A., and Britt, W. J. (1990). Cytomegalovirus. I n “Virology” (B. N. Fields and D. M. Knipe, eds.), pp. 1981-2010. Raven, New York. Anders, D. G., and Gibson, W.(1988). Location, transcript analysis, and partial nucleotide sequence of the cytomegalovirus gene encoding a n early DNA-binding protein with similarities to ICP8 of herpes simplex virus type 1. J. Virol. 62, 1364-1372. Anders, D. G., and Punturieri, S. M. (1991). Multicomponent origin of cytomegalovirus lytic-phase DNA replication. J. Virol. 65, 931-937. Anders, D. G., Kacica, M. A., Pari, G., and Punturieri, S. M. (1992). Boundaries and structure of human cytomegalovirus oriLyt, a complex origin for lytic-phase DNA replication. J. Virol. 66, 3373-3384. Anders, K. H., Guerra, W. F., Tomiyasu, U., Verity, M. A., and Vinters, H. V. (1986). The neuropathology of AIDS. UCLA experience and review. A m . J.Pathol. 124,537-558. Apperley, J. F., Dowding, C . , Hibbin, J., Buiter, J., Matutes, E., Sissons, P. J.,Gordon, M., and Goldman, J . M. (1989). The effect of cytomegalovirus on hemopoiesis: In uitro evidence for selective infection of marrow stromal cells. Exp. Hematol. 17, 38-45. Aqel, N. M., Tanner, P., Drury, A., Francis, N. D., and Henry, K. (1991). Cytomegalovirus gastritis with perforation and gastrocolic fistula formation. Histopathology 18, 165168. Arbustini, E., Grasso, M., Diegoli, M., Percivalle, E., Grossi, P., Bramerio, M., Campana, C., Goggi, C., Gavazzi, A., and Vigano, M. (1992). Histopathologic and molecular profile of human cytomegalovirus infections in patients with heart transplants. A m . J. Clin. Pathol. 98, 205-213. Arlt, H., Lang, D., Gebert, S., and Stamminger, T. (1994). Identification of binding sites for the 86-kilodalton IE2 protein of human cytomegalovirus within a n IE2-responsive viral early promoter. J.Virol. 68, 4117-4125. Arnold, J. C., Portmann, B. C., O’Grady, J. G., Naoumov, N. V., Alexander, G. J., and Williams, R. (1992). Cytomegalovirus infection persists in the liver graft in the vanishing bile duct syndrome. Hepatology (Baltimore) 16, 285-292. Arya, S. K., and Sethi, A. (1990). Stimulation of the human immunodeficiency virus type 2 (HIV-2) gene expression by the cytomegalovirus and HIV-2 transactivator gene. AIDS Res. Hum. Retroviruses 6, 649-658. Aukrust, P., Farstad, I. N., Froland, S. S., and Holter, E. (1992). Cytomegalovirus (CMV) pneumonitis in AIDS patients: The result of intensive CMV replication? Eur. Respir. J. 5, 362-364. Balthesen, M., Messerle, M., and Reddehase, M. J. (1993). Lungs are a major organ site of cytomegalovirus latency and recurrence. J. Virol. 67, 5360-5366. Balthesen, M., Dreher, L., Lucin, P., and Reddehase, M. J. (1994). The establishment of cytomegalovirus latency in organs is not linked to local virus production during primary infection. J. Gen. Virol. 75, 2329-2336. Bamborschke, S., Wullen, T., Huber, M., Neveling, M., Baldamus, C . A., Korn, K., and Jahn, G. (1992). Early diagnosis and successful treatment of acute cytomegalovirus encephalitis in a renal transplant recipient. J.Neurol. 239, 205-208. Baracchini, E., Glezer, E., Fish, K., Stenberg, R. M., Nelson, J. A., and Ghazal, P. (1992). An isoform variant of the cytomegalovirus immediate early auto repressor functions as a transcriptional activator. Virology 188, 518-529. Barkholt, L. M., Ehrnst, A., and Veress, B. (1994). Clinical use of immunohistopatho-

HUMAN CYTOMEGALOVIRUS

241

logic methods for the diagnosis of cytomegalovirus hepatitis in human liver allograft biopsy specimens. Scan. J . Gastroenterol. 29, 553-560. Barnes, P. D., and Grundy, J. E. (1992). Down-regulation of the class I HLA heterodimer and beta 2-microglobulin on the surface of cells infected with cytomegalovirus. J. Gen. Virol. 73, 2395-2403. Behar, R., Wiley, C., and McCutchan, J. A. (1987). Cytomegalovirus polyradiculoneuropathy in acquired immune deficiency syndrome. Neurology 37, 557-561. Benirschke, K., Mendoza, G. R., and Bazeley, P. L. (1974).Placental and fetal manifestations of cytomegalovirus infection. Virchows Arch. B 16, 121-139. Biegalke, B. J., and Geballe, A. P. (1991). Sequence requirements for activation of the HIV-1 LTR by human cytomegalovirus. Virology 183, 381-385. Bitsch, A., Kirchner, H., Dennin, R., Hoyer, J., Fricke, L., Steinhoff, J., Sack, K., and Bein, G. (1993). The long persistence of CMV DNA in the blood of renal transplant patients after recovery from CMV infection. Transplantation 56, 108-113. Blanton, R. A,, and Tevethia, M. J. (1981). Immunoprecipitation of virus-specific immediate-early and early polypeptides from cells lytically infected with human cytomegalovirus strain AD 169. Virology 112, 262-273. Boland, G. J., de-Gast, G. C., Hene, R. J., Jambroes, G., Donckerwolcke, R., The, T. H., and Mudde, G. C. (1990). Early detection of active cytomegalovirus (CMV) infection after heart and kidney transplantation by testing for immediate early antigenemia and influence of cellular immunity on the occurrence of CMV infection. J . Clin. Microbiol. 28, 2069-2075. Boldogh, I., AbuBakar, S., and Albrecht, T. (1990). Activation of proto-oncogenes: An immediate early event in human cytomegalovirus infection. Science 247, 561-564. Boldogh, I., Fons, M. P., and Albrecht, T. (1993). Increased levels of sequence-specific DNA-binding proteins in human cytomegalovirus-infected cells. Biochem. Biophys. Res. Commun. 197, 1505-1510. Bonkowsky, H. L., Lee, R. V., and Klatskin, G . (1975). Acute granulomatous hepatitis. Occurrence in cytomegalovirus mononucleosis. JAMA, J . A m . Med. Assoc. 233,12841288. Boshart, M., Weber, F., Jahn, G., Dorsch-Hasler, K., Fleckenstein, B., and Schaffner, W. (1985). A very strong enhancer is located upstream of an immediate early gene of human cytomegalovirus. Cell (Cambridge, Mass.) 41, 521-530. Braun, R. W., and Reiser, H. C. (1986). Replication of human cytomegalovirus in human peripheral blood T cells. J . Virol. 60, 29-36. Britt, W. J., and Vugler, L. (1987).Structural and immunological characterization of the intracellular forms of a n abundant 68,000 M, human cytomegalovirus protein J.Gem Virol. 68, 1897-1907. Buhler, B., Keil, G. M., Weiland, F., and Koszinowski, U. H. (1990). Characterization of the murine cytomegalovirus early transcription unit e l that is induced by immediateearly proteins. J . Virol. 64, 1907-1919. Buonaguro, F. M., McDougall, J. K., and Galloway, D. A. (1987). Characterization of the integration site of the CMV mtr in a tumor cell line. Virology 156, 74-83. Caswell, R. C., Hagemeier, C., Hayhurst, G. P., Kouzarides, T., and Sinclair, J. H. (1993a). Mechanisms of transactivation of cellular promotors by the HCMV major immediate early proteins involve direct interaction with cellular transcription factors. In “Multidisciplinary Approach to Understanding Cytomegalovirus Disease” (S. Michelson and S. A. Plotkin, eds.), pp 13-16. Elsevier, Amsterdam. Caswell, R., Hagemeier, C., Chiou, C. J., Hayward, G., Kouzarides, T., and Sinclair, J. (1993b).The human cytomegalovirus 86K immediate early (IE)2 protein requires the basic region of the TATA-box binding protein (TBP) for binding, and interacts with

242

BOD0 PLACHTER et al.

TBP and transcription factor TFIIB via regions of IE2 required for transcriptional regulation. J. Gen. Virol. 74, 2691-2698. Chandler, S. H., Holmes, K. K., Wentworth, B. B., Gutman, L. T., Wiesner, P. J., Alexander, E. R., and Handsfield, H. H. (1985). The epidemiology of cytomegaloviral infection in women attending a sexually transmitted disease clinic. J. Infect. Dis. 152,597605. Chang, C. P., Malone, C. L., and Stinski, M. F. (1989). A human cytomegalovirus early gene has three inducible promoters that are regulated differentially at various times after infection. J. Virol. 63, 281-290. Chang, Y. N., Crawford, S., Stall, J., Rawlins, D. R., Jeang, K. T., and Hayward, G. S. (1990). The palindromic series I repeats in the simian cytomegalovirus major immediate-early promoter behave as both strong basal enhancers and cyclic AMP response elements. J. Virol. 64, 264-277. Chee, M. S., Bankier, A. T., Beck, S., Bohni, R., Brown, C. M., Cerny, R., Horsnell, T., Hutchison, C. A., Kouzarides, T., Martignetti, J. A., Preddie, E., Satchwell, S. C., Tomlinson, P., Weston, K. M., and Barrell, B. G. (1990). Analysis of the protein-coding content of the sequence of human cytomegalovirus strain AD169. Curr. Top. Microbiol. Immunol. 154, 125-169. Cherrington, J. M., and Mocarski, E. S. (1989). Human cytomegalovirus iel transactivates the alpha promoter- enhancer via an 18-base-pair repeat element. J. Virol. 63, 1435-1440. Cherrington, J. M., Khoury, E. L., and Mocarski, E. S. (1991). Human cytomegalovirus ie2 negatively regulates alpha gene expression via a short target sequence near the transcription start site. J. Virol. 65, 887-896. Chiou, C. J., Zong, J., Waheed, I., and Hayward, G. S. (1993). Identification and mapping of dimerization and DNA-binding domains in the C terminus of the IE2 regulatory protein of human cytomegalovirus. J. Virol. 67, 6201-6214. Chou, S. W. (1986). Acquisition of donor strains of cytomegalovirus by renal-transplant recipients. N . Engl. J. Med. 314, 1418-1423. Chou, S. W. (1987). Cytomegalovirus infection and reinfection transmitted by heart transplantation. J . Infect. Dis. 155, 1054-1056. Chou, S. (1992). Molecular epidemiology of envelope glycoprotein H of human cytomegalovirus. J. Infect. Dis. 166, 604-607. Chou, S. W., and Dennison, K. M. (1991). Analysis of interstrain variation in cytomegalovirus glycoprotein B sequences encoding neutralization-related epitopes. J. Infect. Dis. 163, 1229-1234. Clarke, L. M., Daidone, B. J.,Inghida, R., Kirwin, M., and Sierra, M. F. (1992). Differential recovery of cytomegalovirus from cellular and supernatant components of bronchoalveolar lavage specimens. A m . J. Clin. Puthol. 97,313-317. Colberg-Poley, A. M., Santomenna, L. D., Harlow, P. P., Benfield, P. A., and Tenney, D. J. (1992). Human cytomegalovirus US3 and UL36-38 immediate-early proteins regulate gene expression. J. Virol. 66, 95-105. Cole, R., and Kuttner, A. G. (1926). A filterable virus present in the submaxillary glands of guinea pigs. J . Exp. Med. 44, 8885-8873. Collins, T., Pomeroy, C., and Jordan, M. C. (1993). Detection of latent cytomegalovirus DNA in diverse organs of mice. J. Infect. Dis. 168, 725-729. Compton, T. (1993). An immortalized human fibroblast cell line is permissive for human cytomegalovirus infection. J. Virol. 67, 3644-3648. Compton, T., Nowlin, D. M., and Cooper, N. R.'(1993). Initiation of human cytomegalovirus infection requires initial interaction with cell surface heparan sulfate. Virology 193, 834-841.

HUMAN CYTOMEGALOVIRUS

243

Crump, J. W., Geist, L. J., Auron, P. E., Webb, A. C., Stinski, M. F., and Hunninghake, G. W. (1992). The immediate early genes of human cytomegalovirus require only proximal promoter elements to upregulate expression of interleukin-1 beta. Am. J. Respir. Cell Mol. Biol. 6, 674-677. Daicker, B. (1988). Cytomegalovirus panuveitis with infection of corneo-trabecular endothelium in AIDS. Ophthalmologica 197, 169-175. Dankner, W. M., McCutchan, J. A., Richman, D. D., Hirata, K., and Spector, S. A. (1990). Localization of human cytomegalovirus in peripheral blood leukocytes in in situ hybridization. J. Infect. Dis. 161, 31-36. D’Aquila, R. T., Hayward, G. S., and Summers, W. C. (1989). Physical mapping of the human cytomegalovirus (HCMV) (Towne) DNA polymerase gene: DNA-mediated transfer of a genetic marker for an HCMV gene. Virology 171, 312-316. Davis, L. E., Rarey, K. E., Stewart, J. A., and McLaren, L. C. (1987). Recovery and probable persistence of cytomegalovirus in human inner ear fluid without cochlear damage. Ann. Otol. Rhinol. Laryngol. 96,380-383. Delgado, R., Lumbreras, C., Alba, C., Pedraza, M. A,, Otero, J. R., Gomez, R., Moreno, E., Noriega, A. R., and Paya, C. V. (1992). Low predictive value of polymerase chain reaction for diagnosis of cytomegalovirus disease in liver transplant recipients. J. Clin. Microbiol. 30, 1876-1878. Del-Val, M., Hengel, H., Hacker, H., Hartlaub, U., Ruppert, T., Lucin, P., and Koszinowski, U. H. (1992). Cytomegalovirus prevents antigen presentation by blocking the transport of peptide-loaded major histocompatibility complex class I molecules into the medial-Golgi compartment. J. Exp. Med. 176, 729-738. DeMarchi, J. M. (1983). Nature of the block in the expression of some early virus genes in cells abortively infected with human cytomegalovirus. Virology 129, 287-297. DeMarchi, J. M., Schmidt, C. A., and Kaplan, A. S. (1980). Patterns of transcription of human cytomegalovirus in permissively infected cells. J. Virol. 35, 277-286. Dorsch-Hasler, K., Keil, G. M., Weber, F., Jasin, M., Schaffner, W., and Koszinowski, U. H. (1985). A long and complex enhancer activates transcription of the gene coding for the highly abundant immediate early mRNA in murine cytomegalovirus. Proc. Natl. Acad. Sci. U S A . 82, 8325-8329. Dworsky, M., Yow, M., Stagno, S., Pass, R. F., and Alford, C. (1983). Cytomegalovirus infection of breast milk and transmission in infancy. Pediatrics 72, 295-299. Einhorn, L., and Ost, A. (1984). Cytomegalovirus infection of human blood cells. J. Infect. Dis. 149, 207-214. Einhorn, L., Gadler, H., and Wahren, B. (1982). Adsorption of purified human cytomegalovirus and induction of early antigens in different cells. J. Med. Virol. 10,225234. Einsele, H., Steidle, M., Muller, C. A., Fritz, P., Zacher, J., Schmidt, H., and Saal, J. G. (1992). Demonstration of cytomegalovirus (CMV)DNA and anti-CMV response in the synovial membrane and serum of patients with rheumatoid arthritis. J. Rheumutol. 19,677-681. el-Beik, T., Razzaque, A., Jariwalla, R., Cihlar, R. L., and Rosenthal, L. J . (1986). Multiple transforming regions of human cytomegalovirus DNA. J. Virol. 60,645-652. Ertl, P. F., and Powell, K. L. (1992). Physical and functional interaction of human cytomegalovirus DNA polymerase and its accessory protein (ICP36) expressed in insect cells. J. Virol. 66, 4126-4133. Escudero-Fabre, A., Cummings, O., Kirklin, J. K., Bourge, R. C., and Aldrete, J. S. (1992). Cytomegalovirus colitis presenting a hematochezia and requiring resection. Arch. Surg. 127, 102-104. Farber, S., and Wolbach, S. G. (1932). Intranuclear and cytoplasmic inclusions

244

B OD0 PLACHTER et al.

(“protozoan-like bodies”) in the salivary glands and other organs of infants. Am. J. Pathol. 8, 123-135. Fernandes, B., Brunton, J.,and Koven, I. (1986). Ileal perforation due to cytomegaloviral enteritis. Can. J. Surg. 29, 453-456. Fioretti, A., Furukawa, T., Santoli, D., and Plotkin, S. A. (1973). Nonproductive infection of guinea pig cells with human cytomegalovirus, J. Virol. 11, 998-1003. Francis, N. D., Boylston, A. W., Roberts, A. H., Parkin, J. M., and Pinching, A. J. (1989). Cytomegalovirus infection in gastrointestinal tracts of patients infected with HIV-1 or AIDS. J. Clin. Pathol. 42, 1055-1064. Frank, D., and Raicht, R. F. (1984). Intestinal perforation associated with cytomegalovirus infection in patients with acquired immune deficiency syndrome. Am. J. Gastroenterol. 79, 201-205. Fries, B. C., Chou, S., Boeckh, M., and “orok-Storb, B. (1994). Frequency distribution of cytomegalovirus envelope glycoprotein genotypes in bone marrow transplant recipients. J.Infect. Dis. 169, 769-774. Furnari, B. A., Poma, E., Kowalik, T. F., Huong, S.M., and Huang, E. S. (1993). Human cytomegalovirus immediate-early gene 2 protein interacts with itself and with several novel cellular proteins. J. Virol. 67, 4981-4991. Furukawa, T. (1979a). Cell cycle-dependent chronic infection of human cytomegalovirus in human osteogenic sarcoma cells. J. Gen. Virol. 45, 81-88. Furukawa, T. (1979b). Persistent infection with human cytomegalovirus in a lymphoblastoid cell line. Virology 94, 214-218. Furukawa, T., Tanaka, S., and Plotkin, S. A. (1975a). Stimulation of macromoleculear synthesis in guinea pig cells by human CMV. Proc. SOC.Exp. Biol. Med. 148,211214. Furukawa, T., Tanaka, S., and Plotkin, S. A. (197513). Restricted growth of human cytomegalovirus in UV-irradiated WI-38 human fibroblasts. Proc. SOC.Exp. Biol. Med. 148, 1249-1251. Furukawa, T., Sakuma, S., and Plotkin, S. A. (1976). Human cytomegalovirus infection of WI-38 cells stimulates mitochondria1 DNA synthesis. Nature (London) 262, 414416. Garcia, A. G., Fonseca, E. F., Marques, R. L., and Lobato, Y. Y. (1989). Placental morphology in cytomegalovirus infection. Placenta 10, 1-18. Garcia-Blanco, M. A,, and Cullen, B. R. (1991). Molecular basis of latency in pathogenic human viruses. Science 254,815-820. . Gartenhaus, R. B., Wong-Staal, F., and Klotman, M. E. (1991). The promoter of human T-cell leukemia virus type-I is repressed by the immediate-early gene region of human cytomegalovirus in primary blood lymphocytes. Blood 78, 2956-2961. Geballe, A. P., Leach, F. S., and Mocarski, E. S. (1986). Regulation of cytomegalovirus late gene expression: Gamma genes are controlled by posttranscriptional events. J. Virol. 57, 864-874. Geist, L. J., Monick, M. M., Stinski, M. F., and Hunninghake, G. W. (1991). The immediate early genes of human cytomegalovirus upregulate expression of the interleukin-2 and interleukin-2 receptor genes. Am. J. Respir. Cell Mol. Biol. 5, 292-296. Geist, L. J., Monick, M. M., Stinski, M. F., and Hunninghake, G. W. (1993). Immediate early genes of cytomegalovirus antagonize the inhibitory effects of cyclosporin A on interleukin-2 gene transcription. Chest 103, 1335-134s. Geist, L. J., Monick, M. M., Stinski, M. F., and Hunninghake, G. W. (1994). The immediate early genes of human cytomegalovirus upregulate tumor necrosis factor-alpha gene expression. J. Clin. Invest. 93, 474-478. Gerna, G., Zipeto, D., Percivalle, E., Parea, M., Revello, M. G., Maccario, R., Peri, G., and

HUMAN CYTOMEGALOVIRUS

245

Milanesi, G. (1992a). Human cytomegalovirus infection of the major leukocyte subpopulations and evidence for initial viral replication in polymorphonuclear leukocytes from viremic patients. J.Infect. Dis. 166, 1236-1244. Gerna, G., Revello, M. G., Percivalle, E., and Morini, F. (199213).Comparison of different immunostaining techniques and monoclonal antibodies to the lower matrix phosphoprotein (pp65) for optimal quantitation of human cytomegalovirus antigenemia. J . Clin. Microbiol. 30, 1232-1237. Ghazal, P., Young, J., Giulietti, E., DeMattei, C., Garcia, J., Gaynor, R., Stenberg, R. M., and Nelson, J . A. (1991). A discrete cis element in the human immunodeficiency virus long terminal repeat mediates synergistic trans activation by cytomegalovirus immediate-early proteins. J. Virol. 65, 6735-6742. Gibson, W. (1981). Immediate-early proteins of human cytomegalovirus strains AD 169, Davis, and Towne d i r e r in electrophoretic mobility. Virology 112, 350-354. Gibson, W. (1993). Molecular biology of human cytomegalovirus. In “Molecular Aspects of Human Cytomegalovirus Disease” (Y. Becker, G. Darai, and E.-S. Huang, eds.), pp. 303-329. Springer-Verlag, Berlin. Gibson, W., and Irmiere, A. (1984). Selection of particles and proteins for use as human cytomegalovirus subunit vaccines. Birth Defects 20, 305-324. Gibson, W., Murphy, T. L., and Fbby, C. (1981). Cytomegalovirus-infected cells contain a DNA-binding protein. Virology 111, 251-262. Gilbert, G. L., Hayes, K., Hudson, I. L., and James, J . (1989). Prevention of transfusionacquired cytomegalovirus infection in infants by blood filtration to remove leucocytes. Neonatal Cytomegalovirus Infection Study Group. Lancet 1, 1228-1231. Gilbert, M. J., Riddell, S. R., Li, C. R., and Greenberg, P. D. (1993). Selective interference with class I major histocompatibility complex presentation of the major immediateearly protein following infection with human cytomegalovirus. J. Virol. 67, 34613469. Gnann, J. W. J.,Ahlmen, J., Svalander, C., Olding. L., Oldstone, M. B., and Nelson. J. A. (1988). Inflammatory cells i n transplanted kidneys are infected by human cytomegalovirus. Am. J. Puthol. 132, 239-248. Gonczol, E., Andrews, P. W., and Plotkin, S. A. (1984). Cytomegalovirus replicates in differentiated but not in undifferentiated human embryonal carcinoma cells. Science 224, 159-161. Gonczol, E., Danczig, E., Boldogh, I., Toth, T., and Vaczi, L. (1985). In vivo model for the acute, latent and reactivated phases of cytomegalovirus infection. Actu Microbiol. Hung. 32, 39-47. Goodman, M. D., and Porter, D. D. (1973). Cytomegalovirus vasculitis with fatal colonic hemorrhage. Arch. Puthol. !#6, 281-284. Grafe, M. R., and Wiley, C. A. (1989). Spinal cord and peripheral nerve pathology in AIDS: The roles of cytomegalovirus and human immunodeficiency virus. Ann. Neurol. 25, 561-566. Grefte, J. M., van-der-Gun, B. T., Schmolke, S., van-der-Giessen, M., van-Son, W. J., Plachter, B., Jahn, G., and The, T. H. (1992a). The lower matrix protein pp65 is the principal viral antigen present in peripheral blood leukocytes during an active cytomegalovirus infection. J . Gen. Viral. 73, 2923-2932. Grefte, J. M., van-der-Gun, B. T., Schmolke, S., van-der-Giessen, M., van-Son, W. J., Plachter, B., Jahn, G., and The, T. H. (1992b). Cytomegalovirus antigenemia assay: Identification of the viral antigen as the lower matrix protein pp65. J.Infect. Dis. 166, 683-684. Grefte, A., Blom, N., van der Giessen, M., van Son, W., and The, T. H. (1993a). Ultrastructural analysis of circulating cytomegalic cells in patients with active cyto-

246

BOD0 PLACHTER et al.

megalovirus infection: Evidence for virus production and endothelial origin. J . Infect. Dis. 168, 1110-1118. Grefte, A., van-der-Giessen, M., van-Son, W., and The, T. H. (1993b). Circulating cytomegalovirus (CMV)-infected endothelial cells in patients with an active CMV infection. J. Infect. Dis. 167, 270-277. Grefte, J. M. M., Harmsen, M. C., van-der-Giessen, M., Knollema, S., van-Son, W. J., and The, T. H. (1994). Presence of human cytomegalovirus (HCMV) immediate early mRNA but not ppUL83 (lower matrix protein) mRNA in polymorphonuclear and mononuclear leukocytes during active HCMV infection. J. Gen. Virol. 75,1989-1998. Grundy, J. E. (1990). Virologic and pathogenetic aspects of cytomegalovirus infection. Rev. Infect. Dis. 12(Suppl. 7), S711LS719. Grundy, J. E., and Downes, K. L. (1993). Up-regulation of LFA-3 and ICAM-1 on the surface of fibroblasts infected with cytomegalovirus. Immunology 78, 405-412. Grundy, J. E., Shanley, J. D., and Griffiths, P. D. (1987). Is cytomegalovirus interstitial pneumonitis in transplant recipients an immunopathological condition? Lancet 2, 996-999. Hagemeier, C., Walker, S., Caswell, R., Kouzarides, T., and Sinclair, J. (1992a). The human cytomegalovirus 80-kilodalton but not the 72-kilodalton immediate-early protein transactivates heterologous promotors in an TATA box-independent mechanism and interacts directly with TFIID. J. Virol. 66, 4452-4456. Hagemeier, C., Walker, S. M., Sissons, P. J., and Sinclair, J. H. (199213).The 72K IE1 and 80K IE2 proteins of human cytomegalovirus independently trans-activate the c-fos, c-myc and hsp70 promoters via basal promoter elements. J. Gen. Virol. 73, 23852393. Hagemeier, C., Caswell, R., Hayhurst, G., Sinclair, J., and Kouzarides, T. (1994). Functional interaction of the HCMV IE2 transactivator and the retinoblastoma protein. EMBO J. 13,2897-2903. Hamzeh, F. M., Lietman, P. S., Gibson, W., and Hayward, G. S. (1990). Identification of the lytic origin of DNA replication in human cytomegalovirus by a novel approach utilizing ganciclovir-induced chain termination. J . Virol. 64,6184-6195. Handsfield, H. H., Chandler, S. H., Caine, V. A., Meyers, J. D., Corey, L., Medeiros, E., and McDougall, J. K. (1985). Cytomegalovirus infection in sex partners: Evidence for sexual transmission. J. Infect. Dis. 151, 344-348. Haymaker, W., Girdany, B. R., Stephens, J., Lillie, R. D., and Fetterman, G. H. (1954). Cerebral involvement with advanced periventircular calcification in generalized cytomegalic inclusion disease in the newborn. A clinicopathological report of a case diagnosed during life. J. Neuropathol. Exp. Neurol. 13, 562-586. Hayward, J. C., Titelbaum, D. S., Clancy, R. R., and Zimmerman, R. A. (1991). Lissencephaly-pachygyria associated with congenital cytomegalovirus infection. J. Child Neurol. 6, 109-114. Heieren, M. H., van-der-Woude, F. J., and Balfour, H. H. J. (1988a). Cytomegalovirus replicates efficiently in human kidney mesangial cells. Proc. Natl. Acad. Sci. U S A . 85,1642-1646. Heieren, M. H., Kim, Y. K., and Balfour, H. H. J. (1988b). Human cytomegalovirus infection of kidney glomerular visceral epithelial and tubular epithelial cells in culture. Transplantation 46,426-432. Heilbronn, R., Jahn, G., Burkle, A., Freese, U. K., Fleckenstein, B., and zur-Hausen, H. (1987). Genomic localization, sequence analysis, and transcription of the putative human cytomegalovirus DNA polymerase gene. J. Virol. 61, 119-124. Helweg-Larsen, S., Jakobsen, J., Boesen, F., and Arlien-Soborg, P. (1986). Neurological complications and concomitants of AIDS. Acta Neurol. Scand. 74,467-474.

HUMAN CYTOMEGALOVIRUS

247

Hendrix, M. G., Salimans, M. M., van-Boven, C. P., and Bruggeman, C. A. (1990). High prevalence of latently present cytomegalovirus in arterial walls of patients suffering from grade I11 atherosclerosis. Am. J. Pathol. 136, 23-28. Hendrix, M. G., Daemen, M., and Bruggeman, C. A. (1991).Cytomegalovirus nucleic acid distribution within the human vascular tree. Am. J. Pathol. 138, 563-567. Henry, S. C., and Hamilton, J. D. (1993). Detection of murine cytomegalovirus immediate early 1 transcripts in the spleens of latently infected mice. J. Infect.Dis. 167,950954. Hermiston, T. W., Malone, C. L., Witte, P. R., and Stinski, M. F. (1987).Identification and characterization of the human cytomegalovirus immediate-early region 2 gene that stimulates gene expression from an inducible promoter. J. Virol. 61, 3214-3221. Hermiston, T. W., Malone, C. L., and Stinski, M. F. (1990). Human cytomegalovirus immediate-early two protein region involved in negative regulation of the major immediate-early promoter. J . Virol. 64, 3532-3536. Ho, M. (1990). Epidemiology of cytomegalovirus infections. Rev. Infect. Dis. 12(Suppl. 7), S701-S710. Ho, M. (1991). “Cytomegalovirus Biology and Infection.” Plenum Medical, New York. Honess, R. W., and Roizman, B. (1974). Regulation of herpesvirus macromolecular synthesis. I. Cascade regulation of the synthesis of three groups of viral proteins. J. Virol. 14, 8-19. Honess, R. W., Gompels, U. A,, Barrell, B. G., Craxton, M., Cameron, K. R., Staden, R., Chang, Y. N., and Hayward, G. S. (1989). Deviations from expected frequencies of CpG dinucleotides in herpesvirus DNAs may be diagnostic of differences in the states of their latent genomes. J . Gen. Virol. 70, 837-855. Horn, M., Schlote, W., Herrmann, G., Gungor, T., and Jacobi, G. (1992). Immunocytochemical characterization of cytomegalovirus (CMV) infected giant cells in perinatal acquired human immunodeficiency virus (HIV) infection. Acta Histochem. Suppl. 42, 115-122. Horwitz, C. A., Henle, W., and Henle, G. (1979). Diagnostic aspects of the cytomegalovirus mononucleosis syndrome in previously healthy persons. Postgrad. Med. 66, 153-158. Hosenpud, J. D., Chou, S. W., and Wagner, C. R. (1991).Cytomegalovirus-induced regulation of major histocompatibility complex class I antigen expression in human aortic smooth muscle cells. Transplantation 52, 896-903. Huang, E. S., and Kowalik, T. F. (1993). The pathogenicity of human cytomegalovirus: An overview. In “Molecular Aspects of Human Cytomegalovirus Disease” (Y.Becker, G. Darai, and E.3. Huang, eds.), pp. 3-45. Springer-Verlag, Berlin. Huang, E. S., Alford, C. A., Reynolds, D. W., Stagno, S., and Pass, R. F. (1980). Molecular epidemiology of cytomegalovirus infections in women and their infants. N. Engl. J. Med. 303,958-962. Hudson, J. B. (1979). The murine cytomegalovirus as a model for the study of viral pathogenesis and persistent infections. Arch. Virol. 62, 1-29. Hudson, J. B., Chantler, J. K., Loh, L., Misra, V., and Muller, M. T. (1979).Model systems for analysis of latent cytomegalovirus infections. Can. J . Microbiol. 25, 245-253. Humbert, M., Devergne, O., Cerrina, J.,Rain, B., Simonneau, G., Dartevelle, P., Duroux, P., Galanaud, P., and Emilie, D. (1992). Activation of macrophages and cytotoxic cells during cytomegalovirus pneumonia complicating lung transplantations. Am. Rev. Respir. Dis. 145, 1178-1184. Hunninghake, G. W., Monick, M. M., Liu, B., and Stinski, M. F. (1989). The promoterregulatory region of the major immediate-early gene of human cytomegalovirus re-

248

BOD0 PLACHTER et al.

sponds to T-lymphocyte stimulation and contains functional cyclic AMP-response elements. J. Virol. 63, 3026-3033. Ibanez, C. E., Schrier, R., Ghazal, P., Wiley, C., and Nelson, J. A. (1991). Human cytomegalovirus productively infects primary differentiated macrophages. J. Virol. 65, 6581-6588. Irmiere, A., and Gibson, W. (1983). Isolation and characterization of a noninfectious virion-like particle released from cells infected with human strains of cytomegalovirus. Virology 130, 118-133. Iwamoto, G. K., Monick, M. M., Clark, B. D., Auron, P. E., Stinski, M. F., and Hunninghake, G. W. (1990). Modulation of interleukin 1 beta gene expression by the immediate early genes of human cytomegalovirus. J. Clin. Invest. 85, 1853-1857. Jacobson, M. A,, and Mills, J. (1988a). Serious cytomegalovirus disease in the acquired immunodeficiency syndrome (AIDS). Clinical findings, diagnosis, and treatment. Ann. Intern. Med. 108,585-594. Jacobson, M. A., and Mills, J. (1988b). Cytomegalovirus infection. Clin. Chest Med. 9, 443-448. Jahn, G., and Mach, M. (1990). Human cytomegalovirus phosphoproteins and glycoproteins and their coding regions. Curr. Top. Microbiol. Immunol. 154,171-185. Jahn, G., Knust, E., Schmolla, H., Sarre, T., Nelson, J. A., McDougall, J . K., and Fleckenstein, B. (1984). Predominant immediate-early transcripts of human cytomegalovirus AD 169. J. Virol. 49, 363-370. Jenkins, D. E., Martens, C. L., and Mocarski, E. S. (1994). Human cytomegalovirus late protein encoded by ie2: A trans-activator as well as a repressor of gene expression. J. Gen. Virol. 75, 2337-2348. Jensen, 0. A., and Klinken, L. (1989). Pathology of brain and eye in the acquired immune deficiency syndrome (AIDS). A comparison of lesions in a consecutive autopsy material. APMZS 97, 325-333. Jesionek, A., and Kiolemenoglou, B. (1904). Uber einen Befund von protozoen-artigen Gebilden in den Organen eines hereditar-luetischen Foetus. Muench. Med. Wochenschr. 15, 1905-1907. Jordan, M. C. (1983). Latent infection and the elusive cytomegalovirus. Reu. Infect. Dis. 6, 205-215. Jordan, M. C., and Mar, V. L. (1982). Spontaneous activation of latent cytomegalovirus from murine spleen explants. Role of lymphocytes and macrophages in release and replication of virus. J. Clin. Inuest. 70, 762-768. Jordan, M. C., Shanley, J. D., and Stevens, J . G. (1978). Immunosuppression reactivates and disseminates latent murine cytomegalovirus. IARC. Sci. Publ., 769-774. Jordan, M. C., Takagi, J. L., and Stevens, J . G. (1982). Activation of latent murine cytomegalovirus in uiuo and in uitro: A pathogenetic role for acute infection. J.Infect. Dis. 145,699-705. Jordan, M. C., Jordan, G. W., Stevens, J. G., and Miller, G. (1984). Latent herpesviruses of humans. Ann. Intern. Med. 100,866-880. Jupp, R., Hoffmann, S., Depto, A., Stenberg, R. M., Ghazal, P., and Nelson, J . A. (1993a). Direct interaction of the human cytomegalovirus IE86 protein with the cis repression signal does not preclude TBP from binding to the TATA box. J. Virol. 67,5595-5604. Jupp, R., Hoffmann, S., Stenberg, R. M., Nelson, J. A., and Ghazal, P. (1993b). Human cytomegalovirus IE86 protein interacts with promoter-bound TATA-binding protein via a specific region distinct from the autorepression domain. J. Virol.67,7539-7546. Keay, S., Merigan, T. C., and Rasmussen, L. (1989). Identification of cell surface receptors for the 86-kilodalton glycoprotein of human cytomegalovirus. Proc. Natl. Acad. Sci. U.S.A.86, 10100-10103.

HUMAN CYTOMEGALOVIRUS

249

Keil, G. M., Ebeling-Keil, A,, and Koszinowski, U. H. (1984). Temporal regulation of murine cytomegalovirus transcription and mapping of viral RNA synthesized at immediate early times after infection. J. Virol. 50, 784-795. Keil, G. M., Fibi, M. R., and Koszinowski, U. H. (1985). Characterization of the major immediate-early polypeptides encoded by murine cytomegalovirus. J. Virol. 54,422428. Keil, G. M., Ebeling-Keil, A., and Koszinowski, U. H. (1987a). Immediate-early genes of murine cytomegalovirus: Location, transcripts, and translation products. J. Virol. 61, 526-533. Keil, G. M., Ebeling-Keil, A., and Koszinowski, U. H. (1987b). Sequence and structural organization of murine cytomegalovirus immediate-early gene 1.J. Virol. 61, 19011908. Kennedy, P. G., Newsome, D. A., Hess, J., Narayan, O., Suresch, D. L., Green, W. R., Gallo, R. C., and Polk, B. F. (1986). Cytomegalovirus but not human T lymphotropic virus type III/lymphadenopathy associated virus detected by in situ hybridisation in retinal lesions in patients with the acquired immune deficiency syndrome. Br. Med. J. Clin. Res. 293, 162-164. Klages, S., Riiger, B., and Jahn, G. (1989). Multiplicity dependent expression of the predominant phosphoprotein pp65 of human cytomegalovirus. Virus Res. 12, 159-168. Klatt, E. C., and Shibata, D. (1988). Cytomegalovirus infection in the acquired immunodeficiency syndrome. Clinical and autopsy findings. Arch. Pathol. Lab. Med. 112, 540-544. Klemola, E., Von-Essen, R., Henle, G., and Henle, W. (1970). Infectious-mononucleosislike disease with negative heterophil agglutination test. Clinical features in relation to Epstein-Barr virus and cytomegalovirus antibodies. J. Infect. Dis. 121, 608-614. Kline, J. N., Geist, L. J., Monick, M. M., Stinski, M. F., and Hunninghake, G. W. (1994). Regulation of expression of the IL-1 receptor antagonist (IL-lra) gene by products of the human cytomegalovirus immediate early genes. J. Immunol. 152, 2351-2357. Klucher, K. M., Sommer, M., Kadonaga, J. T., and Spector, D. H. (1993). In vivo and in vitro analysis of transcriptional activation mediated by the human cytomegalovirus major immediate-early proteins. Mol. Cell. Biol. 13, 1238-1250. Knox, G. E., Pass, R. F., Reynolds, D. W., Stagno, S., and Alford, C. A. (1979). Comparative prevalence of subclinical cytomegalovirus and herpes simplex virus infections in the genital and urinary tracts of low-income, urban women. J. Infect. Dis. 140, 419422. Koskinen, P. K. (1993). The association of the induction of vascular cell adhesion molecule-1 with cytomegalovirus antigenemia in human heart allografts. Transplantation 56, 1103-1108. Koszinowski, U. H. (1991). Molecular aspects of immune recognition of cytomegalovirus. Transplant Proc. 23, 70-73. Koszinowski, U. H., Keil, G. M., Volkmer, H., Fibi, M. R., Ebeling-Keil, A., and Munch, K. (1986). The 89,000-Mr murine cytomegalovirus immediate-early protein activates gene transcription. J. Virol. 58, 59-66. Koszinowski, U. H., Del-Val, M., and Reddehase, M. J. (1990). Cellular and molecular basis of the protective immune response to cytomegalovirus infection. Curr. Top. Microbiol. Immunol. 154, 189-220. Koszinowski, U. H., Reddehase, M. J., and Del-Val, M. (1992). Principles of cytomegalovirus antigen presentation in vitro and in vivo. Semin. Immuaol. 4, 71-79. Kouzarides, T., Bankier, A. T., Satchwell, S. C., Weston, K., Tomlinson, P., and Barrell, B. G. (1987). Sequence and transcription analysis of the human cytomegalovirus DNA polymerase gene. J. Virol. 61, 125-133.

250

BOD0 PLACHTER et al.

Kowalik, T. F., Yurochko, A. D., Rinehart, C. A., Lee, C. Y., and Huang, E. S. (1994). Productive infection of human endometrial stromal cells by human cytomegalovirus. Virology 202, 247-257. Kumar, M. L., Nankervis, G. A., and Gold, E. (1973). Inapparent congenital cytomegalovirus infection. A follow-up study. N . Engl. J. Med. 288, 1370-1372. Lafemina, R. L., and Hayward, G. S. (1983). Replicative forms of human cytomegalovirus DNA with joined termini are found in permissively infected human cells but not in non-permissive Balb/c-3T3 mouse cells. J. Gen. Virol. 64, 373-389. Lafemina, R., and Hayward, G. S. (1986). Constitutive and retinoic acid-inducible expression of cytomegalovirus immediate-early genes in human teratocarcinoma cells. J . Virol. 58, 434-440. Lafemina, R. L., Pizzorno, M. C., Mosca, J. D., and Hayward, G. S. (1989). Expression of the acidic nuclear immediate-early protein (IE1) of human cytomegalovirus in stable cell lines and its preferential association with metaphase chromosomes. Virology 172, 584-600. Landini, M. P., Musiani, M., Zerbini, M., Falcieri, E., and La-Placa, M. (1979). Inhibition of a complete replication cycle of human cytomegalovirus in actinomycin pre-treated cells. J . Gen. Virol. 42, 423-428. Lang, D. J., and Kummer, J. F. (1975). Cytomegalovirus in semen: Observations in selected populations. J. Infect. Dis. 132, 472-473. Lang, D., and Stamminger, T. (1993). The 86-kilodalton IE-2 protein of human cytomegalovirus is a sequence-specific DNA-binding protein that interacts directly with the negative autoregulatory response element located near the cap site of the IE1/2 enhancer-promoter. J . Virol. 67, 323-331. Lang, D., Fickenscher, H., and Stamminger, T. (1992).Analysis of proteins binding to the proximal promoter region of the human cytomegalovirus IE-1/2 enhancer/promoter reveals both consensus and aberrant recognition sequences for transcription factors Spl and CREB. Nucleic Acids Res. 20,3287-3295. Lathey, J. L., and Spector, S. A. (1991). Unrestricted replication of human cytomegalovirus in hydrocortisone-treated macrophages. J. Virol. 65, 6371-6375. Leach, C. T., Detels, R., Hennessey, K., Liu, Z., Visscher, B. R., Dudley, J. P., and Cherry, J. D. (1994). A longitudinal study of cytomegalovirus infection in human immunodeficiency virus type 1-seropositive homosexual men: Molecular epidemiology and association with disease progression. J. Infect. Dis. 170, 293-298. Leatham, M. P., Witte, P. R., and Stinski, M. F. (1991). Alternate promoter selection within a human cytomegalovirus immediate-early and early transcription unit (UL119-115) defines true late transcripts containing open reading frames for putative viral glycoproteins. J . Virol. 65, 6144-6153. Lehner, R., Stamminger, T., and Mach, M. (1991). Comparative sequence analysis of human cytomegalovirus strains. J . Clin. Microbiol. 29, 2494-2502. Li, C. R., Greenberg, P. D., Gilbert, M. J., Goodrich, J. M., and Riddell, S. R. (1994). Recovery of HLA-restricted cytomegalovirus (CMV)-specificT-cell responses after allogeneic bone marrow transplant: Correlation with CMV disease and effect of ganciclovir prophylaxis. Blood 83, 1971-1979. Liebowitz, D., and Kieff, E. (1993). Epstein-Barr virus. In “The Human Herpesviruses” (B. Roizman, R. J. Whitley, and C. Lopez, eds.), pp. 107-172. Raven, New York. Liu, B., and Stinski, M. F. (1992). Human cytomegalovirus contains a tegument protein that enhances transcription from promoters with upstream ATF and AP-1 cis-acting elements. J. Virol. 66, 4434-4444. Liu, B., Hermiston, T. W., and Stinski, M. F. (1991). A cis-acting element in the major immediate-early (IE) promoter of human cytomegalovirus is required for negative regulation by IE2. J. Virol. 65, 897-903.

HUMAN CYTOMEGALOVIRUS

25 1

Ljungman, P., Biron, P., Bosi, A., Cahn, J. Y., Goldstone, A. H., Gorin, N. C., Link, H., Messina, C., Michallet, M., and Richard, C. (1994). Cytomegalovirus interstitial pneumonia in autologous bone marrow transplant recipients. Infectious Disease Working Party of the European Group for Bone Marrow Transplantation. Bone Marrow Transplant. 13, 209-212. Loh, L., and Hudson, J. B. (1980). Immunosuppressive effect of murine cytomegalovirus. Infect. Zmmun. 27, 54-60. Loh, L., and Hudson, J. B. (1982). Murine cytomegalovirus-induced immunosuppression. Infect. Immun. 36, 89-95. Lowenstein, C. (1907). Uber protozoenartige Gebilde in den Organen von Kindern. Zentralbl. Allg. Pathol. Pathol. Anat. 18, 513-518. Lubon, H., Ghazal, P., Henninghausen, L., Reynolds-Kohler, C., Lockshin, C., and Nelson, J. (1989). Cell-specific activity of the modulator region in the human cytomegalovirus major immediate-early gene. Mol. Cell. Biol. 9, 1342-1345. Luby, J. P., Burnett, W., Hull, A. R., Ware, A. J., Shorey, J. W., and Peters, P. C. (1974). Relationship between cytomegalovirus and hepatic function abnormalities in the period after renal transplant. J . Infect. Dis. 129, 511-518. Lukac, D. M., Manuppello, J . R., and Alwine, J. C. (1994). Transcriptional activation by the human cytomegalovirus immediate-early proteins: Requirements for simple promoter structures and interaction with multiple components or the transcription complex. J . Virol. 68, 5184-5193. Macasaet, F. F., Holley, K. E., Smith, T.F., and Keys, T. F. (1975). Cytomegalovirus studies of autopsy tissue. 11. Incidence of inclusion bodies and related pathologic data. A m . J . Clin. Pathol. 63, 859-865. McCarthy, M., Resnick, L., Taub, F., Stewart, R. V., and Dix, R. D. (1991). Infection of human neural cell aggregate cultures with a clinical isolate of cytomegalovirus. J. Neuropathol. Exp. Neurol. 50, 441-450. McCracken, G. H., Shinefield, H. R., Cobb, K., Rausen, A. R., Dische, M. R., and Eichenwald, H. F. (1969). Congenital cytomegalic inclusion disease. Am. J. Dis. Child 117, 522-539. McDonough, S. H., and Spector, D. H. (1983). Transcription in human fibroblasts permissively infected by human cytomegalovirus strain AD169. Virology 125, 31-46. Macias, M. P., and Stinski, M. F. (1993). An in uitro system for human cytomegalovirus immediate early 2 protein (IE2bmediated site-dependent repression of transcription and direct binding of IE2 to the major immediate early promoter. Proc. Natl. Acad. Sci. U S A . 90, 707-711. Maciejewski, J . P., Bruening, E. E., Donahue, R. E., Mocarski, E. S., Young, N. S., and StJeor, S. C. (1992). Infection of hematopoietic progenitor cells by human cytomegalovirus. Blood 80, 170-178. Malone, C. L., Vesole, D. H., and Stinski, M. F. (1990). Transactivation of a human cytomegalovirus early promoter by gene products from the immediate-early gene IE2 and augmentation by IE1: Mutational analysis of the viral proteins. J. Virol. 64, 1498-1506. Martignetti, J. A., and Barrell, B. G. (1991). Sequence of the Hind111 T fragment of human cytomegalovirus, which encodes a DNA helicase. J. Gen. Virol. 72,1113-1121. Matthiessen, L., Marche, C., Labrousse, F., Trophilme, D., Fontaine, C., and Vedrenne, C. (1992). Neuropathology of the brain in 174 patients who died of AIDS in a Paris hospital 1982-1988. Ann. Med. Znt. Paris 143,43-49. Mayo, D., Armstrong, J. A., and Ho, M. (1978). Activation of latent murine cytomegalovirus infection: Cocultivation, cell transfer, and the effect of immunosuppression. J . Infect. Dis. 138, 890-896. Mazeron, M. C., Jahn, G., and Plachter, B. (1992). Monoclonal antibody E-13 (McAb 810)

252

BOD0 PLACHTER et al.

to human cytomegalovirus recognizes an antigen encoded by exon 2 of the major immediate-early gene. J. Gene. Virol. 73,2699-2703. Mercer, J. A., Wiley, C. A., and Spector, D. H. (1988).Pathogenesis of murine cytomegalovirus infection: Identification of infected cells in the spleen during acute and latent infections. J . Virol. 62,987-997. Messerle, M., Keil, G. M., and Koszinowski, U. H. (1991).Structure and expression of murine cytomegalovirus immediate-early gene 2. J. Virol. 65,1638-1643. Messerle, M., Buhler, B., Keil, G. M., and Koszinowski, U. H. (1992).Structural organization, expression, and functional characterization of the murine cytomegalovirus immediate-early gene 3. J. Virol. 66,27-36. Meyer, H., Sundqvist, V. A,, Pereira, L., and Mach, M. (1992).Glycoprotein gp116 of human cytomegalovirus contains epitopes for strain-common and strain-specific antibodies. J. Gen. Virol. 73,2375-2383. Miceli, M. V., Newsome, D. A., Novak, L. C., and Beuerman, R. W. (1989).Cytomegalovirus replication in cultured human retinal pigment epithelial cells. Curr. Eye Res. 8,835-839. Michelson-Fiske, S., Horodniceanu, F., and Guillon, J.4. (1977).Immediate-early antigens in human cytomegalovirus infected cells. Nature fLondon) 270,615-617. Michelson, S., Horodniceanu, F., Kress, M., and Tardy-Panit, M. (1979). Human cytomegalovirus-induced immediate early antigens: Analysis in sodium dodecyl sulfate-polyacrylamide gel electrophoresis after immunoprecipitation. J. Virol. 32, 259-267. Millett, R., Tomita, T., Marshall, H. E., Cohen, L., and Hannah, H. (1991).Cytomegalovirus endomyocarditis in a transplanted heart. A case report with in situ hybridization. Arch. Pathol. Lab. Med. 115, 511-515. Minton, E.J.,Qsoe, C., Sinclair, J.H., and Sissons, J. G. (1994).Human cytomegalovirus infection of the monocyte/macrophage lineage in bone marrow. J. Virol. 68, 40174021. Mocarski, E. S.(1993).Cytomegalovirus biology and replication. Zn “The Human Herpesviruses” (B. Raizman, R. J. Whitley, and C. Lopez, eds.), pp. 173-226. Raven, New York. Mocarski, E. S., and Stinski, M. F. (1979).Persistence of the cytomegalovirus genome in human cells. J. Virol. 31, 761-775. Mocarski, E.S.,Pereira, L., and Michael, N. (1985).Precise localization of genes in large animal virus genomes: Use of lambda gtll and monoclonal antibodies to map the gene for a cytomegalovirus protein family. Proc. Natl. Acad. Sci. U S A . 82, 12661270. Mocarski, E. S. J., Abenes, G. B., Manning, W. C., Sambucetti, L. C., and Cherrington, J. M. (1990).Molecular genetic analysis of cytomegalovirus gene regulation in growth, persistence and latency. Curr. Top. Microbiol. Zmmunol. 154, 47-74. Mocarski, E. S.,Bonyhadi, M., Salimi, S., McCune, J. M., and Kaneshima, H. (1993). Human cytomegalovirus in a SCID-hu mouse: Thymic epithelial cells are prominent targets of viral replication. Proc. Natl. Acad. Sci. U.S.A.90, 104-108. Monick, M. M., Geist, L. J., Stinski, M. F., and Hunninghake, G. W. (1992).The immediate early genes of human cytomegalovirus upregulate expression of the cellular genes myc and fos. Am. J. Respir. Cell Mol. Biol. 7,251-256. Montgomery, R., Youngblood, L., and Medearis, D. N. J. (1972).Recovery of cytomegalovirus from the cervix in pregnancy. Pediatrics 49,524-531. Morgello, S.,Cho, E. S., Nielsen, S., Devinsky, O., and Petito, C. K. (1987).Cytomegalovirus encephalitis in patients with acquired immunodeficiency syndrome: An autopsy study of 30 cases and a review of the literature. Hum. Pathol. 18, 289-297.

HUMAN CYTOMEGALOVIRUS

253

Muhlemann, K., Miller, R. K., Metlay, L., and Menegus, M. A. (1992). Cytomegalovirus infection of the human placenta: An immunocytochemical study. Hum. Pathol. 23, 1234-1237. Muller, M. T., and Hudson, J. B. (1977). Cell cycle dependency of murine cytomegalovirus replication in synchronized 3T3 cells. J.Virol. 22, 267-272. Myerson, D., Hackman, R. C., Nelson, J. A., Ward, D. C., and McDougall, J. K. (1984). Widespread presence of histologically occult cytomegalovirus. Hum. Pathol. 15,430439. Nabeshima, K., Sakaguchi, E., Inoue, S., Eizuru, Y., Minamishima, Y., and Koono, M. (1992). Jejunal perforation associated with cytomegalovirus infection in a patient with adult T-cell leukemia-lymphoma. Acta Pathol. Jpn. 42, 267-271. Nelson, J. A., Fleckenstein, B., Jahn, G., Galloway, D. A., and McDougall, J. K. (1984). Structure of the transforming region of human cytomegalovirus AD169. J. Virol. 49, 109-115. Nelson, J. A,, Reynolds-Kohler, C., and Smith, B. A. (1987). Negative and positive regulation by a short segment in the 5'-flanking region of the human cytomegalovirus major immediate-early gene. Mol. Cell. Biol. 7, 4125-4129. Nelson, J. A., Reynolds-Kohler, C., Oldstone, M. B., and Wiley, C. A. (1988). HIV and HCMV coinfect brain cells in patients with AIDS. Virology 165, 286-290. Nelson, J. A., Gnann, J. W. J., and Ghazal, P. (1990). Regulation and tissue-specific expression of human cytomegalovirus. Cum. Top. Microbiol. Zmmunol. 154,75-100. Nowlin, D. M., Cooper, N. R., and Compton, T. (1991). Expression of a human cytomegalovirus receptor correlates with infectibility of cells. J. Virol. 65, 3114-3121. Ogura, T., Tanaka, J., Kamiya, S., Sato, H., Ogura, H., and Hatano, M. (1986). Human cytomegalovirus persistent infection in a human central nervous system cell line: Production of a variant virus with different growth characteristics. J. Gen. Virol. 67, 2605-2616. Olding, L. B., Jensen, F. C., and Oldstone, M. B. (1975).Pathogenesis of cytomegalovirus infection. I. Activation of virus from bone marrow-derived lymphocytes by in uitro allogenic reaction. J . Exp. Med. 141, 561-572. Pantoni, L., Inzitari, D., Colao, M. G., De-Mayo, E., Marini, P., and Mazzota, F. (1991). Cytomegalovirus encephalitis in a non-immunocompromised patient: CSF diagnosis by in situ hybridization cells. Acta Neurol. Scand. 84, 56-58. Pari, G. S., and Anders, D. G. (1993). Eleven loci encoding trans-acting factors are required for transient complementation of human cytomegalovirus oriLyt-dependent DNA replication. J. Virol. 67, 6979-6988. Pari, G. S., Kacica, M. A., and Anders, D. G. (1993). Open reading frames ULU, IRSUTRSl, and UL36-38 are required for transient complementation of human cytomegalovirus oriLyt-dependent DNA synthesis. J.Virol. 67, 2575-2582. Pass, R. F., Stagno, S., Myers, G. J., and Alford, C. A. (1980). Outcome of symptomatic congenital cytomegalovirus infection: Results of long-term longitudinal follow-up. Pediatrics 66, 758-762. Paya, C. V., Virelizier, J. L., and Michelson, S. (1991). Modulation of T-cell activation through protein kinase C- o r A-dependent signalling pathways synergistically increases human immunodeficiency virus long terminal repeat induction by cytomegalovirus immediate-early proteins. J. Virol. 65, 5477-5484. Perlman, J. M., and Argyle, C. (1992). Lethal cytomegalovirus infection in preterm infants: Clinical, radiological, and neuropathological findings. Ann. Neurol. 31,64-68. Perot, K., Walker, C. M., and Spaete, R. R. (1992). Primary chimpanzee skin fibroblast cells are fully permissive for human cytomegalovirus replication. J. Gen. Virol. 73, 3281-3284.

254

BOD0 PLACHTER et al.

Pisano, G. (1910). Su di un reperto istologico raro in feto scleromatoso. Gazz. Osp. Clin. 31,249-260. Pizzorno, M. C., and Hayward, G. S. (1990). The IE2 gene products of human cytomegalovirus specifically down-regulate expression from the major immediate-early promoter through a target sequence located near the cap site. J. Virol. 64,6154-6165. Pizzorno, M. C., O’Hare, P., Sha, L., Lafemina, R. L., and Hayward, G. S. (1988). Transactivation and autoregulation of gene expression by the immediate-early region 2 gene products of human cytomegalovirus. J. Virol. 62, 1167-1179. Pizzorno, M. C., Mullen, M. A., Chang, Y. N., and Hayward, G. S. (1991). The functionally active IE2 immediate-early regulatory protein of human cytomegalovirus is an 80-kilodalton polypeptide that contains two distinct activator domains and a duplicated nuclear localization signal. J . Virol. 65,3839-3852. Plachter, B., Traupe, B., Albrecht, J., and Jahn, G (1988). Abundant 5 kb RNA of human cytomegalovirus without a major translational reading frame. J. Gen. Virol. 69, 2251-2266. Plachter, B., Klages, S., Hagelmann, S., Britt, W., Landini, M. P., and Jahn, G. (1990). Procaryotic expression of phosphorylated tegument protein pp65 of human cytomegalovirus and application of recombinant peptides for immunoblot analyses. J. Clin. Microbiol. 28, 1229-1235. Plachter, B., Britt, W.J., Vornhagen, R., Stamminger, T., and Jahn, G. (1993).Analysis of proteins encoded by IE-regions 1 and 2 of human cytomegalovirus using monoclonal antibndies generated against recombinant antigens Virology 193, 642-652. Poland, S. D., Costello, P., Dekaban, G. A., and Rice, G. P. (1990). Cytomegalovirus in the brain: In vitro infection of human brain-derived cells. J. Infect. Dis. 162, 1252-1262. Pomeroy, C., Hilleren, P. J., and Jordan, M. C. (1991). Latent murine cytomegalovirus DNA in splenic stromal cells of mice. J. Virol. 65, 3330-3334. Porter, K. R., Starnes, D. M., and Hamilton, J. D. (1985). Reactivation of latent murine cytomegalovirus from kidney. Kidney Int. 28,922-925. Pritchett, R. F. (1980). DNA nucleotide sequence heterogeneity between the Towne and AD169 strains of cytomegalovirus. J. Virol. 36, 152-161. Puchtler, E., and Stamminger, T. (1991). An inducible promotor mediates abundant expression from the immediate-early 2 gene region of human cytomegalovirus at late times after infection. J. Virol. 66, 6301-6306. Pulliam, L. (1991). Cytomegalovirus preferentially infects a monocyte derived macrophage/microglial cell in human brain cultures: Neuropathology differs between strains. J. Neuropathol. Exp. Neurol. 50, 432-440. Rawlinson, W. D., and Barrell, B. G. (1993). Spliced transcripts of human cytomegalovirus. J. Virol. 67, 5502-5513. Reddehase, M. J., Balthesen, M., Rapp, M., Jonjic, S., Pavic, I., and Koszinowski, U. H. (1994). The conditions of primary infection define the load of latent viral genome in organs and the risk of recurrent cytomegalovirus disease. J.Exp. Med. 179,185-193. Reiser, H., Kuhn, J., Doerr, H. W., Kirchner, H., Munk, K., and Braun, R. (1986). Human cytomegalovirus replicates in primary human bone marrow cells. J. Gen. Virol. 67, 2595-2604. Revello, M. G., Percivalle, E., Di-Matteo, A., Morini, F., and Gerna, G. (1992). Nuclear expression of the lower matrix protein of human cytomegalovirus in peripheral blood leukocytes of immunocompromised viraemic patients. J. Gen. Virol. 73, 437-442. Reynolds, D. W., Stagno, S., Hosty, T. S., Tiller, M., and Alford, C. A. J. (1973).Maternal cytomegalovirus excretion and perinatal infection. N. Engl. J. Med. 289, 1-5. Ribbert, H. (1904). iiber protozoenartige Zellen in der Niere eines syphilitischen Neugeborenen und in der Parotis von Kindern. Zentralbl. Allg. Pathol. Pathol. Anat. 15, 945-948.

HUMAN CYTOMEGALOVIRUS

255

Rice, G. P., Schrier, R. D., and Oldstone, M. B. (1984). Cytomegalovirus infects human lymphocytes and monocytes: Virus expression is restricted to immediate-early gene products. Proc. Natl. Acad. Sci. U.S.A.81,6134-6138. Rixon, F. J. (1993).Structure and assembly of herpes viruses. Semin. Virol. 4, 135-144. Roberts, W. H., Sneddon, J. M., Waldman, J., and Stephens, R. E. (1989).Cytomegalovirus infection of gastrointestinal endothelium demonstrated by simultaneous nucleic acid hybridization and immunohistochemistry. Arch. Pathol. Lab. Med. 113, 461-464. Roby, C., and Gibson, W. (1986).Characterization of phosphoproteins and protein kinase activity of virions, noninfectious enveloped particles, and dense bodies of human cytomegalovirus. J. Virol. 59, 714-727. Roizman, B.,and Sears, A. E. (1990).Herpes simplex viruses and their replication. In “Virology” (B. N. Fields and D. M. Knipe, eds.), 2nd Ed., pp. 1795-1841.Raven, New York. Roizman, B., and Sears, A. E. (1993).Herpes simplex viruses and their replication. In “The Human Herpesviruses” (B. Roizman, R. J., Whitley, and C. Lopez, eds.), pp. 1168.Raven, New York. Rosenthal, L. J., Panitz, P. J., Crutchfield, D. B., and Chou, J . Y. (1981).Cytomegalovirus replication in primary and passaged human placental cells. Intervirology 16,168-175. Rowe, W. P., Hartley, J. W., and Waterman, S., Turner, H. C., and Huebner, R. J. (1956). Cytopathogenic agent resembling human salivary gland virus recovered from tissue cultures of human adenoids. Proc. SOC.Exp. Biol. Med. 92,418-424. Rummelt, V., Rummelt, C., Jahn, G., Wenkel, H., Sinzger, C., Mayer, U. M., and Naumann, G. 0. (1994).“riple retinal infection with human immunodeficiency virus type 1, cytomegalovirus, and herpes simplex virus type 1. Light and electron microscopy, immunohistochemistry, and in situ hybridization. Ophthalmology 101, 270-279. Sambucetti, L. C., Cherrington, J. M., Wilkinson, G. W., and Mocarski, E. S. (1989). NFkappa B activation of the cytomegalovirus enhancer is mediated by a viral transactivator and by T cell stimulation. EMBO J. 8, 4251-4258. Schickedanz, J., Philipson, L., Ansorge, W., Pepperkok, R., Klein, R., and Koszinowski, U. H. (1988).The 89,000-Mrmurine cytomegalovirus immediate-early protein stimulates c-fos expression and cellular DNA synthesis. J. Virol. 62, 3341-3347. Schirm, J., Timmerije, W., van-der-Bij, W., The, T. H., Wilterdink, J. B., Tegzess, A. M., van-Son, W. J., and Schroder, F. P. (1987).Rapid detection of infectious cytomegalovirus in blood with the aid of monoclonal antibodies. J. Med. Virol. 23,31-40. Schmader, K. E., Rahija, R. J., Porter, K. R., and Hamilton, J . D. (1991).Murine cytomegalovirus gene amplification and culture after submaxillary salivary gland biopsy. Lab. Anim. Sci. 41, 396-400. Schmidbauer, M.,Budka, H., Ulrich, W., and Ambros, P. (1989).Cytomegalovirus (CMV) disease of the brain in AIDS and connatal infection: A comparative study by histology, immunocytochemistry and in situ DNA hybridization. Acta Neuroputhol. 79, 286-293. Schmolke, S., Drescher, P., Jahn, G., and Plachter, B. (1995).Nuclear targeting of the tegument protein pp65 (UL83) of human cytomegalovirus An unusual bipartite nuclear localization signal functions together with other port ons of the protein to mediate its efficient nuclear transport. J . Virol. 69, 1071-1078. Schrier, R. D.,Nelson, J. A., and Oldstone, M. B. (1985).Detection of human cytomegalovirus in peripheral blood lymphocytes in a natural infection. Science 230, 1048- 1051. Schwartz, D. A,, Khan, R., and Stoll, B. (1992).Characterization of the fetal inflammatory response to cytomegalovirus placentitis. An immunohistochemical study. Arch. Pathol. Lab. Med. 116, 21-27.

256

BOD0 PLACHTER et al.

Schwartz, R., Sommer, M. H., Scully, A., and Spector, D. H. (1994). Site-specific binding of the human cytomegalovirus IE2 86-kilodalton protein to an early gene promoter. J. Virol. 68, 5613-5622. Scott, D. M., Rodgers, B. C., Freeke, C., Buiter, J., and Sissons, J. G. (1989). Human cytomegalovirus and monocytes: Limited infection and negligible immunosuppression in normal mononuclear cells infected in uitro with mycoplasma-free virus strains. J. Gen. Virol. 70, 685-694. Shanley, J. D., Jordan, M. C., Cook, M. L., and Stevens, J. G. (1979). Pathogenesis of reactivated latent murine cytomegalovirus infection. Am. J. Pathol. 95, 67-80. Shelbourn, S. L., Kothari, S. K., Sissons, J. G., and Sinclair, J. H. (1989). Repression of human cytomegalovirus gene expression associated with a novel immediate early regulatory region binding factor. Nucleic Acids Res. 17, 9165-9171. Shintaku, M., Inoue, N., Sasaki, M., Izuno, Y., Ueda, Y., and Ikehara, S. (1991). Cytomegalovirus vasculitis accompanied by an exuberant fibroblastic reaction in the intestine of an AIDS patient. Acta Pathol. Jpn. 41, 900-904. Simmons, P., Kaushansky, K., and Torok-Storb, B. (1990). Mechanisms of cytomegalovirus-mediated myelosuppression: Perturbation of stromal cell function versus direct infection of myeloid cells. Proc. Natl. Acad. Sci. U.S.A.87, 1386-1390. Sinclair, J. H., Baillie, J., Bryant, L. A., Taylor-Wiedeman, J. A., and Sissons, J. G. (1992). Repression of human cytomegalovirus major immediate early gene expression in a monocytic cell line. J. Gen. Virol. 73,433-435. Sinzger, C . , Plachter, B., Stenglein, S., and Jahn, G. (1993a). Immunohistochemical detection of viral antigens in smooth muscle, stromal, and epithelial cells from acute cytomegalovirus gastritis. J.Infect. Dis. 167, 1427-1432. Sinzger, C., Muntefering, H., Loning, T., Stoss, H., Plachter, B., and Jahn, G. (1993b). Cell types infected in human cytomegalovirus placentitis identified by immunohistochemical double staining. Virchows Arch. A Pathol. Anat. Histopathol. 423, 249-256. Sinzger, C., Grefte, A., Plachter, B., Gouw, A. S. H., The, T. H., and Jahn, G. (1995a). Fibroblasts, epithelial cells, endothelial cells and smooth muscle cells are major targets of human cytomeglovirus infection in uiuo. J . Gen Virol. 76, 741-750. Sinzger, C., Plachter, B., Grefte, A., The, T. H., and Jahn, G. (1995b). Tissue macrophages are infected by the human cytomegalovirus in uiuo. J . Infect. Dis., in press. Skolnik, P. R., Pomerantz, R. J., de-la-Monte, S. M., Lee, S. F., Hsiung, G. D., Foos, R. Y., Cowan, G. M., Kosloff, B. R., Hirsch, M. S., and Pepose, J. S. (1989). Dual infection of retina with human immunodeficiency virus type 1 and cytomegalovirus. Am. J . Ophthulmol. 107,361-372. Smith, J. D. (1986). Human cytomegalovirus: Demonstration of permissive epithelial cells and nonpermissive fibroblastic cells in a survey of human cell lines. J. Virol. 60, 583-588. Smith, J. D., and de-Harven, E. (1973). Herpes simplex virus and human cytomegalovirus replication in WI-38 cells. I. Sequence of viral replication. J . Virol. 12, 919-930. Smith, K. L., Cobain, T., and Dunstan R. A. (1993a). Removal of cytomegalovirus DNA from donor blood by filtration. Br. J . Huematol. 83, 640-642. Smith, K. L., Kulski, J. K., Cobain, T., and Dunstan, R. A. (1993b). Detection of cytomegalovirus in blood donors by the polymerase chain reaction. Transfusion 33,497503. Smith, M. (1956). Propagation in tissue cultures of a cytopathogenic virus from human salivary gland virus (SGV) disease. Proc. SOC.Exp. Biol. Med. 92, 424-430. Snider, W. D., Simpson, D. M., Nielsen, S., Gold, J. W., Metroka, C. E., and Posner, J. B. (1983). Neurological complications of acquired immune deficiency syndrome: Analysis of 50 patients. Ann. Neurol. 14, 403-418.

HUMAN CYTOMEGALOVIRUS

257

Soderberg, C., Larsson, S., Bergstedt-Lindqvist, S., and Moller, E. (1993). Definition of a subset of human peripheral blood mononuclear cells that are permissive to human cytomegalovirus infection. J. Virol. 67, 3166-3175. Somogyi, T., Michelson, S., and Masse, M. J. (1990). Genomic location of a human cytomegalovirus protein with protein kinase activity (PK68). Virology 174, 276-285. Sopko, J., and Anuras, S. (1978). Liver disease in renal transplant recipients. Am. J. Med. 64, 139-146. Spaete, R. R., and Mocarski, E. S. (1985). Regulation of cytomegalovirus gene expression: Alpha and beta promoters are trans activated by viral functions in permissive human fibroblasts. J. Virol. 56, 135-143. Span, A. H., van-Boven, C. P., and Bruggeman, C. A. (1989). The effect of cytomegalovirus infection on the adherence of polymorphonuclear leucocytes to endothelial cells. Eur. J. Clin. Invest. 19, 542-548. Span, A. H., Mullers, W., Miltenburg, A. M., and Bruggeman, C. A. (1991). Cytomegalovirus induced PMN adherence in relation to a n ELAM-1 antigen present on infected endothelial cell monolayers. Immunology 72, 355-360. Spector, D. J., and Tevethia, M. J. (1994). Protein-protein interactions between human cytomegalovirus IE2-580aa and pUL84 in lytically infected cells. J. Virol. 68, 75497553. Spector, D. H., Klucher, K. M., Rabert, D. K., and Wright, D. A. (1990). Human cytomegalovirus early gene expression. Curr. Top. Microbiol. Immunol. 154, 21-45. Spector, S. A., Hirata, K. K., and Newman, T. R. (1984). Identification of multiple cytomegalovirus strains in homosexual men with acquired immunodeficiency syndrome. J. Infect. Dis. 150, 953-956. Speir, E., Modali, R., Huang, E. S., Leon, M. B., Shawl, F., Finkel, T., and Epstein, S. E. (1994). Potential role of human cytomegalovirus and p53 interaction in coronary restenosis. Science 265, 391-394. Stagno, S., and Cloud, G. A. (1994). Working parents: The impact of day care and breastfeeding on cytomegalovirus infections in offspring. Proc. Natl. Acad. Sci. U.S.A.91, 2384-2389. Stagno, S., Reynolds, D., Tsiantos, A,, Fuccillo, D. A., Smith, R., Tiller, M., and Alford, C. A. J. (1975). Cervical cytomegalovirus excretion in pregnant and nonpregnant women: Suppression in early gestation. J. Infect. Dis. 131, 522-527. Stagno, S., Reynolds, D. W., Pass, R. F., and Alford, C. A. (1980). Breast milk and the risk of cytomegalovirus infection. N . Engl. J. Med. 302, 1073-1076. Stamminger, T., and Fleckenstein, B. (1990). Immediate-early transcription regulation of human cytomegalovirus. Curr. Top. Microbiol. Immunol. 154, 3-19. Stamminger, T., Fickenscher, H., and Fleckenstein, B. (1990). Cell type-specific induction of the major immediate early enhancer of human cytomegalovirus by cyclic AMP. J. Gen. Virol. 71, 105-113. Stanier, P., Kitchen, A. D., Taylor, D. L., and e m s , A. S. (1992). Detection of human cytomegalovirus in peripheral mononuclear cells and urine samples using PCR. Mol. Cell Probes 6, 51-58. Stasiak, P. C., and Mocarski, E. S. (1992). Transactivation of the cytomegalovirus ICP36 gene promoter requires the alpha gene product TRSl in addition to IE1 and IE2. J. Virol. 66, 1050-1058. Stenberg, R. M., and Stinski, M. F. (1985). Autoregulation of the human cytomegalovirus major immediate-early gene. J. Virol. 56, 676-682. Stenberg, R. M., Thomsen, D. R., and Stinski, M. F. (1984). Structural analysis of the major immediate early gene of human cytomegalovirus. J. Virol. 49, 190-199. Stenberg, R. M., Witte, P. R., and Stinski, M. F. (1985). Multiple spliced and unspliced transcripts from human cytomegalovirus immediate-early region 2 and evidence for a common initiation site within immediate-early region 1. J. Virol. 56, 665-675.

258

BOD0 PLACHTER et al.

Stenberg, R. M., Depto, A. S., Fortney, J., and Nelson, J. A. (1989).Regulated expression of early and late RNAs and proteins from the human cytomegalovirus immediateearly gene region. J . Virol. 63, 2699-2708. Stenberg, R. M., Fortney, J., Barlow, S. W., Magrane, B. P., Nelson, J. A., and Ghazal, P. (1990). Promoter-specific trans activation and repression by human cytomegalovirus immediate-early proteins involves common and unique protein domains. J. Virol. 64, 1556-1565. Sterner, G., Agell, B. O., Wahren, B., and Espmark, A. (1970).Acquired cytomegalovirus infection in older children and adults. A clinical study of hospitalized patients. Scand. J . Infect. Dis. 2, 95-103. Stinski, M. F. (1978). Sequence of protein synthesis in cells infected by human cytomegalovirus: Early and late virus-induced polypeptides. J. Virol. 26, 686-701. Stinski, M. F. (1990). Cytomegalovirus and its replication. In “Virology” (B. N. Fields and D. M. Knipe, eds.), pp. 1959-1980. Raven, New York. Stinski, M. F. (1991). Molecular biology of cytomegalovirus replication. In “Cytomegalovirus Biology and Infection,” pp. 7-36. Plenum Medical, New York. Stinski, M. F., and Roehr, T. J. (1985). Activation of the major immediate early gene of human cytomegalovirus by cis-acting elements in the promoter-regulatory sequence and by virus-specific trans-acting components. J. Virol. 55, 431-441. Stinski, M. F., Thomsen, D. R., Stenberg, R. M., and Goldstein, L. C. (1983). Organization and expression of the immediate early genes of human cytomegalovirus. J. Virol. 46, 1-14. Stinski, M. F., Malone, C. L., Hermiston, T. W., and Liu, B. (1991). Regulation of cytomegalovirus transcription. In “Herpesvirus Transcription and Its Control” (E. K. Wagner, ed.), pp. 245-260. CRC Press, Boca Raton, Florida. Straws, M. (1990). Human cytomegalovirus labyrinthitis. Am. J. Otolaryngol. 11,292298. Studahl, M., Ricksten, A., Sandberg, T., Elowson, S., Herner, S., Sall, C., and Bergstrom, T. (1994). Cytomegalovirus infection of the CNS in non-immunocompromized patients. Actu Neurol. Scand. 29, 451-457. Sullivan, V., Talarico, C. L., Stanat, S. C., Davis, M., Coen, D. M., and Biron, K. K., (1992). A protein kinase homologue controls phosphorylation of ganciclovir in human cytomegalovirus-infectedcells. Nature (London) 358, 162-164. Sullivan, V., Biron, K. K., Talarico, C., Stanat, S. C., Davis, M., Pozzi, L. M., and Coen, D. M. (1993).A point mutation in the human cytomegalovirus DNA polymerase gene confers resistance to ganciclovir and phosphonylmethoxyalkyl derivatives. Antimicrob. Agents Chemother. 37, 19-25. Tanaka, J., Ogura, T., Sato, H., and Hatano, M. (1987). Establishment and biological characterization of an in uitro human cytomegalovirus latency model. Virology 161, 62-72. Tanaka, J., Sadanari, H., Sato, H., and Fukuda, S. (1991). Sodium butyrate-inducible replication of human cytomegalovirus in a human epithelial cell line. Virology 165, 271-280. Tanaka, S., Furukawa, T., and Plotkin, S. A. (1975).Human cytomegalovirus stimulates host cell RNA synthesis. J. Virol. 16, 297-304. Taylor, H. P., and Cooper, N. R. (1989). Human cytomegalovirus binding to fibroblasts is receptor mediated. J . Virol. 63, 3991-3998. Taylor, H. P., and Cooper, N. R. (1990). The human cytomegalovirus receptor on fibroblasts is a 30-kilodalton membrane protein. J . Virol. 64, 2484-2490. Taylor-Wiedeman, J., Sissons, J. G., Borysiewicz, L. K., and Sinclair, J. H. (1991). Monocytes are a major site of persistence of human cytomegalovirus in peripheral blood mononuclear cells. J.Gen. Virol. 72, 2059-2064.

HUMAN CYTOMEGALOVIRUS Taylor-Wiedeman, J., Hayhurst, G. P., Sissons, J. G. P., and Sinclair, J. H. (1993). Polymorphonuclear cells are not sites of persistence of human cytomegalovirus in healthy individuals. J. Gen. Virol. 74, 265-268. Taylor-Wiedeman, J. A., Sissons, J. G. P., and Sinclair, J. H. (1994). Induction of endogenous human cytomegalovirus gene expression after differentiation of monocytes from healthy carriers. J. Virol. 68, 1597-1604. Tenney, D. J., and Colberg-Poley, A. M. (1990). RNA analysis and isolation of cDNAs derived from the human cytomegalovirus immediate-early region at 0.24 map units. Zntervirology 31, 203-214. Tenney, D. J., and Colberg-Poley, A. M. (1991a). Expression of the human cytomegalovirus UL36-38 immediate early region during permissive infection. Virology 182, 199-210. Tenney, D. J., and Colberg-Poley, A. M. (1991b). Human cytomegalovirus UL36-38 and US3 immediate-early genes: Temporally regulated expression of nuclear, cytoplasmic, and polysome-associated transcripts during infection. J. Virol. 65, 67246734. Tenney, D. J., Santomenna, L. D., Goudie, K. B., and Colberg-Poley, A. M. (1993). The human cytomegalovirus US3 immediate-early protein lacking the putative transmembrane domain regulates gene expression. Nucleic Acids Res. 21, 2931-2937. Tevethia, M. J., Spector, D. J., Leisure, K. M., and Stinski, M. F. (1987). Participation of two human cytomegalovirus immediate early gene regions in transcriptional activation of adenovirus promoters. Virology 161, 276-285. The, T. H., van-der-Bij, W., van-den-Berg, A. P., van-der-Giessen, M., Weits, J., Sprenger, H. G., and van-Son, W. J. (1990). Cytomegalovirus antigenemia. Rev. Infect. Dis. 12(Suppl. 7), s734-s744. Theise, N. D., Conn, M., and Thung, S. N. (1993). Localization of cytomegalovirus antigens in liver allografts over time. Hum. Pathol. 24, 103-108. Thomsen, D. R., Stenberg, R. M., Goins, W. F., and Stinski, M. F. (1984). Promoterregulatory region of the major immediate early gene of human cytomegalovirus. Proc. Natl. Acad. Sci. U.S.A.81, 659-663. Tocci, M. J., and St-Jeor, S. C. (1979). Persistence and replication of the human cytomegalovirus genome in lymphoblastoid cells of B and T origin. Virology 96, 664668. Toorkey, C. B., and Carrigan, D. R. (1989). Immunohistochemical detection of a n immediate early antigen of human cytomegalovirus in normal tissues. J. Infect. Dis. 160, 741-751. Tooze, J., Hollinshead, M., Reis, B., Radsak, K., and Kern, H. (1993). Progeny vaccinia and human cytomegalovirus particles utilize early endosomal cisternae for their envelopes. Eur. J. Cell Biol. 60, 163-178. Topilko, A., and Michelson, S. (1994). Hyperimmediate entry of human cytomegalovirus virions and dense bodies into human fibroblasts. Res. Virol. 145, 75-82. Tumilowicz, J. J. (1990). Characteristics of human arterial smooth muscle cell cultures infected with cytomegalovirus. In Vitro Cell Deu. Biol. 26, 1144-1150. Q l e r , K. L., Gross, R. A,, and Cascino, G. D. (1986). Unusual viral causes of transverse myelitis: Hepatitis A virus and cytomegalovirus. Neurology 36, 855-858. Urban, M., Britt, W., and Mach, M. (1992). The dominant linear neutralizing antibodybinding site of glycoprotein gp86 of human cytomegalovirus in strain specific. J. Virol. 66, 1303-1311. Vinters, H. V., Kwok, M. K., Ho, H. W., Anders, K. H., Tomiyasu, U., Wolfson, W. L., and Robert, F. (1989). Cytomegalovirus in the nervous system of patients with the acquired immune deficiency syndrome. Brain 112, 245-268. Waldman, W. J., Sneddon, J. M., Stephens, R. E., and Roberts, W. H. (1989). Enhanced

260

BOD0 PLACHTER et al.

endothelial cytopathogenicity induced by a cytomegalovirus strain propagated in endothelial cells. J.Med. Virol. 28, 223-230. Waldman, W. J., Roberts, W. H., Davis, D. H., Williams, M. V., Sedmak, D. D., and Stephens, R. E. (1991). Preservation of natural endothelial cytopathogenicity of cytomegalovirus by propagation in endothelial cells. Arch. Virol. 117, 143-164. Waldman, J. W., Knight, D. A., and Sedmak, D. D. (1993). Monocyte-mediated transmission of infectious cytomegalovirus among human endothelial monolayers: An in uitro model for dissemination. I n "Multidisciplinary Approach to Understanding Cytomegalovirus Disease" (S. Michelson and S. A. Plotkin, eds.), pp. 91-94. Excerpta Medica, Amsterdam. Walker, S., Hagemeier, C., Sissons, J. G., and Sinclair, J. H. (1992). A 10-base-pair element of the human immunodeficiency virus type 1long terminal repeat (LTR) is an absolute requirement for transactivation by the human cytomegalovirus 72kilodalton I E l protein but can be compensated for by other LTR regions in transactivation by the 80-kilodalton IE2 protein. J. Virol. 66, 1543-1550. Waner, J. L., and Weller, T. H. (1974). Behavior of human cytomegalovirus in cell cultures of bovine and simian origin. Proc. Soc. Exp. Biol. Med. 145, 379-384. Ware, A. J., Luby, J. P., Hollinger, B., Eigenbrodt, E. H., Cuthbert, J. A., Atkins, C. R., Shorey, J., Hull, A. R., and Combes, B. (1979). Etiology of liver disease in renaltransplant patients. Ann. Intern. Med. 91, 364-371. Warren, A. P., Ducroq, D. H., Lehner, P. J., and Borysiewicz, L. K. (1994). Human cytomegalovirus-infected cells have unstable assembly of major histocompatibility complex class I complexes and are resistant to lysis by cytotoxic T lymphocytes. J. Virol. 68, 2822-2829. Wathen, M. W., and Stinski, M. F. (1982). Temporal patterns of human cytomegalovirus transcription: Mapping the viral RNAs synthesized at immediate early, early, and late times after infection. J. Virol. 41, 462-477. Wathen, M. W., Thomsen, D. R., and Stinski, M. F. (1981).Temporal regulation of human cytomegalovirus transcription at immediate early and early times after infection. J. Virol. 38, 446-459. Weber, T., Beck, R., Stark, E., Gerhards, J., Korn, K., Haas, J., Liier, W., and Jahn, G. (1994). Comparative analysis of intrathecal antibody synthesis of DNA amplification for the diagnosis of cytomegalovirus infection of the central nervous system in AIDS patients. J.Neurol. 241, 407-414. Weinshenker, B. G., Wilton, S., and Rice, G. P. (1988). Phorbol ester-induced differentiation permits productive human cytomegalovirus infection in a monocytic cell line. J. Immunol. 140,1625-1631. Weller, T. H., and Henshaw, J. B. (1962). Virologic and clinical observations on cytomegalic inclusion disease. N.Engl. J. Med. 266, 1233-1244. Weller, T. H., Macauley, J. C., Craig, J. M., and Wirth, P. (1957). Isolation of intranuclear inclusion producing agents from infants with illnesses resembling cytomegalic inclusion disease. Proc. Soc. Exp. Biol. Med.94, 4-12. Weston, K. (1988). An enhancer element in the short unique region of human cytomegalovirus regulates the production of a group of abundant immediate early transcripts. Virology 162, 406-416. Wiley, C. A., and Nelson, J. A. (1988). Role of human immunodeficiency virus and cytomegalovirus in AIDS encephalitis. Am. J. Puthol. 133, 73-81. Wilkinson, G. W., Akrigg, A., and Greenaway, P. J. (1984). Transcription of the immediate early genes of human cytomegalovirus strain AD169. Virus Res. 1, 101-106. Wilson, E. J., Medearis, D. N. J., Barrett, L. V., and Rubin, R. H. (1985). Activation of latent murine cytomegalovirus in cardiac explant and cell cultures. J.Infect. Dis. 152, 625-626.

HUMAN CYTOMEGALOVIRUS

261

Winkler, M., Rice, S. A., and Stramminger, T.(1994). UL69 of human cytomegalovirus, adn open reading frame with homology to ICP27 of herpes simplex virus, encodes a transactivator of gene expression. J . Virol. 68, 3943-3954. Wise, T. G., Manischewitz, J . E., Quinnan, G. V.,Aulakh, G. S., and Ennis, F. A. (1979). Latent cytomegalovirus infection of BALB/c mouse spleens detected by a n explant culture technique. J . Gen. Virol. 44,551-556. Wolff, D., Jahn, G., and Plachter, B. (1990). Generation and effective enrichment of selectable human cytomegalovirus mutants using site-directed insertion of the neo gene. Gene 130, 167-173. Wolff, D., Sinzger, C., Drescher, P., Jahn, G., and Plachter, B. (1994). Latent infection of the glioblastoma cell line U138-MG with selectable recombinants of human cytomegalovirus leads to reduced IE2 expression and to shut-down of early and late viral genes. Virology 204, 101-113. Wright, J. F., Kurosky, A., and Wasi, S. (1994). An endothelial cell-surface form of annexin I1 binds human cytomegalovirus. Biochem. Biophys. Res. Cornmun. 198,983989. Wu, J., Jupp, R., Stenberg, R. M., Nelson, J . A., and Ghazal, P. (1993). Site-specific inhibition of RNA polymerase I1 preinitiation complex assembly by human cytomegalovirus IE86 protein. J. Virol. 67, 7547-7555. Wu, T. C., Hruban, R. H., Ambinder, R. F., Pizzorno, M., Cameron, D. E., Baumgartner, W. A., Reitz, B. A., Hayward, G. S., and Hutchins, G. M. (1992). Demonstration of cytomegalovirus nucleic acids in the coronary arteries of transplanted hearts. A m . J. Pathol. 140,739-747. Yee, R. W., Sigler, S. C., Lawton, A. W., Alderson, G. L., Trinkle, J. K., and Lum, C. T. (1991). Apparent cytomegalovirus epithelial keratitis in a cardiac transplant recipient. Transplantation 51, 1040-1043. Younes, M., Podesta, A., Helie, M., and Buckley, P. (1991). Infection of T but not B lymphocytes by cytomegalovirus in lymph node. An immunophenotypic study. A m . J. Surg. Pathol. 15, 75-80.

This Page Intentionally Left Blank

ADVANCES IN VIRUS RESEARCH, VOL. 46

VARICELLA-ZOSTER VIRUS: ASPECTS OF PATHOGENESIS AND HOST RESPONSE TO NATURAL INFECTION AND VARICELLA VACCINE Ann M. Arvin, Jennifer F. Moffat, and Rebecca Redman Stanford University School of Medicine Stanford, California 94305

I. Introduction 11. Varicella-Zoster Virus 111. Cell-Associated Viremia in the Pathogenesis of Varicella-Zoster Virus Infection A. Viremia during Primary VZV Infection B. Viremia during VZV Reactivation C. VZV Infection of Peripheral Blood Cells in Vitro D. VZV Tropism for T Lymphocytes in the Severe Combined Immunodeficient hu Mouse IV. Cell-Mediated Immune Response to Varicella-Zoster Virus A. Methods for Assessing Cell-Mediated Immunity to VZV B. Primary Cell-Mediated Immune Response to VZV C. Memory T-Lymphocyte Responses to VZV D. Mechanisms for Maintaining Cell-Mediated Immunity to VZV E. Cell-Mediated Immunity to VZV in the Guinea Pig Model V. Summary References

I. INTRODUCTION Varicella-zoster virus (VZV) is a human herpesvirus that causes varicella (chickenpox) in susceptible individuals, establishes latent infection in neural ganglion cells, and reactivates as herpes zoster (shingles) (Arvin, 1995; Whitley, 1990). The virus is maintained in the human population because episodes of herpes zoster in adults provide continued opportunities for infection of susceptible children. Most individuals who live in the United States or other areas with temperate climates acquire primary VZV infection during childhood as a result of varicella epidemics that occur annually. These epidemics are facilitated by the fact that, in contrast to the other human herpesviruses, VZV is transmissible in aerosolized respiratory secretions from individuals with varicella. Primary VZV infection is usually uncomplicated in children, but it can be severe in healthy adults. Varicella and herpes zoster may be life-threatening in immunocompromised pa263

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

264

ANN M. ARVIN et al.

tients, including those being treated for malignancy, transplant recipients, and patients with inherited immunologic disorders or acquired immunodeficiencies, especially human immunodeficiency virus (HIV) infection. Herpes zoster also causes significant morbidity in otherwise healthy elderly individuals. The medical significance of VZV-related disease provides the rationale for development of antiviral drugs that inhibit its replication during primary or recurrent infection (Whitley, 1990). Prevention of primary VZV infection is an important goal because varicella is lifethreatening in some patients despite antiviral therapy. Varicellazoster virus is the first of the human herpesviruses for which a vaccine has been licensed (Gershon, 1992). The live attenuated varicella vaccine is prepared from VZV strain Oka, which was isolated from the cutaneous lesion of a Japanese child with varicella and attenuated by passage in guinea pig embryo fibroblasts. The varicella vaccine is now recommended for universal administration to children and to susceptible adults in the United States (American Academy of Pediatrics, 1995). The development of effective strategies for the prevention and treatment of VZV infections requires an understanding of important events in pathogenesis and of the essential components of the host response to this pathogen, which infects almost all of the human population. The capacity of the virus to cause cell-associated viremia is fundamental to the pathogenesis of primary and recurrent VZV infections, and the ability of the host t o prevent or terminate the infection of peripheral blood cells is critical in reducing the risk of VZV-related morbidity or mortality.

11. VARICELLA-ZOSTER VIRUS Varicella-zoster virus is a member of the Alpha herpesvirus subgroup of the genus Herpesvirus (Cohen and Straus, 1995). Like other herpesviruses, the VZV virion consists of a linear, double-stranded DNA genome, a nucleocapsid surrounding the DNA-containing core, and a proteinaceous tegument between the capsid and the lipid envelope; the virion envelope incorporates the major viral glycoproteins, which are also expressed on the surface of VZV-infected cells. The virus is the smallest of the human herpesviruses, but it has open reading frames (ORFs) that permit the synthesis of at least 69 gene products (Davison and Scott, 1986). The DNA of VZV consists of approximately 125,000 base pairs; like herpes simplex type 1 (HSV-11, which is the prototype of this subgroup, the genome exhibits a linear arrangement of long unique and short unique segments with terminal repeat regions.

VARICELLA-ZOSTER VIRUS

265

Putative functions for some VZV proteins have been deduced from homologies with HSV-1 on the basis of similarities between the linear sequence of ORFs, and complementation has been demonstrated for some genes. Nevertheless, there are several obvious differences between the HSV and VZV genomes that may eventually explain their distinctive biological characteristics. For example, the product of ORF62 is the major immediate early protein of VZV that initiates replication and constitutes a major tegument protein. This gene product has functions resembling those of the HSV a-TIF, but there is no direct VZV homolog for a-TIF (Kinchington et al., 1992). Other VZV proteins that have regulatory and/or structural functions include the products of ORFs 4, 10, 61, and 63. The ORF29 product is the major DNA-binding protein, which is made only during replication. The virus has ORFs that encode at least six glycoproteins, designated gp I-gp V, but it lacks a homolog for gD, which is an essential protein of HSV (Grose, 1990). The terminology for the VZV glycoproteins is now revised to be consistent with HSV nomenclature: gE (gp I), gB (gp 111, gH (gp III), gI (gp IV), gC (gp V), and gL (Davison et al., 1986; Cohen and Straus, 1995). Like HSV, VZV produces a viral thymidine kinase, which makes the virus susceptible to inhibition by acyclovir, but this gene is not essential for replication in tissue culture. Varicella-zoster virus isolates do not exhibit any virologic or antigenic characteristics that suggest major strain differences. Restriction endonuclease profiles can be used to document the epidemiological relatedness of VZV isolates and to prove that recurrent disease episodes in a single patient are caused by the same virus strain (Straus et al., 1984). The DNA of the Oka strain, which is used t o manufacture the varicella vaccine, shows some differences in restriction endonuclease patterns when compared to VZV isolates from the United States, but it is similar to isolates that are more closely related in geographic origin. No specific genetic markers have been identified to demonstrate a molecular basis for the clinical attenuation of the Oka strain.

111. CELL-ASSOCIATED VIREMIAIN THE PATHOGENESIS OF VARICELLAZOSTERVIRUSINFECTION A . Viremia during Primary VZV Infection The capacity of VZV to cause viremia is fundamental to the pathogenesis of primary infection, as is evident clinically from the appearance of many discrete, cutaneous vesicles during the course of the illness. This widespread exanthem distinguishes varicella from the

266

ANN M. ARVIN et al.

localized mucocutaneous lesions typical of primary HSV infections (Arvin, 1995; Whitley, 1990). Primary VZV infection is usually initiated by direct inoculation of respiratory mucosal sites with virus from a close contact who has varicella or herpes zoster. Inoculation is followed by a 10 to 21-day incubation period during which the virus is presumed to move to regional lymph nodes and then t o other reticuloendothelial sites during a primary viremic phase. This initial viremia has not been proved by virologic methods, but cell-associated viremia can be documented late in the incubation period and during the first few days after the appearance of the cutaneous rash. The secondary viremia permits the dissemination of infectious virus to cutaneous epithelial cells and coincides with the appearance of activated T lymphocytes in the circulation (Arvin et al., 1986b). Unless virus spread is terminated by the host response, VZV is transported in infected peripheral blood cells to lungs and other internal organs, resulting in progressive life-threatening infection. Despite the essential role of viremia in VZV pathogenesis, detecting the virus in peripheral blood cells and identifying the subpopulations of cells required to sustain viremia is technically difficult. The virus can be isolated from peripheral blood mononuclear cells (PBMC) using standard tissue culture methods, but viremia is detectable in only 1124% of healthy children with acute varicella despite prolonged cocultivation of PBMC with permissive cell lines (Asano et al., 1985; Ozaki et al., 1986). On the basis of in situ hybridization using a VZV cDNA probe, VZV gene sequences are present in PBMC from 75% of subjects when PBMC are obtained within 48 hours after the onset of the varicella rash. (Fig. 1)(Koropchak et al., 1989). The frequency of positive cells ranges from 0.01 to 0.001% of circulating PBMC, which is consistent with the need to culture at least 2 x 105 PBMC t o recover the virus in tissue culture (Arvin et al., 198613). Viremia is also detected using polymerase chain reaction (PCR) methods to test PBMC from healthy subjects with varicella (Fig. 2) (Koropchak et al., 1991; Sawyer, et al., 1992). In our experiments, the virus was present in PBMC from 67% of subjects who were tested at intervals of 2.5 to 20 hr after the onset of varicella, using oligonucleotide sequences complementary t o regions of VZV ORF31 (gB). The PBMC that harbor VZV gene seFIG.1. Demonstration of VZV gene sequences in PBMC during primary VZV infection. PBMC obtained from subjects 24 hours after the onset of acute varicella (A,B) and from a guinea pig 5 days after inoculation with 9 x 105 pfu of VZV, (C) are shown after hybridization with the 3H-labeled VZV DNA probe ( x 63 magnification);PBMC from one subject with varicella are shown after hybridization with the pBR322 vector DNA (D). Reprinted with permission from Koropchak et al. (1989).

VARICELLA-ZOSTER VIRUS

267

268

ANN M. ARVIN et al.

FIG. 2. Detection of VZV in PBMC from healthy subjects with varicella. The left section of the figure illustrates the dot hybridization results when clinical samples were tested after VZV PCR amplification. Sample 1A is an oropharyngeal sample. Samples 1B-1E and 3D represent VZV PCR testing of PBMC from VZV immune subjects who were tested at least 20 years after primary VZV infection. The results of VZV PCR testing of single or multiple aliquots of PBMC (2 x 105 cells/sample) from four subjects with acute varicella are shown in 2A; 2B and 2C; 2D and 2E; and 3A-3C. The VZV DNA control is shown in 3E. As shown in the right section of this figure, no PCR product was detectable by dot hybridization using the 32P-labeled CK1 probe after amplification of HSV-2 (sample lG), CMV (sample lH), or EBV (sample 2F) DNA in the presence of the synthetic oligonucleotides TK4 and TK5 that corresponded to sequences coding for VZV gp I1 (gene 31). Sample 1F represents VZV infected cell DNA and sample 2G was amplified from EBV-transformed B-lymphocytes that were superinfected with VZV. Reprinted with permission from Koropchak et al. (1991).

quences are lymphocytes, judging from their morphology in in situ hybridization experiments. However, the fragility of PBMC from individuals with acute varicella causes extensive cell loss during fluorescence-activated cell sorting (FACS), interfering with experiments to determine whether the infected cells are T or B lymphocytes. Viremia is a transient event in the pathogenesis of primary VZV infection among immunocompetent individuals. It usually persists for less than 24-72 hours, judging from testing of PBMC by viral culture, in situ hybridization, or PCR methods. In contrast, immunocompromised patients with varicella have life-threatening complications as a result of failure to clear cell-associated viremia. Malignant progressive varicella is accompanied by persistence of the virus in the peripheral blood, continued formation of new cutaneous lesions, and dissemination of the virus to the lungs, liver, adrenal glands, and other internal organs (Myers, 1979). Prolonged viremia is demonstrated by viral culture of PBMC from these patients, and the number of circulating VZV-infected cells is probably much higher than in healthy subjects with varicella. The administration of antiviral drugs to immunocompromised children with varicella blocks this altered pathogenic process by terminating the cell-associated viremia and compensating for the diminished host response (Whitley, 1990).

VARICELLA-ZOSTER VIRUS

269

Early studies of VZV pathogenesis document that inoculating susceptible children with vesicular fluid from cutaneous VZV lesions, which contains high titers of cell-free virus, causes varicella (Kundratitz, 1925). Administration of the live attenuated Oka-Merck vaccine by subcutaneous inoculation causes no clinical signs of primary VZV infection in healthy children even when the vaccine preparation contains 9000 plaque-forming units (pfu) of infectious virus (Gershon, 1992). Cell-associated viremia is not detected following immunization of healthy children. However, whether the altered pathogenic potential of the Oka-Merck vaccine virus strain is due to genetic mutations that diminish its capacity to cause cell-associated viremia is not known. The attenuation of the vaccine strain is achieved using traditional methods, by passage of a clinical isolate in nonhuman cells in tissue culture. This process may have introduced stable mutations that reduce virulence. Alternatively, other factors, such as the route of inoculation or the initial host response, may limit replication of the vaccine virus and reduce its pathogenicity. In fact, the clinical experience with administering the varicella vaccine to children who have leukemia in remission suggests that the Oka strain retains virulence genes required to infect PBMC (Gershon et al., 1984). These children derive important benefit from vaccination because vaccine-induced immunity prevents the life-threatening complications of natural VZV infection. However, some of these moderately immunocompromised patients develop cutaneous lesions that are distant from the initial site of inoculation, indicating hematogenous spread of the virus. The recovery of Oka strain virus from these lesions demonstrates that the vaccine virus can infect circulating peripheral blood cells.

B . Viremia during VZV Reactivation The virus infects cells of the sensory ganglia, apparently without causing cell damage, during the course of primary VZV infection and establishes latency. By analogy with HSV, the virus is presumed to move along neural pathways from cutaneous sites of viral replication to the corresponding cervical, thoracic, or lumbosacral dorsal root ganglia; viremia may facilitate the spread of VZV to multiple ganglion sites. Viral nucleic acid sequences are detected in human ganglia many years after primary VZV infection by hybridization methods (Croen et al., 1988; Mahalingam et al., 1990). In contrast to HSV, VZV seems to persist in nonneuronal cells adjacent to neurons in the dorsal root ganglia, rather than within neurons, and many fewer cells harbor the virus (Croen and Straus, 1991). Latency of VZV is not associated

270

ANN M.ARVIN et al.

with the detection of antisense transcripts (LATs) characteristic of HSV; instead, several VZV genes appear to be actively transcribed during latency, including the IE62 gene and ORF29. There is no evidence that VZV persists in PBMC during latency. Symptomatic reactivation of VZV usually causes herpes zoster, which is characterized by acute pain and a unilateral vesicular rash within a single cutaneous dermatome. The dermatomal involvement reflects the sensory nerve distribution from one of the dorsal root ganglia, indicating that the virus reactivates in ganglion cells and migrates along axons t o infect epithelial cells. Most healthy individuals with acute herpes zoster do not have evidence of cell-associated viremia, which can be diagnosed clinically by the appearance of scattered lesions outside of the involved dermatome. However, the prolonged period of latency in ganglion sites probably does not affect the capacity of VZV to infect PBMC. Reactivated virus can cause subclinical viremia during acute herpes zoster in the healthy individual, judging from detection of VZV DNA in PBMC from some patients, and there is some evidence that the virus persists in circulating cells for several months in individuals who develop the syndrome of postherpetic neuralgia (Vafai et al., 1988). The immediate boosting of the host response induced by VZV reactivation probably prevents or limits infection of PBMC in most otherwise healthy individuals, but herpes zoster is associated with viremia in some immunocompromised patients. Reactivation of VZV in immunodeficient patients is often accompanied by scattered cutaneous lesions outside of the primary dermatome and, in some cases, by lifethreatening dissemination to visceral organs. Infectious virus can be detected in circulating PBMC from these patients by tissue culture methods (Feldman and Epp, 1976). Reactivation of HSV is also common among immunocompromised patients; however, the episodes are rarely life-threatening, whereas the capacity of VZV to cause cellassociated viremia significantly enhances its pathogenicity during recurrent disease. In some clinical circumstances, VZV reactivation occurs without causing the usual clinical symptoms of herpes zoster. Bone marrow transplant recipients are at high risk for VZV reactivation and dissemination. Viremia with transport of the virus to lungs, liver, and other organs can produce fatal infection without any cutaneous manifestations in these patients (Locksley et al., 1985). Our studies of bone marrow transplant recipients demonstrate that VZV reactivation remains asymptomatic in some patients, even though PBMC become infected with the virus (Table I). Transient episodes of cell-associated viremia were detected in 19% of patients who were tested by PCR at a

VARICELLA-ZOSTER VIRUS

271

TABLE I THERELATIONSHIPBETWEEN EPISODES OF SUBCLINICAL VARICELLA-ZOSTER VIREMIA, THE INTERVALAFTER TRANSPLANTATIONAND WHETHERTHE PATIENT RECEIVEDAN AUTOLOGOUS OR ALLOGENEIC TRANSPLANT" Type of transplant Interval after transplant

Autologous transplant

Allogeneic transplant

Total

17-99 days

2/15 (13%)

3/11 (27%)

5/26 (19%)

100-183 days

117 (14%)

119 (11%)

2/16 (12%)

3/22

4/20

Total

&printed with permission; Wilson et al. (1992)

mean interval of 94 days after transplant (Wilson et al., 1992). The viremia that occurs in immunodeficient patients who have symptomatic recurrent VZV infection could result from transfer of the virus to the PBMC that traffic through the site of local cutaneous replication. However, the occurrence of subclinical, cell-associated viremia in bone marrow transplant patients suggests that the virus also may be taken up directly by mononuclear cells at sites of viral reactivation in dorsal root ganglia, without requiring a phase of cutaneous replication. The susceptibility of bone marrow transplant patients to cell-associated VZV viremia may be enhanced by the persistence of activated lymphocytes in the circulation for a prolonged period after transplant (Forman et al., 1982). Unless latency is maintained much more efficiently for VZV than for any of the other herpesviruses, episodes of subclinical reactivation probably also occur in otherwise healthy individuals, but prolonged surveillance with repeated testing of PBMC samples would be necessary to document their occurrence.

C . VZV Infection of Peripheral Blood Cells in Vitro Although clinical observations prove the importance of VZV interactions with PBMC in the pathogenesis of primary and recurrent disease, the lymphotropism of VZV is difficult to investigate in uitro because high titers of cell-free infections VZV cannot be prepared for inoculation. In our experience, infection of PBMC can be achieved by

272

ANN M. ARVIN et al.

cocultivation of infected fibroblasts with human lymphocytes (Koropchak et al., 1989). Mitogen stimulation, which enhances lymphocyte infection by other viruses, allows the maintenance of T lymphocytes long enough to follow the cultures for the relatively long VZV replication cycle and reproduces the activation of T lymphocytes observed during primary VZV infection in uiuo. Under these conditions, in situ hybridization demonstrates the entry of the virus into T lymphocytes, with cytoplasmic grains visible in more than 75% of cells by 5 days after inoculation. Staining is most prominent just within the cytoplasmic membrane in most cells, although only 0.01% of T lymphocytes have more than 20 grains over the nucleus (Fig. 3). The breakdown of viral DNA in the cytoplasm by cellular DNases occurs by 10 days, when only 25-40% of the cells have cytoplasmic grains. Nevertheless, signal localized to the nucleus remains detectable in 0.01% of cells a t the late time point, and infectious VZV is recovered from phytohemagglutinin (PHA)-stimulated T lymphocytes at titers of 1.85.0 x 104 pfu/ml for more than 14 days. The observation that VZV DNA reaches the nucleus and replicates in a small percentage of T lymphocytes in vitro is consistent with the observation that VZV viremia in uiuo is due to infection of a only few lymphocytes. Replication of VZV is not detectable in mitogenstimulated B-lymphocyte cultures although it occurs in B lymphocytes transformed by Epstein-Barr virus (EBV). Cell membrane expression of VZV proteins is present in 60-75% of these cells by immunofluorescence using a fluorescein isothiocyanate (FITOconjugated monkey polyclonal antibody to VZV and with monoclonal antibodies to VZV gE (gp I) and the IE62 protein. The nonpermissiveness of most lymphocytes for VZV, despite the uptake of virion DNA, obviously benefits the host during primary infection, but the mechanism for this restriction is not known. The in situ hybridization data suggest that the limited replication in human lymphocytes is not caused by impaired viral entry. One possibility is that adsorbed VZV virions enter most lymphocytes by endocytosis using a pathway that does not involve fusion of the viral envelope and the plasma membrane of the cell and produces abortive infection (Gabel et al., 1989; Grose, 1990). FIG.3. Demonstration of VZV gene sequences in PBMC infected with VZV in vitro. B- and T-lymphocyte cultures were infected with an ultracentrifuged preparation of sonicated VZV-infected melanoma cells at 1.5 x 104 pfu/lO5 cells. After 2-4 hours, the cells were washed and resuspended in RPMI with 15%FCS, 10%11-2, and mitogen. In situ hybridization with the VZV DNA probe is shown for EBV-transformed B-lymphocytes at 3 days (A) and for PHA-stimulated T-lymphocytesat 5 days (B and C) after VZV infection in vitro. Reprinted with permission from Koropchak et al. (1989).

VARICELLA-ZOSTER VIRUS

273

274

A N N M.ARVIN et al.

D . VZV Tropism for T Lymphocytes in the Severe Combined Zmmunodeficient hu Mouse Efforts to develop animal models for studies of VZV pathogenesis are hampered significantly because other mammalian species are resistant to infection with the virus. Varicella-zoster virus causes no disease in strain 2 and Hartley guinea pigs, but a viremic phase occurs in some animals after subcutaneous inoculation with VZV adapted to replicate in guinea pig fibroblasts (Myers et al., 1985; Lowry et al., 1992a). The guinea pig model has limited usefulness for analyzing cell-associated viremia because recovery of infectious virus from PBMC after infection of the animals is sporadic, PBMC are positive by in situ hybridization only at 3-5 days after infection, and the frequency of VZV-infected PBMC is very low (0.001-0.002%). Our work demonstrates that severe combined immunodeficient (SCID) hu mice implanted with human fetal thymus/liver tissue constitute a useful model to investigate the tropism of VZV for human mononuclear cells (Moffat et al., 1995). Because of their immunodeficiency, SCID mice can be used to maintain grafts of various human tissues that differentiate to contain the usual human cell populations in their expected structural organization. Human tissue implants in SCID-hu mice permit studies of the replication of viruses that are highly species-specific, such as human immunodeficiency virus and cytomegalovirus (Namikawa et al., 1988, 1990). In our experiments, thymudliver implants are inoculated with a low-passage clinical isolate of VZV or the Oka strain, after two passages in human foreskin fibroblasts and human lung fibroblasts. Replication of both wild-type VZV and the Oka strain is demonstrated by infectious focus assay of dispersed cells from the thymus/liver implants. During a period of 221 days after inoculation, hematoxylin and eosin-stained sections of the implants reveal progressive cytopathology and extensive ingestion of lymphocytes that express VZV proteins by macrophages (Fig. 4). The lymphoid lobes of the thymus show areas of fibrosis, interstitial hemorrhage, and lymphocyte depletion within 7 days, and necrosis is pronounced by 14 days. In situ hybridization and immunohistochemical staining reveal extensive viral DNA and protein expression in thymic lymphocytes (Fig. 4). Zn situ hybridization signal and viral protein expression are distributed evenly throughout the lymphoid lobes of the implant, but little or no viral DNA or protein synthesis is detected in epithelial cells composing the thymic stroma or capsular structures. Infectious virus is not recovered from implants infected with wild-type VZV or Oka strain by 21 days; the loss of infectivity correlates with a severe depletion of both CD4+, and CD8+ T lymphocytes. These experiments confirm the lymphotropism of VZV suggested by

VARICELLA-ZOSTER VIRUS

275

FIG.4. Histological analysis of thymus/liver implants in SCID-hu mice infected with VZV. Infected implants were fixed in formalin, paraffin embedded, and cut into 3-mm sections before staining. Immunohistochemical staining on day 2 (A, magnification x 1062) showed VZV antigen associated with lymphocytes that had been engulfed by macrophages (arrowheads). Hematoxylin and eosin (H&E)staining (B, magnification x536) revealed lymphocyte depletion, fibrosis, and infiltration of red blood cells by day 7 postinfection. Mock-infected implants stained with H&E had a completely normal appearance (B', magnification X536). By day 14 postinfection, in situ hybridization detected VZV DNA predominantly in lymphocytes (right side of figure) and to a limited extent in the stroma (C, magnification ~ 3 5 4 )At . this time, VZV antigen was associated with most of the cells in the lymphoid lobes (D, magnification ~ 3 5 4while ) the epithelial stromal cells were not a major site of VZV antigen production (left side of figure). Reprinted with permission from Moffat et al. (1995).

the in situ hybridization studies of PBMC from individuals with acute varicella (Koropchak et al., 1989). To identify the cell populations that support VZV replication, the implant cells were stained with antiCD4, anti-CD8, and polyclonal VZV immune serum and analyzed by

276

ANN M.ARVIN et al.

FIG.5. FACS analysis of VZV-infected T-cells from SCID-hu mice. FACS analysis was performed on lymphocytes obtained from infected implants at day 7 postinfection. The cells were treated with antibodies and fluorescent conjugates to CD4, CD8, and VZV proteins. The percentage of VZV-positive T-cells of each subpopulation are shown for implants infected with either wildtype or Oka VZV strains. The negative controls are cells from uninfected implants treated with VZV immune serum and the background fluorescence of infected cells treated with nonimmune human serum. Reprinted with permission from Moffat et al. (1995).

FACS. Viral protein synthesis is evident in all T-lymphocytesubpopulations, including CD4+, CD8+,and dual positive CD4+/CD8+cells (Fig. 5). The frequency of VZV-infected cells among the T-lymphocyte subpopulations is 1-3 per 10,000 cells by infectious focus assay. Viral protein expression reaches a peak of 10-30% in each T-lymphocyte subpopulation by 7 days after infection. The total number of lymphocytes in infected thymus/liver implants decreases more than 10-foldby 7 days, although the relative proportion of each T-lymphocyte subpopulation remains unchanged. This analysis demonstrates equivalent lytic infection of all human T-lymphocyte subpopulations in the implants and indicates that VZV is tropic for both CD4+ and CD8+ T lymphocytes. In the SCID-hu mouse, as in the severely immunocompromised host, there is no immunologic restriction of VZV replication; the virus destroys almost all T lymphocytes in the implant by 21 days. As suggested by the clinical experience with administration of the live attenuated varicella vaccine to moderately immunosuppressed children, experiments in the SCID-hu mouse model indicate that the

277

VARICELLA-ZOSTER VIRUS

0 0 0

b 8 0

0

7

14

21

Days post infection FIG.6. Replication of VZV in thymus/liver implants from SCID-hu mice. The number of infectious foci per implant was determined by titration of infected lymphocytes on Vero cell monolayers. The average titer of the inoculum of infected MRC-5 cells from three experiments was calculated to be 3.2 x 103 2 1.1 x 103 (mean 2 SE) for the wildtype strain and 6.5 x 103 2.9 x 103 for the Oka strain. The scattergraph (A) shows the total number of infectious foci for each thy/liv implant infected with wildtype (open squares) and Oka (closed circles) strains of VZV. B shows the geometric means of the scatter data; the bars represent the standard error of the mean. Only positive samples were used to calculate the data shown for day 14.Dotted lines represent the level of detection of the assay. Reprinted with permission from Moffat et al. (1995).

*

278

ANN M. ARVIN et al.

Oka vaccine strain of VZV retains the virulence genes required to infect PBMC. Although there is some variation between implants, the peak titers after inoculation with wild-type and Oka strains are comparable (Fig. 6). A comparison of the geometric means of the titers shows that wild-type VZV replicates to a peak of 5.0 x 104 by 7 days whereas the Oka strain replicates somewhat more slowly, reaching a peak titer of 8.0 x 104 after 14 days, which is significantly higher than the wild-type strain titer of 1.1 x 104 a t the same time point ( p = 0.04). In FACS studies, the percentage of infected T lymphocytes within each subpopulation is equivalent for wild-type and Oka strains. Experiments in the SCID-hu mouse model provide important new evidence that the infection of lymphocytes with VZV in uiuo is accompanied by the release of infectious virus. The highly cell-associated nature of VZV replication in tissue culture is a well-known characteristic of the virus which has interfered with the analysis of VZV virion structure and replication by standard virologic methods (Cohen and Straw, 1995). Tissue culture cells do not release infectious VZV particles regardless of the cell type tested or the extent of cytopathic effect visible in the monolayer. In contrast, using transwell cultures to test T lymphocytes obtained after 7 days of VZV replication in thymus/liver implants in uiuo, we observed transfer of infectious virus across the membrane to a Vero cell monolayer in three of three samples containing 105 lymphocytes from VZV-infected thymus/liver implants (Moffatt et al., 1995).This observation is significant with regard to VZV pathogenesis since it suggests that infected T lymphocytes may release newly synthesized virions as they traffic through major organs, facilitating the transfer of the virus to lungs, liver, brain, or other sites. IV. CELL-MEDIATED IMMUNE RESPONSETO VARICELLA-ZOSTER VIRUS The pathogenic potential of VZV to cause cell-associated viremia and disseminated infection is restricted primarily by the cell-mediated immune component of the host response. Lymphopenia, with low numbers of circulating lymphocytes, is characteristic of the viremic phase of primary VZV infection, but a marked lymphocytosis occurs within 24-72 hours and is accompanied by cessation of viremia in the healthy individual. Progressive varicella among high-risk patients is associated with absolute lymphocyte counts below 500 cells/ml, persistent lymphopenia, and prolonged viremia. The CD4+ and CD8+ T-lymphocyte populations that are targets of VZV infection also mediate protection against severe disease unless the host response fails to eliminate infected lymphocytes from the circulation. The critical role of cell-mediated immunity in controlling VZV infection is most evident among patients who are receiving immunosup-

VARICELLA-ZOSTER VIRUS

279

pressive therapies for malignancy or other diseases or who have congenital or acquired immunodeficiency disorders (Arvin, 1992; Locksley et al., 1985; Whitley, 1990). These clinical conditions are associated with diminished T-lymphocyte recognition of VZV antigens and with enhanced susceptibility to progressive primary or recurrent VZV infections (Table 11). The capacity of VZV to infect lymphocytes, coupled with poor cell-mediated immune responses that fail to prevent or terminate cell-associated viremia, result in the transport of infectious virus to major organs and the occurrence of life-threatening complications such as pneumonia, fulminant hepatitis, and encephalitis. In contrast, patients with congenital deficiencies of immunoglobulin synthesis are not susceptible to severe VZV infections. After primary infection, induction of memory T lymphocytes specific for VZV is essential to protect the individual against symptomatic reactivation of endogenous, latent VZV as well as to prevent new infections after later exposures t o the virus. Effective vaccine-induced immunity also requires the induction of VZV-specific memory T lymphocytes.

A . Methods for Assessing Cell-Mediated Immunity to VZV The methods used to measure cell-mediated immunity to VZV include T-lymphocyte proliferation, cytokine production, and T-lymphocyte cytotoxicity (Arvin, 1992). In the proliferation assay, T-lymphocyte recognition of VZV antigens is detected by stimulating PBMC cultures with a solubilized extract of virus-infected cells or an uninfected cell control. The VZV antigen preparation contains a mixture of the major viral glycoproteins as well as regulatory and virion structural proteins that are produced during lytic infection of tissue culture cells, and the responding T-lymphocyte population is predominantly in the CD4+ T-cell subset. Antigen-specific T-lymphocyte proliferation is measured as [SHIthymidine incorporation by PBMC after 5 to 7 days in culture. Proliferation is quantitated by calculating the stimulation index (SI)from the ratio of counts per minute (cpm) between antigenstimulated and control wells; the SI is expected to be at least 3.0 in cultures of PBMC from healthy individuals who are immune to VZV. Cytokines, including interleukin 2 (IL-2) and gamma interferon (IFN-y), are produced by CD4+ T lymphocytes from immune subjects after in uitro stimulation with VZV antigen for 2 to 5 days. Cytokines released by PBMC can also be detected in serum from individuals with acute varicella and at local cutaneous sites in those with herpes zoster. The viral protein specificity of CD4+ T-lymphocyte recognition is demonstrated by using purified VZV proteins or synthetic peptides to stimulate PBMC in the proliferation assay (Arvin, 1992; Bergen et al., 1990). Herpes viral glycoproteins are considered likely to be major

280

ANN M. ARVIN et al.

targets of the primary host response because these proteins are expressed on the membranes of virus-infected cells as well as on the virion surface. Immunity elicited by the IE62 protein is of interest because IE62 is a major component of the virion tegument and functions as the major trans-activating factor for viral replication after the virus enters permissive cells. (Kinchington et al., 1992). In our experiments, immunoaffinity purification from VZV-infected cell extracts is used to generate gE (gpI), gH (gp III), and IE62 protein reagents for T-lymphocyte stimulation. Monoclonal antibodies to gE (gpI), gH (gp 1111, or the IE62 protein are coupled to cyanogen bromide-activated Sepharose 4B and incubated with a solubilized extract of VZV-infected cells overnight, and bound proteins are eluted with KSCN, concentrated, and analyzed by polyacrylamide gel electrophoresis. The purified protein reagents are diluted in medium and added to PBMC cultures. Proliferation is measured by pulsing with PHIthymidine after 6 days; an SI value greater than or equal to 2.0 is considered positive. The standard T-lymphocyte proliferation assay tests whether the individual has any circulating T lymphocytes than recognize VZV antigen or specific VZV proteins. The responder cell frequency modification of the proliferation assay provides a quantitative assessment of how many T lymphocytes are programmed to recognize some component of the whole VZV antigen preparation or a particular VZV protein. To quantitate responder cell frequencies, PBMC are cultured at cell concentrations of 0 to lo5 cells/well, with 24 replicates at each cell concentration. After pulsing with [3H]thymidine, responder wells are identified as those with cpm greater than the mean cpm f 3 SD for the corresponding control wells. Responder cell frequencies are calculated from limiting dilution plots, based on the number of negative wells at each cell concentration. Cytotoxic T lymphocytes (CTL) eliminate virus-infected cells by recognizing foreign antigens complexed with class I or class I1 major histocompatibility (MHC) antigens on the cell surface. Demonstrating VZVspecific cytolysis requires incubating effector T lymphocytes with antigen-expressing target cells that share class I or class I1alleles. The CD8+ T lymphocytes are the classic antiviral cytotoxic T lymphocytes that recognize viral peptides in the context of class I MHC proteins (Shaw and Biddison, 1979). In the case of VZV and other herpesviruses, cytotoxic T-lymphocyte function is also mediated by CD4+ T lymphocytes that recognize viral peptides associated with class I1 MHC proteins (Torpey et al., 1989). Lysis of VZV-infected target cells mediated by cytotoxic T lymphocytes is demonstrated using effector T lymphocytes generated by secondary in vitro stimulation with VZV. Target cells are produced by infecting autologous lymphoblastoid cells, which express

VARICELLA-ZOSTER VIRUS

281

both class I and class I1 MHC antigens, or MHC-matched fibroblasts, which express class I MHC antigen only. Autologous target cells are infected with VZV or recombinant vaccinia strains constructed to produce a single VZV protein (Arvin et al., 1991; Diaz et al., 1989; Hayward et al., 1986; Hickling et al., 1987). Our experiments to define the protein specificity of VZV cytotoxic T-lymphocyte function use vaccinia recombinants that express the glycoproteins gE (gp I) and gI (gp IV), IE62 protein, or the major DNA-binding protein (ORF29 gene product). Performing cytotoxicity assays under limiting dilution conditions allows quantitation of circulating cytotoxic T lymphocytes that are sensitized to recognize and lyse VZV-infected cells. Using a vaccinia recombinant that expresses one VZV protein to infect the target cells makes it possible to define the frequencies of cytotoxic T lymphocytes specific for the protein. Effector T lymphocytes are prepared by stimulation with VZV antigen under limiting dilution conditions for 14-18 days, with addition of irradiated feeder cells, IL-2, and fresh antigen a t intervals during the incubation period (Arvin et al., 1991). Antigenspecific lysis of targets is measured by chromium release with replicate wells at each of the effector T-lymphocyte concentrations being scored as positive if cpm are over 3.0 SD above the mean cpm for control wells. Immunofluorescence staining with CD4, CD8, and CD16 monoclonal antibody (MAb) and flow cytometry analysis is used to assess the phenotypes of the effector T-lymphocyte subpopulations that are present when the chromium release assay is done. Cytotoxic T-lymphocyte frequencies and the protein specificity of cytotoxic T lymphocytes within the CD4+ and CD8+ subsets can be determined by setting up limiting dilution cultures after an initial step to separate the cell subpopulations. Purified CD4+ and CD8+ T-lymphocyte preparations are made by flow cytometry purification after incubating PBMC with FITC-conjugated MAb to surface markers; each fractionated cell preparation contains over 99.5% CD4+ or CD8+ cells; purified T-lymphocyte subsets are then stimulated with VZV and evaluated as effector cells in the cytotoxicity assay.

B . Primary Cell-Mediated Immune Response to VZV Susceptible individuals lack VZV-specific T-lymphocyte responses. Primary cell-mediated immunity to VZV is elicited by natural infection with wild-type virus or by immunization with live attenuated varicella vaccine. Our studies of primary cell-mediated immunity focus on analyzing the kinetics of the initial response to VZV in healthy and immunocompromised patients and identifying the immunodominant VZV proteins that elicit helper and CTL functions in natural and vaccine-induced immunity.

282

ANN M. ARVIN et al.

1 . Natural Infection

The kinetics of acquisition of virus-specific cell-mediated immunity during primary VZV infection correlates with how extensive the cutaneous exanthem becomes as well as with the risk of visceral dissemination (Arvin et al., 198613; Giller et al., 1986). Healthy children who have detectable T-lymphocyte proliferation immediately after the appearance of the varicella exanthem have mild primary VZV infection, whereas immunodeficient children who fail to acquire VZV-specific T-lymphocyte proliferation are at risk for progressive, disseminated varicella (Fig. 7). In our experiments, the mean stimulation index to VZV antigen was 7.5 ? 10.43 SD within 72 hours among immunocompetent individuals with varicella who developed fewer than 100 cuta-

TI

FIG.7. Cellular immunity to VZV in healthy subjects with acute varicella. Lymphocyte transformation is expressed as the ratio of the mean cpm in the antigen-stimulated wells to uninfected cell control wells (transformation index, TI). Laboratory assay results are plotted against lesiondm2 as a measure of the clinical severity of varicella. Lesions were counted 48 hours after no new lesions had appeared; lesiondm2 was calculated by body surface area based on height and weight of the subject. Subjects with normal immune function (1); immunocompromised subjects (0); immunocompromised subjects who had fatal VZV infection (m). Reprinted with permission from Arvin et al. (1986b).

VARICELLA-ZOSTER VIRUS

283

*

neous lesions/m2, compared with 1.4 1.85 SD for those with more than 400 lesions/m2 ( p < 0.05). Only one (7.7%) of 13 immunocompromised patients had an immediate CD4+ T-lymphocyte response to VZV compared with 19 of 45 (42%) healthy subjects (p < 0.05). Production of IFN-y, as detected by serum concentrations, was higher among healthy children with more extensive varicella; the mean concentration was 1.36 U/ml in those with more than 400 lesions//m2 compared to 0.59 U/ml in those with less than 100 lesions/m2. The early production of immunoglobulin G (IgG) or IgM antibodies to VZV did not correlate with milder clinical varicella in healthy or immunocompromised patients. Healthy individuals who have uncomplicated acute varicella develop T lymphocytes specific for VZV glycoproteins gE (gp I) and gH (gp 111) as well as the IE62 protein, but the kinetics of acquisition of T lymphocytes specific for each protein is variable among individuals (Fig. 8) (Arvin et al., 1986a). Sixty-seven percent of the subjects had early proliferative responses to gE (gp I), 71% to gH (gp 1111, and 57% to IE62 protein. These observations suggest that the progression of primary VZV infection can be restricted by cell-mediated immune mechanisms involving T-lymphocyte recognition of one or more of several different viral proteins rather than being highly specific for a particular antigen. An effective primary cytotoxic T-lymphocyte response to VZV is likely to be an important component of the host response because clearance of infectious virus correlates with the induction of T lymphocytes that mediate lysis of virus-infected target cells in other viral infections. We detected specific killing of VZV-infected targets by T lymphocytes obtained as soon as 48 hours after the onset of acute varicella, using T lymphocytes expanded by culture with VZV antigen, IL-2, and feeder cells. ,Because antiviral cytotoxic T lymphocytes recognize processed peptides, their targets can represent short amino acid sequences from internal structural or regulatory proteins, as illustrated by cytotoxic T-lymphocyte recognition of the VZV IE62 protein, as well as surface glycoproteins. T lymphocytes that lysed targets expressing gE (gp I) or IE62 protein were detected in two individuals who were tested 5 and 9 days after the resolution of primary VZV infection; the responder cell frequencies of CTL specific for IE62 protein and gE (gp I) were 1:69,000 and 157,000 and 1:173,000 and 1:166,000, respectively (Fig. 9) (Sharp et al., 1992). 2. Varicella Vaccine

The kinetics and viral protein specificity of cell-mediated immunity following immunization with the live attenuated varicella vaccine re-

2 1 :260000

1 :65536

1:16384 1 :4096 IgG TITER

1:1024 1 :256

1:64 1:16 2 1 1:260000 :<1:16 65536 -

gpl18/--;-:i

L

1:16384 l g ~ 1:4096-

1:10241 :256-

1:64-

1

1:16<1:16

1

2 1 :260000

1 :65536

1:16384 l g ~

1:4096 1:1024 1 :256

1:64 1:16 <1:16

FIG.8. Memory T-lymphocyte recognition of VZV proteins. This figure shows the lymphocyte proliferation responses of PBMC from healthy subjects stimulated with immunoaffinity purified preparations of gE (gp I;gp90/58), gH (gp III;gpllW, or IE62 protein ( ~ 1 7 0 )The . results are expressed as the A cpm, which is the difference between mean cpm for VZV protein stimulated wells and control wells; line connect results for the same individual tested during the acute and convalescent phases of varicella. Reprinted with permission from Arvin et al. (1986a).

285

VARICELLA-ZOSTER VIRUS

>.

u z a

LL J

a

w

0

z

0

w

a 0

I

I

IE62 gpl

IE62 gpl

IE62 gpl

10 d

3-4 wk

8 wk

FIG.9. Kinetics of acquisition of cytotoxic T-lymphocytes specific for the varicellazoster virus immediate early protein (IE62) and gE (gp I) following immunization with varicella vaccine. VZV specific cytotoxic T-lymphocyte (CTL) frequency estimates (vertical axis) are shown for individual subjects who tested after one dose of varicella vaccine at intervals as indicated on the horizontal axis. Solid circles indicate CTL frequencies for 10 susceptible vaccine recipients; data for two vaccinees who had preexisting naturally acquired VZV immunity are shown with open circles. CTL frequencies in two subjects with recent acute varicella are indicated by triangles. The points connected by lines indicate CTL frequencies for IE62 and gp I targets as determined in the same subject. %printed with permission from Sharp et al. (1992).

semble the host response to natural VZV infection (Arbeter et al., 1986; Bergen et al., 1990; Diaz et al., 1988; Watson et al., 1990). More than 95%of varicella vaccine recipients have T lymphocytes that recognize VZV antigens by 2 to 6 weeks. Circulating T lymphocytes specific for VZV antigens are elicited before VZV antibodies can be measured; skin test reactivity t o VZV antigens is detected as early as 4 days after immunization (Kamiya et al., 1979). Susceptible household contacts who are vaccinated within 3 days of exposure do not develop varicella, indicating that this early immune response provides active protection in uiuo (Asano et al., 1977). The immediate induction of T lymphocytes that recognize VZV antigens may block VZV replication at sites that are important early in pathogenesis, such as the regional lymph nodes, or may abort the primary viremic phase of infection. The efficient induction of cytotoxic T-lymphocyte responses is considered one potential advantage of immunization with live viral vaccines. Our studies show that administration of the varicella vaccine elicits primary VZV-specific cytotoxicity detectable within 2 weeks after the initial dose (mean specific lysis of 32.4%k 5.5 SE) (Diaz et al.,

286

ANN M. ARVIN et al.

1989). The kinetics and viral protein specificity of the primary cytotoxic T-lymphocyte response were evaluated in 10 vaccinees at 3-4 weeks (Fig. 9) (Sharp et al., 1992).The frequency of T lymphocytes that recognized IE62 protein was 1:156,000 zk 50,000 SE and 1:175,000 2 63,000 SE for gE (g I), which is comparable to natural immunity. There was no preferential early recognition of the immediate early protein as opposed to the glycoprotein; the acquisition of cytotoxic T lymphocytes specific for gE (gp I) occurred in parallel with the development of IE62 protein recognition in individual vaccinees. Five vaccinees were also evaluated shortly after immunization to detect cytotoxic T lymphocytes specific for IE62 protein or gE (gp I) in cultures of purified CD4+ and CD8 T-lymphocyte subpopulations. The specific lysis of autologous targets expressing IE62 protein by CD4+ as well as by CD8+ T lymphocytes was detected by 10-16 days after vaccination in two subjects; neither of these vaccinees had early CD4+ or CD8+ T lymphocytes specific for gE (gpI) targets. A third subject tested at 8 weeks had CD8+-mediated recognition of gE (gp I) but no lysis of IE62 protein targets by CD4+ or CD8+ T-lymphocyte subpopulations. W o subjects who were tested at 12 weeks had cytotoxic T-lymphocyte recognition of gE (gp I) and IE62 protein by CD4+ T lymphocytes, CDS+ T lymphocytes, or both. The initial cellular immune response to varicella vaccine appears t o be influenced in part by noninfectious viral antigens contained in the vaccine preparation. The VZV antigen content of the preparation may act to prime the immune system, with a subsequent phase of further clonal expansion of VZV-specific T lymphocytes following replication of the infectious virus component. In our studies, 96% of children given varicella vaccine with 1140-1145 pfd1.6-1.7 relative antigen content had VZV-specific T-lymphocyte proliferation within 6 weeks compared to only 58%of vaccinees who were tested 8 weeks after being given vaccine with 950 pfu/l.O relative antigen content (p < 0.001) (Bergen et al., 1990).The mean stimulation index was 28.0 k 5.5 SE in first cohort compared to 6.0 zk 1.0 SE in the second group ( p < 0.001, t test). At 6 weeks, T-lymphocyte recognition of the IE62 protein was detected in 83% of recipients given vaccine with 1140-1145 pfuhelative antigen content compared with 20% of those given vaccine with a lower relative antigen content ( p = 0.001). The difference in infectious virus titers between the vaccine preparations was less than 200 pfuldose, which would not be expected to influence immunogenicity based on seroconversion studies. Stimulation by noninfectious antigens may also play a role in the initial host response to natural VZV infection as replication is associated with extensive production of defective particles and synthesis of viral proteins that are not incorpo+

VARICELLA-ZOSTER VIRUS

287

rated into infectious virions within infected cells. Early sensitization elicited by these antigens may favor the host in the host-virus interaction, accounting for the very mild clinical disease that occurs in some individuals with varicella. Similarly, the finding that inoculation with the Oka vaccine strain does not cause disease in healthy children could be related to a biphasic immune response, in which replication of the infectious virus component df the vaccine is limited by the initial priming elicited by noninfectious antigens in the vaccine. Varicella is a severe illness in many healthy adults (Arvin, 1995; Whitley, 1990). Termination of the viremic phase and of viral replication at localized cutaneous sites during primary VZV infection often fails to occur efficiently, and VZV pneumonia is much more common than it is in children. Direct comparisons of cellular immunity in adults and children with natural varicella have not been made, but our analysis of cell-mediated immunity to VZV following immunization with live attenuated varicella vaccine demonstrates that there are agerelated differences in the primary T-lymphocyte response to VZV (Nader et al., 1995). The varicella vaccine elicits protective immunity in susceptible adults as well as children; however, adults require two doses to achieve seroconversion rates over 90%, and adults with vaccine-induced immunity have a higher incidence of mild breakthrough varicella with exposure to wild-type virus. Circulating peripheral blood T lymphocytes from adults have a diminished capacity to recognize VZV antigens following primary sensitization. In our studies, two cohorts of children tested 6 weeks after one dose of varicella vaccine had stimulation indices of 28.6 6.21 SE and 21.1 f 3.81, whereas adult vaccinees had a stimulation index of 9.6 f 1.16 ( p = 0.04) (Fig. 10). Cell-mediated immunity to VZV in adults tested after their second dose of vaccine increased significantly, to a mean stimulation index of 30.5 9.12 SE,correlating with the high levels of seroconversion observed with the two-dose regimen. The numbers of antigen-specific T cells inducible by the initial dose of varicella vaccine in VZV-susceptible adults may be diminished because relatively fewer T lymphocytes are uncommitted, “naive” CD45RA+ T cells. The immunologic basis for the age-associated differences in T-lymphocyte responses to VZV antigens may relate to the increasing predominance with age of memory T cells having the CD45RO+ phenotype (Miller, 1991; Thoman and Weigle, 1989). Agerelated changes in the functional capacities of T-lymphocytes to produce cytokines or to recognize specific viral proteins also could account for the limited response of adults to primary VZV infection or to immunization with varicella vaccine (Hobbs et al., 1993). The varicella vaccine elicits cell-mediated immunity less reliably in

*

*

288

ANN M. ARVIN et al.

FIG.10. The initial T-lymphocyte proliferation response to VZV antigen in children and adults given live attenuated varicella vaccine. The T-lymphocyte proliferation responses to VZV antigen, expressed as the mean stimulation index (SI)+ SE, were compared between children enrolled in vaccine studies A (solid circles) or B (open circles) and the adult participants in vaccine study D (solid squares). Vaccinees were tested for T cell recognition of VZV antigen immediately before and a t 6 weeks after the first dose of live attenuated varicella vaccine was administered. The shaded area indicates SI responses below the threshold of 3.0.Reprinted with permission from Nader (1995).

leukemic children, which is expected because immunocompromised children have diminished T-lymphocyte responses to VZV following natural varicella (Arvin et al., 1986b; Gershon et al., 1984). Children with leukemia who have vaccine-induced immunity also have a higher risk of breakthrough varicella after exposure to wild-type virus (Gershon and Steinberg, 1989). Nevertheless, primary immunity can be induced with two doses of vaccine in most of these children, and significant protection against natural infection is achieved with a two-dose regimen, as it is in healthy adults. The modification of breakthrough infection to mild illness in these populations indicates that even partial primary sensitization induces immunity that is beneficial to the host.

C . Memory T-Lymphocyte Responses to VZV Viral virulence factors are required to establish and maintain VZV latency, but host factors have a major impact on whether the individual remains asymptomatic in spite of lifelong, persistent infection with the virus. Most individuals are protected against symptomatic reac-

VARICELLA-ZOSTER VIRUS

289

tivation or reinfection with VZV for decades following primary VZV infection, but high rates of VZV recurrence are associated with all conditions that cause prolonged suppression of cell-mediated immunity, such as malignancy or human immunodeficiency virus infection. Early studies of high-risk populations, including patients with lymphoma and other malignancies and cardiac, renal, or bone marrow transplant recipients, demonstrated that low or absent T-lymphocyte proliferation to VZV was a necessary but not a sufficient condition for the occurrence of herpes zoster (Arvin, 1992). Elderly adults who are otherwise healthy have diminished VZV T-lymphocyte proliferation responses, and the incidence of herpes zoster increases with age (Hayward and Herberger, 1987). The frequencies of circulating T lymphocytes that recognize VZV antigens are 1:10,000 to 1:30,000PBMC in adults with natural immunity, but the number of responder cells declines in the elderly (Levin et al., 1994). Disseminated VZV infections occur in the most severely immunocompromised patients, such as bone marrow transplant recipients, who lose cell-mediated immunity to VZV, indicating that an active cell-mediated immune response is required to limit the transfer of infectious virus to peripheral blood mononuclear cells during reactivation. 1 . Natural Infection

Virus-specific T-lymphocytes persist for many years after primary VZV infection in otherwise healthy individuals, as detected by the T-lymphocyte proliferation assay (Table 111). Our analysis of the viral protein specificity of the memory T-lymphocyte response showed that T-lymphocyte recognition of gE (gp I) and the IE62 protein was detectable in all subjects for more than 20 years; 86% of these individuals had T-lymphocyte proliferation to gH (gp 111) (Fig. 8) (Arvin et al., 1986a). These individuals had acquired varicella in widely separated geographic areas during different epidemic years, but their T lymphocytes responded to VZV antigens and proteins made from other unrelated VZV strains. The degree of antigenic variability between gE (gp I), gH (gp III), and IE62 produced by epidemiologically distinct strains of VZV did not have a measurable effect on the cellular immune response, as assessed by CD4+ T-lymphocyte proliferation. This observation is consistent with the fact that the exposure of healthy immune subjects to new VZV strains rarely causes clinical disease. Antigen-specific T-lymphocyte responses require binding of the viral peptide-MHC complex to the T-lymphocyte receptor (Table 11). We have used stimulation with synthetic peptides corresponding to sequences of gE (gp I) and the IE62 protein to further examine the immunogenicity of these major VZV proteins (Bergen et al., 1991). The

290

ANN M. ARVIN et al. TABLE I1 BETWEEN CELL-MEDIATED IMMUNITY AND CORRELATIONS VARICELLA-ZOSTER VIRUSDISEASED

Host Healthy susceptible Immunodeficient susceptible Healthy immune Healthy immune Immunodeficient immune a

VZV T-cell proliferation

Clinical Varicella Varicella Varicella exposure Age > 65 years

Early acquisition Delayed or no acquisition Enhanced response Low or absent remonse Low or absent response

Correlations Mild infection Risk of dissemination No disease Risk of herpes zoster Risk of herpes zoster and dissemination

Reprinted with permission; Arvin (1992).

gE and IE62 proteins were analyzed for amino acid sequences that conformed to algorithms for potential T-lymphocyte epitopes, which often form an ci helix with hydrophilic amino acids on one side and hydrophobic residues on the other, or consist of regions having a charged amino acid or glycine followed by two or three hydrophobic residues and a polar amino acid or glycine, that can be present in an

TABLE 111 AND PEPTIDES RECOGNIZED BY VARICELLA-ZOSTER VIRUSPROTEINS T-LYMPHOCYTES FROM IMMUNE INDIVIDUALS~ Assay

T-cell subset

T-cell proliferation Proteins Peptides T-cell cytotoxicity Induction of clonal expansion by secondary stimulation Recognition of target cells expressing VZV proteins

a

Reprinted with permission; Arvin (1992).

Viral proteins

VARICELLA-ZOSTER VIRUS

291

amphipathic helix or in p strands (Fbthbard and Taylor, 1988; Margalit et al., 1987). In our experiments, all 12 VZV immune subjects had T proliferation to at least two of ten gE (gp I) peptides; an average of six of the ten peptides were recognized by T lymphocytes from individual subjects. Six of the gE (gp I) peptides were recognized by T lymphocytes from 67-92% of the VZV immune donors; the frequency of donors responding to the other gE (gp I) peptides ranged from 42 to 58%. All VZV immune donors had memory T-lymphocyte responses to at least two of ten synthetic IE62 peptides. The mean number of IE62 peptides recognized by T lymphocytes from VZV immune donors was seven. Five of the ten IE62 peptides stimulated T lymphocytes from 7 5 4 3 % of the VZV immune donors and the other five IE62 peptides were recognized by T lymphocytes from 42-67% of subjects. Overall, one gE (gp I) peptide in combination with either of two other gE (gp I) peptides induced proliferation of T lymphocytes from all immune subjects. Similarly, a combination of two amino acid sequences of the IE62 protein was recognized by memory T lymphocytes from all 12 VZV immune donors, regardless of the MHC type. The gE (gp I) and the IE62 proteins serve very different functions in viral replication and infectivity, but the immunologic studies indicate that VZV infection in uiuo results in the processing of these two VZV proteins by antigen-binding cells t o yield multiple short amino acid sequences that can elicit memory T lymphocytes in most subjects. The response to multiple regions of the gE (gp I) and IE62 protein suggests that the memory T-lymphocyte repertoire for VZV peptide recognition is quite diverse in individuals with natural immunity to VZV. This finding, like other studies using synthetic peptides to investigate antiviral immunity, suggests that combinations of a relatively few epitopes of viral proteins can be identified that will induce memory T lymphocytes in most individuals with varying MHC phenotypes (Schrier et al., 1989). Memory T-lymphocyte responses are usually maintained ,against several regions of major VZV proteins (Giller et al., 1989; Hayward, 1990). Long-term immunity to VZV is associated with the persistence of cytotoxic T lymphocytes that recognize virus-infected cells. Memory CD4+ cytotoxic T lymphocytes lyse VZV-infected autologous lymphoblastoid cells, and CD8+-mediated cytotoxic T-lymphocyte activity is observed using VZV-infected fibroblasts that express only class I MHC antigen (Bowden et al., 1985; Diaz et al., 1989; Hayward et al., 1986; Hickling et al., 1987). Our studies of memory cytotoxic T lymphocytes were designed to allow statistical comparisons of the immune recognition of different classes of viral proteins, including the glycoproteins and the regulatory/structural proteins like the IE62 gene product (Arvin et al., 1991). Performing cytotoxicity assays using autologous lym-

292

ANN M. ARVIN et al.

phoblastoid targets infected with VZV-vaccinia recombinants makes it possible to quantitate cytotoxic T-lymphocyte frequencies with relative efficiency. These experiments showed that gE (gp I) and the IE62 protein, which elicit memory T lymphocytes as measured by the proliferation assay, were also recognized by VZV-specific cytotoxic T lymphocytes. The comparative precursor frequency data for cytotoxic T-lymphocyte recognition of gE (gp I) and the IE62 protein targets indicate that glycoprotein and regulatory /structural proteins that are expressed in initial viral replication are equivalent as targets for VZVspecific memory CTL. Among immune donors tested at least 20 years after primary VZV infection, the mean frequency of cytotoxic T-lymphocyte precursors specific for gp I in 11 subjects was 1:121,000 k 86,000 SD, with a range of 1:15,000 to 1:228,000; the mean precursor frequency for T lymphocytes that recognized the IE62 protein was 1:105,000 k 85,000 SD, with a range of 1:13,000 to 1:231,000 (Fig. 11).

IE PROTEIN

gp I PROTEIN

FIG.11. Frequencies of cytotoxic T-lymphocytes specific for the immediate early (IE62) protein and gE (gp I) of varicella-zoster virus. T-lymphocytes from VZV immune donors were incubated with inactivated VZV antigen in limiting dilution cultures, with initial T-cell concentrations ranging from 0 , l x 103,5 X 1 0 3 , l x 104,5 X 104, and 1 x 105, and tested for lysis of autologous LCL infected with a vaccinia virus recombinant that expressed the VZV IE62 protein or gp I protein, or with a vaccinia control recombinant. Precursor frequency estimates were derived by statistical analysis of standard limiting dilution plots. T-lymphocytes from five donors were tested for CTL activity against the IE62 protein or against the gp I target only. T-cells from six donors were evaluated against both VZV protein targets in parallel; the points connected by lines indicate the precursor CTL frequencies from these assays. The mean +SD for precursor frequency estimates are indicated next to the data points generated in individual assays. All of the donors had primary infection with VZV a t least 20 years before evaluation. Reprinted with permission from Arvin et al. (1991).

VARICELLA-ZOSTER VIRUS

293

Cytotoxic T lymphocytes specific for each of the VZV proteins were detected in all immune donors with equivalent frequencies ( p = 0.68). When aliquots of T lymphocytes that had been derived in the same limiting dilution culture were tested, there was no preferential recognition of the IE62 protein compared to gE (gp I) ( p = 0.12). Virusspecific memory CD4+ T lymphocytes that recognize gp I1 or gp IV, some of which express B-cell helper function as well as cytotoxic T-lymphocyte activity, are also detected in naturally immune individuals (Huang et al., 1992). On the basis of initial studies indicating that both major T-lymphocyte populations have cytotoxic activity against VZV-infected cells, we compared the capacity of purified CD4+ and CD8+ T-lymphocyte subpopulations to recognize and lyse targets that expressed the same VZV protein (Arvin et al., 1991). Quantitative analysis by limiting dilution showed that the numbers of circulating, memory cytotoxic T lymphocytes within the CD4+ and CD8+ subsets were comparable (Fig. 12). The VZV immune subjects had memory cytotoxic T lymphocytes that recognized the IE62 protein orgE (gp I) within each of the major T-lymphocyte populations. In the CD4+ population, the mean frequency of T lymphocytes specific for IE62 protein was 1:108,000 compared to 1:74,000 in the CD8+ subpopulation; the distribution of CD4+ and CD8+ cytotoxic T lymphocytes specific for the IE62 protein in VZV immune subjects was equivalent ( p = 0.17). The mean frequency of cytotoxic T lymphocytes specific for gp I targets was 1:119,000 in the CD4+ population compared to 1:31,000 in CD8+ T-lymphocyte cultures ( p = 0.97). No relative predominance of CD4+ or CD8+ T lymphocyte recognition of the IE62 protein as compared to gE (gp I) was observed. Although CD8+ T lymphocytes are identified as the major effector cell population in the classic description of antiviral immunity, the inducibility of class I1 MHC expression on the tissue cell types that are commonly infected with VZV, including epithelial cells and skin fibroblasts, means that CD4+-mediated, class II-restricted lysis could be important in uiuo, as has been suggested in HSV immunity (Cunningham et al., 1985; Schmid, 1986). The CD4+ T lymphocytes that infiltrate cutaneous herpes lesions immediately after their development may mediate cytotoxicity a s well as having helper cell function to recruit CD8+ cytotoxic T lymphocytes. The potential clinical significance of CD4+ T-lymphocyte-mediated lysis of VZV-infected cells is suggested by the observation that the incidence of herpes zoster increases during the phase of human immunodeficiency virus infection when absolute CD4+ T-lymphocyte numbers have declined (Whitley, 1990).

ANN M. ARVIN et al.

294 ~

1:250,000

t

1:150,000

1:50,000

t

L

\-

-

I CD4

CD8

IE PROTEIN

CD4

CD8

gp I PROTEIN

FIG. 12. Frequencies of cytotoxic T-lymphocytes specific for the immediate early (IE62) protein and gE (gp I) of varicella-zoster virus within the CD4+ and CD8+ T-lymphocyte populations. CD4+ and CD8+ T-lymphocyte populations were separated from peripheral blood by FACS under sterile conditions and incubated with inactivated VZV antigen in limiting dilution cultures; the initial T-cell concentrations ranged from 0 , l x 103,2 x 103, 1 x 104,5 x 104, and 1 x 105. After 12 days, the effector cells were tested for lysis of autologous LCL infected with a vaccinia virus recombinant that expressed the VZV IE62 protein or gp I protein, or with a vaccinia control recombinant. Precursor frequency estimates were derived by statistical analysis of standard limiting dilution plots. T-lymphocytes from one donor was tested for CTL activity against the IE62 protein only and cells from two donors were tested with the gp I target only. T-cell subpopulations from four donors were evaluated against VZV IE62 protein or gp I targets in parallel; the points connected by lines indicate the precursor CTL frequencies from these assays. Reprinted with permission from Arvin et al. (1991).

2 . Varicella Vaccine

Immunization with the varicella vaccine elicits memory T lymphocytes that recognize VZV antigens and immunoaffinity-purified gE (gp I) and IE62 protein. The responses are comparable to those following natural infection. However, factors such as the antigen content of the vaccine preparation or the age of the vaccine recipient may influence memory as well as primary cell-mediated immunity. In our experiments, the relative antigen content of the vaccine correlated with the detection of memory T lymphocytes at 1 year; 77% of the cohort given vaccine with 1140-1145 pfd1.6-1.7 relative antigen content had detectable T-lymphocyte proliferation responses compared to 40% of those given vaccine containing 950 pfu/l .O relative antigen content ( p = 0.03) (Bergen et al., 1990). The frequency of memory T lymphocytes

VARICELLA-ZOSTER VIRUS

295

specific for gE (gp I) was significantly lower in those given the vaccine with 950 pfWl.0 relative antigen content ( p = 0.02).When the two vaccine groups were tested for T-lymphocyte recognition of the IE62 protein, significant differences in mean stimulation index and in the number of vaccinees with a stimulation index of at least 2.0were also observed. Our vaccinees who received the less immunogenic preparation have had a higher incidence of varicella after exposure to wildtype virus, indicating that the immune response elicited was not sufficient to block cell-associated viremia. However, these breakthrough infections were modified, causing few cutaneous lesions and no systemic symptoms. T lymphocytes sensitized to VZV probably persisted although the frequencies are below the threshold detectable in the proliferation assay. The varicella vaccine that is now licensed for clinical use contains the higher relative antigen content. The infectious virus content of the varicella vaccine does not correlate directly with the persistence of VZV-specific memory T lymphocytes (Nader et al., 1995).At 1 year, the mean SI among children who were given vaccine containing 9000 pfu per dose was 10.2 f 2.16 SE compared to 15.6 f 4.40 SE among vaccinees who received vaccine with 3625 or 3315 pfu per dose. When the responses of all children and adults who received the very high pfu vaccine and those given the lower pfu vaccine were compared, the mean SI values were 10.1 f 1.16 SE and 15.5 k 3.24 SE a t 1 year (not significant). Clinically effective immunity is elicited in children given one dose of varicella vaccine (Gershon, 1992).However, the comparison of cellmediated immune responses to VZV antigen at 1 year suggest that higher numbers of memory T lymphocytes may be elicited by the administration of a two-dose regimen. Cell-mediated immune responses to VZV antigen at 1 year were significantly higher in children who were randomized t o receive two doses of the vaccine. The mean SI at 1 year was 22.2 f 6.42 SE for the two-dose subgroup compared to 9.3 f 1.39 SE for the one-dose subgroup ( p = 0.03).These subgroups included equal number of vaccinees receiving vaccine with the very high infectious virus content. Age at primary exposure to VZV also appears to influence the memory of T-lymphocyte response to VZV (Nader et al., 1995).Cell-mediated immune responses at 1 year were significantly higher among children compared to adults (Fig. 13).The mean SI was 15.6 ? 1.77SE among all children compared to 10.0 2 1.13 SE for adult vaccinees ( p = 0.03). This difference was observed even though only 18% of the children were given two doses of varicella vaccine whereas all adults had received two doses of the vaccine. The mean SI values at 1 year were 16.4 f 2.20 SE and 14.6 f 2.96 SE for the pediatric cohorts, respectively,

296

ANN M. ARVIN et al.

FIG.13. Comparison of the VZV specific cell-mediated immune responses of children and adult vaccinees evaluated a t l year. T-lymphocyte proliferation to VZV antigen was assessed following the administration of live attenuated varicella vaccine to children in vaccine groups A and B (solid circles) and adults in vaccine groups C and D (open circles). The T-cell proliferation response to VZV antigen is expressed as the mean stimulation index (SI) +SE. Mean SI are shown for each vaccine cohort tested immediately before the first dose of vaccine and a t 1 year. The shaded area indicates SI responses below the threshold of 3.0.Reprinted with permission from Nader et al. (1995).

*

compared to 7.7 1.30 and 11.6 k 2.15 SE for the adult groups. Although the responses were lower, 94% of adults immunized with varicella vaccine had a detectable T-lymphocyte response to VZV, as did 97% of children. When responder cell frequencies of memory T lymphocytes were measured, vaccinees who were tested at a mean of 3.5 years had more circulating T lymphocytes that recognized VZV antigens than naturally immune subjects who had primary VZV infection at least 20 years earlier; the frequencies of VZV-specific T lymphocytes were 1:18,000t 2000 SE compared to 1:39,000 t 3000 SE ( p = 0.001) (Sharp et al., 1992).The higher frequencies of VZV-specific T lymphocytes in the vaccinees suggests that numbers of memory T lymphocytes decline in relation to the interval from the primary exposure of the host to VZV. The frequencies of memory T lymphocytes specific for gE (gp I) and the IE62 protein were similar in healthy adult recipients of varicella vaccine and adults with naturally acquired immunity to vzv. Immunization with varicella vaccine also elicits memory T lymphocytes that have VZV-specific cytotoxic function (Fig. 14)(Sharp et al.,

VARICELLA-ZOSTER VIRUS

297

FIG.14. Responder cell frequencies of T-lymphocytes specific for VZV antigen, immediate early protein (IE62), and gE (gp I) in healthy adults with vaccine-induced or natural immunity to VZV. Varicella-zoster virus (VZV) specific responder cell frequencies + standard error (vertical axis) to whole VZV antigen, IE62, and gp I are shown in relation to whether VZV immunity was naturally acquired (NI), indicated by the shaded bars, or induced by immunization with varicella vaccine (vacc), indicated by the open bars, as shown on the horizontal axis. Five subjects were tested in each group. Reprinted with permission from Sharp et al. (1992).

1992). Recognition of gE (gp I) gI (gp IV), gC (gp V), and the IE62 protein is detected in most vaccinees. The frequencies of IE62 proteinspecific cytotoxic T lymphocytes were comparable in vaccinees and naturally immune individuals (1:131,000 k 40,000 SE versus 1:104,000 19,000 SE). The frequencies of memory cytotoxic T lymphocytes specific for gI (gp IV) were also equivalent (1:155,000 ? 60,000 SE versus 1:170,000 71,000 SE). The mean frequency of cytotoxic T lymphocytes recognizing gC (gp V) was 1:208,000 k 53,000 SE in subjects with vaccine-induced immunity compared to 1:108,000 2 27,000 SE in naturally immune subjects. The gC (gp V) specific CTL frequencies were less than 1:200,000 in all of the five naturally immune subjects but were over 1:200,000 in three of six vaccine recipients. This trend toward a lower mean frequency of CTL specific for gC (gp V) in vaccinees compared to naturally immune subjects may be related to the apparent variability in gC (gp V) production by some isolates of VZV strain Oka compared to wild-type VZV strains, and it correlates with the observation of low gC (gp V) antibodies in some vaccinees. As in the case of naturally acquired immunity, the clinical experience is that memory immune responses elicited by the varicella

*

*

298

ANN M. ARVIN et al.

vaccine strain, Oka, usually protect against disease caused by wildtype virus, regardless of the epidemic year or geographic origin of the contact virus.

D . Mechanisms for Maintaining Cell-Mediated Immunity to VZV Whether virus-specific T lymphocytes persist for the life of the host following primary sensitization remains an unresolved issue in viral immunology. Even if some memory T lymphocytes persist, a gradual decline in the numbers of VZV-specific T lymphocytes may occur after primary sensitization. As we observed, circulating T lymphocytes that recognized VZV antigen were significantly more frequent in vaccinees who were immunized 3.5 years earlier than in naturally immune adults who had varicella in childhood (Sharp et al., 1992).In the case of VZV, there are natural mechanisms for maintaining the memory T-lymphocyte responses through restimulation of the immune system by reexposure or reactivation of latent virus. 1 . Reexposure to VZV

The annual epidemics of varicella cause reexposures to the virus through contact with infected individuals. The role of this mechanism in enhancing VZV immunity was evaluated in studies of immunocompetent VZV immune mothers who had household exposure to children with varicella. Seventy-one percent of these women developed increases in T-lymphocyteproliferation to VZV antigens without signs of clinical infection; the mean stimulation index increased from 7.8 t 1.3 SE within 14 days after the onset of varicella in the child compared to 15.3 f 2.56 SE by 3 to 4 weeks (Fig. 7).This restimulation mechanism is supported by the detection of VZV by PCR testing of nasopharygeal secretions from household contacts of children with varicella (Connelly et al., 1993).Herpes zoster occurring in close contacts of immune individuals may also boost their memory immune responses to VZV. 2 . Reactivation of Latent VZV

Reactivation of endogenous latent virus is the second potential mechanism for maintaining persistent immunity to VZV. Both immunocompetent and immunocompromised patients have a marked increase in T-lymphocyte proliferation to VZV antigens following episodes of herpes zoster. However, the persistence of VZV-specific memory T lymphocytes is clearly not dependent on symptomatic VZV reactivations, as it can be demonstrated in immune subjects with no recent VZV-related disease. Subclinical VZV reactivations may be an important stimulus for sustaining cell-mediated immunity to VZV. By

VARICELLA-ZOSTER VIRUS

this hypothesis, helper and cytotoxic T lymphocytes specific for some VZV proteins, such as the IE62 protein, may be maintained by periodic restimulation in uiuo by abortive viral replication in latently infected cells. The induction of helper and cytotoxic T lymphocytes specific for viral glycoproteins would be expected during primary infection with herpes viruses because of their abundant expression on the surface of infected epithelial cells (Cohen and Straus, 1995). Restimulation of glycoprotein-specific immunity due to subclinical reactivation would require that replication is carried to the phase of late viral protein synthesis. It is also possible that memory immunity to different VZV proteins is maintained by varying mechanisms. For example, the persistence of T-lymphocyte recognition of glycoproteins may depend more on exogenous reexposures, whereas subclinical reactivation may be more important for sustaining immunity to early viral proteins. Evidence indicates that the IE62 gene is transcribed in latently infected cells (Croen and Straus, 1991). If transcription proves to be associated with protein synthesis, intermittent boosting of the host response to IE62 protein could be important for preventing progression of subclinical reactivations to clinically apparent episodes of herpes zoster. Although it is difficult to prove in healthy individuals, the contribution of subclinical reactivation to the maintenance of cellular immunity is supported by studies of bone marrow transplant recipients. Clinically, the risk of herpes zoster declines significantly by 1 year after bone marrow transplant, and most long-term survivors of bone marrow transplantation recover VZV-specific cellular immunity (Meyers et al., 1980; Wilson et al., 1992). T-lymphocyte recognition of VZV antigens was detected by proliferation assay in 89% of bone marrow transplant recipients who had herpes zoster, but 52% of patients who did not have VZV recurrences also recovered cellular immunity to VZV (Wilson et al., 1992).In our studies, the mean stimulation index to VZV was 1.0 2 0.42 SD at less than 100 days after bone marrow transplantation compared to 12.0 % 6.03 SD at more than 100 days ( p = 0.003). Among patients tested at more than 100 days, 63% of patients with detectable T-lymphocyte proliferation had subclinical or symptomatic VZV reactivation compared to none of those who lacked VZV T-lymphocyte responses. The documentation of cell-associated VZV viremia before the reconstitution of T-lymphocyte recognition of VZV antigen suggests that subclinical reactivation stimulates the induction or clonal reexpansion of VZV-specific T lymphocytes. Subclinical VZV reactivation may also enhance the recovery of VZVspecific cytotoxic T-lymphocyte function. Fifty percent of the bone marrow transplant patients had VZV-specific cytotoxicity at a mean of 155 k 98 SD days after transplant, and the recovery of VZV-specific

300

ANN M. ARVIN et al.

CTL function was demonstrated as early as 48 days. Among bone marrow transplant recipients, the mean precursor frequency of T lymphocytes that recognized IE62 protein was 1:233,000 k 162,000 SD; 66% of patients had detectable responses to the IE62 protein. The mean frequency of cytotoxic T lymphocytes specific for gE (gp I) was 1:277,000 2 142,000 SD; 60% of bone marrow transplant patients had T lymphocytes that recognized gE (gp I) targets. Nevertheless, bone marrow transplant recipients had significantly fewer circulating cytotoxic T lymphocytes that recognized the IE62 protein or gE (gp I) than healthy VZV immune subjects. The distribution of peripheral blood lymphocyte phenotypes after bone marrow transplantation is characterized by an inversion of the CD4+/CD8+ T-lymphocyte ratio that can last for several years (Forman et al., 1992). The lymphocyte subpopulations that were detected after stimulating limiting dilution cultures from bone marrow transplant recipients with whole VZV antigen showed a significant reduction in the outgrowth of CD4+ cells when compared to healthy individuals. It is obvious that in order for endogenous reexposure by VZV reactivation to sustain protective immunity, the immune system must have the functional capacity to respond to restimulation. The failure of the CD4+ T-lymphocyte population to respond to secondary reexposure to VZV antigen may cause a significant deficiency in the capacity of bone marrow transplant recipients to restrict VZV replication by cytotoxic mechanisms, and it may account for the prolonged susceptibility to herpes zoster.

3. Varicella Vaccine The capacity of reexposure to VZV antigens to boost memory T-cell immunity is demonstrated directly by the responses of individuals with preexisting immunity to VZV who are given varicella vaccine. Immunization of healthy VZV immune elderly individuals increases the number of circulating peripheral blood T lymphocytes that recognize VZV antigens significantly (Levin et al., 1994). In the course of our vaccine studies, four children and five adults who were considered susceptible were later found to have IgG antibodies and T-lymphocyte proliferation to VZV in samples taken just before immunization (Nader et al., 1995). All of these vaccines had a marked increase in T-lymphocyte proliferation to VZV antigen after immunization. The prevaccine baseline stimulation index was 7.9 5 2.18 SE compared to 36.3 t 12.22 SE after immunization (Fig. 9). The mean stimulation index among these vaccinees was 12.0 ? 3.12 SE after 1 year, which was not statistically different from the baseline responses. Cytotoxic T-lymphocyte responses t o VZV are also boosted by the

VARICELLA-ZOSTER VIRUS

301

administration of varicella vaccine to immune individuals. The highest frequencies of VZV-specific cytotoxic T lymphocytes that we have observed were in two vaccinees who had preexisting naturally acquired immunity to VZV. The cytotoxic T-lymphocytefrequencies were 15000 for IE62 protein and 1:8000 for gp I in one subject 3 weeks after receiving varicella vaccine; a second subject, who was tested 8 weeks after immunization, had cytotoxic T-lymphocyte frequencies of 1:14,000for IE62 protein and 1:40,000for gE (gp I). This evidence that cytotoxic T-lymphocyte responses can be boosted further supports the hypothesis that immunization of otherwise healthy elderly individuals could reduce the risk of symptomatic VZV reactivation. Our studies of bone marrow transplant patients show that the reconstitution of virus-specific cellular immunity is often delayed and may not occur until after the patient has experienced recurrent VZV infection. Whether VZV-specific cell-mediated immunity will be maintained during subsequent immunosuppressive therapy is also unpredictable. Inactivated or subunit herpes viral vaccines could provide an effective substitute for “natural” resensitization by viral reactivation and a means to boost the host response periodically after bone marrow transplantation or other immunosuppressive therapy. Our preliminary studies indicate that administering one or more doses of an heat-inactivated preparation of the live attenuated varicella vaccine enhances the recovery of cell-mediated immunity in bone marrow transplant recipients (Redman et al., 1995). Demonstrating the capacity of varicella vaccine to boost memory T-lymphocyte responses of healthy VZV immune individuals is particularly relevant now that the vaccine is licensed for clinical use. A reduction of the annual varicella epidemics is expected to follow the introduction of universal vaccination against VZV infection in childhood. If reexposure to close contacts who have varicella is an important mechanism for maintaining VZV immunity, it may be necessary to substitute for the diminishing frequency of these exposures by giving additional doses of the vaccine to individuals who have vaccineinduced immunity at later ages. Long-term surveillance to determine the persistence of immunity to VZV among vaccinees is planned to address this issue.

E . Cell-Mediated Immunity to VZV in the Guinea Pig Model Viral proteins that constitute major targets of the host response to VZV can be identified in studies of individuals with natural or vaccineinduced immunity by comparing the frequencies of T lymphocytes that recognize specific proteins, but these experiments do not define the

302

A N N M.ARVIN et al.

protective efficacy of immunity by a single viral protein. Guinea pigs develop T-lymphocyte recognition and cytotoxic T-lymphocyte responses to VZV that can be detected with assays that are useful for assessing the human immune response to VZV (Arvin et al., 1987; Hayward et al., 1991). In our initial experiments to assess the immunogenicity of specific VZV proteins, strain 2 guinea pigs were inoculated with vaccinia virus recombinants expressing gE (gp I), gI (gp IV), gC (gp V), or the IE62 protein (Lowry et al., 199213). All of the VZVvaccinia recombinants elicited VZV-specific T-lymphocyte proliferation. The gE (gp I) recombinant induced the highest response, namely, a stimulation index of 3.8 ? 0.9 SE 3 weeks after a second injection, which persisted at 15 to 18 weeks. Inoculation with gC (gp V) elicited the lowest initial response at 3 weeks, with a stimulation index of 2.5 f 1.1SE. Priming of VZV-specific T-lymphocyte proliferation was observed in all animals immunized with VZV-vaccinia virus recombinants following inoculation with infectious VZV (Fig. 15). The peak stimulation index was 10.6 ? 3.1 SE for the gE (gp I) group, 14.1 k 4.1 SE for those given gI (gp IV), 11.2 +- 4.5 SE after gC (gp V) inoculation, and 10.1 f 2.7 SE after priming with IE62 protein (Fig. 9). These experiments indicate that any of several VZV proteins are effective for priming the host response to VZV. The finding that virus-specific cellular immunity was induced using IE62 protein suggests that this major tegument protein of VZV, which is a target of persistent helper and cytotoxic T-lymphocyte immunity in the human host, could be a useful component for vaccine development. Whether viral proteins expressed in infectious virus vectors, such as vaccinia, are more effective than purified protein preparations in inducing cell-mediated immunity is an important issue in designing vaccines to prevent herpesvirus infections. In our experiments, immunization of guinea pigs with immunoaffinity-purified gE (gp I) or IE62 protein elicited T-lymphocyte responses that were comparable to those in guinea pigs inoculated with vaccinia recombinants that expressed the same proteins. High concentrations of purified gE (gp I) or IE62 protein were required to elicit T-lymphocyte responses equivalent to those observed in animals inoculated with vaccinia recombinants expressing gE (gp I) or IE62 protein, and T-lymphocyte proliferation was not detected by 12-18 weeks after purified protein immunization. Protection is difficult to evaluate in the guinea pig model because VZV infection causes no signs of illness except for transient rash in some euthymic, hairless animals (Myers et al., 1991).We have used the production of IgG antibodies to the IE62 protein to assess whether challenge virus replicated in animals immunized with vaccinia recombinants expressing gE (gpI), gI (gp IV), or gC (gpV). Antibodies to IE62

303

VARICELLA-ZOSTER VIRUS

6.0

t

/-

7

A

L

**O

t

01

I

I

I

Pre-VZV

6 to 12 weeks

15 to 18 weeks

FIG.15. VZV-specific T-lymphocyte proliferation in strain 2 guinea pigs immunized with vaccinia recombinants expressing gE (gp I), gI (gp IV), gC (gp V),or IE62 protein after challenge with infectious VZV. This figure shows VZV-specificT-lymphocyte proliferation responses, expressed as the mean SI + SE, in animals immunized with vac-gp I, vac-gp IV, vac-gp V, vac-IE-62 or vac-c followed by inoculation with guinea pig celladapted, infectious VZV. The animals were tested before and at intervals of 6-12 weeks and 15-18 weeks after challenge with infectious VZV. Reprinted with permission from Lowry et al. (1992a).

protein were not detected in any of four animals immunized with gE (gp I) or gI (gp IV), but all of those given gC (gp V) had antibodies to IE62 protein 3 weeks after challenge with infectious VZV. This evidence of challenge virus replication correlated with the somewhat lower cell-mediated immune responses of animals immunized with the gC (gp V) recombinant. In other experiments, we used PCR detection of VZV to assess the protective efficacy of immunization with the IE62 protein or the major DNA-binding protein, ORF29 protein, in Hartley guinea pigs (Sabella et al., 1993). Protection against cell-associated viremia and spread of VZV to dorsal root ganglia sites, as determined by PCR assay, was defined as effective immunity (40). All eight animals immunized with the vaccinia recombinant expressing the ORF29 protein had cell-

304

ANN M. ARVIN et al.

associated viremia 3 days after challenge compared to only 2 of 11 animals given vaccinia recombinant expressing the IE62 protein ( p = 0.005). Virus was detected by PCR in dorsal root ganglia tissue from 3 of 8 animals (38%)immunized with the ORF29 protein and in 4 of 9 control animals (44%) compared with only 1 of 11 (9%) animals immunized with the IE62 protein. This animal also had cell-associated viremia despite immunization. As observed in earlier experiments, guinea pigs inoculated with the IE62 protein had a significant boost in cell-mediated immunity when they were challenged with infectious VZV; the mean stimulation index increased from 3.4 k 0.77 SE to 20.7 k 6.25 SE ( p = 0.005). T-lymphocyte recognition of VZV antigens was not primed in animals immunized with the ORF29 major DNAbinding protein (3.9 1.12 versus 4.8 ? 0.91). Lack of priming of cellmediated immunity was associated with cell-associated viremia and a higher frequency of ganglia infection in these animals. These results are consistent with the evidence that the IE62 protein is an important target of the human cell-mediated immune response to VZV.

*

V. SUMMARY Events in the pathogenesis of infection and the host response to VZV are very closely linked. Our experiments demonstrate that CD4+ and CD8+ T-lymphocyte populations that are targets of cell-associated VZV viremia also mediate protection against severe infection. Diminished cell-mediated immunity predisposes the host to progressive primary or recurrent VZV disease because infected lymphocytes persist in the circulation and carry the virus t o major organs, causing pneumonitis, hepatitis, or other life-threatening complications. The live attenuated varicella vaccine induces cell-mediated immunity and protects against or significantly reduces the morbidity associated with primary VZV infections. The universal administration of varicella vaccine is likely to generate new insights about host-virus interactions, particularly in relation to how VZV immunity is maintained, that will be relevant to the design of vaccines for other human herpesviruses. ACKNOWLEDGMENTS Studies in Dr. Arvin’s laboratory were supportedby US.Public Health Service Grants A1 20459, A1 22280, and A1 18449 from the National Institute of Allergy and Infectious Disease, by Merck & Co., Inc., and by the National Cancer Institute (CA 49605). The authors acknowledge the contributions of Celine Koropchak, Margart Sharp, Alec E. Wittek, Pamela S. Diaz, Randy Bergen, Camille Sabella, Philip Lowry, and Sonia Nader

VARICELLA-ZOSTER VIRUS

305

and of our collaborators John Hay, William Ruyechan, and Paul Kinchington to the work described in this review.

REFERENCES American Academy of Pediatrics. (1995). Recommendations for the use of varicella vaccine. Pediatrics 95, 791-796. Arbeter, A. M., Starr, S. E., Weibel, R. E. and Plotkin, S. A. (1986). Varicella vaccine studies in healthy children and adults. Pediatrics 78 (Suppl.), 748-756. Arvin, A. M. (1992). Cell-mediated immunity to varicella-zoster virus. J.Infect. Dis. 166, (suppl. 11, 35-41. Arvin, A. M. (1995). Varicella-zoster virus. In “Virology” (B. N. Fields et al., eds.), 3rd Ed., in press. Raven, New York. Arvin, A. M., Koropchak, C. M., and Wittek, A. E. (1983). Immunologic evidence of reinfection with varicella-zoster virus. J. Infect. Dis. 148, 200-205. Arvin, A. M., Kinney-Thomas, E., Shriver, K., Grose, C., Koropchak, C. M., Scranton, E., Wittek, A. E., and Diaz, P. S. (1986a). Immunity to varicella-zoster viral glycoproteins, gp I (90/58) and gp 111 (gp 118) and to a nonglycosylated protein, p170. J. Immunol. 137, 1346-1351. Arvin, A. M., Koropchak, C. M., Williams, B. R. G., Grumet, F. C., and Foung, S. K. (1986b). Early immune response in healthy and immunocompromised subjects with primary varicella-zoster virus infection. J. Infect. Dis. 154, 422-429. Arvin, A. M., Solem, S. M., Koropchak, C. M., Kinney-Thomas, E., and Paryani, S. G. (1987). Humoral and cellular immunity to varicella-zoster virus glycoprotein, gp I, and to a non-glycosylated protein, p170, in the strain 2 guinea pig. J. Gen. Virol. 68, 2449-2454. Arvin, A. M., Sharp, M., Smith, S., Koropchak, C. M., Diaz, P. S., Kinchington, P., Ruyechan, W., and Hay, J. (1991). Equivalent recognition of a varicella-zoster virus immediate early protein (IE62) and glycoprotein I by cytotoxic T-lymphocytes of either CD4+ or CD8+ phenotype. J. Immunol. 146,257-264. Asano, Y., Nakayama, H., Yazaki, T., et al. (1977). Protection against varicella in family contacts by immediate inoculation with live varicella vaccine. Pediatrics 59, 3-7. Asano, Y., Itakura, N., Hiroishi, Y. et al. (1985). Viral replication and immunologic responses in children naturally infected with varicella-zoster and in varicella vaccine recipients. J. Infect. Dis. 152, 863-868. Bergen, R. E., Diaz, P. S., and Arvin, A. M. (1990). The immunogenicity of the Oka/Merck varicella vaccine in relation to infectious varicella-zoster virus and relative viral antigen content. J.Infect. Dis. 162, 1049-1054. Bergen, R. E., Sharp, M., Sanchez, A., Judd, A. K., and Arvin, A. M. (1991). Human T-cells recognize multiple eptiopes of a major tegument/immediate early protein (IE62) and glycoprotein I of varicella-zoster virus. Viral Immunol. 4, 151-166. Bowden, R. A,, Levin, M. S., Giller, R. H., Tubergen, D. G., and Hayward, A. R. (1985). Lysis of varicella zoster virus infected cells by lymphocytes from normal humans and immunosuppressed pediatric leukaemic patients. Clin. Exp. Immunol. 60, 387-95. Cohen, J. I. and Straus, S. E. (1995). Varicella-zoster virus and its replication. In “Virology” (B. N. Fields et al., eds.), 3rd Ed., in press. Raven, New York. Connelly, B. L., Stanberry, L. R., and Bernstein, D. I. (1993). Detection of varicella-zoster virus DNA in nasopharyngeal secretions of immune household contacts of varicella. J. Infect. Dis. 168, 1253-1255. Croen, K., and Straus, S. (1991). Varicella-zoster virus latency. Annu. Rev. Microbiol. 4, 265-282.

306

ANN M. ARVIN et al.

Croen, K.,Ostrove, J., Dragovic, L., and Straus, S. (1988).Patterns of gene expression and sites of latency in human nerve ganglia are different for varicella-zoster and herpes simplex viruses. Proc. Natl. Acad. Sci. U S A . 86, 9773-9777. Cunningham, A. L., Turner, R. R., Miller, A. C., Para, M. F., and Merigan, T.C. (1985). Evolution of recurrent herpes simplex virus lesions. An immunohistologic study. J. Clin. Invest. 76, 226. Davison, A. J.,and Scott, J. E. (1986).The complete DNA sequence of varicella zoster virus. J. Gen. Virol. 67, 597-611. Davison, A. J., Edson, C. M., Ellis, R. W., Forghani, B., Gilden, D., Grose, C., Keller, P. M., Vafai, A., Wroblewska, Z., and Yamanishi, K. (1986).A new nomenclature for the glycoprotein genes of varicella-zoster virus and their glycosylated products. J. Virol. 67, 1195-1197. Diaz, P.S.,Smith, S., Hunter, E., and Arvin, A. M. (1988).Immunity to varicella-zoster virus antigen, glycoprotein I (gpI) and a non-glycosylated protein (p 170):Relation to the immunizing regimen of live attenuated varicella vaccine. J. Infect. Dis. 168, 1245-52. Diaz, P. S.,Hunter, E. L., Smith, S. M., and Arvin, A. M. (1989).T lymphocyte cytotoxicity with natural varicella-zoster virus infection and after immunization with live attenuated varicella vaccine. J.Immunol. 142,636-641. Feldman, S., and Epp, E. (1976).Isolation of varicella-zoster virus from blood. J.Pediatr. 88,265-268. Forman, S. J., Nocker, P., Gallagher, M., Zaia, J., Wright, C., Bolen, J., Mills, B., and Hecht, T. (1982).Pattern of T cell reconstitution following allogenic bone marrow transplantation for acute hematological malignancy. Transplantation 34, 96-98. Gabel, C. A., Dubey, L., Steinberg, S. P., Sherman, D., Gershon, M. D., and Gershon, A. A. (1989).Varicella-zoster virus glycoprotein oligosaccharides are phosphorylated during posttranslational maturation. J. Virol. 83,42644276. Gershon, A. A. (1992).Varicella vaccine: The American experience. J.Infect. Dis. 166, (Suppl. l), 63-68. Gershon, A. A., and Steinberg, S. P. (1989).Persistence of immunity to varicella in children with leukemia immunized with live attenuated varicella vaccine. N . Eng. J. Med. 320,892-897. Gershon, A. A., Steinberg, S. P., Gelb, L., and the NIAID Collaborative Varicella Vaccine Study Group. (1984).Live attenuated varicella vaccine: Efficacy for children with leukemia in remission. JAMA, J. Am. Med. Assoc. 262,355-362. Giller, R.H., Bowden, R. A., Levin, M. J., Walker, L. J., Tubergen, D. G., and Hayward, A. R. (1986).Reduced cellular immunity to varicella zoster virus during treatment for acute lymphoblastic leukemia of childhood: In uitro studies of possible mechanisms. J. Clin. Immunol. 6,472-80. Giller, R. H., Winistorfer, S., and Grose, C. (1989).Cellular and humoral immunity to varicella zoster virus glycoproteins in immune and susceptible human subjects. J. Infect. Dis. 160, 919-22. Grose, C. (1990).Glycoproteins encoded by varicella-zoster virus: Biosynthesis, phosphorylation, and intracellular trafficking. Annu. Reu. Microbiol. 44,59-80. Hayward, A. R. (1990).T-cell responses to predicted amphipathic peptides of varicellazoster virus glycoproteins I1 and IV. J. Virol. 64, 651-55. Hayward, A. R.,and Herberger, M. (1987).Lymphocyte responses to varicella-zoster virus in the elderly. J. Clin. Immunol. 7 , 174-178. Hayward, A. R.,Pontesilli, O., Herberger, O., Laszlo, M., and Levin, M. (1986).Specific lysis of varicella zoster virus-infected B lymphoblasts by human T-cells. J. Virol. 58, 179-184.

VARICELLA-ZOSTER VIRUS

307

Hayward, A. R., Burger, R., Schleper, T., and Arvin, A. M. (1991). Major histocompatibility restriction of T-cell responses to varicella-zoster virus in guinea pigs. J . Virol. 65, 1491-1495. Hickling, J. K., Borysiewicz, L. K., and Sissons, J. G. P. (1987). Varicella-zoster virus specific cytotoxic T lymphocytes (Tc): Detection and frequency analysis of HLA class I-restricted Tc in human peripheral blood. J . Virol. 61, 3463-3469. Hobbs, M. V., Weigle, W. O., Noonan, D. J., Torbett, B. E., McEvilly, R. J., Koch, R. J., Cardenas, G. J., and Emst, D. N. (1993). Patterns of cyotokine gene expression by CD4+ T cells from young and old mice. J. Immuml. 160, 3602-3614. Huang, Z., Vafai, A., and Hayward, A. R. (1992). Specific lysis of targets expressing VZV gpI or gpIV by CD4+ clones. J. Virol. 66, 2664-2669. Kamiya, H., Ihara, T., Hattori, A., Iwasa, T., Sakurai, M., Izawa, T., Yamada, A., and Takahashi, M. (1979). Diagnostic skin test reactions with varicella-zoster virus antigens and clinical application of the test. J. Infect. Dis. 136, 784-791. Kinchington, P. R., Houghland, J., Awin, A. M., Ruyechan, W., and Hay, J. (1992). Varicella zoster virus IE62 protein is a major virion component. J. Virol.66,359-366. Koropchak, C. M., Diaz, P. S., and Awin, A. M. (1989). Investigation of varicella-zoster virus infection of lymphocytes by in situ hybridization. J. Virol. 63, 2392-2395. Koropchak, C. M., Graham, G., Palmer, J., et al. (1991). Investigation of varicella-zoster virus infection by polymerase chain reaction in the immunocompetent host with acute varicella. J . Infect. Dis. 1016-1022. Kundratitz, K. (1925). Experimentelle ubertrsgungen von herpes zoster auf menschen und die beriehungen von herpes zoster zu varizellen. 2.Kinderheilkd. 39, 379-387. Levin, M. J., Murray, M., Rotbart, H. A., Zerbe, G. O., White, C. J., and Hayward, A. F. (1994). The immune response of elderly individuals to a live attenuated varicella vaccine. J. Infect. Dis. 166, 253-259, 1992. Locksley, R. M., Flournoy, N., Sullivan, K. M., and Meyers, J. D. (1985). Infection with varicella zoster virus after bone marrow transplantation. J. Infect. Dis. 152, 11721180. Lowry, P. W., Solem, S., Watson, B. N., Koropchak, C. M., Thackray, H. M., Kinchington, P. R., Ruyechan, W. T., Ling, P., Hay, J., and Arvin, A. M. (1992a).Immunity in strain 2 guinea pigs inoculated with vaccinia virus recombinants expressing varicella-zoster virus glycoproteins, I, IV, V, or the protein product of the immediate early gene 62. J. Gen. Virol. 73, 811-819. Lowry, P. W., Sabella, C., Koropchak, C. M., Watson, B. N., Thackray, H. M., and Arvin, A. M. (1992). Investigation of the pathogenesis of varicella-zoster virus infection in guinea pigs using polymerase chain reaction. J. Infect. Dis. 167, 78-83. Mahalingam, R., Wellish, M., Wolf, W., Dueland, A. N., Cohrs, R., Vafai, A., and Gilden, D. (1990). Latent VZV DNA in human trigeminal and thoracic ganglia. N. Engl. J. Med. 323,627-631. Margalit, H., Spouge, J. L., Cornette, J. L., Cease, K. B., DeLisi, C., and Berzofsky, J. A. (1987). Prediction of immunodominant helper T-cell antigenic sites from the primary sequence. J . Immunol. 138, 2213. Meyers, J. D., Flournoy, N., and Thomas, E. D. (1980). Cell-mediated immunity to varicella-zoster virus after allogeneic marrow transplant. J.Infect.Dis. 141,479-487. Miller, R. A. (1991). Aging and immune function. Znt. Rev. Cytol. 124, 187-194. Moffat, J. F., Stein, M. D., Kaneshima, H., and Awin, A. M. (1995). Tropism of varicellazoster virus for human CD4+ and CD8+ T-lymphocytes and epidermal cells in SCID-hu mice. J. Virol.69, 5236-5242, 1995. Myers, M. (1979). Viremia caused by varicella-zoster virus: Association with malignant progressive varicella. J. Infect. Dis. 140, 229-232.

ANN M. ARVIN et al. Myers, M. G., Stanbury, L. R., and Edmond, B. J. (1985). Varicella-zoster virus infection of strain 2 guinea pigs. J . Infect. Dis. 151, 106-113. Myers, M. G., Connelly, B. L., and Stanberry, L. R. (1991). Varicella in hairless guinea pigs. J . Infect. Dis. 163, 746-751. Nader, S., Bergen, R., Sharp, M., and Arvin, A. M. (1995).Age-related differences in cellmediated immunity to varicella-zoster virus among children and adults immunized with live attenuated varicella vaccine. J . Infect. Dis. 171, 13-17. Namikawa, R., Kaneshima, H., Lieberman, M., Weissman, I. L., and McCune, J. M. (1988). Infection of the SCID-hu mouse by HIV-1. Science 242, 1684-1686. Namikawa, R., Weilbaecher, K. M., Kaneshima, J., Yee, E. J., and McCune, J. M. (1990). Long term human hematopoiesis in the SCID-hu mouse. J . Exp. Med. 172, 1055-63. Ozaki, T., Ichikawa, T., Matsui, Y., Kondo, H., Nagai, T., Asano, Y., Yamanishi, K., and Takahashi, M. (1986). Lymphocyte-associated viremia in varicella. J . Med. Virol. 19, 249-253. Redman, R., Nader, S., Zerboni, L., and Arvin, A. M. (1995). Evaluation of inactivated varicella vaccine in patients receiving bone marrow transplantation. Abstract Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco. Fbthbard, J. B., and Taylor, W. R. (1988).A sequence pattern common to T cell epitopes. EMBO J . 7 , 93. Sabella, C., Lowry, P. W., Abbruzzi, G. M., Kinchington, P. R., Sadeghi-Zadeh, M., Ruyechan, W. T., Hay, J., and Arvin, A. M. (1993). Immunization with the immediate earlyhegument protein (ORF 62) of varicella-zoster virus protects guinea pigs against virus challenge. J. Virol. 67, 1613-1676. Sawyer, M. H., Wu, Y. N., Chamberlin, C. J., et al. (1992). Detection of varicella-zoster virus DNA in the oropharynx and blood of patients with varicella. J. Infect. Dis. 166, 885-888. Schmid, D. S. (1986). The human MHC restricted cellular response to herpes simplex virus type 1 is mediated by CD4+, CD8- T cells and is restricted to the DR region of the MHC complex. J . Immunol. 140,3610-3615. Schrier, R. D., Gann, J. W., Jr., Landes, R., Lockshin, C., Richman, D., McCutchan, A., Kennedy, C., Oldstone, M. B. A., and Nelson, J. A. (1989). T cell recognition of HIV synthetic peptides in a natural infection. J . Immunol. 142, 1166. Sharp, M., Terada, K., Wilson, A., Nader, S., Kinchington, P. E., Ruyechan, W., Hay, J., and Arvin, A. M. (1992). The kinetics and protein specificity of T-lymphocyte cytotoxicity in healthy adults immunized with live attenuated varicella vaccine. J . Infect. Dis. 165, 852-858. Shaw, S., and Biddison, W. E. (1979). HLA-linked genetic control of the specificity of human cytotoxic T-cell responses to influenza virus. J. Exp. Med. 149, 565. Straus, S., Reinhold, W., Smith, H., et al. (1984). Endonuclease analysis of viral DNA from varicella and subsequent zoster in the same patient. N . Engl. J . Med. 311,13261328. Thoman, M. L., and Weigle, W. 0. (1989). The cellular and subcellular bases of immunosenescence. Adu. Immunol. 46, 221-231. Torpey, 111, D. J., Lindsley, M. D. and Rinaldo, C. R., Jr. (1989).HLA-restricted lysis of herpes simplex virus-infected monocytes and macrophages mediated by CD4' and CD8+ T-lymphocytes. J . Zmmunol. 142, 1325-1332. Vafai, A., Wellish, M., and Gilden, D. H. (1988). Expression of varicella-zoster virus DNA in blood mononuclear cells of patients with postherpetic neuralgia. Proc. Natl. Acad. Sci. U.S.A. 85, 2161-2110. Watson, B., Keller, P. M., Ellis, R. W., and Starr, S. E. (1990). Cell-mediated immune

VARICELLA-ZOSTER VIRUS

309

responses after immunization of healthy seronegative children with varicella vaccine: Kinetics and specificity. J. Infect. Dis.162,794-799. Whitley, R. J. (1990).Varicella-zoster virus. In “Antiviral Agents and Viral Diseases of Man” (G. Galasso, R. Whitley, T. C. Merigan, eds.), pp. 235-264. Raven, New York. Wilson, A,, Sharp, M., Koropchak, C. M., Ting, S. F., and Arvin, A. M. (1992).Subclinical varicella-zoster virus viremia, herpes zoster and recovery of T-lymphocyte responses to varicella-zoster viral antigens after allogeneic and autologous bone marrow transplantation. J . Infect. Dis.165, 119-126.

This Page Intentionally Left Blank

ADVANCES IN VIRUS RESEARCH, VOL. 46

ANATOMY OF VIRAL PERSISTENCE: MECHANISMS OF PERSISTENCE AND ASSOCIATED DISEASE Juan Carlos de la Torre and Michael B. A. Oldstone The Scripps Research Institute Department of Neuropharmacology La Jolla, California 92037

I. Introduction 11. Requirements for Establishment of Viral Persistence A. Virus-Induced Changes in Cells Allowing Evasion from Immune Surveillance B. Nonlytic Strategy of Replication 111. Virus-Induced Alterations of Host Cellular Differentiated Functions in Absence of Cytolysis A. Experimental Evidence B. Growth Hormone Deficiency Syndrome Caused by Lymphocytic Choriomeningitis Virus Persistent Infection C. Virus-Induced Neuroendocrine Dysfunctions in Absence of Cytolysis and Inflammation D. Interactions between Virus and Cytoskeleton Affecting Cell Function IV. Conclusions References

I. INTRODUCTION Studies by Ivanovski and Beijerinck with the causative agent of the tobacco mosaic disease (Beijerinck, 1899;Ivanovski, 18991, together with investigations conducted by Loeffler and Frosch on the causative agent of the bovine foot-and-mouth disease (LoeMler and Frosch, 1898) during the closing years of the nineteenth century, provided the basis for defining viruses as subcellular entities that caused distinct forms of tissue destruction. A different mechanism by which tissue injury occurs during a viral infection was described for the first time in the early 1950s by Wallace Rowe on the basis of observations that lymphocytic choriomeningitis virus (LCMV) induced cellular destruction not directly caused by the infectious agent, but rather by the immunological response of the host against the virus (Rowe, 1954).Evidence accumulated during the years that followed to the present, as more viruses were discovered, firmly established that diseases with a proven viral cause were frequently associated with the hallmarks of cellular necrosis, tissue destruction, and inflammatory infiltrates. In many cases, 311

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

312 J U A N CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE the virally induced damage to the cellular and tissue architecture is specific enough to indicate the infectious agent involved. Examples are destruction of anterior horn neurons in poliomyelitis, the presence of Negri or Lyssa bodies in neurons with rabies, and cytomegalic cell clustering in cytomegalovirus infection. In contrast, other manifestations such as encephalitis may be caused by multiple viruses. When the classic histopathological hallmarks of cell necrosis and inflammation were absent, it was frequently assumed that the particular disease process was not caused by a virus. However, it is now clear that viruses can establish persistent infections in the absence of lymphoid cell infiltration and lysis of virally infected cells, the “classic” hallmarks of virus infection. Such viral infections can remain unnoticed because they are not associated with easily identifiable manifestations of acute infections (i.e., cell lysis and lymphoid infiltration). However, these persistent infections are relatively common and can frequently alter the function of the infected cells. Thus, a nonlytic persistent infection can be associated with the loss of a physiologic function of a cell such as synthesis of a hormone, cytokine, or neurotransmitter. This, in turn, depending on the specific cell type involved and the extent of the damage, can disrupt the host homeostasis and lead to disease, despite the fact that the infected cells and tissues maintain their normal anatomic architecture. Hence, disturbances in differentiated systems such as endocrine, immune, nervous, heart, and muscle in a variety of diseases currently of unknown

CLASSIC VIEW

:ELL NECROSIS AND INFLAMMATlOl

IMMUNE RESPONSE

DIRECT VIRAL

NEW CONCEPT PERSISTENT INFECTION WITHOUT HALLMARKS OF NECROSIS AND INFLAMMATION

4 4

EFFECTS ON DIFFERENTIATED FUNCTIONS

INFLAMMATION

CYTOLYSIS

\

r

H

CELWISSUE DAMAGE

DISTURBANCES OF THE HOST’S HOMEOSTASIS

FIG.1. New perspective on viral pathogenesis.

VIRAL PERSISTENCE

313

etiology may have a viral etiology (Oldstone, 1984,1985;de la Torre et al., 1991;Fig. 1). This chapter is concerned with review of data that support the conclusions stated above. First, we discuss some basic requirements for the establishment of viral persistence. Second, we describe the experimental evidence supporting the concept that viruses can perturb host cell differentiated functions in the absence of lysis of the infected cell and inflammation in the affected organ or tissue. Then, we detail examples of these kinds of virus-host interactions to illustrate the very diverse mechanisms viruses use to interfere with cell differentiated functions, and last we present some general remarks for future studies.

11. REQUIREMENTS FOR ESTABLISHMENT OF VIRALPERSISTENCE For persistence to occur, the virus must possess two essential characteristics. First, the virus by any of several mechanisms must escape TABLE I ESSENTIAL REQUIREMENTS FOR VIRALPERSISTENCE

I. Avoidance of immunological surveillance A. Removal of recognition molecules from surface of infected cells 1. Alteration of viral protein expression, antiviral antibody response inducing capping and modulation of viral gene expression 2. Alteration of MHC expression a. Direct viral effect on MHC expression b. Modulation of MHC expression by virally-induced cytokine production c. Negligible expression in cells that do not have potential for regeneration (i.e., neurons) B. Infection of thymus prior to or during immune system development (central tolerance or hyporesponsiveness) establishing persistent infection C. Infection of peripheral immune system (peripheral tolerance, energy, immunological ignorance) 1. Abrogation of lymphocytic, macrophage, and antigen-presenting function a. generalized immunosuppression b. Selective immunosuppression c. Lymphocyte/monocyte/macrophage/cytokine-inducedalteration of gene translation D. Generation of antibody or cytotoxic T-lymphocyte escape variants 11. Noncytolytic Replication A. Generation of viral variants 1. Defective interfering (DI) viral particles 2. Single or limited amino acid mutation B. Diminished expression of viral gene product

314 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE

from being recognized as foreign by the host’s immunological system. Second, the virus must adapt a strategy of nonlytic replication in the cells it infects (Oldstone, 1984, 1991; Table I). In addition, cells that support viral persistence must adopt strategies that prevent their usual suicidal tendency (apoptosis) when confronted with a viral infection.

A . Virus-Induced Changes in Cells Allowing Evasion from Immune Surveillance A main role of the immune system is the recognition of nonself foreign materials like bacteria and viruses. Once the materials are recognized as foreign, the host mounts immune responses, usually both humoral and cell-mediated, whose function is to eliminate the infectious agent and the cells they infect. Viruses that persist have evolved a variety of mechanisms to escape immune surveillance and thereby avoid clearance by the host immune response (Oldstone, 1991). One common strategy is that all known viruses that persist are also able to infect cells of the immune system (reviewed McChesney and Oldstone, 1987). Several studies of such virus infection of immune cells including measles human immunodeficiency virus (HIV), and cytomegalovirus (CMV) indicate that these infections frequently interfere with the function of the infected immune cell to make antibody or act as a cytotoxic lymphocyte but without lysing the infected cell (McChesney and Oldstone, 1989). In other instances virus infects specialized cells of the immune system whose main function is to present viral antigens to initiate an antiviral immune response that will then recognize such immunocompetent cells as foreign and destroy them. The ability to process and present foreign antigens can then be lost, and a generalized suppression of immunity follows. 1 . Avoidance of Recognition of Infected Cells by Specific Immune Response

Frequently, the immune response recognizes a virus-infected cell so it is a target for destruction. The requirements for the cellular and humoral arms of the immune system for recognizing virally infected cells as foreign are different. Antibodies are targeted primarily against viral structural components; thus, infected cells expressing viral glycoproteins on their surface are good targets for lysis by antibody plus complement or antibody-dependent cell-mediated cytotoxicity (ADCC). Altered glycosylation patterns of host proteins on the surface of the cell induced by the viral infection, can facilitate recognition by natural killer (NK) cells, which play an important role in early and nonspecific defense against infectious agents (Lopez, 1988; Biron

VIRAL PERSISTENCE

315

et al., 1989; Koszinowski et al., 1990). The cellular arm of the immune response, T lymphocytes, recognize structural and nonstructural regulatory viral proteins. However, in contrast to antiviral antibodies, T lymphocytes do not recognize viral components directly, but rather the complex between major histocompatibility complex (MHC) glycoproteins and defined small peptide fragments of viral proteins (Zinkernagel and Doherty, 1974; Bjorkman et al., 1987;Townsend et al., 1986). Such peptides have specific motifs that allow them to bind to a chemically defined groove on the surface of the MHC protein (Riitzschke et al., 1990; Van Bleek and Nathenson, 1990; Fremont et al., 1992; Matsumura et al., 1992). The presence of viral peptide-MHC complex on the surface of the infected cell allows T-cell recognition and activates cytotoxic T lymphocytes (CTL),leading to a series of events that destroy the infected cell. The CTL mediate destruction directly by the release of granules like perforin (Kagi, 1994) or indirectly by releasing cytokines or chemotaxis factor that attract other killer cells (i.e., activated macrophages) to the neighborhood (reviewed by Whitton and Oldstone, 1990; Zinkernagel and Doherty, 1974; Zinkernagel, 1977; Mims and White, 1984; Townsend et al., 1986). Viruses that persist must interfere with these various schemes of killing. Studies of persistent infections with arenaviruses, paramyxoviruses, retroviruses, and rhabdoviruses show a selective downregulation of viral glycoprotein expression at the cell surface as compared to other internal virus proteins. This, in turn, diminishes or aborts antibody recognition and destruction of virally infected cells (Buchmeier and Welsh, 1979; Oldstone and Buchmeier, 1982; Francis and Southern, 1988; Oldstone 1984, 1991). Further, antibody-induced modulation of viral antigens from the cell surface as has been documented to occur in measles virus infection both in tissue culture and in uiuo can also facilitate the establishment of viral persistence (Fujinami and Oldstone, 1984; Liebert et al., 1990). A similar phenomenon has been documented for alphavirus infections (Levine et al., 1991). Herpes simplex virus, like other viruses, employs a double-hit strategy. First, it hides in neurons, cells that normally do not express MHC antigens, and, second, one of its gene products prevents the neuron from being destroyed by apoptosis (Chou and Fbizman, 1992). Antigenic variation is another effective way for viruses to escape neutralizing antibodies. Well-documented examples of virus escaping neutralizing antibodies are the antigenic “shift” and “drift” exhibited by influenza viruses (Palese and Young, 1982; Webster et al., 1982) and the antigenic variation observed in persistent infections of lentiviruses (Clements and Narayan, 1984; Narayan et al., 1988). Avoidance of T-cell recognition utilizes other mechanisms. For exam-

316 J U A N CARLOS DE LA TORRE A N D MICHAEL B. A. OLDSTONE ple, many RNA and DNA viruses persist in cells whose constitutive levels of MHC expression are very low (e.g., neurons; Joly et al., 19911, whereas other viruses are able to interfere directly with antigen presentation and MHC expression. For example, certain adenoviruses early in infection synthesize an integral membrane protein, E3-gp19K. This protein is anchored to the endoplasmic reticulum through the C-terminal domain. E3-gp19K complexes with MHC class I molecules, preventing the MHC from being correctly processed and transported to the cells surface by inhibiting terminal glycosylation of MHC molecules (Wold and Gooding, 1991). Human cytomegalovirus encodes and synthesizes a MHC class I homolog polypeptide, UL18, that competes with native MHC for binding to &-microglobulin, a subunit of MHC class I molecules that is required for the transport of the viral peptideMHC complex to the cell surface (Browne et al., 1990). The HIV Tat protein directly downregulates MHC class I expression (Howcroft et al., 1993). Other viruses, like Epstein-Barr virus, interfere with the expression of cell-surface adhesion molecules, like LFA-3 and ICAM-1, that are necessary for efficient T cell-target interaction (Springer et al., 1987; Martz, 1987; Khanna et al., 1993; Sandrej et al., 1993). Thus, a picture evolves that the antigen-processing pathway is frequently targeted and disrupted by viruses which can facilitate the establishment of viral persistence. 2 . Block of Action of Nonspecific Antiviral Defense

Mechanisms on Infected Cells Viruses also have developed strategies to block several innate nonspecific antiviral effector mechanisms including interferons (IFN), complement lysis, and lysis mediated by tumor necrosis factor (TNF). Viruses counteract the IFN system by two general ways (Sen and Ransohoff, 1993). First, viruses can make products that directly block IFN signaling. This is illustrated by the anti-IFN activity of EIA proteins of adenovirus (Anderson and Fennie, 1987, Ackrill et al., 1991; Gutch and Reich, 1991; Kalvakolanu et al., 1991). Other examples of viral polypeptides that inhibit IFN action are the hepatitis B virus (HBV) polymerase and the Epstein-Barr EBNA-2 protein (Foster et al., 1991; Aman and von Gabain, 1990). The second mechanism is based on the ability of viruses to interfere with the action of IFN-induced enzymes, like the activity of the double-stranded (ds) RNA-dependent protein kinase (P68 kinase) (Sen and Lengyel, 1992). By preventing P68 kinase activation, a virus can block the phosphorylation of the cellular peptide chain initiation factor 2 (eIF-2a); in the presence of nonphosphorylated eIF-acw, initiation of the synthesis of viral polypeptides is more efficient.

VIRAL PERSISTENCE

317

Complement consists of multiple serum proteins, several of which combine to form antibody-viral antigen complex and enhance the activity of antibodies against viruses (reviewed by Oldstone, 1975). Viruses can encode proteins that interact with components of the complement pathway and thereby alter complement activation. For example, herpes simplex virus (HSV) expresses a glycoprotein, gC1, that binds C3b and modulates complement activation by accelerating the decay of CSbBb, the amplification convertase of the alternative complement pathway (Fries et al., 1986). The gC1 protein also prevents C5bC6initiated reactive lysis of C3b-bearing cells. The end result is that gC1 provides protection against complement-mediated viral neutralization (McNearney et al., 1987; Harris et al., 1990). In addition, the gE and gI HSV glycoproteins bind to the Fc domain of immunoglobulin G (IgG) (FcR activity). This HSV FcR activity causes “antibody bipolar bridging,” a process that may help to protect HSV and HSV-infected cells from those host immune defenses mediated by the Fc domain of the antiviral IgG response (Frank and Friedman, 1989; Dubin et al., 1991). Cytolysis of adenovirus-infected cells by TNF is prevented by the expression of E3-14.7K, E3-10.4K/14.5K, and E1B-19K viral polypeptides, which act by a yet to be defined mechanism at a step subsequent to TNF binding to its receptor. By a different means, the poxvirus Shope fibroma virus encodes a polypeptide, TR, that interferes with TNF activity by competitively inhibiting TNF binding to cell surface receptors (Gooding, 1992). 3. Induction of Suppression of Host Immune Response

Virus infection can be associated with a transient or long lasting suppression of the host immune response. By these means viruses can escape clearance and establish a persistent infection (Mims, 1986; Oldstone, 1984; Ahmed and Stevens, 1990). One mechanism by which this occurs is when the thymus is infected in utero or at birth. The thymus is responsible for the proper development of the immune system. To discriminate self from nonself (foreign), self-antigens are bound to the MHC molecules in thymic cells; developing T cells in the thymus that recognize the self-peptide-MHC complex are selected against and destroyed. This process is called negative selection and ensures that antiself T lymphocytes will not enter the peripheral lymphoid organs where they might be stimulated by a self-antigen to differentiate and respond by generating a CTL response against one’s own cells or the generation of anti-self antibodies. Thus, negative selection prevents autoimmune responses and disease. When a virus infects thymic cells, viral antigen is presented, like self-antigens, and precursor immune cells that would normally respond to that particular virus are deleted.

318 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE Thus, virus escapes detection in infected cells because of the lack of precursor cells available in spleen and lymph nodes to respond to the virus. This mechanism is called central tolerance but rarely is complete; although most high-affinity immune cells are removed, some may escape. In addition, low-affinity cells usually pass to the periphery. Thus, immune responses (specifically antibodies) are usually generated during congenital (in utero) or neonatal viral infections. T cells not selected against in the thymus pass to the periphery, poised to respond to any conceivable foreign challenge. The generation of a specific immune response requires the participation of both effector cells, T and B lymphocytes, as well as antigen-presenting cells, such as macrophages and dendritic cells. When a virus infects and kills or interferes with the function of one or several of these immune system cells, peripheral tolerance occurs. Immune suppression can result from destruction of infected immune cells as a direct consequence of virus multiplication, by the antiviral immune response directed against infected cells, or by an autoimmune response induced following virus infection (Weiss, 1993; Odermatt et al., 1991). Suppression also occurs when virus-infected cells of the immune system are not destroyed but the virus disturbs their physiological function. Examples of the latter mechanism are persistent infection of B-cell hybridomas with LCMV leading to a decreased production of immunoglobulin (McChesney and Oldstone, 1987) and the interference in the killing activity of T and NK cells caused by infection with measles virus or cytomegalovirus. Frequently, when dysfunction of or killing of antigen-presenting cells occurs the suppression is general. This is the mechanism of suppression seen with an immunosuppressive variant of LCMV (Borrow et al., 1994), and is likely the basis of suppression associated with HIV (Fauci, 1994). The type of immune response triggered by the virus can also contribute to alterations in the functioning of the immune system. For example, virus-mediated hyperstimulation of T cells can lead to apoptotic cell death (Moskophidiset al., 1993),resulting in generalized or specific immune suppression. Similarly, preferential activation of Th2 or T h l T cells by a virus can influence the cytokine expression pattern and thus also the immune response. Th2 activation leads to depression of cell-mediated immunity and favors antibody formation; in contrast, Thl favors activation of CTL, NK cells, and macrophages. Preferential Th2 activation has been proposed to contribute significantly to the long-term depression of cell-mediated immunity associated with measles virus infection (Griffin and Ward, 1993). Similarly, this preferential Th2 activation is also thought to be an important factor of the many proposed to contribute to the immunosuppression caused by HIV

VIRAL PERSISTENCE

319

in humans (Weiss, 1993). Other viruses may interfere with the cytokine expression patterns by encoding polypeptides with cytokine homologies. For example, Epstein-Barr virus encodes the BCRFl polypeptide, which is a homolog of the inhibiting cytokine interleukin (IL) 10 (Hsu et al., 19901, whereas poxviruses encode both B15R, a protein which binds to IL-1 and inhibits its functions, and clmA, a serpin-like protease which prevents cleavage of IL-lP to its active form, thus inhibiting its immune stimulatory functions (Gooding, 1992).

B . Nonlytic Strategy of Replication To establish a persistent infection successfully, the regulation of the replication and expression of the viral genome should allow for the survival of a critical number of cells, by not interfering with vital functions (i.e., housekeeping genes like respiratory enzymes) required for the survival of the infected cell. Certain viruses have evolved exquisitely to establish persistent infections. For example, most often arenaviruses are characterized by their exclusively noncytolytic phenotype. This makes them well-suited for persistence, and, in fact, most tissue culture cells infected by arenaviruses as well as the natural host, the mouse, develop life-long chronic infections (Buchmeier et al., 1980; Lehmann-Grube, 1984). Viruses that are usually lytic overcome this drawback by either regulating their gene expression, generating variants with nonlytic phenotypes, or generating deletion mutants termed defective interfering (DI) particles that possess the ability to modulate the virulence and pathogenesis of wild-type virus (Barrett and Dimmock, 1986; Roux et al., 1991; Holland, 1992). Productive infection of a cell by herpes simplex virus (HSV) is usually accompanied by cell death. However, by restricting its gene expression HSV is able to establish a persistent infection in neurons (Roizman and Sears, 1987; Stevens, 1989). Similarly, restricted gene expression of cytomegalovirus (CMV) accounts for its ability to cause persistent infection of undifferentiated cells, whereas differentiated cells do not restrict CMV gene expression, and are lysed by the virus infection (Dukto and Oldstone, 1981; Gonczol et al., 1984). Studies on persistent virus infections of tissue culture cells have shown that viral variants can modulate the lytic phenotype of normally cytocidal wild-type viruses (Mahy, 1985; Wechsler and Meissner, 1982; Younger and Preble, 1980; de la Torre et al., 1985). Analysis of these variants has allowed in several instances the viral genes involved in the establishment of persistence t o be defined. Further, such analysis revealed that a single or a limited number of specific changes are responsible to establish and maintain persistence by viruses that

320 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE normally are cytolytic. For example, the low virulence and attenuation displayed by high-passage stocks of reovirus, ordinarily a lytic virus, is due to mutations in the a 3 protein (Ahmed and Fields, 1982),whereas changes in the a1 protein, the viral hemaglutinin, is involved in the maintenance of persistence (Kauffman et al., 1983). Persistence can also result as a consequence of cycling infection, where only a small fraction of cells are infected and killed at any particular time and the progeny virus continue to infect only a small number of permissive cells. Persistence of lactate dehydrogenase virus (LDH) in mice (Mahy, 1985) and adenovirus in humans (Porter, 1985) has been postulated by these authors to be examples of this type of persistence infection. Defective interfering particles are subgenomic deletion mutants generated from a full-length virus genome, often by replicate errors (Holland, 1990; Roux et al., 1991). The DI particles interfere strongly with virus replication, and thus they frequently exhibit a direct cellsparing effect. The particles can also exhibit indirect protective effects, such as induction of IFN and other cytokines and alteration of immune responses. In addition, DI particles can promote reduction of early virus yields or decreased expression of viral antigens at the cell surface. This, in turn, likely facilitates the establishment of persistent infection by decreasing the levels of viral antigens that would otherwise allow recognition and clearance by the host immune response. In other cases, DI particles allow immune abrogation of infection (Barrett and Dimmock, 1986; Roux et al., 1991). Defective interfering particles and satellite RNAs of plants have been shown to modulate the outcome of infection. They cause both reduction of symptoms as well as mild to profound intensification of disease severity (Kaper and Collmer, 1988; Simon, 1988). The DI genomes and other defective genomes can also influence the outcome of animal virus infections. Prevention of lethal infections and their conversion to chronic disease or long-term persistent infections with or without clinical signs have been demonstrated in several animal models. Moreover, generation of DI particles can contribute to inducing more subtle virally induced disturbances of the host biology, such as neurochemical alterations and effects on the immune responses (Barrett and Dimmock, 1986; Holland, 1987; Huang, 1988; Roux et al., 1991).It should be mentioned that host cell type can be a major determinant for the biological effects of DI particles. Variability of DI particle effects within different cell types in uiuo can hinder the assessment of DI genome influences on disease processes (Huang, 1988; Brinton et al., 1984; Cave et al., Kang et al., 1981; Holland, 1987; Gillies and Stollar, 1980; Barrett and Dimmock, 1986).

VIRAL PERSISTENCE

321

111. VIRUS-INDUCED ALTERATIONS OF HOSTCELLULAR DIFFERENTIATED FUNCTIONS IN ABSENCE OF CYTOLYSIS

A . Experimental Evidence Experimental evidence accumulated by virologists studying different viruses indicate that viral infections are generally characterized by cell destruction and tissue morphological injury which is frequently associated with immune cell infiltrates. Until relatively recently it was assumed that virus-induced disease was a consequence of this structural damage, which could be directly caused by the effects of replication and expression of the virus genome or could be mediated by the host immune response directed against infected cells (Fenner et al., 1974;Notkins, 1975). The first indication that this situation might not always be so clearcut was provided by the observation in the mid 1970s that persistent infection of murine neuroblastoma cells with LCMV significantly lowered the intracellular levels of choline acetyltransferase (CAT) and acetylcholine esterase (AChE),enzymes that participate in the synthesis or degradation of the neurotransmitter acetylcholine (Oldstone et al., 1976).Despite these alterations, LCMV-persistently infected neuroblastoma cells produced normal levels of total RNA and protein, as well as vital enzymes. Furthermore, the persistently infected cells were morphologically indistinguishable from control uninfected neuroblastoma cells. Conceptually similar results were provided during studies of Rous sarcoma virus (RSV) infection of myotubes, chondroblasts, and melanoblasts (Holtzer et al., 1975; Roby et al., 1976). Differentiated chick chondroblasts, myotubes, and melanoblasts infected with a temperature-sensitive variant of RSV produced normal levels of differentiated cell products during infection at nonpermissive temperature for viral replication; however, at temperatures that allowed virus replication, chondroblasts stopped manufacturing sulfated proteoglycans, and melanoblasts did not make melanin. In contrast to the observations with LCMV-persistent infection of neuroblastoma cells, these alterations of cell differentiated functions caused by RSV infection were associated with morphological changes, but both the vital functions of the cell and the cell survival rate remained unchanged. Over the years, studies conducted with a variety of RNA and DNA viruses affecting humans and other species provided conclusive evidence that many viruses have the ability to establish persistent infections, both in tissue culture and in uivo, and cause alterations in the

322 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE differentiated functions of a cell without perturbation of its vital functions. Depending on what specific differentiated function of the host is affected, and its degree of alteration, different pathological manifestations will be observed. A compilation, by no means complete, of this kind of virus-host interactions is presented in Table 11. Common to all the examples listed is the observation that, despite replication and expression of the viral genome, the infected cells and tissues remain free of structural damage. Examples listed in Table I1 include different viruses as well as target cells with remarkably different biological properties, suggesting that very different mechanisms are likely operating in each particular virus-host interaction. The molecular mechanisms by which most of these different viruses persist and interfere with cell differentiated functions in the absence of cytolysis remain largely unknown. The growth hormone (GH) deficiency syndrome associated with LCMV persistent infection in the mouse represents an exception. Here, a significant amount of information has been obtained that provides a framework for understanding the molecular basis whereby viruses can subtly interfere with specific host cellular differentiated functions leading to disturbances in host homeostasis and disease.

B. Growth Hormone Deficiency Syndrome Caused by Lymphocytic Choriomeningitis Virus Persistent Infection Lymphocytic choriomeningitis virus, the prototype member of the arenavirus family (Fenner, 19761, has been extensively studied as a paradigm for understanding the mechanisms and biological consequences of viral persistence (Buchmeier et al., 1980; Lehmann-Grube, 1984; Oldstone, 1991). The virus has a negative single-stranded RNA genome composed of two segments, L and S, with an ambisense coding strategy. Transcription and replication of the viral genome occur in the cytoplasm and involve a viral encoded RNA-dependent RNA polymerase. The S segment of approximately 3.4 kb codes for the nucleoprotein (NP) and the glycoprotein (GP), which is processed into GP1 and GP2 (Riviere et al., 1985a).The L segment of approximately 7.2 kb codes for the viral polymerase of L protein and a small polypeptide termed Z whose function is presently unknown (Salvato and Shimomaye, 1989). The GP and Z gene products are coded in the RNA genome polarity, whereas L and NP are coded in the antigenomic polarity. The virus can infect its natural host, the mouse, either acutely or persistently, depending on the immune status of the mouse, the route of infection, and the strain of virus used. Infection of C3H/ST mice at

VIRAL PERSISTENCE

323

birth with LCMV Armstrong strain leads to the establishment of a persistent tolerant infection that is associated with a retarded growth syndrome manifested as decreased body weight and length (Oldstone et al., 1982, 1984). This is frequently accompanied by hypoglycemia. Despite high levels of virus replication in GH-producing cells of the anterior pituitary, there is no evidence of structural damage or inflammatory infiltrates, yet production of GH mRNA and protein is significantly diminished (Oldstone et al, 1982, 1984; Rodriguez et al., 1983a; Valsamakis et al., 1987; Klavinskis and Oldstone, 1989). The role of GH in growth is mostly mediated through insulin-like growth factors (IGF), and levels of IGF-1 in serum of LCMV persistently infected mice are significantly reduced, whereas cortisol and insulin serum levels are normal (Oldstone et al., 1984). The central role played by GH in this virally induced metabolic disorder was demonstrated by the restoration of normal growth in LCMV persistently infected mice treated with exogenous GH (Oldstone et al., 1984). Results from detailed studies on the biochemistry and genetics of the LCMV-induced GH deficiency syndrome are summarized in Table I11 (Oldstone et al., 1985; Riviere et al., 198513; Valsamakis et al., 1987; Klavinskis and Oldstone, 1989; Tishon and Oldstone, 1990).However, the complex physiological regulation of GH biology, involving the immune hypothalamicpituitary-adrenal axis (Bateman et al., 19891, made it difficult to determine whether the reduction in GH synthesis was directly caused by LCMV replication within the somatotrope cells. Studies conducted on a tissue culture model using a cell line derived from anterior pituitary (PC cells) that expresses GH and PL (Chomczynski et al., 1988) provided valuable information regarding the molecular mechanisms whereby LCMV persistent infection turns down GH mRNA synthesis without impairment of cellular vital functions. Steady-state levels of GH mRNA in LCMV-infected PC cells was significantly reduced, but no differences were observed between uninfected and LCMV-infected cells in the mRNA levels of several housekeeping cellular genes including actin, cyclophylin, and glyceraldehyde-phosphate dehydrogenase mRNA (de la Torre and Oldstone, 1992). Several mechanisms might underlie a reduction in the steadystate level of a specific host cellular mRNA occurring as a consequence of a viral infection. These include (i) decrease in the half-life of the particular mRNA, which could be due to an active process of virusmediated mRNA degradation or an indirect mechanism triggered by the viral infection; (ii) interference by the viral infection with the nucleocytoplasmic transport of the mRNA; (iii) altered processing of pre-mRNA into mature mRNA; and (iv) direct interference with the initiation of transcription of cellular genes. There are documented

TABLE I1 VIRALINFECTIONS INTERFERING WITH DIFFERENTIATED CELL.FUNCTION LEADING TO DISORDERED HOMEOSTASIS AND DISEASE Syndrome phenotype

w

52

A. Tissue culture Decreased acetylcholinesterase and choline acetyltransferase Altered synthesis of melanin, mucosin, and sulfated proteoglycan Alered lipid metabolism

Cell tissue target

Virus

Reference

Neuroblastoma

Lymphocytic choriomeningitis virus (LCMV)

Oldstone et al. (1976)

Melanoblastoma, myoblastoma, chondroblasts

RSVkhicken cells

Holtzer et al. (1975), Roby et al. (1976)

BGM cells

Measles virus (MV)

Changes in phospholipid methylation

Glioma cells

Canine distemper virus (CDV), MV

Altered NK activity and immunoglobulin synthesis Altered MHC expression

NK and B cells

Cytomegalovirus (CMV), MV

Wild et al. (1981), Anderson et al. (1983) Balbach and Koschel (1979), Koschel and Muntel (1980), Muntel and Koschel(1982) Casali et al. (1984)

MHC expressing cells susceptible to adenovirus Somatotropes

Adenovirus

Wold and Gooding (1991)

CMV Moloney leukemia virus (MMuLV)

Krajcsi et al. (1992) Schrier and Oldstone 1986)

Diminished CTL activity Altered hormone secretion

Fibroblasts

Foot-and-mouth disease virus (FMDV)

de la Torre et al. (1988)

Growth retardation: reduced growth hormone (GH) mRNA and protein levels Obesity: reduced levels of norephinephrine and dopamine Deformed whiskers

Growth hormone-producing cells in anterior pituitary

LCMV/mouse

Oldstone et 01 (1982, 1984)

Hypothalamus involvement

CDVlmouse

Lyons et al. (19821, Bernard

Unknown

&we (1983)

Diabetes: Hyperglycemia, abnormal glucose tolerance test

p-cells in islets of Langerhans

Hypothyroidism: reduced levels of T3 and T4, decreased levels of thyroglobulin mRNA Neurochemical and neurobehavioral alterations

Thyroid follicular cells

MCF-recombinant virus/mouse Venezuelan encephalitis virus (VEV)/hamster, LCMV/mouse LCMV/mouse

Semliki forest virus (SFV)/mouse, LCMV/mouse, Rabies virus (RV)/rat/mouse, Borna disease virus (BVDVrat

Barrett et al. (1986), Hotchin and Seegal (1977), Lipkin and Oldstone, Gold et al. (19941, Tsiang (1992). Dittrich et al. (19891, Bautista el al. (1994)

Changes in cell morphology and cell gene expression

B. In vivo

u1 N

et al.

Neurons, astrocytes, other CNS cell types

(1983)

Rayfield et al. (19811, Oldstone et al. (1984) Klavinskis and Oldstone ( 1986)

326 J U A N CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE TABLE I11 CHARACTERISTICS OF GROWTH HORMONE SYNDROME CAUSED BY LYMPHOCYTIC CHORIOMENINGITIS VIRUS Comment

Parameter Growth development

50% reduction in size

Rate of survival (30th day)

5% LCMV-infected mice; 95% uninfected mice

Hypoglycemia

Fivefold lower blood glucose level

Biochemistry

Decrease in GH mRNA and GH protein levels in the pituitary gland

Viral replication

LCMV replicated in GH-producing cells, but no structural damage detected

GH effect

Adoptive transfer of GH-producing cells to LCMV persistently infected mice restores wild-type phenotypes

Host genetics

No linkage to MHC haplotype or sex; multiple genes involved

examples of viral infections affecting all these steps (Schroder et al., 1988; Inglis, 1982; Black and Lyles, 1992; Beloso et al., 1992). Investigation of the biology of GH mRNA in uninfected and LCMVinfected PC cells revealed that LCMV infection specifically interferes with the initiation of GH transcription. Run-on experiments indicated that the decrease in GH mRNA steady-state level correlated with a reduction in the initiation of GH transcription, whereas the half-life and rate of nucleocytoplasmic transport of GH mRNA were similar in LCMV-infected and uninfected PC cells. Moreover, studies of the expression of the reporter gene chloramphenicol acetyltransferase (CAT) under control of the GH promoter in uninfected and LCMV-infected PC cells revealed that LCMV infection caused a significant decrease in CAT activity when the reporter gene was expressed under control of the GH promoter (GH-CAT) (de la Torre and Oldstone, 1992; Fig. 2). This decrease was not due to a general and nonspecific impairment of cellular transcription but was specific for the GH promoter, as illustrated by the similar levels of CAT activity obtained in infected and uninfected cells when expression of the same CAT reporter gene was driven by either a cytomegalovirus (CMV) immediate early promotor or simian virus (SV40) promoter. Results from dose-response studies using increasing amounts of the GH-CAT plasmid DNA indicated that a decrease in the amount of

VIRAL PERSISTENCE

327

functionally active factors involved in GH promoter activation likely mediated the impairment of GH transcription caused by LCMV infection (de la Torre and Oldstone, 1992). GHFl (Pit l), a tissue-specific Pou domain transcription factor, is known to be required for activation of GH, Prolactin (PL), as well as GHFl promoters (Lefevre et al., 1987; Dana and Karin, 1989; Bodner and Karin, 1987; Bodner et al., 1988; Lie et al., 1990; Fox et al., 1990; Dalle et al., 1990; Chen et al., 1990; Ingraham et al., 1988; Mangalam et al., 1989; McCormick et al., 1988). Moreover, expression of GHFl in HeLa cells, which do not express GH, was sufficient to allow expression of a cotransfected reporter gene under the control of the GH promoter (Ingraham et al., 1988; Mangalam et al., 1989). Interestingly, levels of GHFl protein were decreased in LCMV-infected cells. Furthermore, LCMV infection severely limited trans-activation of GHp-CAT by GHFl in HeLa cells. The basal transcriptional activity of a GH promoter lacking GHFl binding sites

FIG.2. Infection with LCMV causes a specific decrease i n the GH promoter activity. Uninfected o r LCMV-infected PC cells (GH+,PL+) were transfected with the CAT reporter gene under control of the indicated promoters. Expression of CAT under control of the GH promoter, but not under control of the SV40 or CMV promoters, was significantly reduced in LCMV-infected PC cells.

328 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE was not affected by LCMV infection (de la Torre and Oldstone, 1992). Extracts from LCMV-infected cells provided approximately 10-fold less GH promoter activity than extracts from uninfected PC cells when compared in in uitro transcription assays. In contrast, extracts from uninfected or infected PC cells provided similar levels of CMV promoter activity. In addition, extracts from uninfected PC cells, but not from HeLa cells, which do not produce GHF1, were able to complement in trans extracts from LCMV-infected PC cells and restored normal levels of GH promoter activity (de la Torre and Oldstone, 1992). These results indicate that replication of LCMV in GH-producing cells leads to a decrease in the amount of functionally active GHF1, resulting in lower levels of GHFl transcription. This, in turn, causes a further decrease in levels of functional GHF1. The decrease in levels of functional GHFl protein eventually causes a decrease in the transcriptional activity of the GH promoter. C3H/ST mice persistently infected with LCMV have normal PL levels despite the fact that GHFl has been implicated in PL expression. This can be explained by the lack of virus replication in PL cells in uiuo (Oldstone et al., 1984). The impairment in GH transcription requires replication and/or expression of the LCMV RNA genome because viral infection of PC cells with UV-inactivated LCMV does not affect GH transcription, and PC cells cured of the LCMV infection by ribavirin therapy exhibited normal GH mRNA levels (de la Torre and Oldstone, 1992). Genetic studies using different mouse strains and strains of LCMV that cause or do not cause the GH deficiency syndrome revealed a strict correlation between the ability of the particular LCMV strain to replicate in GH-producing cells of the anterior pituitary and the development of the GH deficiency syndrome (Oldstone et al., 1985; Buesa et al., 1994; Teng et al., 1994).These studies also showed that not only viral factors, but also the particular genetic makeup of the corresponding host determines the pathological consequences of a virus persistent infection. The use of reassortant viruses between LCMV Armstrong strain, which causes the GH disease, and LCMV WE strain, which does not cause the disease, mapped the ability to cause the GH deficiency syndrome to the S segment of the LCMV genome (Riviere et al., 1985b). The S segment codes for two viral gene products, NP and GP. The contribution individually made by each of these two viral gene products was assessed using recombinant vaccinia viruses expressing either NP (VV-NP) or GP (VV-GP)to infect PC cells. Initiation of GH transcription was significantly reduced in VV-NP-infected PC cells but not in PC cells infected with VV-GP or with the control vaccinia recombinant virus VV-SC11 that expressed p-galactosidase. In contrast, the initiation of transcription of two other cellular mRNAs, actin and cy-

329

VIRAL PERSISTENCE

clophylin, was similar in PC cells infected with VV-NP, VV-GP, or VVsc11. These results suggest a key role of LCMV-NP in the effect on GH transcription caused by LCMV infection. However, other genetic studies indicate that the viral GP also plays an essential role in determining the ability of LCMV to cause the GH deficiency syndrome (Buesa et al., 1994; Teng et al., 1994). Closely related LCMV-WE clones isolated from the same WE parental clonal-pool population (disease nil) displayed differences in abilities to replicate in GH-producing cells and thus to cause or not cause the GH deficiency syndrome. Sequence analysis revealed that a single specific amino acid change (Ser Phe) in position 153 of the GP1 determines whether the virus will be able to replicate in GH-producing cells in the anterior pituitary and induce the GH deficiency syndrome. The parental LCMV-WE (disease nil) does not replicate in GH-producing cells; it has a Ser in position 153, whereas those LCMV-WE variants containing Phe at amino acid position 153 in GP1 replicate in GH-producing cells and cause the GH deficiency syndrome (disease positive). These results illustrate how different viral determinants play a role in the ability of a virus to display a particular disease phenotype, in this instance the GH deficiency syndrome. A scenario for the interaction between LCMV and GH-producing cells leading to the impairment of GH transcription is described in Fig. 3. To cause the GH disease, LCMV has first to recognize and enter the target cell, in this case the GH-producing cells. There is experimental evidence supporting GP1 as the ligand that interacts with the LCMV receptor localized at the cell surface (Borrow and Oldstone, 1992). Thus, a LCMV carrying a GP1 that is unable to recognize the corresponding viral cellular receptor will not be competent to infect the cells. Once LCMV replicates within the GH-producing cells, the viral NP synthesized in the infected cells is likely to be responsible for the interference with GHFl leading to a decrease in GH transcription. The exact mechanism of this interference is presently unknown. However, the steady-state level of GHFl mRNA expressed under control of the RSV promoter in HeLa cells was not affected by LCMV infection, suggesting that the nucleocytoplasmic transport and stability of GHFl mRNA are not perturbed during LCMV infection. These findings favor the notion that the LCMV effect on GH transcription is caused by interfering with GHFl protein activity. It is worth emphasizing that virus variants with the ability to cause the GH deficiency syndrome are hidden within the wild-type LCMVWE clonal population (GH-disease nil). This finding illustrates an important biological concept that applies to most, if not all, RNA viruses.

-

330 JU A N CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE

FIG.3. Interaction between LCMV and GH-producing cells leading to decreased levels of functionally active GHF-1 protein. %cognition of LCMV and entry into GH target cells is mediated by interaction between the virus receptor (Rc) at the cell surface and the virus GP1. Viruses carrying a specific version of GPl(Y) are unable to recognize and infect GH-producing cells. A single specific amino acid change in the virus g p l ( 0+ 0 )may allow LCMV to interact correctly with Rc present at the surface of GH-producing cells. Within the infected cells, the LCMV NP (U) gene product may interfere with GHFl (0)activity by blocking its translocation to the nucleus or by presenting posttranslational modifications (+) required for GHFl activity. The virus NP could also complex with GHF1, rendering GHFl nonactive. Reduction in levels of nuclear functionally active GHFl protein will cause a decrease in GHFl transcription. This, in turn, will cause a further decrease in GHFl protein levels, which will lead to reduced GH promoter transcriptional activity and corresponding reduction in GH synthesis.

Small genomic size and high mutation frequency, facilitated by the error-prone nature of RNA-dependent RNA polymerases, mandate that even a clonal pool of RNA virus will consist of a complex mixture of related mutants differing from one another at one, two, or several nucleotide positions in the genome, a population structure termed quasi-species. This, in turn, determines that within quasi-species swarms, variants with unique biological properties may occasionally be generated but be hidden within the population (Holland, 1992; de la Torre and Holland, 1990).During bottleneck transmissions, these variants may have the opportunity to establish productive infections with pathological manifestations that can differ significantly from those

VIRAL PERSISTENCE

33 1

previously observed with the particular infectious agent. In addition, these variants can be selected out and dominate the population as a consequence of changes in the tissue and cellular environment due to variations in the physiobiolgical and immunological status of the host. Thus, the concept advanced by many virologists and physicians that a particular RNA virus will generally cause a highly specific disease, while generally true, requires a good deal of flexibility in thought. Single or very few mutations may cause profound alterations in virulence, pathogenesis, or host cell tropism (Holland, 1992). Viral infections, both acute and persistent, can be viewed as a succession of invasions modulated by the response of the host. The acute and chronic effects of infections will differ not only because of the genetic, physiological, and immunological differences among hosts, but also because each host experiences a unique array of quasi-species challenges during an infection. These concepts are frequently overlooked because disease syndromes are often similar for each type of virus. Furthermore, variants with remarkably different biological phenotypes regarding their effects on host biology can maintain the same biochemical and antigenic characteristics, creating additional difficulties for the clinician who relies on immunobiochemical test results to determine whether a particular virus is responsible, or associated with, the disease.

C . Virus-Induced Neuroendocrine Dysfunctions in Absence of Cytolysis and Inflammation Failure to thrive and the wasting syndrome associated with a growth dysfunction have been described during rabies virus (RV) and feline leukemia virus (FeLV) infections (Torres-Anjel et al., 1988).Studies on these systems indicate the participation of the pituitary/hypothalamic /thymic axis (PHTA) in the pathogenesis of RV and FeLV infections. However, contrary to the previously described GH-deficiency syndrome caused by LCMV, very significant cell and tissue damage is found during RV and FeLV infections (Torres-Anjel et al., 1988). Moloney murine leukemia virus (M-MuLV) has been shown to modulate hormonal secretion in somatotrope cells (Nsiah and Turner, 1990). Perhaps more relevant to the topic discussed in this review is the decreased growth, both length and weight, exhibited by rats infected at birth with Borna disease virus (BDV),which develop a persistent tolerant infection (BDV-PI). The BDV-PI rats have levels of GH, both mRNA and protein, as well as IGF-1, similar to those found in sex- and age-matched uninfected controls. Preliminary studies have suggested that alterations in self-feeding behavior that take place in the absence

332 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE of overt histopathological symptoms are responsible for the decreased growth observed in BDV-PI rats (Bautista et al., 1994). Table I1 illustrates several neuroendocrine dysfunctions found to be associated with persistent viral infections occurring in the absence of significant cell and tissue structural damage. The consequences of persistent virus infections of the @ cells of the islets of Langerhans deserve comment. A number of viruses induce diabetes in animal models. These have been well characterized and frequently present many similarities to human insulin-dependent diabetes mellitus (IDDM) (Yoon, 1991). Some viruses are able to cause persistent infection in the absence of any significant structural damage of the pancreas. For example, after inoculation of Golden Syrian hamsters with the TC-83 vaccine strain of Venezuelan encephalitis virus (VEV), a sustained diminution in glucose-stimulated insulin release from @ cells develops. This endocrine dysfunction is associated with the VEV persistent infection of @ cells, which occurs in the absence of cytolysis and inflammation (Rayfield et al., 1981).It appears that VEV infection may interfere with @ cell membrane adenylate cyclase, calmodulin, and/or islet cAMP generation. On the basis of the relationship of glucose and cAMP to insulin release, this could contribute to the abnormalities in insulin release observed in this system. The LCMV persistent infection of @ cells of the islets of Langerhans is also associated with chemical evidence of diabetes indicated by hyperglycemia and abnormal glucose tolerance test (Oldstone et al., 1984). Viral antigens and budding of virus are identified predominantly in @ cells in the islets of Langerhans, but no structural abnormalities of the infected cells are noted. Concentrations of cortisol and IGF-1 in serum are normal, whereas insulin levels are normal or low. The strains of mice used for these studies SWRIS, BALB, and C57BL/6, do not allow replication of LCMV in GH-producing cells. Consequently, LCMV-infected mice have normal GH levels. The virusinfected islet cells showed normal anatomy, and neither cytolysis nor inflammatory infiltrates were observed (Rodriguez et al., 1985). This picture bears similarities to the early stages of human adult-onset diabetes mellitus. Epidemiological studies have implicated different viruses as etiological or contributor factors in human IDDM. Viral candidates include coxsackie B3 and B4 (Yoon et al., 1979),mumps (Sultz et al., 1975; Gamble, 1980), and perhaps retrovirus type 3 (Yoon, 1991).In addition, congenital rubella syndrome is associated with a high incidence of IDDM (McEvoy et al., 1988). Human cytomegalovirus virus (HCMV) sequences have been found in pancreatic tissues of patients with noninsulin-dependent diabetes mellitus, suggesting a possible association of HCMV with type I1 diabetes (Lohr and Oldstone, 1990).

VIRAL PERSISTENCE

333

The central nervous system (CNS) is an environment ideally suited for the establishment of virus persistent infections (Johnson, 1982; ter Meulen et al., 1984; Kristensson and Norrby, 1986). This is because (i) metabolic features of neuroectodermal cells exert special host-cell restrictions on virus replication; (ii) a complex network exists of intercellular connections with unusual separation of one cell from another (connections by long axons or dendritic processes), with the connections functioning as highways for virus traffic and facilitating spread of the infection over long distances; and (iii) cells of the CNS are protected by the blood-brain barrier from peripheral environmental influences. T lymphocytes do cross the blood-brain barrier, but CNS cells like astrocytes and oligodendrocytes express low levels whereas neurons normally express nondetectable levels of MHC I class molecules, which are required for CD8+ T-cell recognition and activation. Several lines of experimental evidence indicate that these virus persistent infections of the CNS can induce slowly progressive neurological disorders, which are associated with diverse neurochemical abnormalities and pathological findings that not uncommonly occur in the absence of encephalitis and only minor or nonexisting cell necrosis within the CNS (Gilden and Lipton, 1989; ter Meulen, 1991; Mohammed et al., 1993; Lipkin et al., 1988). Rabies virus causes a slowly progressive infection of the brain, with behavioral disturbances frequently being the major or only disease sign (Warrell, 1976; Tsiang, 1992). The virus is found widespread in brain necropsy samples, but neuronal destruction and inflammatory cell infiltrations are uncommon or negligible, suggesting that rabies severely disturbs the normal functioning of structurally intact neurons (Tsiang, 1992). With rabies infection, death follows. However, many persistent infections perturb CNS functions without causing death of the infected host. For example, inoculation of newborn rats with the neurotropic agent BDV causes a persistent tolerant infection (PT-NB). The picture is one of an absence of an inflammatory response but the presence of distinct and reproducible behavioral and physiological alterations (Dittrich et al., 1989; Carbone et al., 1991; Bautista et al., 1994). These include diminished taste (especially to salt), impaired learning abilities, loss of spatial discrimination, enhanced activity in open field tests and mazes, and reduced growth despite normal levels of GH and IGF-1. Interestingly, when tree shrews, a subhuman primate, are persistently infected with BDV, they develop marked alterations in their normal expected social behaviors; these examples include the loss of normal male and female roles, disturbances in the breeding behavior of females, and improper social adjustments (Sprankel et al., 1978). Perinatal infection of rats with BDV interferes with normal development of the cerebellum and hippo-

334 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE campus, resulting in a significant reduction of granular cells (Carbone et al., 1991). Borna disease virus appears to interfere with granular cell division and migration, resulting in a loss of normal interneuron formation. This may contribute to an impaired CNS performance that occurs in the absence of inflammation and significant cytolysis. Studies indicate BDV has a nonsegmented negative strand (NNS) RNA genome whose organization is characteristic of the Mononegauirales. However, BDV has the property, unique among NNS RNA animal viruses, that its replication and transcription occur in the nucleus of infected cells, indicating that BDV represents the prototype of a new group of animal RNA viruses (Cubitt and de la Torre, 1994; Cubitt et al., 1994; Briese et al., 1994). Interestingly, serum and cerebrospinal fluid (CSF) of human patients with specific psychiatric disorders are reported to have antibodies that recognize BDV antigens (Amsterdam et al.,1985; Bode et al.,1992, 1994; Fu et al.,1993; Rott et al., 1991; VandeWoude et al., 19901, and BDV-specific antigens have been detected in macrophage monocytes from psychiatric patients but not in samples obtained from healthy control individuals (Bode et al.,1994). This raises the possibility that BDV, or a related virus, may play a role in human mental disorders, specifically those characterized by neurochemical and behavioral disturbances in the absence of inflammation and significant cytolysis. Alterations in behavior and learning have also been found to be associated with persistent LCMV infection of neurons in otherwise healthy mice (Hotchin and Seegal, 1977; Gold et al.,1994). Such infected mice are less likely to show explorative behavior in an open field, and they show a deficit in avoidance performance using a Y-maze spatial discrimination test. Pharmacological studies indicated an enhanced sensitivity to the cholinergenic antagonist scopolamine, suggesting an effect of LCMV on the cholinergenic system (Gold et al., 1994). These in viuo findings complement and extend the early in vitro experiments discussed above in which neuroblastoma cells persistently infected with LCMV have altered levels of choline acetyltransferase and acetycholinesterase, key enzymes of the cholinergenic system. Despite the replication and expression of LCMV genome within cholinergic and other neurons, no signs of structural injury or neuronal dropout occur (Rodriguez et aZ.,1983b). Studies (J. C. de la Torre and E. Mashlia, unpublished data) have revealed that synaptic density and axonal organization are not altered, but expression of the growth associated protein 43 (GAP-43) is decreased, especially in the hippocampus. The GAP-43 phosphoprotein is associated with the membrane of the presynaptic terminals. It is involved primarily in neuritic outgrowth during development and in plasticity and long-term potentia-

VIRAL PERSISTENCE

335

tion in the adult brain; both of these processes are involved in learning and memory (Gispen et al., 1991; Strittmatter et al., 1992). The inner molecular layer of the hippocampus showed the most significant deficiency of GAP-43. This region contains the dendrites of the granular cells of the dentate gyrus which receive projections from the entorhinal cortex and septal region, two circuits known to be involved in the formation of memory. In addition, the septal region provides a strong cholinergenic input. Thus, LCMV-induced alterations of GAP-43 may interfere with the molecular mechanisms involved in learning and memory, which is likely to contribute to the behavioral alterations observed in mice persistently infected with LCMV.

D . Interactions between Virus and Cytoskeleton Affecting Cell Function The cell cytoskeleton system constitutes an intricate network of protein filaments that include actin microfilaments, microtubules, and intermediate filaments. Increasing evidence indicates that many animal viruses use the cytoplasmic and/or nuclear skeleton structure of cells during their multiplication within infected cells. Therefore, not surprisingly, many viruses induce changes or disruptions of the cytoskeletal fiber systems (Penman, 1985; Knipe, 1990). Some of these effects are responsible for virus-induced cytopathic effects. In other cases, viral infections induce changes in the cytoskeleton organization that do not lead to cell death but may have important consequences for cell properties such as shape, movement, and intracellular transport, all regulated by the cytoskeleton. More recent evidence indicates that, based on its association with the second messenger system as well as its regulation of the localization of regulatory molecules, the cytoskeleton may exert an important role on the regulation of gene expression (Ben-Ze’ev, 1985, 1991). Consequently, persistent infections of viruses whose replication process involves interactions between viral products and components of the cytoskeleton may also determine changes in cell properties. Evidence for such an interaction has been provided in tissue culture cells (Penman, 1985; Knipe, 1990), but the implications during natural infections remain to be established. The schematic presented in Fig. 4 summarizes findings during a persistent infection by foot-and-mouth disease virus (FMDV) in tissue culture (de la Torre et al., 1988) that also illustrates some interesting aspects of viral persistence. The picornavirus FMDV is usually highly cytolytic; however, by mechanisms not well defined, FMDV is capable of establishing a persistent infection that is associated with significant cellular phenotypic changes including alterations in cell shape and cell

336 J U A N CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE

FIG. 4. Virus persistent infections can be associated with changes in cell shape and reorganization of cell gene program expression. The picornavirus foot-and-mouth disease virus (FMDV) is normally highly cytolytic in BHK-21 cells. However, FMDV is also able to establish a persistent infection in BHK-21 cells that is associated with significant changes in cell morphology. Changes in cell morphology are accompanied by changes in the levels of expression of several host cellular genes. Cells cured of the infection at early stages restore the wild-type cell phenotype displayed by uninfected control cells. However, cells cured after a long-term persistent infection did not restore a wild-type phenotype. The virus that initiated the persistent infection (V,) also evolves during persistence, generating viruses with different phenotypes (VL)including increased virulence, changes in host range, and temperature sensitivity among others.

gene expression program. Cells cured of the persistent infection are restored to the wild-type phenotype, thus supporting a direct role for the virus. Cells persistently infected for long periods do not revert to the wild-type phenotype after being cured of the virus infection, indicating that virally induced cellular changes can remain after the virus has long disappeared. It is thus conceivable that a virus persistent infection can initiate a disease process that will be manifested in the absence of any vestiges of the infectious agent. In addition, it is worth noting that changes not only in cell properties but also in the phenotype of virus that initiated the infection can be associated with the persistent infection, leading to a virus-cell coevolution process (Ahmed and Fields, 1982; de la Torre et al., 1988).

IV. CONCLUSIONS The knowledge that viruses can persist is as important finding in virology. Indeed, the puzzle for a virus to solve is how to live within a

VIRAL PERSISTENCE

337

host over the lifetime of the host. A successful solution would necessitate the absence of cell lysis and inflammation, signs that we usually associate with viral infection. In the past and present, virologists have focused attention on agents that have failed to solve this problem because they display destructive (lytic) behavior, frequently causing disease, which results in making them easy to detect. As a result of their short-sighted abuse of the host, these viruses must continually seek new cells and/or new hosts to infect. It is the study of such viruses that has dominated the first 100 years of virology. Viruses that persist in a host are ultimately successful by virtue of their abilities to (1) survive within cells that provide their sustenance and (2) avoid recognition by the host immune system. Viruses that persist must, first, remain within a cell for a prolonged time without disturbing the transcription or translation of genes necessary for survival of the infected cell o r altering lysosomal or plasma membranes. Second, such viruses must interfere with antigen presentation, MHC restriction, CTL activation, and/or CTL and antibody activity. However, unless the host-virus relationship is completely symbiotic for both partners, which often it is not, the host cell pays a price to ensure its survival during viral persistence. It has become clear since the 1980s that many persistent virus infections can interfere subtly with the ability of cells to produce differentiated products (hormones, neurotransmitters, cytokines, and immunoglobulins) without disrupting vital cellular functions (respiratory enzymes, lysosomal and plasma membrane integrity, etc.). By this means, the virus can replicate in cells that appear histologically normal by light or high-resolution electron microscopy, although the function of the cell is altered. Further, because the intruder viruses have evolved strategies to escape immunological surveillance, the ordinarily expected T-lymphocyte and monocyte infiltration in their immediate neighborhoods does not occur. Despite viral replication, such infected cells maintain their normal anatomical architecture, yet the virus can induce disorders in differentiated cellular functions, often leading to disturbances in homeostasis and eventually disease. We believe that a number of current diseases affecting differentiated systems like the nervous, endocrine, immune, and cardioskeletal muscle systems as yet of unknown etiology may likely be caused by infectious agents like viruses which have evolved to persist and replicate in differentiated cells without causing lysis of the cell they infect. With the availability of highly sensitive molecular techniques to identify limited amounts of materials, this hypothesis can be adequately tested in the coming decades. From the evidence that is evolving it is likely that the study of such persistent viruses will dominate virology in the twenty-first century.

338 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE

ACKNOWLEDGMENTS Research conducted in the authors’ laboratories is supported in part by US. Public Health Service Grants AGO4342 (J.C.T.) and NS12428 (M.B.A.O.). We thank Jody Anderson for excellent editorial assistance.

REFERENCES Ackrill, A., Foster, G., Laxton, C., Flavell, D., Stark, G., and Kerr, I. (1991).Nucleic Acids Res. 19, 4387-4393. Ahmed, R., and Fields, B. N. (1982).Cell (Cambridge, Mass.) 28,605-612. Ahmed, R., and Stevens, J. G. (1990).Zn (B. N. Fields, D. M. Knipe, R. M. Chanock, J. L. Melnick, M. S. Hirsch, T. G. Monath, and B. Roizman, eds.), 2nd Ed., pp. 241-265. Raven, New York. Aman, P., and von Gabain, A. (1990).EMBO J. 9,147-152. Amsterdam, J. D., Winokur, A., Dyson, W., Henog, S., Gonzalez, F., Rott, R., and Koprowski, H. (1985).Arch. Gen. Psychiatry 42, 1093-1096. Anderson, K., and Fennie, E. (1987).J. Virol. 61, 787-795. Anderson, P., Wild, T. F., and Zwingelstein, G. (1983).Biochem. J. 214,665-670. Balbach, M.,and Koschel, M. (1979).J. Gen. Virol. 42, 615-619. Barrett, A. D., and Dimmock, N. J. (1986).Curr. Top. Microbiol. Zmmunol. 128, 55-84. Barrett, A.D., Cross, A. J., Crow, T. J., Johnson, J. A., Guest, A. R., and Dimmock, N. J. (1986)J . Gen. Virol. 67, 1727-1732. Bateman, A., Singh, A., Kral, T., and Solomon, S. (1989).Endocr. Rev. 10,92-112. Bautieta, J. R., Schwartz, G. J., de la Torre, J. C., Moran, T. H., and Carbone, K. (1994). Brain Res. Bull. 34, 31-36. Beijerinck, M. W. (1899).Zentralbl. Bakteriol. Parasitenkde. Znfektionskrankh. 5, 310311. Beloso, A., Martinez, C., Valcarcel, J., Fernandez-Santaren, and Ortin, J. (1992).J. Gen. Virol. 73, 575-581. Ben-Ze’ev, A. (1985).Trends Biochem. Sci. 11,478-481. Ben-Ze’ev, A. (1991).Bioessays 13,207-212. Bernard, S.,Wild, T. F., and Tripier, M. F. (1983).J. Gen. Virol. 64, 1571-1579. Biron, C. A.,Byron, K. S., and Sullivan, J. L. (1989).N . Engl. J.Med. 320, 1731-1735. Bjorkman, P.J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., and Wiley, D. C. (1987).Nature (London) 329,512-518. Black, B. L., and Lyles, D. S. (1992).J. Virol. 66,4058-4064. Bode, L., Riegel, S., Lange, W., and Ludwig, H. (1992).J. Med. Virol. 26,309-315. Bode, L.,Steinbach, F., and Ludwig, H. (1994).Lancet 343,297-298. Bodner, M., and Karin, M. (1987).Cell (Cambridge, Mass.) 50, 267-275. Bodner, M., Castrillo, J. L., Theill, L., Deerinck, T., Ellisman, M., and Karin, M. (1988). Cell (Cambridge, Mass.) 55, 505-518. Borrow, P., and Oldstone, M. B. A. (1992).J. Virol. 66, 7270-7281. Borrow, P.,Evans, C. F., and Oldstone, M. B. A. (1994).J. Virol. in press. Briese, T., Schneemann, A., Lewis, A., Park, Y., Kim, S., Ludwig, H., and Lipkin, I. (1994).Proc. Natl. Acad. Sci. U S A . 91, 4382-4366. Brinton, M. A., Blank, K. J., and Nathanson, N. (1984).Zn “Concepts in Viral Pathogenesis’’ (A. L. Notkins and M. B. A. Oldstone, eds.), pp. 71-78. Springer-Verlag, New York. Browne, H., Smith, G., Beck, S., and Minsou, T. (1990).Nature (London) 347,770-772. Buchmeier, M.J., and Welsh, R. M. (1979).Virology 96,503-515.

VIRAL PERSISTENCE Buchmeier, M. J., Welsh, R. M., Dutko, F. J., and Oldstone, M. B. A. (1980).Adv. Immunol. 30, 275-331. Buesa, J., Oldstone, M. B. A,, and de la Torre, J. C. (1994).“Virology” submitted. Carbone, K. M., Park, S. W., Rubin, S. A., Waltrip, 111, R. W., and Vogelsang, G. B. (1991). J . Virol. 65, 6154-6164. Casali, P., Rice, G. P. A., and Oldstone, M. B. A. (1984).J. Exp. Med. 159, 1322-1337. Cave, D. R., Hendrickeon, F. M., and Huang, A. S. (1985).J. Virol. 55,366-373. Chen, R., Ingraham, H. A., Teacy, M. N., Albert, V. R., Wilson, L., and Rosenfeld, M. G. (1990).Nature (London) 346, 583. Chomczynski, P., Brar, A., and Frohman, L. A. (1988).Endocrinology (Baltimore) 123, 2276-2283. Chow, J., and Roizman, B. (1992).Proc. Natl. Acad. Sci. U.S.A.89, 3266-3270. Clements, J. E., and Narayan, 0. (1984).In “Concepts in Viral Pathogenesis” (A. L. Notkins, and M. B. A. Oldstone, eds.), pp. 152-157. Springer-Verlag, New York. Cubitt, B., and de la Torre, J. C. (1994).J. Virol. 68,1371-1381. Cubitt, B., Oldstone, C., and de la Torre, J. C. (1994).J. Virol. 68, 1382-1396. Dalle et al. (1990). Dana, S., and Karin, M. (1989).Mol. Endocrinol. 3,815-821. de la Torre, J. C., and Holland, J. J. (1990). J. Virol. 64,6278-6281. de la Torre, J. C., and Oldstone, M. B. A. (1992).Proc. Natl. Acad. Sci. U.S.A.89,99399943. de la Torre, J. C., Davila, M., Sobrino, F., Ortin, J., and Domingo, E. (1985). Virology 145, 24-35. de la Torre, J . C., Martinez-Salan, E., Diez, J., Villaverde, A., Gebauer, F., Rocha, E., Davila, M.,and Domingo, E. (1988).J. Virol. 62, 2050-2058. de la Torre, J. C., Borrow, P., and Oldstone, M. B. A. (1991).Br. Med. Bull, 47,838-851. Dittrich, W., Bode, L., Ludwig, H., Kao, M., and Schneider, K. (1989). Biol. Psychiatry 26, 181-828. Dolle, P., Castillo, J. L., Theill, L. E., Deerinck, T., Ellisman, M., and Karin, M.(1990). Cell (Cambridge, Mass.) 60, 809-820. Dubin, G., Socolof, E., Frank, I., and Friedman, H. M. (1991).J. Virol. 65,7046-7050. Dukto, F. J., and Oldstone, M. B. A. (1981).J. Exp. Med. 154, 1636-1651. Fauci, A. (1994).Communication, presented NIMH meeting, Chantilly, Virginia. Fenner, F., McAuslan, B. Mims, C., Sambrook, J., and White, D. (1974).In “The Biology of Animal Viruses” 2nd Ed. Academic Press, New York. Fenner, P. (1976).Intervirology 7, 1. Foster, G.,Ackrill, A., Goldin, R., Kerr, I., Thomas, H., and Stark, G. (1991).Proc. Natl. Acad. Sci. U.S.A. 88,2888-2892. Fox, S . R., Jorg, M. T. C., Casianover, J., Ye, S. F., Stanley, F., and Sammuels, H. M. (1990).Mol. Endocrinol. 4, 1069-1080. Francis, S.J.,and Southern, P. J. (1988).J. Virol. 62, 1251-1257. Frank, I., and Friedman, H. M. (1989).J. Virol. 63,4479-4488. Fremont, D. M., Matsurnuza, M., Stura, E. A,, Peterson, P. A., and Wilson, I. A. (1992). Science 257, 919-927. Fries, L. F., Friedman, H. M., Cohen, G. H., Eisenberg, R. J., Hammer, C. M., and Frank, M. M. (1986).J.Immunol. 137,1636-1641. Fu, Z. F., Amsterdam, J. D., Kao, M., Shankar, V., Koprowski, H., and Dietzschold, B. (1993).J . Affective Disord. 27, 61-68. Fujinami, R. S., and Oldstone, M. B. A. (1979).Nature (London) 279, 529-530. Fujinami, R. S.,and Oldstone, M. B. A. (1984).In “Concepts in Viral Pathogenesis” (A. L. Notkins and M. B. A. Oldstone, eds.), pp. 187-193. Springer-Verlag, New York.

340 JUAN CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE Gamble, D. R. (1980).Br. Med. J . 281, 99-101. Gilden, D. H., and Lipton, H. L. (1989).In “Clinical and Molecular Aspects of Neurotropic Virus Infection” (D. H. Gilden and H. L. Lipton, eds.), Kluwer, Norwell. Gillies, S., and Stollar, J. (1980).Virology 107, 264-267. Gispen, W. H., Neilander, H. B., DeGraan, P. N. E., Oestreicher, A. B., Schrama, L. H., and Schotman, P. (1991).Mol. Neurobiol. 5, 61-85. Gold, L. H., Brot, M. D., Polis, I., Schroeder, R., Tishon, A., de la Torre, J. C., Oldstone, M. B. A,, and Koob, G. F. (1994).Behav. Neural Biol. 62, 100-109. Gonczol, E., Andrews, P. W., and Plotkin, B. A. (1984).Science 224, 159-161. Gooding, L. R. (1992).Cell (Cambridge, Mass.) 71,5-7. Griffin, D. E.,and Ward, B. I. (1993).J. Infect. Dis. 168,275-281. Gutch, M., and Reich, N. (1991).Proc. Natl. Acad. Sci. U.S.A.88, 7913-7917. Harris, S. L., Frank, I., Yee, A., Cohen, G. H., Eisenberg, R. J., and Friedman, H. M. (1990).J. Infect. Dis. 162,331-337. Holland, J. J. (1987).In “The Rhabdoviruses” (R. R. Wagner, ed.), pp. 297-360.Plenum, New York. Holland, J. J. (1990).In “Virology” (B. N. Fields, D. M. Knipe, R. M. Chanock, J. L. Melnick, M. S. Hirsch, T. G. Monath, and B. Roizman, eds.), 2nd Ed., pp. 151-165. Raven, New York. Holland, J. J., ed. (1992).Curr. Top. Microbiol. Immunol. 176. Holtzer, H., Biehl, J., Yeoh, G., Meganathan, R., and Waji, A. (1975).Proc. Natl. Acad. Sci. U.S.A. 72, 4051-4055. Hotchin, J., and Seegal, R. (1977).Science 196, 671-674. Howcroft, T. K., Strebel, K., Martin, M. A., and Singer, D. S. (1993).Science 260, 13201321. Hsu, D.H., de Waal, M., Fiorentino, D. F., Darg, M. N., Viera, P., de Vries, J., Spits, H., Mosmann, T. R., and Moore, K. W. (1990).Science 250, 830-832. Huang, A. S. (1988).In “RNA Genetics” (E. Domingo, J. J. Holland, and R. Ahlquist, eds.), Vol. 3,pp. 195-208.CRC Press, Boca Raton, Florida. Inglis, S. C. (1982).Mole. Cell. Biol. 2, 1644-1648. Ingraham, H. A,, Chen, R., Mangalam, H. J., Elsholtz, H. P., Flynn, S. E., Lin, C. R., Simmons, D. M., Swanson, L., and Rosenfeld, M. G. (1988).Cell (Cambridge, Mass.) 55,519-529. Ivanovski, D. I. (1899).Zentralbl. Bakteriol. Parasitenkde. Infektionskrankh. 5,250-254. Johnson, R. T. (1982).“Viral Infections of the Nervous System.” Raven, New York. Joly, E., Mucke, L., and Oldstone, M. B. A. (1991).Science 253, 1283-1285. Kagi, D. (1994).Nature London 369, 31-37. Kalvakolanu, D., Bandyopadhyay, S., Harter, M., and Sen, G. (1991).Proc. Natl. Acad. Sci. U S A . 88, 7459-7463. Kang, C. Y., Weide, L. G., and Tischfield, J. A. (1981).J. Virol. 40, 946-952. Kaper, J. M., and Collmer, C. W. (1988).In “RNA Genetics” (E. Domingo, J. J. Holland, and P. Ahlquist, eds.), Vol. 3,pp. 171-194. CRC Press, Boca Raton, Florida. Kauffman, R. S., Ahmed, R., and Fields, B. N. (1983).Virology 131,79-87. Khanna, R., Burrows, S. R., Suhrbier, A., Jacob, C. A., Griffin, H., Misko, I. S., Sueller, T. B., Rowe, M., Rickinson, A. B., and Moss, D. J. (1993).Immunology 150,5154-5162. Klavinskis, L. S.,and Oldstone, M. B. A. (1989).Virology 168,232-235. Knipe, D. M. (1990).In “Virology” (B. N. Fields, D. M. Knipe, R. M. Chanock, J. L. Melnick, M. S. Hirsch, T. G. Monath, and B. Roizman, eds.), 2nd Ed., pp. 278-281. Raven, New York. Koschel, K., and Munzel, P. (1980).J. Gen. Virol. 47,513-517,

VIRAL PERSISTENCE

341

Koszinowski, U. H., Val, M. D., and Reddehase, M. J. (1990).Curr. Top. Microbiol. Zmmunol. 154, 189-220. Krajesi, P., Tollefson, A,, Anderson, C. W., and Wold, W. S. M. (1992).J. Virol. 66, 16541673. Kristensson, K., and Norrby, E. (1986). Annu. Rev. Microbiol. 40, 159-184. Lefevre, C., Inayawa, M., Dana, S., Grindlay, J., Bodner, M., and Karin, M. (1987). EMRO J . 6,971-981. Lehmann-Grube, F. (1984). Zntervirology 22, 121-145. Levine, B., Hardwick, J. M., Trapp, B. D., Crawford, T. O., Bollinger, R. C., and Griffin, D. E. (1991). Science 264,856-860. Li, S., Crenshaw 111, E. B., Rawson, E. J., Simmons, D. M. I., Sawnson, L. W., and Rosenfeld, M. B. (1990). Nature (London) 347, 528-533. Liebert, U. G., Schneider-Schaulies, S., Baczko, K., and ter Meulen, V. (1990). J. Virol. 64,706-713. Lipkin, I., Battenberg, E. L. F., Bloom, F. E., and Oldstone, M. B. A. (1988).Brain Res. 451,333-339. Loeffler, F., and Frosch, P. (1898). Zentbl. Bakt. Parasitkde 23, 371-391. Lohr, J. M., and Oldstone, M. B. A. (1990). Lancet 2, 644-648. Lopez, C. (1988). Zn “Immunobiology and Pathogenesis of Persistent Virus Infections” (C. Lopez, ed.), pp. 39-53. American Society for Microbiology, Washington, D.C. Lyons, M. J., Faust, I. M., Hemmls, R. B., Buskirk, D. R., Hirsch, J., and Zubriskie, J. P. (1982). Science 216, 82-85. McChesney, M. B., and Oldstone, M. B. A. (1987). Annu. Rev. Zmmunol. 5,279-304. McChesney, M. B, and Oldstone, M. B. A. (1989).Adv. Zmmunol. 45,335-380. McCormick et al. (1988). McEvoy, R. E., Fedun, B., Cooper, L. Z., Thomas, N. M., Rodriguez de Cordoba, S., Rubistein, P., and Ginsberg-Fellner, F. (1988).Adv. Exp. Med. Biol. 246, 221-227. McNearney, T. A,, Odell, C., Holers, M., Spear, P. G., and Atkinson, J. P. (1987). J . Exp. Med. 166, 1525-1535. Mahy, B. W. J. (1985).Br. Med. Bull. 41, 50-55. Mangalam, H. J., Albert, V. R., Ingraham, H. A., Kapiloff, M., Wilson, L., Nelson, C., Elsholtz, H., and Rosenfeld, M. G. (1989). Genes Dev. 3, 946-958. Martz, E. (1987). Hum. Zmmunol. 16,3-37. Matsumura, M., Fremont, D. H., Peterson, P. A., and Wilson, I. A. (1992). Science 257, 927-934. Mims, C. A. (1986). Clin. Exp. Zmmunol. 66, 1-16. Mims, C. A., and White, D. 0. (1984).Zn “Viral Pathogenesis and Immunology.” Blackwell, Oxford. Mohammed, A,, Norrby, E., and Kristensson, K. (1993).Rev. Neurosci. 4, 267-286. Moskophidis, D., Laine, E., and Zinkernagel, R. (1993).Eur. J . Zmmunol. 23,3306-3311. Munzel, P., and Koschel, K. (1982). Proc. Natl. Acad. Sci. U S A . 79, 3692-3696. Narayan, O., Zink, M. C., Huso, D., et al. (1988). Microbiol. Pathogenesis 5, 149-157. Notkins, A. (1975). “Viral Immunology and Immunpathology.” Academic Press, New York. Nsiah, Y.A., and Turner, W. (1990). Zntervirology 31, 265-275. Odermatt, B., Eppler, N., Leist, T.P., Hergartner, H., and Zinkernagel, R. (1991). Proc. Natl. Acad. Sci. U.S.A. 88, 8252-8256. Oldstone, M. B. A. (1975). Prog. Med. Virol. 19, 84-119. Oldstone, M. B. A. (1984). In “Concepts in Viral Pathogenesis” (A. L. Notkins, and M. B. A. Oldstone, eds.), pp. 269-276. Springer-Verlag, New York.

342 J U A N CARLOS DE LA TORRE AND MICHAEL B. A. OLDSTONE Oldstone, M. B. A. (1985). Cell (Cambridge, Mass.) 56, 517-520. Oldstone, M. B. A. (1991). J. Virol. 65, 6381-6386. Oldstone, M. B. A., and Buchmeier, M. J. (1982).Nature (London) 300,360-362. Oldstone, M. B. A., Holmstoen, J., and Welsh, R. M. (1976).J.Cell. Physiol. 91,459-472. Oldstone, M. B. A., Fujinami, R. S., and Lampert, P. W. (1980).Prog. Med. Virol.26,4593. Oldstone, M. B. A., Sinha, Y. N., Blount, P., Tishon, A,, Rodriguez, M., von Wedel, R., and Lampert, P. W. (1982).Science 218, 1125-1127. Oldstone, M. B. A., Rodriguez, M., Daughaday, W. H., and Lampert, P. W. (1984).Nature (London) 307,278-281. Oldstone, M. B. A,, Ahmed, R., Buchmeier, M. J., Blount, P., and Tishon, A. (1985). Virology 142, 158-174. Palese, P., and Young, J. V. (1982). Science 215, 1468-1474. Penman, S. (1985). In “Virology.” (B. N. Fields and D. M. Knipe, eds.), 1st Ed., pp. 169182. Raven, New York. Porter, D. D. (1985). In “Medical Microbiology.” (S. Baron, ed.), 2nd Ed., pp. 784-790. Addison Wesley, Menlo Park, California. Rayfield, E., Seto, Y., Walsh, S., and McEvoy, R. (1981). J. Clin. Invest. 68, 1172-1181. Riviere, Y., Ahmed, R., Southern, P., and Oldstone, M. B. A. (1985a). Virology 142, 175182. Riviere, Y., Ahmed, R., Southern, P., Buchmeier, M. J., Dutko, F. J., and Oldstone, M. B. A. (1985b).J. Virol. 53, 966-968. Roby, K., Boettinger, D., Pacifici, M., and Holtzer, H. (1976).Am. J.Anut. 174,401-405. Rodriguez, M., Buchmeier, M. J., Oldstone, M. B. A., and Lampert, P. W. (1983a).Am. J. Pathol. 110, 95-100. Rodriguez, M., von Wedel, R. J., Garrett, R. S., Lampert, P. W., and Oldstone, M. B. A. (1983b).Lab. Invest. 49, 48-53. Rodriguez, M., Garret, R. S., Raitt, M., Lampert, P. W., and Oldstone, M. B. A. (1985). Am. J.Pathol. 121, 497-504. Roizmann, B., and Sears, A. E. (1987). Annu. Rev. Microbiol. 41, 543-571. Rott, R., Henog, S., Bechter, H., and Freese, K. (1991).Arch. Virol. 118, 143-149. Riitzschke, O., Falk, K., Deres, K., Schild, H., Norda, M., Metzger, J., Jung, G., and Rammensee, H. G. (1990). Nature (London) 348,252-254. R o n , L., Simon, A. E., and Holland, J. J. (1991). Adv. Virus Res. 40, 181-211. Row, W. P. (1954). “Naval Medical Research Institute Report,” Vol. 12, pp. 167-220. National Naval Medical Center, Bethesda, Maryland. Rowe, W. P. (1983). Science 221, 562-563. Salvato, M., and Shimomaye, E. M. (1989). Virology 173, 1-10. Sandrej, K. B., Hamilton-Dutoit, S., and Patterson, G. (1993). Leukemia Lymphoma 9, 95-101. Schrier, R. D., and Oldstone, M. B. A. (1986). J. Virol. 59, 127-131. Schroder, H. C., Sarin, P. S., Rottman, M., Wenger, R., Maidhof, A., Renneisen, K., and Muller, W. E. G. (1988).Biochem. Pharmacol. 37, 3947-3952. Schroder, H. C., Falke, D., Weise, K., Bachman, Carmo-Fonseca, M., Zaubitzer, T., and Muller, W. E. G. (1989). Virus Res. 13, 61-78. Sen, G., and Lengyel, P. (1992). J. Biol. Chem. 267, 5017-5020. Sen, G. C., and Ransohoff, R. M. (1993). Adv. Virus Res. 42, 57-102. Simon, A. E. (1988). Plant Mol. Biol. Rep. 6, 240-252. Sprankel, H., Richarz, K., Ludwig, H., and Rott, R. (1978). Med. Microbiol. Immunol. 165, 1-18.

VIRAL PERSISTENCE

343

Springer, T. A,, Dustin, M. L., Kishimoto, T. K., and Marlin, S. 0. (1987). Annu. Reu. Immunol. 5,223-252. Stevens, J. G. (1989). Microbiol. Rev. 53, 318-332. Strittmatter, S. M., Vartanian, T., and Fishman, M. C. (1992).J.Neurobiol. 23,507-520. Sultz, H. A,, Hart, B., Zielezny, M., and Schlesinger, E. R. (1975).J.Pediatr. 86,654-656. Teng, M. N., Oldstone, M. B. A,, and de la Torre, J. C. (1994). J. Virol. submitted. ter Meulen, V. (1991).Virus-cell interactions in the nervous system. In “Seminars in the Neurosciences.” (ter Meulen, ed.), Vol. 3. ter Meulen, V., Carter, M. J., Wege, H., and Watanabe, R. (1984).Ann. N.Y. Acad. Sci. 436,86-97.

Tishon, A,, and Oldstone, M. B. A. (1990). A m . J.Pathol. 137, 965-969. Torres-Anjel, M., Volz, D., Torres, M. J. R., Turk, M., and Tshikuka, J. G. (1988). Reu. Infect. Dis. 10, (Suppl. 4), 710-725. Townsend, A. R. M., Rothbard, J.,Gotch, F. M., Bahadur, G., Wraith, D., and McMichael, A. J. (1986). Cell (Cambridge, Mass.) 46, 959-968. Tsiang, H. (1992). Adu. Virus Res. 42, 375-412. Valsamakis, A,, Riviere, Y., and Oldstone, M. B. A. (1987). Virology 156, 214-220. VanBleek, G. M., and Nathenson, S. G. (1990).Nature (London) 348, 213-216. VandeWoude, S., Richt, J. A., Zink, M. C., Rott, R., Narayan, O., and Clements, J. E. (1990). Science 250, 1278-1281. Warrell, D. A. (1976). Trans. R . SOC.Trop. Med. Hyg. 70, 188-195. Webster, R. G., Laver, W. G., Air, G. M., and Schied, G. C. (1982).Nature (London) 296, 115-121.

Wechsler, S., and Meissner, H. C. (1982).Prog. Med. Virol. 28, 65-95. Weiss, R. A. (1993). Science 260, 1273-1279. Whitton, J. L., and Oldstone, M. B. A. (1990).In “Virology” (B. N. Fields, D. M. Knipe, R. M. Chanock, J. L. Melnick, M. S. Hirsch, T. G. Monath, and B. Roizman, eds), 2nd Ed., Chapter 15, pp. 369-381. Raven, New York. Wild, T. F., Bernard, A., and Greenland, T. (1981).Arch. Virol. 67, 297-308. Wold, W. S. M., and Gooding, L. R. (1991). Virology 184, 1-8. Yoon, J. W. (1991). Ann. Med. 23, 437-445. Yoon, J. W., Austin, M., Onodera, T., and Notkins, A. L. (1979). N . Engl. J. Med. 300, 1173-1179.

Younger, J. S., and Preble, 0. T. (1980). In “Comprehensive Virology” (M. FrankelConrat and R. Wagner, eds.), pp. 73-135. Plenum, New York. Zinkernagel, R. M., and Doherty, P. C. (1974).Nature (London) 248, 701-702.

This Page Intentionally Left Blank

ADVANCES IN VIRUS RESEARCH, VOL. 46

THE IRI DOVlRUSES Trevor Williams ECOSUR-El

Colegio de la Frontera Sur

30700 Tapachula, Chiapas, Mexico

I. Introduction 11. Classification

111.

IV.

V. VI. VII.

A. Current System B. Comparative Studies of Iridoviruses C. New Nomenclature for Iridescent Viruses D. Alternative Approaches E. Suggested Changes to Current Classification Structure A. Physicochemical Properties B. Capsid C. Lipid Membrane D. Core E. Iridescence Phenomenon Replication A. Cell Penetration and Uncoating B. Shutdown of Host Macromolecular Synthesis C. DNA Replication D. Methylation of Viral DNA E. Transcription F. Translation G. Packaging of Virions H. Cytoskeletal Manipulation I. Enzymatic Activities Molecular Biology A. Virus Genes B. Repetitive DNA Ecology A. Transmission B. Persistence Future Directions for Iridoviruses References

I. INTRODUCTION In March 1954, on his first trip from the laboratory of Kenneth Smith in Cambridge, Claude Rivers went in search of crane fly larvae (Tipula spp.) infected with the recently discovered hemocyte polyhedrosis virus. He applied St. Ives fluid to pasture land in the grounds 345

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

346

TREVOR WILLIAMS

of Plowden Hall, Shropshire, England. As the tipulid larvae wriggled to the surface to escape the irritating phenols, Rivers was amazed to see larvae with brilliant patches of iridescent blue color showing beneath the epidermis. Later, he returned with Nick Xeros and collected more infected material, which resulted in the note to Nature and which marked the beginning of research on the iridescent viruses (Xeros, 1954). At the time, the discovery caused great excitement and controversy because the structure and host range of the iridescent virus was in marked contrast to that of the polyhedrosis viruses discovered previously, and this brought into focus much of Smith’s earlier work on the host specificity of insect viruses. Subsequently, a number of other arthropods were diagnosed with patent iridescent virus infections. Because of the occurrence of iridescent virus diseases in soma important pest and vector species, the viruses have attracted attention as potential agents for biological control. For laboratory studies they have a number of advantages in that many can be grown in massive quantities in insect larvae and several isolates are highly amenable to cell culture. Vertebrate iridoviruses are found in fish, amphibians, and reptiles. For the amphibian iridoviruses, initial interest was sparked by finding one isolate in association with a renal carcinoma of the leopard frog, Rana pipiens (Granoff et al., 1966).The association was later shown to be coincidental. Much of the known biology of iridoviruses comes from work with this frog isolate (named frog virus 3),which has proved to be highly amenable to manipulation in cell culture, a more important reason for its popularity. For the fish iridoviruses, the wartlike lesions symptomatic of “lymphocystis disease” have been known for over a century although the disease-causing agent was not recognized until much later (Walker, 1962).Certain iridovirus infections of fish may be inapparent, whereas others produce overt, sometimes lethal disease, and thus have economic significance, but research has been hampered because of difficulties with cell culture, especially for the lymphocystis disease viruses. Iridoviruses are large icosahedral viruses, 120 to 300 nm in diameter, that assemble in the cytoplasm of host cells (Fig. 1).The exact size measurement is highly dependent on the method of measurement and on the isolate; fish isolates tend to be larger (200-300 nm) than the amphibian or the invertebrate viruses (120-200 nm). Measurements made of particles in the hydrated state give significantly higher size values than measurements of dehydrated particles (e.g., Klug et al., 1959).Particles comprise a capsid, intermediate lipid membrane, and core. Icosahedral form was demonstrated by double shadowing (Williams and Smith, 1958;Wrigley, 1969,1970;Devauchelle, 1977).The

IRIDOVIRUSES

347

most abundant polypeptide is the major capsid protein (MCP) which is an important feature in comparative studies. The genome comprises double-stranded DNA (dsDNA), typically of 100-210 kbp. For all of the iridoviruses examined to date, the genome is circularly permuted and terminally redundant; that is, each virus particle contains a complete genome (plus about 10% redundant DNA), but the terminal sequences are different for each particle in a population (circular permution).

FIG.1. Iridescent virus from a mosquito assembling in the cytoplasm of an infected host cell. Arrays of virus particles are surrounded by viroplasmic stroma (virus assembly sites) characteristic of infection by iridoviruses ( x 14,000). (Photograph supplied by D. B. Stoltz.)

348

TREVOR WILLIAMS

Vertebrate and invertebrate iridoviruses now attract little attention. There are probably no more than seven research groups worldwide active in this field, the majority of which comprise one or two individuals. This is probably set to change somewhat in the future as the relationships of the established viruses in the family are now becoming clearer and as it becomes feasible to study the impact of pathogens in natural animal populations following the development of powerful diagnostic and genetic analysis methods such as PCR (polymerase chain reaction) and DNA sequencing. This review is general in nature, and I have tried to emphasize developments since the 1980s, particularly in comparative studies, molecular biology, and ecological aspects of iridoviruses. The standard abbreviation IV is used here for invertebrate iridescent virus. The word iridovirus is a coverall term for viruses of this family and may be qualified by reference to a particular host or group of hosts.

11. CLASSIFICATION A . Current System There are currently five recognized genera within the family (Table I). The iridescent viruses from invertebrates are named according to the host species and are given a type number according to the sequence of discovery following an interim system recommended when the number of IV isolates being reported was increasing rapidly (Tinsley and Kelly, 1970). The small iridescent viruses from invertebrates (genus Zridouirus) have been isolated from a diverse selection of invertebrate taxa, mostly insect orders. They are united by their size, some 120-140 nm in ultrathin sections. The type species for the genus is IV type 6 , from Chilo suppressah. This isolate has received the most attention probably because of the agricultural significance of the host from which it was isolated, a stem-boring lepidopteran, and for the ease with which it replicates in cell culture. The large iridescent viruses (genus Chloriridouirus) are some 180 nm in diameter in ultrathin sections. They have been isolated only from Diptera, and virtually all of these records are from mosquitoes. The type species for the genus is the first large IV to be discovered, IV type 3 from Aedes taeniorhynchus. Nomenclature of the vertebrate isolates differs from that of the invertebrate isolates. There are many iridovirus isolates from fish and frogs, but attention has focused on just two isolates: frog virus 3 (FV3), the type virus for the genus Ranauirus, and lymphocystis virus type 1 (LCDV-l), the type virus of the Lymphocystiuirus genus. The patho-

TABLE I CURRENT CLASSI~~CATION OF Zridoviridae Genus

Zridovirus

Chloriridovirus

Ranavirus

Lymphocystivirus Goldfish virus-like a

Vernacular name Small iridescent insect virus

Host species

Member of genusa

Tipula paludosa IV (IV1) Sericesthis pruirwsa IV (N2) Chilo suppressalis IV (IV6)T Wiseana cervinata IV (IV9) Witlesia sabulosella IV (IV10) Costelytra zealandica IV (IV16) Pterosticus mudidus IV (IV17) Opogonia sp. IV (IV18) Odontriu striata IV (IV19) Simocephalus expinosus IV (IV20) Helicoverpa armigera IV (IV21) Simulium sp. IV (IV22) Heteronychus arator IV (IV23) Apis cerana IV (IV24) Tipula sp. N (IV25) Ephemopteran (IV26) Nereis diversicolor (IV27) Lethocerus columbiae IV (IV28) Tenebrio molitor IV (IV29) Helicoverpa zea IV (IV30) Armadillidium vulgare IV (IV31) Porcellio scaber IV (IV32) Large irides- Diptera Aedes tueniorhynchus IV (IV3)T cent insect (mosquitoes) Aedes cantans IV (IV4) virus Aedes annulzpes IV (IV5) Simulium ornatum 1V (IV7) Culicoides sp. IV (IV8) Aedes stimulans (IV11) Aedes cantans (IV12) Corethralla brakeleyi IV (IV13) Aedes detritus IV (IV14) Aedes detritus IV (IV15) Chironomus plumosus IV (probable member) Frog virus Frog Virus 1 , 2 Amphibia Frog Virus 3 (FV3) T Frog Virus 5-24 Frog Virus L2, L4, L5 Tadpole Edema Virus Luck6 triturus virus LT1-LT4 Newt Virus T6-T20 Xenopus Virus T21 LymphoMany teleost Lymphocystivirus type 1 (LCDV-1) T cystis dis(fish) fami- Lymphocystivirus type 2 (LCDV-2) ease virus Octopus vulguris disease virus lies (possible member) Goldfish Goldfish Goldfish virus (GFV-1) T virus Goldfish virus (GFV-2)

T denotes type species

Diverse invertebrate taxa (mostly insects)

350

TREVOR WILLIAMS

genicity of the vertebrate viruses varies, but high levels of mortality have been attributed to iridovirus infections in natural and particularly in farmed fish populations (Langdon et al., 1986, 1988; Wolf, 1988; Armstrong and Ferguson, 1989; Ahne et al., 1989; Hendrick et al., 1990; Pozet et al., 1992). Most of the isolates from vertebrates have yet to be characterized to a level sufficient to assign them to genera. The International Committee for the Taxonomy of Viruses (ICTV) has recognized a new genus, described as “Goldfish virus-like,” to account for two isolates detected in swimbladder cell cultures of the goldfish, Carassius auratus (Berry et al., 1983). There are also reports of iridovirus-like agents from molluscs (Comps and Bonami, 1977; Barthe et al., 1984),marine Crustacea (Montanie et al., 1993; Lightner and Redman, 1993), and a reptile (Stebhens and Johnston, 1966), but those isolates have yet to be characterized and assigned to genera.

B . Comparative Studies of Iridoviruses Throughout the late 1960s and 1970s, serological comparison with other established isolates was used to investigate the interrelationships among the iridoviruses, but it was only intermittently used as part of the characterization process when describing a novel isolate. Moreover, this method was not standardized, quantitative, or statistically robust, and vague terms such as “partially related or “showing some relatedness” were used to describe various pairwise comparisons. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) polypeptide profiles were of use, especially when considering the differences between fairly similar isolates that had been grown in the same host, and indeed could even be used to indicate that apparently novel isolates were really variants of established viruses (Elliott et al., 1977). Following a decline in interest in invertebrate IVs in the 1980s, knowledge of the interrelationships among these viruses benefited little from the dramatic advances in molecular biology that have since occurred. The need for comparative genetic studies has been well recognized among those who maintained an interest in the family (Hall, 1985; Kelly, 1985; Willis, 1990; Ward and Kalmakoff, 1991; Stohwasser et al., 1993; Schnitzler and Darai, 1993). Currently, taxonomic problems exist in all genera of the Iridoviridae, particularly in the invertebrate genera. This is mainly due to a lack of broad comparative studies among the various isolates and partly due to the interim system of classification and nomenclature, for invertebrate isolates, based on sequence of discovery and host. As will be explained, the system of naming isolates according to the host can be very misleading when

351

IRIDOVIRUSES

viruses naturally infect more than one species. In addition, previously, a number of uncharacterized and apparently unrelated isolates have been assigned to genera on the basis of common particle size or common host. This is not helpful to understanding the relationships among the various IV isolates. Proposals have now been put forward to change the system of classification and the system of nomenclature of invertebrate isolates in favor of a more revealing and less misleading alternative system. The serological interrelationships among the various IV isolates that had been kept and studied was summarized by Kelly et al. (1979). For the small invertebrate IVs (genus Iridouirus) serologically, there appeared to be three distinct groups (Fig. 2). There was a large interrelated group of 11 isolates, some of which appear to be very closely related or even serologically indistinguishable from one another (e.g., IV21 and IV28). Conversely, IV6 and IV24 did not appear to be serologically related to any of the other small IVs or to one another. Finally, there was one isolate, IV29, which showed partial serological relatedness to some, but not all members of the large serogroup. The chloriridovirus, IV3, from mosquitoes showed no serological affinity to members of the genus Iridouirus (Cunningham and Tinsley, 1968).The FV3 and LCDV-1 isolates appear distinct from other members of the Iridouiridae and from one another (Bellett and Fenner, 1968; McAuslan and Armentrout, 1974; Darai et al., 1983; Williams and Cory, 1994). Within the genus Ranauirus, the following isolates show high levels of serological interrlatedness: FV1, FV2, FV3, LT1, LT3, LT4, L4, L5, TEV, T6, T8, and T15 (Table I). The L2 virus from Rana pipiens was

6

24

Unrelated IVs

Related to some IVs in serogroup

All types related

FIG. 2. Summary of serological relationships among invertebrate iridescent viruses. Viruses are identified by type number. Isolates showing the highest levels of interrelatedness are boxed together.

352

TREVOR WILLIAMS

reported as serologically distinct from all others tested (Lehane et al., 1968; Clark et al., 1969; Kaminski et al., 1969). Sparked by the obvious need for comparative studies, recent and ongoing work has begun to reveal the interrelationships both within and among genera of the Iridouiridae. Comparative genetic studies have examined isolates from all the genera of the family (with the exception of the “Goldfish virus-like” genus). Standard molecular techniques have been used, namely, restriction endonuclease analysis, Southern blot analysis, PCR of the major capsid protein (MCP) gene region, and DNA-DNA hybridization. The results of these studies prompted the proposals for a new classification and nomenclature of invertebrate iridescent viruses (Williams and Cory, 1994; Williams, 1994). These studies are described in order to understand the reasons behind the new proposals. Fourteen isolates were studied from a broad range of invertebrate hosts (IV1, IV2, IV9, IV10, IV18, IV21, IV22, IV24, IV28, IV29, IV30, IV31; see Table I for hosts) including the type species of the Zridouirus genus (IV6) and of the Chloriridouirus genus (IV3). In addition, the type species from the vertebrate genera were included, FV3 and LCDV-1. Restriction profiles (HindIII, EcoRI, and SalI) immediately indicated that several isolates were actually variants of one another (Fig. 3a), namely, IV21 and IV28, with a separate group of variants being IV9, IV10, and IV18. A coefficient of similarity was calculated for all possible pairwise comparisons, representing the proportion of common-sized fragments which any two isolates shared, and this reinforced what was clear from the gels. Coefficients of similarity among IV9, IV10, and IV18 of up to 91.5%, depending on isolate and enzyme, and likewise between IV21 and IV28 coefficient values up to 94.3% were recorded. The mosquito IV (IV3) (Chloriridouirus) and the vertebrate iridoviruses, FV3 and LCDV-1, showed no restriction profile similarities to any other isolates or to one another. Southern blots of the gels were probed with a MCP gene SalI fragment (1.4 kb) from Aberystwyth IV (IV22), an isolate for which the entire MCP gene sequence is known. At high stringency (50% formamide, 37°C) the probe showed clear hybridization to a subset of the invertebrate IVs (IV1, IV2, IV9, IV10, IV18, IV24, IV29, IV30, and of course to IV22) (Fig. 3b). The MCP probe highlighted fragments of common size for the IV9, IV10, and IV18 isolates. At lower stringency (20% formamide, 37”C), in addition to the isolates above the probe hybridized to fragments of common size of IV6, IV21, and IV28 (Fig. 312).This pattern was highly consistent for each of the blots tested. In no blots did the probe show affinity to the vertebrate isolates FV3 and LCDV-1. Probe hybridization to IV31 was notably weak but fairly

IRIDOVIRUSES

353

FIG. 3. (a)HindIII restriction endonuclease profiles of DNA from invertebrate iridoviruses (identified by type number) and isolates from vertebrates, namely, frog virus 3 (FV3) and flounder lymphocystis disease virus (FLCDV = LCDV-1). Note the similarities among IV9, IV10, and IV18 and between IV21 and IV28, as well as the distinct nature of profiles from FV3 and FLCDV. (b) Hybridization of a major capsid protein (MCP) gene fragment of IV22 a t high stringency to a Southern blot of the HindIII gel. The gene probe shows hybridization to only a subset of the invertebrate iridoviruses and no hybridization to the vertebrate viruses (FV3 and FLCDV). (c) Hybridization of the MCP gene probe from IV22 at low stringency, showing additional hybridization to IV6, IV21, and IV28 and weak hybridization to IV31. (From Williams and Cory, 1994.)

consistent, whereas probe hybridization to IV3 was only seen very weekly in a SaZI blot. Under high-stringency conditions PCR amplification of a 719-bp re-

354

TREVOR WILLIAMS

gion of the MCP gene using primers derived from the Aberystwyth IV sequence produced an amplicon only for IV1, IV2, IV9, IV10, IV22, and IV29, which were all members of the subset of isolates to which the Aberystwyth IV (IV22) MCP gene probe showed affinity at high stringency (IV18 and IV28 were not used in the PCR study). DNA-DNA dot-blot hybridization results were very clear and were consistent with the Southern blot analysis. With hybridization conditions of 40%formamide, 37"C, and critical washes in 2 x SSC at 55"C, one large group of interrelated isolates was detected (Table 11) which corresponded to the large serogroup of isolates described above (Fig. 2). The degree of relative hybridization within this group varied between 10 and 90%; the very high values coming from IV9, IV10, and IV18 confirm them to be variants of the same virus. There was a second, smaller group of closely related isolates that comprised IV21 and IV28, which are clearly variants of the same virus, and IV6, the type species

IRIDOVIRUSES

355

of the genus Iridouirus.The isolate from woodlice (IV31) appeared only distantly related to the other IVs in the study, and the relationship of IV3 from mosquitoes to other isolates appeared even more distant, confirming the placing of this isolate in a separate genus (Chloriridouirus).The vertebrate isolates consistently failed to show DNA hybridization to the invertebrate isolates or to one another under the conditions used. Overall, it is possible to see a very clear picture emerging from these DNA studies that closely support the previous serological findings, albeit with a few differences involving the different placing of IV24 and IV21/IV28 to that given by serology. This prompted the proposal to assign isolates to discrete complexes based on DNA and serological characteristics (Fig. 4). Three complexes were identified within the genus Iridouirus. The main group of interrelated isolates were placed in a complex named Polyiridouirus. The second smaller group of isolates comprising IV6 and IV21/IV28 were assigned to a separate complex, the Oligoiridouirus complex. Finally, the isolate from woodlice

TABLE I1 RELATIVETO HOMOLOWUS DNA DNA-DNA DOT-BLOTHYBRIDIZATION VALUES IV1 IV1 IV2 IV3 IV6 IV9 IVlO IV18 IV2 1 IV22 IV24 IV28 IV29 IV30 IV31 FV3 LCDV-1 a

100 41 0 0 20 17 17 1.2 33 13 1 13 11 (1

0 0

IV2

IV3

100 <1 <1 46 41 55 <1 22 24 (1 26 36 1.4 0 0

100 0 0 0 0 0 0 0 0 (1 0 0 0 0

IV6

100 0 0 0 42 0 0 44 1.1 0 2.4 0 0

Data from Williams and Cory (1994).

IV9

IVlO

IV18

100 91 100

100 89

100

0

0

(1

28 22 0 32 33 0 0 0

27 17 0 31 32 0 0 0

20 18 <1 28 32 0 0 0

IV21

100 0

<1 100 0 0 1.6 0 0

FOR

VERTEBRATE AND INVERTEBRATE IRIWVIRUSEW

IV22

IV24

IV28

IV29

IV30

Tv31

FV3

LCDV-1

100 10 <1 15 23 0 0 0

100 1 27 24 0 0 0

100 0 0 1.4 0 0

100 56 0 0 0

100 <1 0 0

100 0 0

100 0

100

IRIDOVIRUSES Genus Indovirus

357 Genus Chlorindrovrrus

Polyindovim Complex T w Plowden IV

Oligoindovirus Complex Type Dazaifu IV

Crustaceoiridovirus Complex Type Rwerside IV

Type: Vero Beach IV

Aherystwyth

(IV22, IV25)

Dazaifu OV6)

Riverside (IV3 I.IV32)

Vero Beach (1~3)

Fort Collins (IV29)

Ntondwe (IV21, IV28)

Nelson (IV9, IVIO, IV18) Plowden (IVI) Srinagnr (IV24) Stoneville (IV30) Tia (IV2) TimnN (IV16, IV19)

Uitenhage (IV23)

FIG.4. Diagram of hybridization complexes in the genera Zridouirus and Chloriridouirus of invertebrate iridoviruses, detected by Williams and Cory (1994). Isolates that are strains of a common virus have been synonymized under new names based on geographical origin. See text for explanation.

(IV31) was assigned a separate complex, Crustaceoiridouirus. These findings did not change the situation regarding the mosquito isolate IV3, the sole representative of the genus Chloriridouirus, or that with respect to the vertebrate isolates FV3 or LCDV-1, as representatives of two separate vertebrate genera. C . New Nomenclature for Iridescent Viruses From these genetic studies and previous serology (Kelly et al., 19791, it has become apparent that one IV may infect several host species, for example, IV9 from Wiseana cervinata, IVlO from Witlesia sabulosella, and IV18 from Opogonia sp. Other studies have reported that apparently different IVs may infect the same host population at the same place and the same time (see Section VI). Thus, using the name of the host to name the virus can be misleading. This system is causing much

358

TREVOR WILLIAMS

confusion in other invertebrate virus families, for example, the Baculoviridae, where an increasing number of isolates from different host species appear to be strains of the same virus. Consequently, Williams and Cory (1994) proposed that the nomenclature of IV isolates should be changed to a neutral form which dissociates virus and host species. They suggested using the geographical origin (nearest large town) of the original isolate as a name. This system is currently used in a number of other virus families, for example, in the Reoviridm, Bunyaviridae, Rhabdoviridae, Arenaviridae, and others. Following this system, all the isolates used in these studies and existing characterized isolates were assigned new names according to their place of origin. Serological and genetic evidence from other studies indicated that IV16 and IV19 were strains of the same virus, as were IV22lIV25, and IV31/IV32, whereas IV23 appeared to be a distinct entity related to the main group (reviewed by Williams and Cory, 1994). Therefore, these viruses appear as tentative members of the polyiridovirus complex (Fig. 4). It is important to note that using a geographical descriptor is in no way supposed to reflect information regarding the distribution or origin of a virus, for, as has been pointed out, diseases tend to travel with their hosts and so exotic hosts introduced to novel locations may harbor pathogens originating far away (J. Kalmakoff, 1995, personal communication). In addition, flexibility and common sense are required when adopting this system, and names should be selected that are essentially convenient to use. Thus, the type species of the Iridovirus genus was named Dazaifu IV (this being an earlier source of infected material mentioned in the original paper) rather than after the place where the authors themselves isolated infected stem borers: Tsukushinomachi (Fukaya and Nasu, 1966). This is discussed further in Section I1,E. To demonstrate the effectiveness of this system of nomenclature, it is adopted here with the historical type numbers used alongside for maximum clarity.

D . Alternative Approaches Schnitzler and Darai (1993) used a PCR technique developed for use with Dazaifu IV [IV6] (Stohwasser et al., 1993) of selecting highly conserved regions of the MCP gene from published sequences (Tajbakhsh et al., 1990a; Cameron, 1990) to construct primers for use in the amplification of the homologous LCDV-1 gene. Gene fragments amplified by PCR were used to probe an LCDV-1 gene library at high

IRIDOVIRUSES

359

stringency. Library fragments showing clear hybridization to the probes were sequenced and compared to the MCP gene sequences of the other isolates. Comparative sequence analysis revealed that the deduced amino acid sequences of LCDV-1 had a sequence identity of 49.1% with Aberystwyth IV [IV22], 50.3% with Plowden IV [IVl], and 53% with Dazaifu IV [IV61. This pattern of interrelatedness is in agreement with the results of Williams and Cory (1994). In addition, MCP sequences were compared to the currently orphaned African swine fever virus (ASFV, previously a member of the Zridouiridae)and found to have a 16%amino acid sequence identity (38.1%similarity). ASFV has clear morphological similarities with iridoviruses but has a distinct genomic organization. The genome of ASFV is not circularly permuted or terminally redundant, but rather is cross-linked and unmethylated and appears more closely related to the poxviruses. ASFV is currently an orphan virus (Willis, 1990) but the sequence comparison results suggest a common ancestry. Current work in progress is following the Schnitzler and Darai (1993) approach for a broad selection of invertebrate IVs. The use of primer sequences from the conserved MCP gene regions is proving to be a successful tool for the direct amplification and sequencing of gene fragments (R. Webby and J. Kalmakoff, 1995, personal communication). This approach has the advantage that it is highly reproducible both within and among different laboratories given standard sets of primers and cycling conditions and is more quantitative than, for example, DNA hybridization or serological comparisons. When completed, such studies should provide essential information on the phylogenetic relationships among a variety of isolates from different hosts. Iridoviruses contain an internal lipid layer which usually differs in composition from that of the host cell. Williams and Thompson (1995) reasoned that the apparently unique lipid composition of these viruses could be used as an indicator for comparative studies. They compared the fatty acid profiles of eight IV species grown in the same lepidopteran host, Galleria mellonella. In all cases the virus fatty acid profiles were markedly different from that of host material. The isolates fell into two main groups of equal size: one containing Plowden IV, Tia IV, Nelson IV, and Riverside IV [IVl, IV2, IV9/10/18, and IV31, respectively] and the other group containing Ntondwe IV, Aberystwyth IV, Stoneville IV, and San Miguel IV [IV21/28, IV22, IV30, and Anticarsia gemmatalis IV, respectively]. Consequently, patterns of similarity in virus fatty acid profiles did not at all resemble the patterns of genetic or serological relatedness observed previously. Lipid composition does not seem to be a useful indicator of IV interrelationships.

TREVOR WILLIAMS

360

E . Suggested Changes to Current Classification The ICTV dictates that a polythetic system be employed for the classification of virus species. Strains of a polythetic species need not have a single defining characteristic in common. Rather, viral entities are united under the term species if they share a large proportion of predefined key features (Van Regenmortel, 1990). The problem then arises of precisely defining the set of characteristics on which the polythetic system is based. Without precise (preferably quantitative) definitions for each characteristic in the system, species definition would depend too heavily on individual interpretations of what should or should not be considered as separate species, leading to a lack of standardization and increased confusion. The flexible nature of polythetic classification would then work against a clearer understanding of virus taxonomic relationships. The definition of precise characteristics for iridoviruses is in its earliest stages. The ICTV recommends that characteristics be included from 10 different domains (structure, genetic properties, antigenic properties, ecology, replication, etc.), but the scarcity of information means that such complete definitions are not possible for iridoviruses. Many other lesser-studied virus families will probably experience the same problems. There appears to be no reason at present to change the existing genera of the Iridoviridue or recognize additional ones. However, changes are necessary in the status of a number of isolates currently assigned to vertebrate and invertebrate genera and, I would argue, in the nomenclature of the invertebrate isolates. 1 . Invertebrate Viruses

Many of the early isolates currently recognized by the ICTV seem no longer to exist. For these “ancient” isolates, data available are insufficient to recognize the moment of their reisolation because they have never been adequately characterized. As in any taxonomic system, a name cannot be assigned to an entity until the features which identify that entity have been determined. Consequently, many of the currently recognized isolates are simply records of IV-susceptible host species. The isolates themselves fall within the ranks of the “undescribed to all intents and purposes and should not be recognized. The isolates for which no more than reports of an isolation exist should only be considered as records of host species susceptible to IV infection (listed in Williams, 1994). In this respect it is remarkable to note, given the interest in insect vector control, that Vero Beach IV [IV31 from Aedes taeniorhynchus seems to be the sole surviving member (and thus type species) of the

IRIDOVIRUSES

361

genus Chloriridovirus. In the ICTV reports of Mathews (19821, Francki et al. (19911, and Murphy et al. (19951, a number of isolates were assigned to the Chloriridouirus genus: IV4, IV5, IV7, IV8, IV11, IV12, IV13, IV14, and IV15 (Table I). Such assignations are inappropriate because (i) there are insufficient characterization data to define these isolates as species (probably none of the original isolates are still in existence, so they cannot be characterized a posteri); (ii) large particle size (-180 nm) is the major criterion for assignation of isolates to the genus, but IV7, IV11, and IV13 are all “small” in size (-130 nm) and only IV12 has been reported to have serological similarity to IV3 (Tinsley et al., 1971); (iii) these isolates are united only by the fact that they all come from Diptera, but host of origin is not one of the criteria for assigning isolates to any of the genera-likewise, there are dipteran isolates which are correctly placed in the genus Zridovirus, for example, Plowden IV [IVl] and Aberystwyth IV [IV22/IV25]. The correct placing of an isolate from Chironomus plumosus as a tentative member of the genus Chloriridovirus is also inappropriate because genetic or serological data on this isolate are not available and the size of the particle (145 nm) would place it in the genus Zridouirus. This isolate was unusual in the abundance of fibrillar structures originating from the viral capsid, and it did not show iridescence (Stoltz et al., 1968). It seems doubtful that this isolate is still in existence (R. Webby, 1995, personal communication; D. B. Stoltz, 1995, personal communication). The genus Iridouirus also currently contains isolates that no longer exist (Table I) and that should be reisolated and characterized in the appropriate manner before being recognized taxonomically. Of the isolates listed by type number in the Murphy et al. (1995) report, the following are not valid: IV17 from Pterosticus madidus (coleopteran), IV 19 from Odontria striata (coleopteran), IV20 from Simocephalus expinosus (daphnid), IV25 from Tipula sp. (dipteran),IV26 from a mayfly nymph (ephemopteran). The status of IV27 from Nereis diuersicolor is not known; no genetic or serological data are available for IV27. (There is evidence that IV19 and IV25 are actually strains of other recognized viruses; see Section 11,B.I Limited characterization information is available for three isolates previously not recognized by the ICTV and not assigned type numbers: Phylophaga anxia IV, Scapteriscus aclectus IV, and Simulium vittatum IV (Poprawski and Yule, 1990; Boucias et al., 1987; Erlandson and Mason, 1990). However, material for study of these isolates does not appear to be available. Consequently, no comparative studies are possible, and it is inappropriate to recognize them as distinct entities in the absence of suitable comparative data. One additional tentative

362

TREVOR WILLIAMS

species has been characterized and found to belong to the main interrelated complex of IVs (polyiridovirus complex). This isolate from Anticarsia gemmatalis was reported to have no restriction profile similarities to previously characterized IVs. The MCP gene probe from Aberystwyth IV hybridized to a Southern blot of Anticarsia gemmatalis IV DNA, and the MCP gene fragment of this isolate was amplified by the PCR primers used previously (Williams and Cory, 1994). Genomic DNA showed significant hybridization only to members of the polyiridovirus complex. This isolate was assigned a geographical name in line with the new nomenclature: San Miguel IV. 2 . Vertebrate Viruses

Similar to the invertebrate genera, the vertebrate genera currently include many viruses for which characterization data are not available. At best, there are some serological studies of relatedness (described earlier). The following members of the Ranauirus genus have not been adequately characterized to merit inclusion in the genus: frog virus 1, frog virus 2, frog virus 5-24, frog virus L2, frog virus L4, frog virus L5, Lucke triturus virus LT2-LT4, Newt virus T6-T20, Xenopus virus T21. Other members, namely, tadpole edema virus (TEV) and Lucke triturus virus 1 (LTl), have been demonstrated to be strains of frog virus 3 by restriction endonuclease analysis and DNA hybridization studies. The restriction profiles were nearly identical among these viruses (Essani and Granoff, 1989). As a consequence, FV3 should be the sole established member (and type species) of the genus Ranauirus. Several studies have examined the relationships among vertebrate iridoviruses not yet recognized by the ICTV. An isolate of epizootic hematopoietic necrosis virus (EHNV), an iridovirus isolated from sheatfish (Silurus glanis), and an iridovirus isolated from catfish (Zctalurus melas) have been reported from fish showing similar symptoms of systemic disease. These isolates had near identical particle sizes when negatively stained and appeared to be strains of one species by cross indirect immunofluorescence assays and by comparison of polypeptide profiles (PAGE). Serologically and by PAGE analysis, the piscine isolates were shown to be related to the type species of the genus Ranauirus, frog virus 3 (FV3) (Hendrick et al., 1992). It is appropriate to synonymize these isolates as one species on a tentative basis pending additional genetic studies. The name epizootic hematopoietic necrosis virus (EHNV) describes the disease and is a well-established name. A iridovirus from the burrowing frog Lymnodynastes ornatus has been characterized in a comparative study with EHNV isolates and compared to published data for FV3. This virus, Bohle virus, was shown to be closely related but distinct from EHNV in terms of particle

IRIDOVIRUSES

363

size, protein, antigenic and genomic characteristics (Hengstberger et al., 1993). Both EHNV and Bohle virus showed characteristics which indicated their similarities to FV3. It is appropriate to assign these two species, Bohle virus and EHNV, to the Ranauirus genus on a tentative basis pending further comparative work with FV3. This change extends the known range of hosts for members of the genus Ranauirus to include piscine hosts. An iridovirus-like agent has been reported from lymphocystis lesions of the common octopus (Runnger et al., 1971). This was previously recognized as a possible member of the genus Lymphocystivirus. However, the isolate has not been characterized or compared genetically or serologically with other members of the genus or with members of other iridovirus genera. There is no basis for recognizing this entity as an iridovirus species, tentatively or otherwise. The genus Lymphocystiuirus should comprise just two species, LCDV-1 (type species) and LCDV-2. Similarly, the genus “Goldfish virus-like” has two members (GFV-1 and GFV-21, but only GFV-1 has been characterized and included in comparative studies (Essani and Granoff, 1989).Nothing is known of GFV-2, and it is not appropriate to recognize this isolate as a distinct species. 3. New Classification Scheme

Recommendations for these changes are now being made. Incorporation of all the changes suggested above would result in a new classification scheme (Table 111). The new scheme would greatly simplify the taxonomic situation across all the recognized genera and would permit novel isolates to be characterized and assigned to genera in a systematic way. The use of predescribed characteristics for defining iridovirus species will greatly assist in this. There are now a number of options available for the nomenclature of invertebrate iridescent viruses. The most conservative measure would be to retain the current system unchanged and wait for additional reports to support or refute the studies described above. A second option would be to retain the current system with some changes: synonymize the type numbers of the isolates in line with recent findings and current work in progress. This would presumably be an interim measure while the ICTV decided the best course to take in IV nomenclature and would allow time for some other system to be proposed, perhaps based on comparative sequence data for the MCP gene (although phylogenetic systems of classification currently fall outside the remit of the ICTV). A third option would be to adopt the proposed system of geographical descriptors and rename the characterized isolates accordingly. This would be the boldest option and, I would argue, the best

364

TREVOR WILLIAMS

TABLE I11 PROPOSED NEWSYSTEMFOR IRIDOVIRUS TAXONOMYGenus

Recognized tentative species6

Tentative members

Iridovirus

Plowden IV [IVll Tia IV [IV2] Dazaifu IV [IV6] Nelson IV [IV9/IV10/IV181 Timaru IV [IV16/IV191 Ntondwe IV [IV21/IV28] Aberystwyth IV [IV22/IV25] Uitenhage IV [IV231 Srinagar IV [IV24] Fort Collins IV [IV291 Stoneville IV [IV30] Riverside IV [IV31/IV321 San Miguel IV (Anticarsia gemmatalis IV)

None

Chloriridouirus

Vero Beach IV [IV3]

None

Ranauirus

Frog virus 3

Epizootic hematopoietic necrosis virus Bohle virus

Ly mphocystivirus

Lymphocystivirus type 1 Lymphocystivirus type 2

None

“Goldfish virus-like”

Goldfish virus type 1

None

a

Following changes outlined in text.

* Previous type numbers, where applicable, are given in brackets.

option. However, others have argued for a more cautious approach to nomenclature changes, pointing out that this is a major step and one that would be difficult to correct if found to be obstructive or unworkable in the future. Perhaps the main potential problem with geographical descriptors is the lack of internationality of place names, but this could be overcome with careful selection of a name from the available local options, as occurs in other virus families. Decisions on the above recommendations rest with the ICTV executive committee. Their report is expected in 1996.

111. STRUCTURE A . Physicochemical Properties The physicochemical properties of members of the family Iridoviridae are fairly well established (see reviews by Bellett, 1968;

IRIDOVIRUSES

365

Goorha and Granoff, 1979; Hall, 1985). Briefly, the molecular weight of intact virions is between 500 and 2000 million with a density of 1.26-1.6 g cm-3 (Aubertin, 1991; Cole and Morris, 1980). The small IVs all have an ssO,+,, of approximately 2200 (Kelly and Robertson, 1973) or up to 4458 for the larger Vero Beach IV [IV31 chloriridovirus (Matta, 1970; Wagner et al., 1973). Some 12-17% of the weight of the particle is dsDNA (Bellett and Inman, 1967;Kalmakoff and Tremaine, 1968; Glitz et al., 1968;Matta, 1970; Stadelbacher et al., 1978).The GC content is typically around 29-32% for the small invertebrate IVs (Glitz et al., 1968; Black et al., 19811, 54%in the chloriridovirus Vero Beach IV [IV3] (Wagner and Paschke, 1977),30.7%for LCDV-1 (Darai et al., 1983),and 53%in FV3 (Smith and McAusland, 1969;Houts et al., 1970). The usual number of polypeptides resolved by one-dimensional PAGE is 20-32 (Krell and Lee, 1974; Barray and Devauchelle, 1979; Elliott et al., 1980a; Cole and Morris, 1980; Black et al., 1981; Flugel et al., 1982; Tajbakhsh and Seligy, 1990), although two-dimensional PAGE has revealed additional polypeptide diversity (Cerutti and Devauchelle, 1985). Polypeptide sizes are usually in the range l l to 200 kDa, although smaller (Tajbakhsh and Seligy, 1990) and larger proteins have been reported (Krell and Lee, 1974; Barray and Devauchelle, 1979; Cerutti and Devauchelle, 1990). Several virionassociated enzymes have been detected, the functions of which are discussed later. The structure of the capsid, lipid membrane, and core are considered in turn.

B . Capsid The capsid comprises an icosahedral lattice of closely packed hexagonal subunits of 7-9 nm diameter and about 7-9 nm height (Wrigley, 1969, 1970; Stoltz, 1971, 1973; Murti et al., 1984). The total number of subunits is probably 1472 for Plowden IV [IVl] and Tia IV [IV21.These subunits are arranged in 20 trisymmetrons (each of 55 subunits) and 12 pentasymmetrons (each of 31 subunits). The corners of the trisymmetrons are not perfectly aligned with those of the pentasymmetron but are offset (skewed) about the pentasymmetron by 3 subunits, at least in Plowden IV [IVlI (Manyakov, 1977). The larger viruses (e.g., an isolate from the midge, Chironomus pZumosus, -165 nm diameter) have larger trisymmetrons, with 78 subunits, giving a total of 1560 subunits per particle (Stoltz, 1971, 1973). The capsid subunits are comprised of a single polypeptide of 48 to 55 kDa (typically 50 kDa), the major capsid protein (MCP), which comprises some 40-45% of the total particle polypeptide (Willis et al., 1977; Moore and Kelly, 1980; Aubertin et al., 1981; Black et al., 1981; Flugel et al., 1982; Davison et al., 1992; Schnitzler and Darai, 1993;

366

TREVOR WILLIAMS

Stohwasser et al., 1993).The MCP protein is highly conserved and has proved to be a valuable indicator in comparative studies. The length of the MCP varies from 459 amino acid residues in LCDV-1 (Schnitzler and Darai, 1993) to 472 amino acids in Aberystwyth IV [IV221 (Cameron, 1990), with intermediate values in Plowden IV [IVlI and Dazaifu IV [IV6] (Tajbakhsh et al., 1990a; Stohwasser et al., 1993). A series of SDS-PAGE and surface labeling studies under different conditions has indicated that the MCP is the basis for two different structures in the capsid of Dazaifu IV. A trimeric form of the MCP, held together (presumably) by hydrogen bonding, is located on the outer surface of the capsid, whereas a covalently bonded trimer of the same protein lies beneath the surface layer (reviewed by Cerutti and Devauchelle, 1990). Tia IV [IV2] is alone in being suspected of having two MCPs of different sizes; one of 53 and one of 55 kDa (Elliott et al., 1977).Whether the two proteins are structurally different or the smaller is produced by cleavage or degradation of the larger is not known. There is also evidence that an infected cell protein (ICP 38) is located externally on the FV3 particle (Chinchar et al., 1984). For a number of isolates, from both vertebrate and invertebrate hosts, the presence of fibrils attached t o capsid subunits has been reported (Fig. 5). Fibrils were particularly evident in an isolate from the midge (Chironomus plumosus (Stoltz et al., 1968; Stoltz, 1971)but also in LCDV (Zwillenberg and Wolf, 1968; Midlige and Malsberger, 1968; Yamamoto et al., 1976). Neutron scattering measurements of FV3 particles indicated that half the volume of the capsid was water, a level of hydration expected in the presence of fibrillar structures (see also Lunger and Came, 1966). In LCDV, the fibrils were measured at 200 nm length and 4 nm width (Zwillenberg and Wolf, 1968), but in other iridoviruses the fibrils are far shorter (-2.5 nm) and appear as a fringe around the edge of the capsid (Willison and Cocking, 1972; Cole and Morris, 1980; Black et al., 1981; Devauchelle et al., 1985a; Flugel, 1985). Short fibrils may have terminal knobs (Stoltz, 1971, 1973). C . Lipid Membrane The lipid layer is intimately associated with the capsid (Stoltz, 1973; Cuillel et al., 1979; Klump et al., 1983). This layer is an essential component for infectivity in the vertebrate iridoviruses (Wolf et al., 1968; Willis and Granoff, 1974; Berry et al., 1983; Langdon et al., 1986; Speare and Smith, 1992).It has become generally accepted in the literature that invertebrate IVs are ether-resistant. Day and Mercer (1964) stated that the infectivity of Tia IV [IV21 was unaffected by treatment with ether or chloroform, although no data were presented in support

IRIDOVIRUSES

367

FIG.5. Trisymmetrons from the capsid of an iridovirus of Chironomus plumosus (Diptera) showing abundant fibrillar structures attached. Bar: 100 nm. (Photograph supplied by D. B. Stoltz.)

of this assertion and, given the biological activity of membrane extracts during the early stages on infection (described later), such statements must be viewed with suspicion. This basic characteristic of the invertebrate viruses clearly requires confirmation. The lipid layer is about 4 nm thick (Cuillel et al., 1979; Kelly, 1985). Complexes of protein pass through the intermediate lipid layer and appear to connect capsid and core polypeptides (Aubertin et al., 1980; Cerutti and Devauchelle, 1982; "ripier-Darcy et al., 1982; Klump et al., 1983). The total lipid content of the particle has been reported as 5.2% (dry weight) of Plowden IV [IVll (Kalmakoff and Tremaine, 1968), 7%

368

.

TREVOR WILLIAMS

for Dazaifu IV [IV6] (Balange-Orange and Devauchelle, 1982), 9% of Dazaifu IV, Tia IV [IV2], and FV3 (Kelly and Vance, 1973; Willis and Granoff, 1974), and 12.5% for purified cores of Vero Beach IV [IV31 (Wagner et al., 1975). Higher values (17%) have been reported for a lymphocystis disease virus isolate, although the assayed material appeared to be a heterogeneous mixture of filled and unfilled virions (Robin et al., 1993). The phospholipid content has been reported as 90% of total lipid for FV3 (Willis and Granoff, 1974), 75% for Dazaifu IV (Balange-Orange and Devauchelle, 1982), and 44% for Dazaifu IV and Tia IV (Kelly and Vance, 1973). The fatty acid composition of these lipids has been reported as nearly identical to the host cell in FV3 (Willis and Granoff, 1974) but markedly different in the invertebrate viruses, Dazaifu IV (Balange-Orange and Devauchelle, 1982), and eight other IVs (Williams and Thompson, 1995).In Dazaifu IV [IV6] the phospholipid composition was not sensitive to the type of host: lymantrid cell line or noctuid larvae (Balange-Orange and Devauchelle, 1982).Such findings and the observation that iridovirus particles do not obviously bud through a particular organelle membrane during synthesis have led to the assumption that iridoviruses acquire their lipid component by d e nouo synthesis, a situation also believed to exist in poxviruses. However, Schmelz et al. (1994) have shown that vaccinia virus inner lipid envelopes are derived from an intracellular compartment between rough endoplasmic reticulum (ER)and the Golgi apparatus. Virally encoded proteins appear in the trans Golgi network shortly after infection and appear to facilitate the sequential wrapping events of vaccinia. The relevance of these observations for iridoviruses is not known. Viruses released by budding may possess an outer envelope (Yule and Lee, 1973; Webb et al., 1976; Braunwald et al., 19791,but this is not an essential component for infectivity (Granoff et al., 1966). Some fish iridoviruses also appear to have an outer envelope in ultrathin sections, which may be double layered (Flugel, 1985).

D . Core The particle core is an electron-dense, highly hydrated entity containing about 80% water in FV3 (Cuillel et al., 1979). Thermodynamic studies were consistent with the presence of a nucleosomal structure in Dazaifu IV [IV61 (Klump et al., 19831, and ultrasonic adsorption studies have supported the concept that there is structural organization among the DNA and protein components of the core (Robach et al., 1983). Freeze-etching of FV3 particles in cells showed randomly oriented rods some 10 nm width to be evident in the core. These rods

IRIDOVIRUSES

369

appear to be part of the DNA-protein complex arranged so as to form a long coiled filament (Tripier-Darcy and Nermut, 1983). Lipidenveloped cores of Vero Beach IV UV31 were reported to have a diameter of 176 nm when negatively stained compared to 224 nm for intact particles (Wagner et al., 1973). Cores were not infectious to mosquito cells or larvae; apparent failure of the cores to attach to cells may have been the cause. Polypeptide profiles of cores were remarkably similar to those of whole particles, the main difference being the absence of the 55-kDa capsid protein (Wagner et ul., 1975). This suggests that much of the polypeptide (structural and enzymatic) complexity of IVs is associated with the core and membrane. Cerutti and Devauchelle (1985) reported that at least six DNA-associated polypeptides are localized within the core of Dazaifu IV [IV6], ranging from 12.5 to 81 kDa, of which the 12.5-kDa protein was the major species. A number of other IVs also have a major polypeptide species close to this mass (Elliott et al., 1977; Carey et al., 1978; Kelly et al., 1979; Cole and Morris, 1980; Tajbakhsh and Seligy, 1990).

E . Iridescence Phenomenon Most, but not all, of the invertebrate iridescent viruses iridesce. The vertebrate isolates do not iridesce in host tissues or as purified pellets of virus, although there is one report to the contrary (R. Walker, personal communication in Stoltz, 1973).It is easy to get transfixed by the iridescence of invertebrate isolates, and many authors, myself included, find it hard to resist a few choice adjectives concerning the eyecatching nature of patent IV infections. The iridescence occurs because, at high densities, IV particles selfassemble and crystallize in host cells following the laws of entropy, that is, crystalline organization is the arrangement with the lowest energy. The spacing between the particles is such that light reflected from the surface of the viral arrays interferes with newly arriving light, resulting in “Bragg” reflections (Klug et al., 1959). For such events to occur, a superabundance of virus particles must be present in the host tissues, so that iridescence is not seen until fairly late in the infection cycle. The colors commonly seen in patent IV infections include lavender, blue, or turquoise for the small IVs (genus Iridouirus) and green-yellow, orange, or red for the large IVs of mosquitoes (genus Chloriridouirus). These are of course the colors of the rainbow, and the Greek word iridos translates as “shining like a rainbow” (Aubertin, 1991). There is a direct relationship between particle size and iridescent color as described in the formula:

TREVOR WILLIAMS

370 A

=

2rn{D2 - [(D/~)cos 3O0I2}

where A is the wavelength of the iridescent color, rn is the refractive index of the material, and D is the particle center-center spacing, which is the same as particle diameter in a close-packed tetrahedral arrangement of particles. Although most of the measurements of iridoviruses are made in ultrathin sections, negative staining may actually yield more realistic results. Negatively stained particle size measurements have been reported at 170 nm for Plowden IV [IVl] (Iridouirus)and 224 nm for Vero Beach IV [IV3] (Chloriridouirus) (DeBlois et al., 1978; Wagner et al., 1973). Using the above measurements in this formula and assuming an average refractive index for biological matrices of 1.52, iridescent colors of blue and orange are correctly predicted for Plowden IV and Vero Beach IV, respectively (Hemsley et al., 1994). Certain isolates (e.g., FV3, LCDV-1, and an isolate from the midge, Chironornous plurnosus) appear to assemble into paracrystaline structures in the cytoplasm but show no iridescence. This could be due to the presence of the fibrillar structures described earlier which may increase the interparticular spacing and prevent the occurrence of Bragg reflections, etc. The relationship between fibrils and iridescence is not clear.

IV. REPLICATION The pathology of invertebrate IVs in host cells is characterized by cell enlargement, rounding and detachment, cell-cell fusion, vesicle production from protrusions at the cell surface, increased vacuolization, nuclear hypertrophy, decondensation of nuclear chromatin, and, most characteristically, the formation of electron-dense viroplasmic centers (virus assembly sites) in the cytoplasm. Not all effects are seen simultaneously, and details depend on the virus and host under study (Hall and Anthony, 1971; Webb et al., 1973; Lee and Brownrigg, 1982; Lea, 1985; Czuba et al., 1994). Kelly and Tinsley (1974) noticed that infection of cells with Tia IV DV21 or Dazaifu IV [IV61caused a marked contraction of mosquito and lepidopteran cells (inuitro) with 72 hours. The cellular changes elicited by components of the Dazaifu IV [IV6] virion have been found to be dramatic both physically, in the formation of syncytia, and physiologically, as the shutdown of host cell macromolecular synthesis (Cerutti and Devauchelle, 1979,1980).Viral particles caused extensive and very rapid cell fusion in permissive and nonpermissive cell lines

IRIDOVIRUSES

371

alike. UV-inactivated virus elicited the same effect. The incidence of cell fusion was sensitive t o the multiplicity of infection. The syncytic effect of virions was even observed in vertebrate (monkey) cells, albeit in a less spectacular fashion. Host hormones can have a profound effect on patterns of replication of IVs (Kloc et al., 1984). In general, the cytopathology of FV3 shows similar features to those above, although the effects on gross pathology at the tissue level are markedly different (e.g., Wolf et al., 1968). A. Cell Penetration and Uncoating

Patterns of replication and the time course of replicative events vary with temperature, cell type, virus, and multiplicity of infection, among other things. Frog virus 3 replicates at temperatures between 12" and 32°C (Gravel1 and Granoff, 1970). Replication of invertebrate IVs is also inhibited at temperatures over about 30°C, which no doubt reflects the damp habitats and cool body temperature of the majority of their poikilothermic hosts (Tanada and Tanabe, 1965; Day and Dudzinski, 1966; Carter, 1975; Witt and Stairs, 1976). In cell culture or in the insect hemocoel (following injection of inocula) virus particles appear to be adsorbed onto the plasma membrane of the host cell and enter by pinocytosis. Direct penetration of cells has also been suspected for FV3 following the observation of particles lying free in the cytoplasm (Houts et al., 1974). Host macromolecular synthesis is rapidly inactivated. In the cytoplasm, virions may be enclosed by a membrane or may exist as free particles during uncoating (Kelly and Tinsley, 1974; Webb et al., 1976). In cells infected by Plowden IV [IVlI, pinocytotic vesicles containing virus aggregate and fuse to form lysosomes full of virus particles, which have been reported as the site of viral uncoating (Mathiesen and Lee, 1981). Uncoating renders the virus sensitive to DNase treatment. Following cell penetration by FV3, approximately 50% of viral DNA became DNase-sensitive within 1 hr. Uncoating appears not to require any form of protein synthesis, as the rate of uncoating was similar in the presence or absence of a protein synthesis inhibitor (Smith and McAusland, 1969).

B . Shutdown of Host Macromolecular Synthesis Whole virions of FV3 have been shown to elicit the shutdown of cellular macromolecular synthesis in permissive and nonpermissive host cells alike. This occurs even if the virions are inactivated by heat or UV light (Maes and Granoff, 1967a; Guir et al., 1970; Willis and

372

TREVOR WILLIAMS

Granoff, 1976) or if a soluble extract of the virions is applied to cells (Aubertin et al., 1973,1976).Goorha and Granoff (1974) superinfected cells with active FV3 following macromolecular shutdown caused by 4-hr exposure to heat-inactivated FV3; they observed a normal cycle of replication with normal kinetics and a normal yield of progeny virus. Studies by Drillien et al. (1977) have indicated that a single FV3 particle can elicit the inhibition of DNA, RNA, and polypeptide synthesis in a cell, which results in cell inactivation (defined as lack of ability to form colonies). The kinetics of cell inactivation were very similar when cells were treated with a soluble extract of FV3. This was interpreted as evidence that single entities (a single polypeptide or group of polypeptides) were responsible for cell macromolecular synthesis shutdown. This polypeptide must be heat stable. Cordier et al. (1981) treated mammalian cells with heat-inactivated FV3 or soluble FV3 extracts and then assayed the rate of protein synthesis in cell-free systems. Exposure to virus or viral extracts greatly diminished the rate of protein synthesis in uitro, and the effect was not reversed by altering the sodium ion concentration (which usually affects the initiation of intracellular protein synthesis). The cause of the inhibition was traced to impaired activity in the ribosomal fraction of the cell lysate. Ribosomal preparations from the livers of mice treated with FV3 displayed similarly reduced transcriptional activity (Elharrar and Kim, 1977). As pointed out previously, this shutdown phenomenon has been extremely useful in the study of viral replication, as host cell synthetic pathways can be completely disrupted by brief exposure to inactivated virus. Subsequent incorporation of labeled compounds following superinfection with viable virus will be due t o viral macromolecular synthesis alone (Willis et al., 1985). Similar shutdown properties have been studied in Dazaifu IV [IV6]. Shutoff of all macromolecular synthesis in mosquito cells was achieved within 1 hr of exposure to virions, and the rate of shutoff was dependent on the multiplicity of infection. A soluble extract of Dazaifu IV [IV61 was shown to have the same properties as intact particles when tested on invertebrate or vertebrate cells under permissive and nonpermissive conditions (Cerutti and Devauchelle, 1980). The lipid content of the virus was extracted and reconstituted as vesicles of trilayered lipid membrane, 50-200 nm in diameter. These vesicles had the same cell fusion and macromolecular shut-down properties as whole virions (Cerutti and Devauchelle, 1982). The vesicles were found to comprise 86%of the viral phospholipid and 4% of total viral protein. The proteins were identified as five entities up to 53 kDa in size, the major species being of 11 kDa (Cerutti and Devauchelle, 1990). Inci-

IRIDOVIRUSES

373

dentally, populations of subgenomic viral DNA fragments (-75 kbp) contaminating the vesicle preparations were later found to be infectious (at very low levels), indicating that recombination had occurred among coinfecting fragments to form viable viral genomes. Viral DNA alone was not infectious (Cerutti et al., 1989).

C . DNA Replication Descriptions of DNA replication can be found in Murti et al. (1985a) and Willis et al. (1985). Most detailed studies on iridovirus replication have used FV3. In this species, replication of virus DNA occurs in two phases: a nuclear phase and a cytoplasmic phase. This two-site DNA replication is unique for animal viruses. A functional nucleus is an essential cellular component for virus replication (Goorha et al., 1977). Following shutdown of host cell macromolecular synthesis by treatment with inactivated FV3, viral DNA synthesis was detected in the nucleus by autoradiography. Purified nuclei were shown to contain viral DNA by hybridization (Goorha et al., 1978). During the first 3 hours postinfection, viral DNA is synthesized solely in the cell nucleus. DNA produced here is genomic or up to twice genomic size. Pulse-chase experiments showed that the majority of viral DNA synthesized in the nucleus moved to the cytoplasm (Goorha, 1982). In the cytoplasm viral DNA exists as long concatamers (more than 10 times genomic length). A temperature-sensitive FV3 mutant held at nonpermissive temperatures showed low levels of DNA production in the nucleus, but concatamers did not appear in the cytoplasm. When shifted to a permissive temperature, large concatamers quickly appeared in the cytoplasm (within 30 minutes), indicating that concatamers are formed as a result of recombination among the lengths of viral DNA produced in the nucleus (as observed in the bacteriophage T4) rather than, as an alternative hypothesis, that concatamers arise by a rolling circle process. The mutant FV3 appears to encode a protein involved in the second stage of DNA replication. The activity of this protein, possibly some 31 kDa in mass (Martin et al., 19841, was not affected by the addition of protein synthesis inhibitor at the moment of switch between nonpermissive and permissive temperatures, indicating that it is the protein itself which is thermosensitive rather than the mechanism of de nouo synthesis (Goorha and Dixit, 1984). The similarities between FV3 and phage T4 in DNA replication may extend to the packaging of DNA into the viral capsid. In T4, concatameric DNA is packaged until the head of the phage is full, and this leads to a genome that is circularly permuted and terminally redun-

374

TREVOR WILLIAMS

dant. Because these are features of the FV3 genome (and several other members of the Iridouiridae), it is assumed that “headful” packaging of DNA also occurs in these viruses (Goorha and Murti, 1982; Darai et al., 1983; Soltau et al., 1987; Ward and Kalmakoff, 1987).The evidence for genomic circular permution and terminal redundancy are outlined in Section V.

D. Methylation of Viral DNA High levels of methylation of cytosine at CpG residues are only seen in iridoviruses from vertebrate hosts, namely, FV3 (Willis and Granoff, 1980), LCDV-1 (Wagner et al., 19851, Goldfish iridovirus (Essani and Granoff, 19891, and epizootic hematopoietic necrosis virus (EHNV) (Eaton et al., 1991), as indicated by treatment of genomic DNA with the endonuclease HpaII, which recognizes CCGG sequences. DNA methylated at CpG residues is not cleaved by HpaII but is cleaved by MspI, which recognizes identical sequences but which is not sensitive to methylation. There are conflicting reports as to the site of DNA methylation in FV3. Eukaryotic DNA methylases are found in the nucleus, whereas pulse-chase studies using radiolabeled uridine (a cytosine precursor) indicated that cytosine of FV3 was methylated in the cytoplasm. Endonuclease studies have supported this; FV3 DNA isolated from the cytoplasm was cleaved by MspI but not HpaII, whereas viral DNA isolated from the nucleus was cleaved by both enzymes. Host DNA methylase activity in the nucleus appears to be inhibited by the infecting virus, and there is evidence of differences in substrate specificities for host and viral DNA methylases. The FV3 DNA methylase showed an affinity for dsDNA and an ability to methylate a broad range of natural and synthetic DNAs in vitro (Willis et al., 1984a). However, Schetter et al. (1993) reported that DNA methylase activity was detected only in the nucleus, not in the cytoplasm of FV3-infected cells, which suggests that DNA is methylated prior to export to the cytoplasm. The FV3 DNA methylase comprises two polypeptides of 18 and 26 kDa, and there is also some evidence for endonuclease activity (possibly related to a polypeptide of 30 kDa) (reviewed by Essani, 1990). Methylation has been speculated as a mechanism to prevent nicking of viral DNA by viral endonucleases. In the T4 bacteriophage, DNA that has been nicked is not packaged. A DNA methylase inhibitor does not appear to influence the rate of macromolecular synthesis but dramatically reduces the rate of production of virions in FV3-infected cells (Goorha et al., 1984).Methylation appears to be important in the packaging of DNA into virions. Apparently, FV3 mutants that lack DNA

IRIDOVIRUSES

375

methylase appear able to survive only because they do not show any viral endonuclease activity (Essani, 1990). The goldfish iridovirus is unique in that cytosines appear to be methylated at CpT as well as CpG sites (Essani and Granoff, 19891, perhaps reflecting a requirement for additional protection against endonucleases in this virus.

E . Transcription Transcription and control of transcription of iridoviruses appear complex and have only been studied in detail in FV3. A clear review has been provided by Willis et al. (1990). Studies of transcription have been assisted by the host macromolecular synthesis shutdown immediately following infection and the fact that most FV3 mRNAs do not have poly(A) tails to protect against degradation (Willis and Granoff, 1976).This allows the resolution of up to 47 mRNA species on denaturing acrylamide or agarose gels (Willis et al., 1977). Transcription of viral DNA in FV3-infected cells is a coordinated sequential process involving the production and regulation of mRNAs that can be classified according to their temporal sequence of synthesis. These classes are “immediate early,” “delayed early,” and “late,” although finer divisions have also been defined by the use of various temperaturesensitive mutants (Goorha and Granoff, 1979). Analysis of such mutants has indicated that transcription is not dependent on successful DNA synthesis for any but the very last mRNAs to appear (-15% of the genome) (Goorha and Granoff, 1979; Goorha et al., 1981). Other studies produced different findings. The presence of various DNA synthesis inhibitors has markedly reduced late gene expression in some studies (Elliott and Kelly, 1980; Elliott et al., 1980b; Chinchar and Granoff, 1984). One explanation may be that template DNA undergoes modifications during the course of infection that permit the stepwise sequence of transcriptional events to progress. 1 . Transcription and Nongenetic Reactivation

Immediate-early mRNAs are those transcribed in the presence of protein synthesis inhibitor (e.g., cycloheximide). The immediate-early mRNAs comprise 10 species, of which 7 with masses of 600 kDa or less can be clearly resolved on polyacrylamide gels (Willis and Granoff, 1978). These sequences represent about 32% of the genome, although probing of Northern blots with radiolabeled viral restriction fragments has suggested that rather more than one-third of the FV3 genome may be transcribed in the presence of cycloheximide (Mesnard et al., 1988). The presence of host RNA polymerase I1 is necessary for early transcription (Goorha, 1981), but this enzyme must be modified

376

TREVOR WILLIAMS

by a trans-acting viral protein(s) to overcome methylation of immediate early promoters. Virions of both FV3 and Dazaifu IV [IV61 possess the property of nongenetic reactivation (Gravell and Naegele, 1970; Gravell and Cromeans, 1971; Cerutti et al., 1989). UV-irradiated or heat-treated particles alone possess no infectivity but show cytocidal effects. However, when in combination, infection occurs and viable virus is produced. The UV-irradiated virus has a nonfunctional genome but possesses functional polypeptides. The heat-treated virus has denatured proteins but possesses a normal genome. The genome of the progeny virus in such coinfections is always derived from the heat-treated inoculum, whereas the UV-irradiated partner provides essential components to initiate replication. Naked “cores” or purified viral DNA can also be rendered infectious in the presence of UV-inactivated (genetically nonfunctional) virus. The titer of progeny virus was proportional to the concentration of transfecting DNA except at very high concentrations, where inhibition was observed (Willis et al., 1979a). In Dazaifu IV [IV6] the entity responsible for nongenetic reactivation was identified as one or more structural viral polypeptide(s) that could be solubilized with detergent (octylglucoside) and reconstituted with lipid vesicles derived from the virus intermediate membrane (Cerutti et al., 1989). Evidence from transcription studies suggests that the solubilized proteids) is the first of a series of trans-acting proteins which facilitate the temporal sequence of viral transcription. Transcription studies have focused on an immediate early FV3 gene (ICR 169). The ICR 169 (infected cell RNA of 169 kDa) codes for a putative protein of 18 kDa of unknown function. This RNA is produced in large quantities throughout the course of the infection cycle. Fathead minnow cells were transfected with a plasmid containing a 78-bp promoter region of the ICR 169 gene inserted upstream of a chloramphenicol acetyltransferase (CAT) reporter gene. The CAT activity was observed only when the cells were subsequently challenged by untreated or UV-irradiated FV3 but not by heat-treated FV3 (Willis and Granoff, 1985). Production of CAT-specific RNA occurred when transfected cells were infected with FV3 in the presence of cycloheximide, demonstrating a viral protein, a trans-acting factor, to be responsible, and that protein synthesis is apparently not required for mRNA synthesis from this early gene. Synthesis of CAT induced by FV3 showed marked inhibition in Chinese hamster ovary cells (CHO) cells in the presence of the polymerase inhibitor, a-amanitin, but inhibition did not occur under identical conditions in CHO cells with an a-amanitin-resistant RNA polymerase I1 (Goorha, 1981). These studies suggest the viral transactivating factor to have one or two possible modes of action. Either the

IRIDOVIRUSES

377

factor interacts with the host polymerase and modifies the enzyme to facilitate the transcription of virus-specific sequences, or the factor binds to a DNA template to enhance the attachment or processing of the sequences by host polymerase (Willis and Granoff, 1985). The latter hypothesis has some support by the fact that two viral proteins in the reconstituted membrane vesicles studied by Cerutti et al. (1989) showed marked DNA-binding properties. The activity of the ICR 169 promoter was further investigated by S1 nuclease mapping and deletion studies using the CAT reporter system. The start site of transcription for the ICR 169 gene was located 29 bases upstream of an (A-T)-rich motif: TATTTTA. Deletion of this motif reduced CAT synthesis by 85%.However, a region of 14 bases 3’ of the A-T motif was identified as of even greater importance in the function of this promoter. Point mutations or deletions in the sequence CAGGGGAATTGAAA dramatically reduced CAT activity (Willis et al., 1984b; Willis, 1987). The central GGGGAAT motif of this region is seen in a number of other viral and cellular enhancers and was demonstrated to bind strongly to a nuclear protein (Willis et al., 1990). A promoter region with significant sequence similarity to the 14base sequence above has been found in another immediate early gene (ICR 489), although at a greater distance from the 5’ end of the message. Transcription of the ICR 489 gene contrasts with that of ICR 169. As with ICR 169, a transacting viral protein(s) is required to elicit transcription. However, in the presence of cycloheximide a massive overproduction of ICR 489 is observed, indicating that transcription is normally downregulated by an early synthesized viral protein (not produced in the presence of cycloheximide). The key regions for the ICR 489 promoter are different from those of ICR 169. Transcription of ICR 489 is dependent on the presence of a CCGCCC and a CCAAT motif, which have been reported to bind cellular transcription factors. The marked differences between the promoter regions of these two immediate early genes indicates that transcription of the two genes is not directed by common signals but is under the control of different processes and different cellular transcription factors (Beckman et al., 1988). There are three delayed early mRNAs. Studies with a temperaturesensitive FV3 mutant and a phenylalanine analog (fluorophenylalanine) indicated that a virus-induced protein(s) was required for expression of delayed early mRNAs (Willis and Granoff, 1978; Goorha et al., 1979). This protein(s) is probably another transacting factor for switch on of delayed early mRNA synthesis, and it appears not to contain phenylalanine at any key sites. The remaining species of mRNA are classed as late. These mRNAs

378

TREVOR WILLIAMS

are usually synthesized after DNA replication has occurred in the nucleus. However, DNA replication is not a prerequisite for production of late proteins (Goorha et al., 1979), and temperature-sensitive mutants have been analyzed that synthesize DNA but do not produce late mRNAs (Willis et al, 1979b). Again, a trans-acting viral-induced protein appears necessary for transcription of late sequences. A 95-bp region 5‘ to the transcriptional start site has been studied using the CAT reporter system. The appearance of CAT activity in transfected cells, subsequently infected by FV3, was consistent with the “late” nature of this mRNA. No CAT mRNA was detected in the presence of cycloheximide. A motif of TATTTTA identical to that of the ICR 169 promoter was seen, but the upstream sequences ( 5 ’ ) were completely different, indicating that the late properties of the promoter reside 5’ to the (A-T)-rich region (Willis et al., 1990). In general, the TATA and CAAT motifs are found upstream of genes in a number of other iridoviruses that have been studied, wherein their function is presumed to follow that found in FV3 (Tajbakhsh et al., 1990a; Home et al., 1990; Cameron, 1990; Schnitzler and Darai, 1993; Sonntag et al., 1994; Schnitzler et al., 1994b). 2. Transcription of Methylated DNA

A CAT reporter gene under the control of a methylated adenovirus promoter was not activated when transfected mammalian cells were subsequently infected by FV3 in the presence of cycloheximide or by UV-inactivated FV3 (Gorman et al., 1982).This indicates that a virally induced protein is required to permit transcription of methylated viral DNA. This is a different situation from the trans-activating factor for the immediate-early ICRs 169/489, which are virion-associated rather than virally induced. If methylated DNA is not transcribed without prior protein synthesis, then immediate early genes must be unmethylated, or the methylation must be restricted to regions which were not crucial for transcription. Mutant promoters were produced in which each CG site was changed to methylated CCGG or CGCG by sitedirected mutagenesis followed by treatment with bacterial methylases. Transcription of the methylated mutants was observed following infection by UV-inactivated FV3 or by normal FV3 in the presence of protein synthesis inhibitor. It appears that methylation of immediate early promoters does not affect transcription, suggesting that the CG motifs appear in regions that are not important for transcription (Thompson et al., 1988). 3. Methylation and Stability of mRNA The degree of methylation of FV3 mRNAs is dependent on the sequence in which they are synthesized. All viral mRNAs were termi-

IRIDOVIRUSES

379

nally blocked and methylated with three common types of nuclear RNA cap structures: m7GpppNlmpN2p and m7GpppNlmpN2mpN3p comprised more than 80% of the total cap types and m7GpppNlp the rest. Between four and seven m6A residues were detected in each early viral mRNA transcript, whereas methylation was not detected in late gene transcripts (Raghow and Granoff, 1980). This may reflect the fact that transcription of early genes occurs in the nucleus, whereas later genes are transcribed in the cytoplasm (possibly from concatameric DNA) by a virally induced RNA polymerase. Evidence for the existence of such a polymerase has come from sequencing studies of Dazaifu IV [IV61. Schnitzler et al. (1994b) reported the presence of an open reading frame (ORF) encoding a putative protein of 1051 amino acids with very significant sequence similarity to the largest subunit of DNA-dependent RNA polymerase 11, although with the C-terminal domain missing. All FV3 mRNAs observed to date have terminal hairpin structures (dyad symmetry) (Willis et al., 198413; Beckman et al., 1988; Schmitt et al., 1990; Rohozinski and Goorha, 1992). Chinchar et al. (1994) examined the stability of transcripts from early and late genes with and without the transcription inhibitor actinomycin D. The level of viral transcripts remained high or increased in untreated cells infected by FV3. In the presence of actinomycin D, however, the level of early and late gene transcripts declined rapidly. Half-lives were estimated to be between 1.75 and 3.25 hours (although these values are affected by the experimental methods used). The mRNA-degrading mechanism is not known although synthesis of a virally encoded RNase is the most appealing option. These results led Chinchar et al. (1994) to suggest that the terminal hairpin structures of FV3 mRNAs play a role in signaling transcript termination rather than affecting transcript stability, as suspected previously (Aubertin et al., 1990; Rohozinski and Goorha, 1992).

F. Translation The mechanisms by which translation is regulated in iridovirusinfected cells is not well understood. Early mRNAs continue to be synthesized but are not translated late in infection (Willis et al., 1977). For example, the immediate early transcript ICR 489 (described above) can be detected long after production of the protein has peaked and declined to undetectable levels. Mesnard et al. (1988) reported the abundant production of two RNAs complementary to the 5’ region of the ICR 489. If the “antisense” RNA hybridizes to ICR 489 to prevent translation, this may be an effective way of limiting production of the early protein later in the replication cycle. Another suspected mecha-

380

TREVOR WILLIAMS

nism for translational control appears to reside in the structure of the 5’ region of the transcript. The length of an immediate early mRNA was found to change from 1.3 kb early in infection to 1.35 kb late in infection. This difference arises from an additional 50-base sequence at the 5’ end of the late transcript (Aubertin et al., 1990). Translation was observed only for the early (unmodified) species (Tondre et al., 1988). The mode of action of 5’ alterations on message viability is not known. AUG codons are signals for the initiation of translation that are detected by the 40s ribosomal subunit. The number and context of AUG codons are different in early and late mRNAs of FV3. Translation of early gene mRNAs starts at the first AUG encountered, in a context for which the consensus sequence is A/GCCAUGGG. However, for the major capsid protein mRNA (late gene), translation starts at the fourth AUG codon encountered, and none of the codons have more than one base in common with the consensus sequence (apart from the AUG codon) (Aubertin et al., 1990). Post-translational modifications appear limited to phosphorylation of some core polypeptides (Aubertin et al., 1980). No evidence of glycosylation has been obtained (Krell and Lee, 1974).

G . Packaging of Virions Different hypotheses have been advanced regarding the sequence of events leading to production of the mature virion. These ideas have arisen from electron microscope observations of particles in various stages of formation in tissue sections. Bird (1961, 1962) studied Plowden IV [IVl] and observed that the viral DNA and nucleoprotein condensed to form the core, which was then enveloped by the capsid. This interpretation found support in studies of Tia IV [IV21 and Dazaifu IV DV61 in cultured insect cells (Kelly and Tinsley, 1974). Yule and Lee (1973) described the capsid of Plowden IV [IVlI assembling face by face and then being filled by DNA and fibrillar structures through a hole left in the shell. Studies with Riverside IV [IV31/321 supported this view (Hess and Poinar, 1985). Similar scenarios were previously proposed to explain the common observation of empty capsids (Smith, 1958; Xeros, 1964), but without explicit reference to DNA entry through a capsid pore. Others have interpreted the sequence of events being more simultaneous than the stepwise mechanisms above. There is coordinated assembly of inner membrane and capsid. As the capsid enlarges it sequesters core material from the assembly site (viroplasmic center). The DNA condenses to form the core proper once capsid construction is complete (Devauchelle, 1977; Federici, 1980; Devauchelle et al., 1985a).

IRIDOVIRUSES

381

There is a striking similarity in the patterns of particle assembly in some of the vertebrate iridoviruses. In these viruses capsids initially appear as numerous angular crescent-shaped bodies in the assembly sites. As the crescents develop into icosahedrons, the early core structure becomes apparent. In some iridoviruses the degree of condensation of core material is clearly heterogeneous (“patchy”in appearance). The electron opacity of the core remains heterogeneous after apparent completion of the capsid, but the core subsequently develops into the even electron-dense structure typical of all iridovirus cores. These observations apply to the frog viruses studied by Lunger and Came (19661, a putative reptilian iridovirus from erythrocytes of a gecko (Stebhens and Johnson, 19661, goldfish iridovirus (Berry et al., 19831, and epizootic hematopoietic necrosis virus (Eaton et al., 1991). The mechanism of particle development is not well understood in FV3 (Goorha and Granoff, 1979). In tadpole edema virus, a strain of FV3 (Essani and Granoff, 19891, crescent-shaped capsid precursors develop into fully formed capsids, but condensation of the core elements does not occur until after completion of the capsid (Wolf et al., 1968). A similar situation appears to exist in assembly of LCDV virions (Zwillenberg and Wolf, 1968) although production and assembly of LCDV may be seasonally affected as indicated by the large proportion of “empty” capsids and low virion production in lesions from flounders sampled in winter months compared to abundant mature virions in lesions of fish caught in the spring (Flugel, 1985).

H . Cytoskeletal Manipulation Frog virus 3 manipulates the host cytoskeleton to assist in the production and release of virions. The three major cytoskeletal components are subject to different forms of manipulation and play different roles at different sites in the infected host cell. Three types of elements are involved: microtubules, intermediate filaments, and microfilaments. Microtubules are tubes of 22-26 nm diameter composed of tubulin. They appear to play no role in virus replication or release, and between 6 and 10 hours postinfection they disappear. This is due to severe inhibition of synthesis of tubulin (along with other host macromolecules) in infected cells (Murti and Goorha, 1983; Murti et al., 198513). The disappearance of microtubules is concurrent with the appearance of virus assembly sites. The morphology of virus assembly sites is different for FV3 in vertebrate cells compared to iridescent viruses in invertebrate cells, wherein they are dense entities often referred to as virogenic stroma or viroplasmic centers (Maes and Granoff, 196713; Yule and Lee, 1973; Devauchelle et al., 1985a; Ward and Kalmakoff, 1991). The assembly sites in cell infected by FV3, epizootic he-

382

TREVOR WILLIAMS

matopoietic necrosis virus, and goldfish iridovirus are reported as being less electron dense than the surrounding cytoplasm and devoid of cellular components such as organelles and ribosomes (Maes and Granoff, 1967b; Murti et al., 1984; Berry et al., 1983; Eaton et d . ,1991). Intermediate filaments are 7-11 nm in diameter composed of vimentin (for cells of mesenchymal origin). These filaments surround and radiate from the perinuclear region into the cytoplasm. Their usual function is uncertain. In FV3-infected cells, at 6-8 hours postinfection filaments retract from the periphery and cluster around the newly forming viral assembly sites. A marked increase in the degree of phosphorylation of the vimentin protein in infected cells is required to trigger this event (Willis et al., 1979b). Inhibition of intermediate filament function by a drug (taxol) results in poorly defined assembly sites and a reduction in virion production of about 80% (Murti et al., 1988). Although not strictly essential for virion assembly, the intermediate filaments are important in the structural integrity of assembly sites and in the efficient production of particles. Early but not late FV3induced proteins also seem to be involved in the formation of assembly sites (Chinchar et al., 1984). Microfilaments are filaments of 4-8 nm diameter comprising subunits of actin. Concurrent with other major changes in the cytoskeleton, at 6 hours postinfection existing bundles of microfilaments disappear and the actin subunits disperse through the cytoplasm. Existing individual microfilaments persist unchanged. At 7-10 hours postinfection the microfilaments re-form at the cell surface and facilitate the budding of virus particles through the cell membrane. At this time the normally smooth cell surface is transformed into numerous microvilli-like projections. Actin continues to be synthesized during this period, although it appears in a biochemically different form. Whether this is a result of postranslational processing or is due to the synthesis of a different isoform of the protein is not known. The presence of microfilament disrupters did not affect the rate of FV3 production, but almost all the virus remained cell-associated, as large accumulations of particles beneath the cell membrane. As a result, the yield of budded virus fell by 80-99% compared to untreated controls (Murti and Goorha, 1990).In a remarkably similar manner to the FV3 studies above, an intimate association between cytoskeletal elements and epizootic hematopoietic necrosis virus has been observed. Intermediate filaments were clearly associated with the periphery of the virus assembly sites, and particles were observed enmeshed within the cytoskeleton of critical point dried cells. The role of the cytoskeletal elements in budding of this virus is not certain; 99% of the virus produced in bluegill cell culture remained cell-associated (Eaton et al., 1991).

IRIDOVIRUSES

383

Kelly and Tinsley (1974) reported a marked reorganization of microtubules in lepidopteran cells infected by Tia IV [IV21. In contrast, for Plowden IV [IVl] replicating in lepidopteran and mosquito cells, immunofluorescence studies indicated that microtubules and microfilaments were not involved in the formation and maintenance of assembly sites (Seagull et al., 1985). Moreover, cytoskeletal disrupting drugs did not prevent the formation of viral assembly sites of Plowden IV UV11 in insect cells (Bertin et al., 1987). However, the presence of nuclear matrix or nuclear matrix-associated proteins within the structure of assembly sites has been reported for this system. Antibodies raised against mammalian lymphocyte nuclear matrix bound to Plowden IV [IVlI assembly sites in viuo and t o purified (fractionated) assembly sites in uitro (Bladon et al., 1986).

I . Enzymatic Activities A variety of enzymatic activities have been detected in association with the purified virions of Dazaifu IV [IV61, FV3, and LCDV-1. The enzymes are similar for each of the viruses, possibly reflecting a common set of requirements during the initial stages of infection; however, not in all cases have the enzymes been shown to be virally encoded, so caution is necessary in the interpretation of some of these results. A nucleotide phosphohydrolase with high affinity for ATP has been reported for all three viruses (Monnier and Devauchelle, 1976; Vilagines and McAuslan, 1971; Flugel et al., 1982). This enzyme hydrolyzed ATP to ADP and in the case of FV3 was localized in the core (Vilagines and McAuslan, 1971), whereas enzymatic activity was detected between the core and envelope of LCDV-1 (Flugel, 1985). A protein kinase (PK) is also common to these viruses. The PK of Dazaifu IF [IV61 phosphorylated low molecular weight proteins (in uitro) and showed a high affinity for basic substrates, especially a 12.5 K core polypeptide with DNA-binding properties (Monnier and Devauchelle, 1980). Phosphorylation changes the solubility of this protein, and it is presumed that endogenous PK activity within the core may result in the decondensation of DNA as a preliminary step toward release of the virus genome during virion uncoating early in infection (Cerutti and Devauchelle, 1990).The PK of FV3 appears different. The enzyme is external (i.e., can be solubilized), of mass 44 kDa, and is virally encoded (Silberstein and August, 1973, 1976). Little is known of the PK of LCDV-1 (Flugel, 1985). A third common enzyme is a DNase. In FV3, pH optima were reported as pH 5 (Aubertin et al., 1971) and pH 7.5 (Kang and McAuslan, 1972). The DNase of Dazaifu IV was apparently able to cleave DNAs from various origins including

384

TREVOR WILLIAMS

homologous viral DNA (unpublished data in Devauchelle et al., 198513; Cerutti and Devauchelle, 1990).DNase activity was also shown associated with LCDV-1 virions (Flugel, 1985). RNase activity has been reported only for LCDV-1 (Flugel, 1985) and FV3 (Kang and McAuslan, 1972). Thymidine kinase (TK) activity has been detected in purified LCDV-1 (Flugel, 1985) and in TK-negative mouse cells following infection by FV3 (Aubertin and Longchamps, 1974). The TK of LCDV-1 has been shown to be virally encoded. Transformation of 3T3 TK- cells into TK+ cells was observed following transfection of cells with LCDV-1 DNA restriction fragments (Scholz et al., 1988).The viral TK gene was mapped and sequenced and found to encode a polypeptide of 318 amino acids (Schnitzler et al., 1991).A protein phosphatase has been reported from FV3 particles (Silberstein and August, 1973). The presence of a virion-associated RNA polymerase has been reported for Tia IV [IV2] and Dazaifu IV [IV6] but has not been supported by the work of others (Monnier and Devauchelle, 1976; Goorha and Granoff, 1979). V. MOLECULAR BIOLOGY Advances in molecular biology of iridoviruses through the 1980s have been summarized by Darai (1990).Developments since that time include the identification of a number of iridovirus genes. A unique feature of iridoviruses is the organization of the genome. The evidence that the genome is circularly permuted and terminally redundant (for all iridoviruses examined to date) comes from the following observations, which are based mainly on FV3 but also on LCDV-1 and Dazaifu IV [IV6]. 1. Treatment with A 5’-exonuclease exposes single-stranded ends to the DNA molecule. Subsequent annealing of FV3 DNA results in the formation of duplex circles with two gaps of single-stranded DNA, indicating that the single-stranded ends are complementary (the gaps result from digestion proceeding beyond the terminal repeats). Circles were not observed in FV3 DNA not pretreated with exonuclease, indicating terminal DNA to be normally blqnt. The length of these terminal repeat sections was estimated by measurement of photomicrographs at approximately 4% of the length of the FV3 genome (Goorha and Murti, 1982). 2. Fully denatured DNA reanneals as duplex circles, each with a pair of single-stranded tails at different positions along the length of the molecule. The tails occur where terminal repeats fail to find com-

IRIDOVIRUSES

385

plementary single-stranded regions available for hybridization (Goorha and Murti, 1982; Darai et al., 1983; Delius et al., 1984). 3. End labeling of FV3 DNA failed to show specific labeled fragments following restriction endonuclease treatment because there are no fixed DNA termini in a population of virions with a circularly permuted genome (Goorha and Murti, 1982). 4. The DNA molecule is linear, yet restriction maps for all iridoviruses mapped to date are circular (Lee and Willis, 1983; Darai et al., 1985; Ward and Kalmakoff, 1987; Soltau et al., 1987; Schnitzler et al., 1987; Davison et al., 1992). 5. In LCDV-1 and Dazaifu IV [IV61, treatment of genomic DNA with 3’- or 5’-exonuclease followed by restriction endonuclease treatment results in the gradual disappearance of all restriction fragments, indicating that the distribution of termini is random across the length of the genome (Delius et al., 1984; Darai et al., 1983). However in FV3, each molecule in a population had a common region comprising about 75% of the total genome; the location of terminal sequences was limited to the remaining 25% (Murti et al., 1982). Estimates for the degree of terminal redundancy in iridoviruses range from 4 to 6% in FV3 (Goorha and Murti, 19821, 12%in Dazaifu IV [IV6] (Delius et al., 1984), and up to a remarkable 50% in LCDV-1 (Darai et al., 1983). It is t o be expected that circularly permuted genomes should have a number of origins of replication. To date, six origins of replication have been reported scattered across the genome of Dazaifu IV [IV61. Each origin was detected by plasmid rescue of genomic library fragments that were amplified following transfection into insect cells infected by Dazaifu IV [IV6]. Three of these origins of replication have been sequenced and found to comprise a 12- to 16-base inverted repeat. The degree of sequence similarity among the three entities was 55-77%, and all were predicted to form a hairpin structure (Sonntag and Darai, 1992). . A Plowden IV [IVll isolate from North America appears unusual in that it has been reported to comprise a number of genomic components: a large component (176-247 kbp) and a small component (10.8 kbp) (Tajbakhsh et al., 1986). The relative abundances of the two components were dependent on the stage of virion assembly. The small component was found in much greater abundance in partially filled capsids. Transcription from the small component was detected in permissive and semipermissive cells (Tajbakhsh et al., 1990b). How common such genomic components are in the Iridoviridueis not known. An Irish strain of Plowden IV [IVlI studied in the same laboratory showed only one component (see review by Tajbakhsh and Seligy, 1990).

386

TREVOR WILLIAMS

A . Virus Genes In addition to the transcription-regulatinggenes of FV3 described previously, sequencing work has provided a small but intriguing catalog of putative viral genes. The approach has been either to sequence genomic fragments from established iridovirus gene libraries or to construct oligonucleotide primers using available information on conserved and variable regions of viral genes. The latter approach has been highly successful in the identification and sequencing of the major capsid protein (MCP) gene, and it appears set to become a standard technique for comparative studies and characterization of novel isolates (R. Webby and J. Kalmakoff, 1995, unpublished data). The MCP gene was first sequenced from Plowden IV [IVlI (Tajbakhsh et al., 1990a) and subsequently from Aberystwyth IV [IV221 (Cameron, 19901, Dazaifu IV [IV6] (Stohwasser et al., 19931, and LCDV-1 (Schnitzler and Darai, 1993). Sequence information generated from these studies has been used for comparative studies, described previously (Schnitzler and Darai, 1993). Other identified genes code for putative proteins with nucleic acid replication or transcription functions. Schnitzler et al., (1994a) detected eight ORFs in a 5.7-kb EcoRI fragment of the Dazaifu IV [IV6] genome. Of these eight, relationships were found to exist with sequences in databases for three of the ORFs. The putative proteins for these genes were as follows: (i) a homolog of eukaryotic nonhistone chromosomal protein (221 amino acids) with a DNA-binding region some 70 amino acids in length; (ii) a polypeptide of 145 amino acids with a single putative zinc finger motif similar to that known as the RING motif, a motif known for its DNA-binding properties and involvement in transcription and repair of DNA damage (another RING motif was also found in second larger polypeptide on another EcoRI fragment in the Dazaifu IV UV61 genome); and (iii) a polypeptide of 127 amino acids with a highly conserved region common to the bacterial antimutator, GTP phosphohydrolase (Mut-TI. This enzyme hydrolyzes the highly mutagenic substrate 8-oxo-GTP to prevent transversions during replication. Another viral antimutator enzyme, uracil DNA glycosylate, has been reported from poxviruses, wherein it is essential for replication (Stuart et al., 1993). In Dazaifu IV [IV6], comparison of sequence data has led to the suggestion that certain iridovirus genes were more recently acquired from eukaryotic hosts than the homologous genes of other cytoplasmic DNA viruses. A putative helicase gene coding for a polypeptide of 606 amino acids and a gene coding for a putative DNA-dependent RNA polymerase I1 enzyme subunit (1051 amino acids) have been reported

IRIDOVIRUSES

387

from different fragments of the gene library of this virus (Sonntag et al., 1994; Schnitzler et al., 1994b).These enzymes appeared more closely related to eukaryotic polymerases and helicases than to the homologous enzymes from vaccinia virus or African swine fever virus. However, these comparisons make important assumptions regarding the rate of evolution of iridoviruses compared to other viral and eukaryotic systems, assumptions for which data do not exist. A putative apoptosis inhibiting gene has been found in Dazaifu IV [IV61 (Birnbaum et aZ., 1994). The ability for cells to undergo programmed death (apoptosis) in response to viral infection may be blocked by viral gene products in a number of different ways. In other invertebrate DNA viruses such as baculoviruses, the mechanism for blocking cell apoptosis appears different from those found in mammalian viruses (adenoviruses or herpesviruses). Several baculoviruses have been found to have a gene (iap, inhibition of apoptosis) for a product of 30 kDa that showed apoptosis blocking activity when cotransfected with DNA from an apoptosis-negative baculovirus (AcNPV p35).The putative IAP protein has a central tandemly repeated motif containing cysteines and histidine, probably a zinc finger with nucleic acid-binding properties, as well as a more widely recognized cysteine/histidine motif; C3HC4, found in approximately 30 other proteins and described from Dazaifu IV [IV6] as being a zinc finger structure (Handermann et aZ., 1992; Sonntag et al., 1994). The iap gene of Dazaifu IV [IV6] is not complete compared to baculovirus homologs because it has only one repeat motif and one C3HC4 motif. Whether this apparent difference is biologically important is not known because the activity of this gene has not been tested (Birnbaum et al., 1994). Home et al. (1990) reported the sequence and transcription map for a late gene of Plowden IV [IVl]. The putative protein was 867 amino acids (96 kDa), rich in serine, proline, and basic residues, hydrophilic, and showed similarities to proteins with known DNA-binding properties, in particular with GAG polyproteins of vertebrate viruses. A role in DNA packaging and core formation was suggested. The gene was unusual in having 3’ polyadenylation signals, structures that are not commonly seen in FV3 transcripts (Willis et al., 1990).Polyadenylation signals were also reported for the major capsid protein gene of Plowden IV [IVl] (Tajbakhsh et al., 1990a), and for genes of unknown function in LCDV-1 (Schnitzler et al., 19901,but the relatively sharp appearance of bands in Northern blots of other iridescent virus mRNAs suggests that poly(A) tails are not ubiquitous. The gene sequence for a putative integrase-recombinase enzyme of 275 amino acids has been reported from FV3. This group of enzymes is responsible for catalyzing strand exchange between DNA molecules

388

TREVOR WILLIAMS

and plays an important role in DNA replication of bacteriophages. The mRNA for this gene was delayed early in nature. The enzyme is believed to be important in the formation of concatameric DNA intermediates or resolution of such entities prior to virion packaging. This finding once more emphasizes the similarities between the replication of FV3 and these bacterial viruses (Rohozinski and Goorha, 1992).

B . Repetitive DNA Restriction endonuclease mapping of FV3, LCDV-1, and several invertebrate IVs has revealed extensive repeat sequences in certain regions of the genome. In the IV9 strain of Nelson IV [IV9/10/181 the repetitive DNA was found to occur in over 25% of the genome (Ward and Kalmakoff, 1987; Kalmakoff et al., 1990). These repeat sequences are distinct from the complementary terminal sequences that arise from the terminal redundancy of iridovirus genomes. Transcription of the repetitive sequences of Nelson IV [IV9/10/181 was clearly restricted to late times postinfection (McMillan and Kalmakoff, 1994), and the coding function of these regions was obscure. The repetitive sequences of LCDV-1, however, were less extensive, comprising two nearly identical EcoRI fragments of some 1400 bp in length. One fragment had strong promoter function when inserted into a plasmid with a CAT reporter gene and transfected into Escherichia coli. Virtually identical repeat sequences were found in LCDV-2, a genetically distinct isolate from dabs (94.9-99.9% homology) (Schnitzler et al., 1990). By sequence analysis the pattern of repetitive DNA in Dazaifu IV [IV61 has been shown to be unusually complex. Initially two boxes (A and B) were found in EcoRI fragment. Box A was 91 bp in length and was complementary t o nine tandemly repeated regions of a PuuII fragment that mapped to coordinates distant from the EcoRI fragment. Box B was 46 bp in length and was complementary to one region of the same PuuII fragment. In addition, a cluster of imperfect but somewhat regular repeat sequences spanning a region of over 4 kb was detected in the PuuII fragment. Four distinct repetitive elements were identified, three of which appeared duplicated with 80-87% homology. The fourth element, some 240 bp, appeared as 12 repeats in three segments (3-5 boxes per segment). The degree of homology among these boxes of the fourth element was 90-98%, although three of the boxes were not complete. In general, the arrangement of boxes was complex and interdigitated. The coding function of these regions was not clear, although a number of ORFs were detected (Fischer et al., 1988a,b, 1990). Inverted repeats were also responsible for a stem-loop structure that was detected in the sequence of one EcoRI fragment. The structure was

IRIDOVIRUSES

389

confirmed by electron microscopy of reannealing DNA, forming a stem of 62 bp and a loop of 692 bp (Fischer et al., 1988a). VI. ECOLOGY Knowledge of the ecology of both vertebrate and invertebrate iridoviruses lags far behind biochemical and molecular laboratory studies. The search for effective agents for the biological control of insect pests has meant that the ecology of iridescent viruses has received more attention than the ecology of the vertebrate isolates. Sadly, the amount of data on the ecology of vertebrate isolates barely merits a small paragraph here. There have been some studies on experimental transmission by cohabitation of infected and uninfected fish (e.g., Hendrick et al., 1990) and some tests of pathogenicity to different amphibian life stages (Tweedell and Granoff, 1968; Wolf et al., 1968; Clark et al., 1968; Granoff et al., 1969), but quantitative studies of iridovirus ecology in vertebrate populations do not appear to exist. What information is available on IV ecology is plagued by using the iridescent phenomenon as the sole criterion for infection; in all but a few cases the possibility of covert IV infections have been ignored. Actually, iridescence may be a trivial characteristic (Kelly, 1985) with no selective advantage to the virus in terms of enhanced virus transmission, for example. Previous authors have speculated on unifying features in the ecology of these viruses, in the mechanisms of transmission and persistence in host populations (Hall, 1985; Kelly, 1985; Ward and Kalmakoff, 1991). As yet, however, there is no evidence to indicate that the strategies of transmission, persistence, or dispersal of iridoviruses show any obvious unification at all. Rather, a spectrum of different strategies may be involved dictated by the biology and life history of the host(s) and the environment that it inhabits. Consequently, the opportunities for transmission and replication available to the virus will differ in each system, and generality in statements concerning the ecology of these viruses may not be possible. The true incidence of transmission and the infectivity of iridescent viruses remain particularly vague, as the importance of covert infection by IVs in host populations is largely unknown. Key parameter values necessary for the development of models of host-iridovirus population dynamics, for example, the rate of transmission, are not known. It is possible to make crude estimates of certain parameters by scanning the available literature. Survival time (the interval between infection and death) is dependent on temperature, route of infection (Carter, 1973b, 1975), and stage of host infected

390

TREVOR WILLIAMS

(Carter, 1974; Sieburth and Carner, 1987). Survival times seem to be fairly independent of dose (Carter, 1974), although when very high doses are administered larvae may become paralyzed and die without developing patent disease (Sieburth and Carner, 1987; Ward and Kalmakoff 1991). This may be related t o the cytotoxic effects of iridoviruses. Survival times following per 0s doses range from 45 days for Plowden IV in Tipula oleracae, about 35 days for Riverside IV [IV311 in terrestrial isopods, approximately 28 days for Stoneville IV [IV301 in Helicoverpa zea, and 21 days for a n IV of the cricket Scapteriscus borellii (Carter, 1973a; Grosholz, 1992, 1993; Sikorowski and Qson, 1984; Fowler, 1989), although survival times as short as 6 days have been reported for late instar Anticarsia gemmatalis larvae orally infected with the homologous virus, San Miguel IV (Sieburth and Carner, 1987). Compared to feeding, survival times were reduced by 30-50% when the inoculum was injected; a reduction in the variability of survival times was also seen (Carter, 1973a,b). The yield of infectious particles per host is massive for arthropods with patent IV infections. Remarkably, 25% (dry weight) of the body of a dead insect may be virus (Williams and Smith, 1957). Using this figure and previously published values for IV particles, namely, weights of approximately 2 x 10-15 g (Thomas, 1961; Glitz et al., 19681, Carter (1973b) calculated that an infected tipulid larva of 200 mg should contain some 2.5 x 1012 particles of Plowden IV [IVl]. Day and Mercer (1964) obtained a similar figure of 2 x 1012 particles per pupa for Tia IV [IV21 in Galleria mellonella. Following these same assumptions, yields for Riverside IV [IV31/321 can be crudely estimated in different isopod species at approximately 2.6 x 1010 particles released from Armadillidium vulgare and 3.1 x 1010 particles from the larger Porcellio scaber. Likewise, a typical Tenebrio molitor pupa would be expected to yield some 1.5 x 1010 particles of Fort Collins IV rIV291 [estimates from data given in Cole and Morris (1980) and Black et al. (1981), respectively]. Clearly, these calculations are open to criticism, but they provide a starting point for the development of models of epizootiology of these diseases. Most of the work on the ecology of IVs comes from four systems for which the hosts are mosquitoes, crane flies (Tipula spp.), terrestrial isopods (woodlice), and blackflies (Simulium spp). With one exception, described later, nothing is known about the rate or mechanisms of IV dispersal, although, as will become apparent, autodispersal via the infected host may be one likely mechanism. Consequently this section is divided into two main parts: transmission and persistence. The persistence section also includes examples of the limited knowledge on the natural host range of IVs.

IRIDOVIRUSES

391

A. Transmission Previous authors have looked for a common transmission strategy for all the known invertebrate isolates and have postulated about possibilities such as the role of cannibalism of infected conspecifics, cuticular lesions, and parasitic vectors (e.g., nematodes) as possible routes of horizontal transmission. Riverside IV (IV31/IV32) is transmitted horizontally via cannibalism of infected conspecifics or by predation of infected individuals of other woodlouse species. Enzymelinked immunosorbent assays (ELISA) indicated that covert IV infections do not occur in this system (<1% of total); all infections appear to be patent and lethal. Interspecific competition between two woodlouse species (Porcellio laevis and P. scaber) had more effect on the incidence of disease in P. scaber than did intraspecific competition: the prevalence of infection was doubled in the presence of the competitor. This was not a direct effect of competition for food, but probably arose from increased levels of aggression when the other species was present. This result would be expected for a pathogen directly transmitted by predation/cannibalism. An increase in the frequency of aggressive encounters and wounding may also increase the probability of virus entry through cuticular lesions (Grosholz, 1992). In a mosquito IV system, Vero Beach IV [IV31, horizontal transmission was demonstrated via cannibalism of patently infected mosquito cadavers (Linley and Nielsen, 1968b; Hall and Anthony, 1971). In this system, it was evident that inapparent IV infections could also occur but at unknown frequencies. Vertical transmission occurred when mosquito larvae were challenged with IV inoculum shortly prior to pupation. The progeny of 19-47% of IV-challenged females developed patent infections (Woodard and Chapman, 1968). Vertical transmission may be an all-or-nothing response. The progeny from 5 egg batches produced by females exposed to IV showed patent infection, whereas the remaining 58 egg batches did not give rise to patently infected larvae (Linley and Nielsen, 1968a). The all-or-nothing theory has found only partial support (Fukuda and Clark, 1975). However, mosquito progeny that failed to develop patent infections in these studies were not assayed for the presence of covert infections, so the true incidence of vertical transmission in this system remains unclear. When wild or cultured mosquito larvae were challenged with increasing doses of IV, it was evident that progeny from the wild mosquitoes developed notably higher frequencies of patent disease (19.6% overall) than their laboratory cultured conspecifics (5.7% overall). The possibility that exposure to IV provoked patent infections in wild mosquitoes that were already covertly infected is an appealing concept and

392

TREVOR WILLIAMS

finds parallel in the latent infections of lepidopteran baculoviruses (Hughes et al., 1993). The transmission of pathway of infection of Plowden IV [IVlI in the crane fly, Tipula oleracue, were studied using iridescence as the criterion for infection. The frequency of inapparent infections in this system is unknown, although when pupae were injected with virus, 20% of the resulting adults developed patent infections while the remaining 80% appeared to be covertly infected as shown by serology (Carter, 1973~).In another assay, the feces of infected larvae were demonstrated to contain viable virus, but in insufficient quantities to cause patent disease in healthy larvae when ingested. Larvae, pupae, and adults that died following the dose of feces were tested for inapparent infections and proved negative; however, the remaining apparently healthy insects were apparently not tested, so the ability of contaminated feces to induce covert infections in Tipula is not known. Transmission of virus by cannibalism of infected cadavers was an efficient mechanism of transmission and resulted in high levels of patent infection in susceptible conspecifics; the doses of virus ingested during cannibalism are massive (Carter, 1973a). The route of infection was also investigated by injecting virus, applying virus to cuticular lesions, applying virus to the spiracles, or by allowing larvae to drink a virus suspension. The most efficient routes of transmission were first infection, followed by abrasion, feeding, and lastly via spiracles. Larvae did not become patently infected when tipulid eggs were allowed to hatch in agar containing virus (Carter, 1973b). Covert infections have also been reported in host range studies where Nelson IV [IV9/IV10/IV181 was injected into various lepidopteran species; the larvae did not develop iridescence, but IV DNA could be detected in tissue extracts by dotblot hybridization (Ward and Kalmakoff, 1991). We may predict that vertical transmission should be common in many invertebrate IV systems due to the rarity of patent IV infections, which suggests a low virulence of these viruses. The majority of infections may be inapparent and somewhat benign in nature (Kelly, 1985; Poprawski and Yule, 1990; Williams, 1993). However, pathogens cannot be sustained in host populations by vertical transmission alone, and they must effect horizontal transmission to some extent (Anderson and May, 1981). How this is achieved in covert infections by IVs is not known. The possibility of prolonged low-level excretion of virus particles by covertly infected hosts appears an appealing concept and finds comparison in nonlethal cytoplasmic polyhedrosis infections of Lepidoptera in which virus particles are continually produced from gut cells and excreted in the feces. Such a method of horizontal transmission would require that IVs display a far higher per 0 s infectivity than

IRIDOVIRUSES

393

previously recognized. Such infectivity is yet to be demonstrated. A microscopic study by Stoltz and Summers (1971) demonstrated that large doses of IV were degraded in the gut of mosquito larvae; the observation of intimate contact between gut cell and virus was not seen, a finding which appears not to support a theory of high infectivity of these viruses. The possibility that subparticle units are infectious (e.g., viral DNA and some viral proteins) was noted by these authors. Given the highly infectious nature of FV3 DNA in the presence of soluble viral proteins (nongenetic reactivation), this possibility merits consideration. If a unifying feature to IV ecology exists, it may be related to transmission of these viruses. The fact that IV infections are mostly limited to hosts in damp or aquatic environments along with the stability of the particles in water might be indicative of transmission of “freeliving” infective stages (sensu Anderson and May, 1981) as being important in the ecology of the viruses. In contrast, if cannibalism/predation of infected hosts were a major transmission route for most of the IVs, we may expect these viruses to be far more common than they appear to be among species in which cannibalism was common (lepidopteran larvae, grasshoppers, etc.) or in predatory arthropods (predatory beetles, wasps, ants, spiders, etc.). This is probably a n oversimplification of a potentially complex area. The matter remains wide open.

B . Persistence 1 . Physical Persistence

The ability of IV particles to physically persist outside of the host is virtually unknown. Temperature may be an important factor in this respect, as IVs are rapidly inactivated a t temperatures over 50°C (Day and Mercer, 1964). Humidity may be more important still, given the moist or aquatic host habitats. The ability of crude preparations of Vero Beach IV [IV31 to elicit patent infections in mosquito larvae has been reported to decline by 50% in 0.8 days in fresh water a t 27°C. In brackish water, virus retained infectivity for 2 days, after which infectivity declined rapidly; on soil, the virus was inactivated even more rapidly (estimated from data in Linley and Nielsen, 1968b). Riverside IV [IV31/321 in the cadaver of a woodlouse apparently remained infective for up to 5 days at ambient laboratory temperatures (Grosholz, 1993). The ability of purified Tia IV [IV2] at 4°C to elicit patent infections of Galleria mellonella larvae fell by 50% after 32 days in a laboratory refrigerator (estimated from data in Day and Gilbert, 1967).

394

TREVOR WILLIAMS

2 . Persistence in Host Populations For one putative iridovirus from the water snail Lymnaea truncatula, the incidence of infection in each of two annual generations was monitored between 1983 and 1984 and again in 1987-1989. Snails were sampled from 11 sites. Virus-induced hypertrophy of blood cells and the presence of an intracytoplasmic inclusion were used as the diagnostic features of infection. The incidence of infection was between 1.6 and 87%, depending on site and time of year. The incidence of infection was reasonably consistent and specific to each site. The site where the incidence of infection was highest, in beds of rushes, consistently had infection frequencies greater than 70%, whereas infection frequencies at certain meadow sites did not exceed 15%.There was no obvious pattern in the incidence of infection in spring versus autumn generations. Data on the density of snail populations at each site were not given (Rondelaud and Barthe, 1992). A handful of IV isolates have been reported as causing epizootics of patent (lethal) disease. One such report comes from Ricou (19751, who sampled a population of Tipula larvae wherein the incidence of disease differed according to host population density and soil moisture. Larval densities were twice as great in damp soil compared to wet or dry soil. The incidence of disease was highest in wet and damp soils, which may indicate moisture to be of great importance in IV persistence. In this study the epizootic developed over the course of a year t o 90% infection levels, with a clear resulting impact on the Tipula population density in the following years. In the woodlouse-Riverside IV (IV31/32) system, Grosholz (1993) presented data to show how changes in the incidence of IV infections were affected by the patchiness of the host population and the withinpatch density of hosts. These factors changed seasonally. During the dry months of late summer, the woodlouse populations were highly aggregated, the within-patch host density was high, and the distance between adjacent patches was large (July mean, 5.75 m). The incidence of IV infection at this time was approximately 1-2%. During the wet winter and spring months, the patchy structure of the population became more homogeneous, the within patch host density declined, and the distance between patches fell (January mean, 2.0 m). These conditions were more suitable for the transmission of the virus, and the incidence of infection rose to approximately 13%.No significant differences were found in the incidence of infection in different sexes of host or in different size classes of host. However, significant species effects were detected: Porcellio laevis suffered the highest frequency of infection (5.7% overall), Armadillidium vulgare had the lowest incidence

IRIDOVIRUSES

395

(1.6%),and P. dilatatus and P. scaber were intermediate. The effect of interpatch spacing on the prevalence of IV infection was supported by field experiments in which the distance between patches was manipulated. The distance between adjacent patches was also shown to be negatively correlated with the probability of isopod dispersal, although the effect of virus infection on the dispersal abilities of isopods was not tested. Clearly, in this system, both population structure and the presence of competing species (see above) are important factors affecting the prevalence of the disease. Moreover, the overall population density appeared vital in determining the persistence of the disease. Viral infections were endemic at high host densities (>2000 isopods m-2) but completely absent in low-density host populations nearby (Grosholz, 1992).With four different hosts involved in the ecology of the virus, this system is a clear demonstration of why IV nomenclature should be uncoupled from that of the host species. Studies of blackfly populations in the River Ystwyth, Wales (UK), have addressed the problem of detecting and describing the incidence of covert IV infections in insects. These studies have demonstrated that the incidence of covert IV infection can be measured using sensitive bioassay and PCR techniques. Larvae of the lepidopteran G . mellonella developed patent IV infections following intrahemocoelomic injections of blackfly homogenates derived from covertly infected blackfly larvae. This assay is extremely sensitive and may have detection limits of below 10 particles as reported for another IV isolate (Day and Mercer, 1964; Day and Gilbert, 1967). The bioassay results were supported by results of PCR analysis using primer sequences targeted at the MCP gene of Aberystwyth IV (IV22), an isolate from blackflies found in the River Ystwyth previously (Batson et al., 1976). Nested sets of primers were designed for high specificity (high GC content and therefore high annealing temperature) rather than to target particularly conserved regions of the MCP gene. Likewise, cycling conditions were highly stringent. An outer set of primers were designed to yield an amplicon of 816 bp. This amplicon was not visible by gel electrophoresis. A sample of this reaction mixture was subjected to further amplification using a second set of primers to amplify a fragment of 719 bp from within the 816-bp fragment. The 719-bp fragment was visible by electrophoresis. Cameron (1990) reported a XhoI site in this region of the MCP gene, and so the identity of the amplicon was confirmed by treatment with XhoI to yield fragments predicted to be 472 and 247 bases in length. Using the bioassay and PCR techniques, the presence of abundant covert infections in the springtime populations of Simulium larvae in the River Ystwyth was detected (Williams, 1993). The frequency of covert infection was highest in the central

396

TREVOR WILLIAMS

section of the river. The nature of the covert infection in blackflies is not known. Certainly the infection does not appear to be latent in the sense usually used for insect viruses (Hughes et al., 19931, because it can be transmitted to other insects by injection. Probably the infection exists as particles within host cells but at very low densities. Following this finding, the bioassay technique was used for screening larval blackfly samples from a monthly sampling program from sites along the River Ystwyth over the reproductive season of the blackflies. This has revealed a complicated interaction involving one blackfly species and three tentative virus species in the river. The incidence of covert infection in blackfly larvae sampled in March (overwintering populations) varied between 17 and 37%depending on site. All of these infections were different variants of the original Aberystwyth IV. In April, the incidence of covert infection declined (0-20% depending on site), and a seemingly new (second) virus was detected. In May, a number of patent infections were observed albeit at an extremely low density, all of which were different variants of Aberystwyth IV (Williams and Cory, 1993). However, at this time, no covert infections could be detected in the blackfly populations. Throughout the remainder of the summer period the incidence of covert infections remained at virtually undetectable levels (<1%).However, in September, covert infections from a third virus species appeared in the populations at levels of up to 20%. This third virus had been isolated from a patently infected larva in September of the previous year. High frequencies of covert infection occurred when host population densities were low. Coefficients of similarity were calculated from restriction profiles following treatment with Hind111 or EcoRI, for both the Aberystwyth IV and the new (September) isolates. Coefficient values within each tentative virus species ranged from 70 to greater than 95%, whereas comparison of isolates from different “species” consistently gave coefficient values of less than 66%.The genetic diversity of variants isolated from covert or patent infections was remarkable. In no instance were identical virus variants found infecting different individual larvae. Just what the adaptive significance of such variation is remains obscure. The marked fluctuations in the incidence of covert IV infections in the blackfly populations and the fact that three apparently different viruses were involved at different times of the year led to speculation about the potential role of other aquatic organisms in the transmission and persistence of these viruses (Williams, 1995). 3. Role of “Alternative”Hosts

Certainly one-virus-multihost systems evidently exist, judging from the observation of patent infections elsewhere. Examples from sym-

IRIDOVIRUSES

397

patric species include Nelson IV [IV9/10/181which infects two lepidopterans and a coleopteran, Riverside IV [IV31/321which infects Armudillidium vulgare, Porcellio scaber, P. laevis, P. dilatatus, and possibly several other isopods (Grosholz, 1993; Schultz et al., 1982), and probably Timaru IV [IV16/191which infects two scarabid species, Costelytra zealandica and Odontria striata (N. McMillan, 1990, personal communication). A possible example from nonsympatric species may be Aberystwyth IV (IV22/25) from a blackfly (Simulium sp.) which has also been isolated from a crane fly larva (Tipula sp.) in the United Kingdom (Elliott et al., 1977). Aquatic organisms of overtly different taxa than arthropods may be involved in the transmission and persistence of IVs, it has been speculated. Such speculation arises mainly from the observation of iridovirus-like particles in other aquatic organisms. In Chlorella-like algae symbiotic with Hydra and Paramecium, icosahedral particles have often been reported that are 150-190 nm diameter in negatively stained preparations, contain 5-10% lipid, possess a major structural protein of approximately 54 kDa, and have a dsDNA genome (Van Etten et al., 1982).However, it is now clear that these algal viruses are not closely related to iridoviruses. The most obvious differences lie in the structure of the genome (which is linear and nonpermuted with hairpin ends) and the fact that some isolates showed high levels of methylation at cytosine and adenine bases, which is not seen in any iridovirus (Schuster et al., 1986; Van Etten et al., 1991). The Chlorella viruses have now been assigned to a new family, the Phycodnaviridae. The observation of IV-like particles in a daphnid, Cerodaphnia dubia, lead to speculation that this may be an alternative host for IVs of two mosquito species found patently infected in the same pools of water (Ward and Kalmakoff, 1991). The role of “alternative” hosts in persistence of IVs is clearly yet another facet of the ecology of these viruses that requires attention. VII. FUTURE DIRECTIONS FOR IRIDOVIRUSES Looking t o the future, it is clear that two important issues require study: the taxonomy and the ecology of iridoviruses. The taxonomic organization within and among the various genera of the lridouzridae requires revision. Genetic studies have begun to clarify the interrelationships of the invertebrate IVs for which distinct complexes have been detected using hybridization and other techniques. These studies require support from studies using complementary techniques. Work is in progress to sequence regions of the major capsid protein gene from a broad range of iridoviruses, from both vertebrate and inverte-

398

TREVOR WILLIAMS

brate genera. This should become the basis for defining a new characteristic by which iridoviruses can be characterized and classified in a rigorous standardized manner. The polythetic system of classification requires that a number of other characteristics would be included in the definition, but the quantitative nature of sequence data puts this character at the top of the list for comparative studies. The discarding of a considerable number of previously recognized but uncharacterized viruses from the family should permit a fresh start with a new set of criteria for iridovirus classification. Characterization of novel isolates should always be made with reference to existing characterized material, and the rationalization of recognized iridovirus species should assist in the selection of key species for use in comparative studies. Possibly, a virologist would perceive a different set of priorities for future research with iridoviruses, but 40 years after IV discovery, knowledge of the ecology of these viruses is still in its earliest stages. This is because most work with iridoviruses has been done by virologists interested in questions different from those that interest the ecologist. Studies of host-pathogen ecology may also be neglected because it involves learning new and rather esoteric techniques from those used by the traditional field ecologist. This situation is now changing as PCR and other molecular techniques have been developed that are routine and simple to master. For the invertebrate IVs, the recognition of widespread covert infection raises a number of intriguing questions. Within individual hosts, the nature of covert infection remains unknown. Are covert infections localized to specific tissue types? Are the gonads infected, and is vertical transmission resulting in covertly infected host progeny a common occurrence? Preliminary evidence suggests that IVs may have a far higher infectivity than previously believed because challenging hosts with IV inocula may lead to few patent infections, yet abundant covert infections (T. Williams, C. Tiley, and J. Cory, 1993, unpublished data). Thus, the incidence of transmission of IVs may previously have been grossly underestimated due to the rarity of patent infections in most host-IV systems. Likewise, the involvement of alternative host species in the transmission and persistence of IVs deserves far more attention than received to date. The involvement of alternative hosts may also be reflected in the high levels of genetic diversity observed in patent and covert IV isolates from blackfly larvae. Here, too, studies are required to understand the distribution of variable regions within particular isolates, the diversity of isolates found in a particular host blackfly, and the diversity of genotypes within the IV population. Are blackflies infected with a similar diversity of IV strains that are selected out during the process of bioassay in Galleria larvae, or does each blackfly

IRIDOVIRUSES

399

harbor distinct, near clonal isolates? The population consequences of such diversity touches on major theories of host-pathogen evolution: Red Queen, arms-race hypotheses (e.g., Dawkins and Krebs, 1979; Anderson and Gordon, 1982; Hamilton, 1982). Basic studies of the ecology of vertebrate iridoviruses are crying out to be done. With animals as easy to culture as frogs/tadpoles and a virus that is as amenable to cell culture as FV3, the system seems perfect for experiments on transmission and persistence. Most of the fish viruses also require quantitative studies of basic aspects of their ecology. An appreciation of the major impact that iridoviruses can have on fish populations will no doubt elicit the required studies in due course. The iridoviruses are a fascinating but neglected family, clearly united by physical and genetic features but showing diverse patterns of pathogenicity, host range, and host exploitation. Advances in understanding the taxonomy and ecology of certain iridoviruses should permit a diversity of further studies. Although IVs have previously been likened to Cinderella-like characters for their striking iridescent hues, a major advance will come from recognizing that invertebrate iridoviruses are more “Jekyll and Hyde” in nature, and the majority of their existence may be as covert infections that must be searched for. This probably also applies to many of the isolates from vertebrates. IN PROOF NOTEADDED

At the time of going to press, reports of extensive mortality in the frog populations of Britain are making the national newspapers (The Times, July 17, 1995). Up to 2000 deaths per site have been reported (Drury et al., 1995; Cunningham et al., 1995).

ACKNOWLEDGMENTS I am grateful to the following people for advice and comments that have helped in formulating some of the ideas in this chapter. David Bishop, Pete Christian, Gholamreza Darai, Rosie Hails, Alex Hyatt, Jim Kalmakoff, Gary Kinard, Lois Miller, Vern Seligy, Don Stoltz (photographs), and Richard Webby. Sima6 Vasconcelos deserves special thanks for assistance.

REFERENCES Ahne, W., Schlotefeldt, H. J., and Ogawa, M. (1989). Fish viruses: Isolation of an icosahedral cytoplasmic deoxyribovirus from sheatfish (Silurus glunis). J. Vet. Med. 36, 333-33 6. Anderson, R. M., and Gordon, D. M. (1982). Processes influencing the distribution of parasite numbers within host populations with special emphasis on parasite-induced host mortalities. Parasitology 85, 373-398.

400

TREVOR WILLIAMS

Anderson, R. M., and May, R. M. (1981). The population dynamics of microparasites and their invertebrate hosts. Philos. Trans. R . Soc. London B 291,451-524. Armstrong, R. D., and Ferguson, H. W. (1989). A systemic viral disease of chromide cichlids, Etropus maculatus Bloch. Dis. Aquat. Org. 7, 155-157. Aubertin, A. M. (1991). Iridoviridae. Arch. Virol. Suppl. 2, 132-136. Aubertin, A. M., and Longchamps, M. 0. (1974). Thymidine kinase induction in FV3infected mouse cells. Virology 58, 111-118. Aubertin, A. M., Palese, P., Tan, K. B., Vilagines, R., and McAuslan, B. R. (1971). Proteins of a polyhedral cytoplasmic deoxyvirus. 111. Structure of FV3 and location of virus associated adenosine triphosphate phosphohydrolase. J. Virol. 8, 643-648. Aubertin, A. M., Hirth, C., Travo, C., Nonnenmacha, H., and Kirn, A. (1973). Preparation and properties of an inhibitory extract from frog virus 3 particles. J. Virol. 11, 694-701. Aubertin, A. M., Travo, C., and Kim, A. (1976). Proteins solubilized from frog virus 3 particles: Effect on transcription. J. Virol. 18, 34-41. Aubertin, A. M., Tondre, L., Martin, J. P., and Kirn, A. (1980).Structural proteins of frog virus 3, phosphorylated proteins. FEBS Lett. 112, 233-237. Aubertin, A. M., Tondre, L., and Descamps, P. (1981).Capping of frog virus 3 antigen on the plasma membrane of infected cells. Ann. Virol. Znst. Pasteur 132E, 195-205. Aubertin, A. M., Tham, T. N., and Tondre, L. (1990). Regulation of protein synthesis in frog virus 3-infected cells. In “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 187-202. Kluwer, Boston. Balange-Orange, N., and Devauchelle, G. (1982). Lipid composition of an iridescent virus type 6 (CIV). Arch. Virol. 73, 363-367. Barray, S., and Devauchelle, G. (1979). Etude des polypeptides de structure du virus iridescent de Chilo suppressalis (iridovirus type 6). Can. J. Microbiol. 25, 841-849. Barthe, D., Rondelaud, D., Faucher, Y., and Vago, C. (1984). Infection virale chez le Mollusque pulmon6 Lymnaea truncatula Muller. C . R . Acad. Sci. Ser. 3 298,513-515. Batson, B. S., Johnston, M. R. L., Arnold, M. K., and Kelly, D. C. (1976). An iridescent virus from Simulium sp. (Diptera; Simuliidae) in Wales. J. Znuertebr. Puthol. 27,133135. Beckman, W., Tham, T. N., Aubertin, A. M., and Willis, D. B. (1988). Structure and regulation of the immediate-early frog virus 3 gene that encodes ICR489.J. Virol. 62, 1271-1277. Bellett, A. J. D. (1968). The iridescent virus group. Adu. Virus Res. 13, 225-241. Bellett, A. J. D., and Fenner, F. (1968). Base sequence homology among some cytoplasmic deoxyriboviruses of vertebrate and invertebrate animals. J. Virol. 2, 13741379. Bellett, A. J. D., and Inman, R. B. (1967). Some properties of deoxyribonucleic acid preparations from Chilo, Sericesthis and Tipula iridescent viruses. J. Mol. Biol. 25, 424-437. Berry, E. S., Shea, T. B., and Gabliks, J. (1983). ‘ h o iridovirus isolates from Carassius auratus (L.). J. Fish Dis.6, 501-510. Bertin, J., Frosch, M., and Lee, P. (1987). Further studies on Tipula iridescent virus and the cytoskeleton of virus-infected cells. Eur. J. Cell Biol. 43, 215-222. Bird, F. T. (1961). The development of Tipula iridescent virus in the crane fly Tipula paludosa Meig. and the wax moth, Galleria mellonella. Can J. Microbiol. 7, 827-830. Bird, F. T. (1962). On the development of the Tipulu iridescent virus particle. Can. J. Microbiol. 8, 533-534. Birnbaum, M. J.,Clem, R. J., and Miller, L. K. (1994).An apoptosis inhibiting gene from a nuclear polyhedrosis virus encoding a polypeptide with Cys/His sequence motifs. J . Virol. 68, 2521-2528.

IRIDOVIRUSES

40 1

Black, P. N., Blair, C. D., Butcher, A., Capinera, J. L., and Happ, G. M. (1981).Biochemistry and ultrastructure of iridescent virus type 29. J. Invertebr. Pathol. 38: 12-21. Bladon, T., Frosch, M., Sabour, P. M., and Lee, P. E. (1986).Association of nuclear matrix proteins with cytoplasmic assembly sites of Tipula iridescent virus. Virology 55,524533. Boucias, D. G., Maruniak, J. E., and Pendland, J. C. (1987). Characterization of an iridovirus isolated from the southern mole cricket, Scapteriscus vicinus. J. Inuertebr. Pathol. 50, 238-245. Braunwald, J., Tripier, F., and Kirn, A. (1979). Comparison of the properties of enveloped and naked frog virus 3 (FV3) particles. J. Gen. Virol. 45, 673-682. Cameron, I. R. (1990). Identification and characterization of the gene encoding the major structural protein of insect iridescent virus type 22. Virology 178, 35-42. Carey, G. P., Lescott, T., Robertson, J. S., Spencer, L. K., and Kelly, D. C. (1978). Three African isolates of small iridescent viruses. Virology 85, 307-309. Carter, J. B. (1973a). The mode of transmission of Tipula iridescent virus. 11. Source of infection. J. Invert. Pathol. 21, 123-130. Carter, J. B. (1973b). The mode of transmission of Tipula iridescent virus 11. Route of infection. J . Invertebr. Pathol. 21, 136-143. Carter, J. B. (1973~). Detection and assay of Tipula iridescent virus by the latex agglutination test. J. Gen. Virol. 21, 181-185. Carter, J. B. (1974). Tipula iridescent virus infection in the developmental stages of Tipula oleracea. J. Invertebr. Pathol. 24, 271-281. Carter, J. B. (1975). The effect of temperature upon Tipula iridescent virus infection in Tipula oleracea. J. Inuertebr. Pathol. 25, 115-124. Cerutti, M., and Devauchelle, G. (1979). Cell fusion induced by an invertebrate virus. Arch. Virol. 61, 149-155. Cerutti, M ., and Devauchelle, G. (1980). Inhibition of macromolecular synthesis in cells infected with an invertebrate virus (iridovirus type 6 or CIV). Arch. Virol. 63, 297303. Cerutti, M., and Devauchelle, G. (1982). Isolation and reconstruction of Chilo iridescent virus membrane. Arch. Virol. 74, 145-155. Cerutti, M., and Devauchelle, G. (1985). Characterisation and localisation of CIV polypeptides. Virology 145, 123-131. Cerutti, M., and Devauchelle, G. (1990).Protein composition of Chilo iridescent virus. In “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 81-112. Kluwer, Boston. Cerutti, M., Cerutti, P., and Devauchelle, G. (1989). Infectivity of vesicles prepared from Chilo iridescent virus inner membrane: Evidence for recombination between associated DNA fragments. Virus Res. 12, 299-314. Chinchar, V. G., and Granoff, A. (1984). Isolation and characterization of a frog virus 3 variant resistant to phosphonoacetate: Genetic evidence for a virus-specific DNA polymerase. Virology 138, 357-361. Chinchar, V. G., Metzger, D., Granoff, A., and Goorha, R. (1984). Localization of frog virus 3 proteins using monoclonal antibodies. Virology 137, 211-216. Chinchar, V. G., Han, J., Mao, J., Brooks, I., and Srivastava, K. (1994). Instability of frog virus 3 mRNA in productively infected cells. Virology 203, 187-192. Clark, H. F., Brennan, J. C., Zeigel, R. F., and Karzon, D. T.(1968). Isolation and characterization of viruses from the kidneys of Rana pipiens with renal adenocarcinoma before and after passage in the red eft (Triturus viridescens). J. Virol. 2, 629-640. Clark, H. F., Gray, G., Fabian, F., Ziegel, R. F., and Karzon, D. T. (1969). Comparative studies of amphibian cytoplasmic virus strains isolated from the leopard frog, bullfrog and newt. In “Recent Results in Cancer Research Special Supplement Biology of Amphibian Tumors” (M. Mizell, ed.), pp. 310-326. Springer-Verlag, Berlin.

TREVOR WILLIAMS Cole, A,, and Morris, T. J. (1980). A new iridovirus of two species of terrestrial isopods, Armadillidium vulgare and Porcellio scaber. Intervirology 14, 21-30. Comps, M., and Bonami, J. R. (1977). Infection viral associbe des mortalit& chez 1’Huitre Crassostrea gigm Thunberg. C. R . Acad. Sci. Paris 285(Ser. D), 1139-1140. Cordier, O., Aubertin, A. M., Lopez, C., and Tondre, L. (1981). Inhibition de la tranduction par le FV3: Action des proteines virales de structure solubilisees sur la synthesis proteique in vivo et in vitro. Ann. Virol. Inst. Pasteur 132E, 25-39. Cuillel, M., Tripper, F., Brunwald, J., and Jacrot, B. (1979).A low resolution structure of frog virus 3. Virology 99, 277-285. Cunningham, A. A., Langton, T. E. S., Bennett, P. M., Lewtin, J. F., Drury, S. E. N., Gough, R. E., and Macgregor, S. K. (1995).Investigations into unusual mortalities of the common frog ( R a w temporaria) in Britain. J. Wildlife Dis. (in press). Cunningham, J. C., and Tinsley, T. W. (1968). A serological comparison of some iridescent nonoccluded insect viruses. J. Gen. Virol. 3, 1-8. Czuba, M., Tajbakhsh, S., Walker, T., Dove, M. J., Johnson, B. F., and Seligy, V. L. (1994). Plaque assay and replication of Tipula iridescent virus in Spodoptera frugiperda ovarian cells. Res. Virol. 145, 319-330. Darai, G. (1990). “Molecular Biology of Iridoviruses,” Kluwer, Boston. Darai, G., Anders, K., Koch, H., Delius, H., Gelderblom, H., Samalecos, C., and Flugel, R. M. (1983). Analysis of the genome of fish lymphocystis disease virus isolated directly from epidermal tumours of pleuronectes. Virology 126, 466-479. Darai, G., Delius, H., Clarke, J., Apfel, H., Schnitzler, P., and Flugel, R. M. (1985). Molecular cloning and physical mapping of the genome of fish lymphocystis disease virus. Virology 146, 292-301. Davison, S., Carne, A., McMillan, N. A. J.,and Kalmakoff, J. (1992). A comparison of the structural polypeptides of three iridescent viruses (types 6, 9, and 16) and the mapping of the DNA region coding for their major capsid polypeptides. Arch. Virol. 123, 229-237. Dawkins, R., and Krebs, J. R. (1979). Arms races between and within species. Proc. R . Soc. London B 205,489-511. Day, M. F., and Dudzinski, M. L. (1966).The effect of temperature on the development of Sericesthis iridescent virus. Aust. J. Biol. Sci. 19, 481-493. Day, M. F., and Gilbert, N. (1967).The number of particles of Sericesthis iridescent virus required to produce infections of Galleria larvae. Aust. J. Biol. Sci. 20, 691-693. Day, M. F., and Mercer, E. H. (1964). Properties of an iridescent virus from the beetle Sericesthis pruinosa. Aust. J. Biol. Sci. 17, 892-902. DeBlois, R. W., Uzgiris, E. E., Cluxton, W. R., and Mazzone, H. M. (1978). Comparative measurements of size and polydispersity of several insect viruses. Anal. Biochem. 90, 273-288. Delius, H., Darai, G., and Flugel, R. M. (1984). DNA analysis of insect iridescent virus 6: Evidence for circular permutation and terminal redundancy. J. Virol. 49, 609-614. Devauchelle, G. (1977). Ultrastructural characterization of an iridovirus from the marine worm, Nereis diversicolor (0.F. Miiller). Virology 81, 237-247. Devauchelle, G., Stoltz, D. B., and Darcy-Tripier, F. (1985a).Comparative ultrastructure of Iridoviridae. C u m Top. Microbiol. Immunol. 116, 1-21. Devauchelle, G., Attias, J., Monnier, C., Barray, S., Cerutti, M., Guerillon J., and Orange-Balange, N. (1985b). Chilo iridescent virus. Curt-. Top. Microbiol. Zmmunol. 116,37-48. Drillien, R., Spehner, D., and Kirn, A. (1977). Cell killing by FV3: Evidence for cell killing by single viral particles or single viral subunits. Biochem. Biophys. Res. Commun. 79, 105-111.

IRIDOVIRUSES

403

Drury, S. E. N., Gough, R. E., and Cunningham, A. A. (1995).Isolation of a n iridoviruslike agent from common frogs (Runu temporaria). Vet. Record 137, 72-73. Eaton, B. T.,Hyatt, A. D., and Hengstberger, S. (1991).Epizootic haematopoietic necrosis virus: Purification and classification. J. Fish. Dis. 14, 157-169. Elharrar, M., and Kim, A. (1977).Effect of frog virus 3 infection on protein synthesis activity of mouse liver ribosomes. FEMS Lett. 1, 13-16. Elliott, R. M., and Kelly, D. C. (1980).Frog virus 3 replication: Induction and intracellular distribution of polypeptides in infected cells. J. Virol. 33, 28-51. Elliott, R. M., Lescott, T., and Kelly, D. C. (1977).Serological relationships of iridescent virus type 25. Virology 81, 309-316. Elliott, R. M., Bravo, R., and Kelly, D. C. (1980a).Frog virus 3 replication: Analysis of structural and non-structural polypeptides in infected BHK cells by acidic and basic two-dimensional gel electrophoresis. J. Virol. 33, 18-27. Elliott, R. M. Bateson, A., and Kelly, D. C. (1980b).Phosphonacetic acid inhibition of frog virus 3 replication. J. Virol. 33, 539-542. Erlandson, M.A., and Mason, P. G. (1990).An iridescent virus from Simulium vittutum (Diptera: Simuliidae) in Saskatchewan. J. Znvertebr. Puthol. 56, 8-14. Essani, K. (1990).The DNA methylase of frog virus 3.In “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 163-172. Kluwer, Boston. Essani, K., and Granoff, A. (1989).Amphibian and piscine iridoviruses: Proposal, for nomenclature and taxonomy based on molecular and biological properties. ZntervirolOgy 30, 187-193. Federici, B. A. (1980).Isolation of a n iridovirus from two terrestrial isopods, the pill bug, Armudillidium vulgare and the sow bug, Porcellio dilututus. J. Znvertebr. Pathol. 36, 373-381. Fischer, M., Schnitzler, P., Delius, H., and Darai, G. (1988a).Identification and characterization of the repetitive DNA elements in the genome of insect iridescent virus type 6.Virology 167,485-496. Fischer, M., Schnitzler, P., Scholz, J., Rosen-Wolff, A., Delius, H., and Darai, G. (1988b). DNA nucleotide sequence analysis of the PvuII DNA fragment L of the genome of insect iridescent virus type 6 reveals a complex cluster of multiple tandem, overlapping and interdigitated repetitive DNA elements. Virology 167,497-506. Fischer, M., Schnitzler, P., Delius, H., Rosen-Wolff, A., and Darai, G. (1990).Molecular biology of insect iridescent virus type 6. In “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 47-80. Kluwer, Boston. Flugel, R. M. (1985).Lymphocystis disease virus. Curr. Top. Microbiol. Zmmunol. 116, 133-150. Flugel, R. M., Darai, G., and Gelderblom, H. (1982).Viral proteins and adenosine triphosphate phosphohydrolase activity of fish lymphocystis disease virus. Virology 122, 48-55. Fowler, H. G. (1989).An epizootic iridovirus of Orthoptera (Gryllotalpidae: Scuptericus borellii) and its pathogenicity to termites (Isoptera: Cryptotermes). Rev. Microbiol. 20, 115-120. Francki, R. I. B., Fauquet, C. M., Knudson, D. L., and Brown, F. (1991).Classification and nomenclature of viruses. Fifth report of the international committee on taxonomy of viruses. Arch. Virol. Suppl. 2, 132-136. Fukaya, M., and Nasu, S. (1966).A Chilo iridescent virus from the rice stem borer, Chilo supressalis Walker (Lepidoptera: Pyrallidae). Appl. Entomol. Zool. 1, 69-72. Fukuda, T.,and Clark, T. B. (1975).Transmission of the mosquito iridescent virus (RMIV) by adult mosquitoes of Aedes tmniorhynchus to their progeny. J. Znvertebr. Puthol. 25, 275-276.

404

TREVOR WILLIAMS

Glitz, D. G., Hills, G. J., and Rivers, C. F. (1968). A comparison of the Tipula and Sericesthis iridescent viruses. J. Gen. Virol. 3, 209-220. Goorha, R. (1981). Frog virus 3 requires RNA polymerase I1 for its replication. J. Virol. 37,496-499.

Goorha, R. (1982). Frog virus 3 replication occurs in two stages. J. Virol. 43, 519-528. Goorha, R., and Dixit, P. (1984). A temperature-sensitive (ts) mutant of frog virus 3 (FV3) is defective in second stage DNA replication. Virology 136, 186-195. Goorha, R., and Granoff, A. (1974). Macromolecular synthesis in cells infected by frog virus 3. I. Virus-specific protein synthesis and its regulation. Virology 80, 237-250. Goorha, R., and Granoff, A. (1979). Icosahedral cytoplasmic deoxyriboviruses. In “Comprehensive Virology” (H. Fraenkel-Conrat and R. R. Wagner, eds.), pp. 347-399. Plenum, New York. Goorha, R., and Murti, K. G. (1982). The genome of frog virus 3, an animal DNA virus, is circularly permuted and terminally redundant. Proc. Natl. Acad. Sci. U.S.A.79,248252.

Goorha, R., Willis, D. B., and Granoff, A. (1977). Macromolecular synthesis in cells infected by frog virus 3. VI. Frog virus 3 replication is dependent on the cell nucleus. J. Virol. 21, 802-805. Goorha, R., Murti, K. G., Granoff, A., and Tirey, R. (1978). Macromolecular synthesis in cells infected by frog virus 3. VIII. The nucleus is the site of frog virus 3 DNA and RNA synthesis. Virology 84, 32-50. Goorha, R., Willis, D. B., and Granoff, A. (1979). Macromolecular synthesis in cells infected by frog virus 3. XII. Viral regulating proteins in transcriptional and posttranscriptional controls. J . Virol. 32, 442-448. Goorha, R., Willis, D. B., Granoff, A,, and Naegele, R. F. (1981). Characterization of a temperature sensitive mutant of frog virus 3 defective in DNA replication. Virology 112,40-48.

Goorha, R., Granoff, A., Willis, D. B., and Murti, K. G. (1984). The role of DNA methylation in virus replication: Inhibition of frog virus 3 replication by 5-azacytidine. Virology 138, 94-102. Gorman, C. M., Moffatt, L. F., and Howard, B. H. (1982). Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cells. Mol. Cell. Biol. 2, 1044-1051.

Granoff, A., Came, P. E., and Breeze, D. C. (1966). Viruses and renal carcinoma ofRana pipiens. I. The isolation and properties of viruses from normal and tumor issue. Virology 29, 133-148. Granoff, A., Gravell, M., and Darlington, R. W. (1969). Studies on the viral etiology of the renal adenocarcinoma of Rana pipiens (Luck6 tumour). In “Recent Results in Cancer Research Special Supplement: Biology of Amphibian Turnours” (M. Mizell, ed.), pp. 279-295. Springer-Verlag, Berlin. Gravell, M., and Cromeans, T. (1971). Mechanisms involved in the non-genetic reactivation of frog polyhedral cytoplasmic deoxyribovirus: Evidence for RNA polymerase in the virion. Virology 46, 39-49. Gravell, M., and Granoff, A. (1970). Viruses and renal carcinoma of Rana pipiens. IX. The influence of temperature and host cell on replication of frog polyhedral cytoplasmic deoxyribovirus (PCDV). Virology 41, 596-602. Gravell, M., and Naegele, R. F. (1970). Non-genetic reactivation of frog polyhedral cytoplasmic deoxyribovirus (PCDV). Virology 40,170-174. Grosholz, E. D. (1992). Interactions of intraspecific, interspecific and apparent competition with host-pathogen population dynamics. Ecology 73, 507-514. Grosholz, E. D. (1993). The influence of habitat heterogeneity on host-pathogen population dynamics. Oecologiu 96, 347-353.

IRIDOVIRUSES

405

Guir, M., Braunwald, J., and Kirn, A. (1970). Inhibition de la synthese du DNA et des RNA cellulaires dans les cellules KB infectee avec le virus 3 de la grenouille (FV3). C. R. Hebd. Acad. Sci. Paris 270, 2605-2608. Hall, D. W. (1985). Pathobiology of invertebrate icosahedral cytoplasmic deoxyriboviruses (Zridouiridae).Zn “Viral Insecticides for Biological Control” (K. Maramorosh and K. E. Sherman, eds.), pp. 163-196. Academic Press, New York. Hall, D. W., and Anthony, D. W. (1971). Pathology of a mosquito iridescent virus (MIV) infecting Aedes taeniorhychus. J . Znuertebr. Pathol. 18, 61-69. Hamilton, W. D. (1982). Pathogens as causes of genetic diversity in their host populations. Zn “Population Biology of Infectious Diseases” (R. M. Anderson and R. M. May, eds.), pp. 269-296. Springer-Verlag, Berlin. Handermann, M., Schnitzler, P., Rosen Wolff, A., Raab, K., Sonntag, K. C., and Darai, G. (1992). Identification and mapping of origins of DNA replication within the DNA sequences of the genome of insect iridescent virus type 6. Virus Genes 6, 19-32. Hemsley, A. R., Collinson, M. E., Kovach, W. L., Vincent, B., and Williams, T.(1994). The role of self-assembly in biological systems: Evidence from iridescent colloidal sporopollenin in Selaginella megaspore walls. Philos. Trans. R. Soc. London B 345, 163173. Hendrick, R. P., Groff, J. M., McDowell, T., and Wingfield, W. H. (1990). An iridovirus infection of the integument of the white sturgeon Acipenser transmontanus. Dis. Aquat. Org. 8, 39-44. Hendrick, R. P., McDowell, T. S., Ahne, W., Torhy, C., and de Kinkelin, P. (1992). Properties of three iridovirus like agents associated with systemic infections of fish. Dis. Aquat. Org. 13, 132-136. Hengstberger, S. G., Hyatt, A. D., Speare, R., and Coupar, B. E. H. (1993). Comparison of epizootic haematopoietic necrosis and Bohle iridoviruses, recently isolated Australian iridoviruses. Dis. Aquat. Org. 15, 93-107. Hem, R. T., and Poinar, G. O., Jr. (1985). Iridoviruses infecting terrestrial isopods and nematodes. Curr. Top. Microbiol. Zmmunol. 116,49-76. Home, W. A., Tajbakhsh, S., and Seligy, V. L. (1990). Molecular cloning and characterization of a late Tipula iridescent virus gene. Gene 94, 243-248. Houts, G. E., Gravell, M.,and Darlington, R. W. (1970).Base composition and molecular weight of DNA from a frog polyhedral cytoplasmic deoxyribovirus. Proc. SOC.Exp. Biol. Med. 135, 232-236. Houts, G. E., Gravell, M., and Granoff, A. (1974). Electron microscopic observation of early events of frog virus 3 replication. Virology 58, 589-594. Hughes, D. S., Possee, R. D., and King, L. A. (1993). Activation and detection of a latent baculovirus resembling Mamestra brassicae nuclear polyhedrosis virus in M . brassicue insects. Virology 194, 608-615. Kalmakoff, J., and Tremaine, J. H. (1968). Physiochemical properties of Tipula iridescent virus. J. Virol. 2, 738-744. Kalmakoff, J., McMillan, N., and Davison, S. (1990). Insect iridescent virus type 9 and type 16. Zn “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 113-136. Kluwer, Boston. Kaminski, S., Clark, H. F., and Karzon, D. T. (1969). Comparative immune response to amphibian cytoplasmic viruses assayed by the complement fixation and gel immunodiffusion techniques. J. Zmmunol. 103, 260-267. Kang, H. S., and McAuslan, B. R. (1972). Virus associated nucleases: Location and properties of deoxyribonucleases and ribonucleases in purified frog virus 3. Virology 58,587-594. Kelly, D. C. (1985). Insect iridescent viruses. Curr. Top. Microbiol. Immunol. 116, 2335.

406

TREVOR WILLIAMS

Kelly, D. C., and Robertson, J. S. (1973). Icosahedral cytoplasmic deoxriboviruses. J.Gen. Virol. Suppl. 20, 17-41. Kelly, D. C., and Tinsley, T. W. (1974). Iridescent virus replication: A microscope study of Aedes uegypti and Antherea eucalypti cells in culture infected with iridescent virus types 2 and 6. Microbios 9, 75-93. Kelly, D. C., and Vance, D. E. (1973). The lipid content of two iridescent viruses. J. Gen. Virol. 21, 417-423. Kelly, D. C., Ayres, M. D., Lescott, T., Robertson, J. S., and Happ, G. M. (1979). A small iridescent virus (type 29) isolated from Tenebrio molitor: A comparison of its proteins and antigens with six other iridescent viruses. J. Gen. Virol. 42, 95-105. Kloc, M., Lee, P. E., and Tajbakhsh, S. (1984). The effect of two insect hormones on the replication of Tipula iridescent virus in Estigmene acrea cells in suspension culture. J.Inuertebr. Pathol. 43, 114-123. Klug, A., Franklin, R. E., and Humphreys-Owen, S. P. F. (1959). The crystal structure of Tipula iridescent virus as determined by Bragg reflection of visible light. Biochim. Bwphys. Acta 32, 203-219. Klump, H., Beaumais, J., and Devauchelle, G. (1983). Structural and thermodynamic investigation of the Chilo iridescent virus (iridovirus type 6). Arch. Virol. 75, 269276. Krell, P., and Lee, P. E. (1974). Polypeptides in Tipula iridescent virus (TIV) and in TIVinfected hemocytes of Galleria Mellonella (L.) larvae. Virology 60, 315-326. Langdon, J. S., Humphery, J. D., Williams, L. M., Hyatt, A. D., and Westbury, H. A. (1986). First virus isolation from Australian fish: An iridovirus-like pathogen from redfin perch, Perca fluviatilis. J.Fish Dis. 9, 263-268. Langdon, J. S., Humphery, J. D., and Williams, L. M. (1988).Outbreaks of a n EHNV-like iridovirus in cultured rainbow trout, Salmo gairdneri Richardson, in Australia. J. Fish Dis. 11, 93-96. Lea, M. S. (1985). A Sericesthis iridescent virus infection of the hemocytes of the waxmoth, Galleria mellonella (Lepidoptera). J.Inuertebr. Pathol. 46, 219-230. Lee, P. E., and Brownrigg, S. (1982). Effect of virus inactivation on Tipula iridescent virus-cell relationships. J.Ultrastruct. Res. 79, 189-197. Lee, M. H., and Willis, D. B. (1983).Restriction endonuclease mapping of the frog virus 3 genome. Virology 126, 317-327. Lehane, D. E., Clark, H. F., and Kanon, D. T. (1968). Antigenic relationships among frog viruses demonstrated by plaque reduction and neutralization kinetics tests. Virology 34,590-595. Lightner, D. V., and Redman, R. M. (1993). A putative iridovirus from the penaeid shrimp Protrachypene precipua Burkenroad (Crustacea: Decapoda). J. Inuertebr. Pathol. 62, 107-109. Linley, J. R., and Nielsen, H. T.(1968a). Transmission of a mosquito iridescent virus in Aedes taeniorhynchus. I. Laboratory experiments. J.Inuertebr. Pathol. 12, 7-16. Linley, J. R., and Nielsen, H. T. (1968b). Transmission of a mosquito iridescent virus in Aedes taeniorhynchus. 11. Experiments related to transmission in nature. J.Inuertebr. Pathol. 12, 17-24. Lunger, P. D., and Came, P. E. (1966). Cytoplasmic viruses associated with Luck6 tumor cells. Virology 30, 116-126. McAuslan, B. R., and Armentrout, R. W.(1974). The biochemistry of icosahedral cytoplasmic deoxyviruses. Curr. Top. Microbiol. Immunol. 68, 77-105. McMillan, N., and Kalmakoff, J. (1994). RNA transcript mapping of the Wiseana iridescent virus genome. Virus Res. 32, 343-352. Maes, R., and Granoff, A. (1967a). Viruses and renal carcinoma of Rana pipzens. IV. Nucleic acid synthesis in frog virus 3-infected BHK 21/13 cells. Virology 33,491-501.

IRIDOVIRUSES

407

Maes, R., and Granoff, A. (196713). Viruses and renal carcinoma ofRanapipiens. 111. The relationship between input multiplicity of infection and inclusion body formation in frog virus 3 infected cells. Virology 33, 137-144. Manyakov, V. F. (1977).Fine structure of the iridescent virus type 1 capsid. J. Gen. Virol. 36, 73-79. Martin, J . P., Aubertin, A. M., Londre, L., and Kirn, A. (1984). Fate of frog virus 3 DNA replicated in the nucleus of arginine-deprived CHO cells. J. Gen. Virol. 65, 721-732. Mathews, R. E. F. (1982). Classification and nomenclature of viruses: Fourth report of the international committee on taxonomy of viruses. Zntervirology 17, 56-58. Mathieson, W. B., and Lee, P. E. (1981). Cytology and autoradiography of Tipula iridescent virus infection of insect suspension cell cultures. J. Ultrastruct. Res. 74, 59-68. Matta, J. F. (1970).The characterization of a mosquito iridescent virus (MIV). 11. Physiochemical characterization. J . Znuertebr. Pathol. 16, 157-164. Mesnard, J. M., Tham, T. N., Tondre, L., Aubertin, A. M., and Kirn, A. (1988). Organization of RNA transcripts from a 7.8-kb region of the frog virus 3 genome. Virology 165, 122-133. Midlige, F. H., and Malsberger, R. G. (1968). In uitro morphology and maturation of lymphocystis virus. J. Virol. 2, 830-835. Monnier, C., and Devauchelle, G. (1976). Enzyme activities associated with a n invertebrate iridovirus: Nucleotide phosphohydrolase activity associated with iridescent virus type 6 (CIV). J. Virol. 19, 180-186. Monnier, C., and Devauchelle, G. (1980). Enzyme activities associated with a n invertebrate iridovirus: Protein kinase activity associated with iridescent virus type 6 (Chilo iridescent virus). J. Virol. 35, 444-450. Montanie, H., Bonami, J. R., and Comps, M. (1993).Irido-like virus infection in the crab Macropipus depurator L. (Crustacea: Decapoda). J. Znuertebr. Pathol. 61,320-322. Moore, N. F., and Kelly, D. C. (1980). A comparative study of the polypeptides of three iridescent viruses by N terminal analysis and surface labeling. J . Invert. Pathol. 36, 415-422. Murphy, F. A., Fauquet, C. M., Bishop, D. H. L., Ghabrial, S. A., Jarvis, A. W., Martelli, G. P., Mayo, M. A,, and Summers, M. D. (1995). Virus Taxonomy: Sixth report of the International Committee on Taxonomy of Viruses. Springer-Verlag, NY. Murti, K. G., and Goorha, R. (1983). Interaction of frog virus 3 with the cytoskeleton. I. Altered organization of microtubules, intermediate filaments and microfilaments. J. Cell Biol. 96, 1248-1257. Murti, K. G., and Goorha, R. (1990). Virus-cytoskeleton interaction during replication of frog virus 3. I n “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 137-162. Kluwer, Boston. Murti, K. G., Goorha, R., and Granoff, A. (1982). Structure of frog virus 3: Genome size and arrangement of nucleotide sequences as determined by electron microscopy. Virology 116,275-283. Murti, K. G., Porter, K. R., Goorha, R., Ulrich, M., and Wray, G. (1984). Interaction of frog virus 3 with the cytoskeleton 2. Structure and composition of the virus assembly site. Exp. Cell Res. 154, 270-282. Murti, K. G., Goorha, R., and Granoff, A. (1985a). An unusual replication strategy of a n animal iridovirus. Adu. Virus Res. 30, 1-19. Murti, K. G., Goorha, R., and Chen, M. (1985b). Interaction of frog virus 3 with the cytoskeleton. Curr. Top. Microbiol. Zmmunol. 116, 107-131. Murti, K. G., Goorha, R., and Klymowsky, M. W. (1988).A functional role for intermediate filaments in the formation of frog virus 3 assembly sites. Virology 162, 264-269. Poprawski, T. J., and Yule, W. N. (1990). A new small iridescent virus from grubs of Phyllophaga anxia (LeConte) (Col.:Scarabidae). J. Appl. Entomol. 110, 63-67.

408

TREVOR WILLIAMS

Pozet, F., Moussa, A., Torhy, C., and de Kinkelin, P. (1992).Isolation and preliminary characterization of a pathogenic icosahedral deoxyribovirus from the catfish Zctalurus melas. Dis. Aquat. Org. 14, 35-42. Raghow, R., and Granoff, A. (1980).Macromolecular synthesis in cells infected by frog virus 3.XIV. Characterization of the methylated nucleotide sequences in viral messenger RNAs. Virology 107,283-294. Ricou, G. (1975).Production de Tipulapaludosa Meig. en prarie en function de l’humiditi6 du sol. Rev. Ecol. Biol. Sol 12, 69-89. Robach, Y.,Michels, B., Cerf, F., Braunwald, J., and Tripier-Darcy, F. (1983).Ultrasonic absorption evidence for structural fluctuations in frog virus 3 and its subparticles. Proc. Natl. Acad. Sci. U.S.A.80,3981-3985. Robin, J., Larviere-Durand, C., and Bernard, J. (1983).The chemical composition of lymphocystis disease virus of fish. Reu. Can. Biol. Exp. 42, 173-176. Rohozinski, J., and Goorha, R. K. (1992).A frog virus 3 gene codes for a protein containing the motif characteristic of the INT family of integrases. Virology 186, 693-700. Rondelaud, D., and Barthe, D. (1992).Observations epidemiologiques sur l’iridovirose de Lymnaea truncatula, mollusque vecteur de Fasciola hepatica. C . R . Acad. Sci. Paris Ser. 3 314,609-612. Runnger, D., Rastelli, M., Braendle, E., and Malsberger, R. G. (1971).A virus-like particle associated with lesions in the muscle of Octopus uulgaris. J.Znuertebr.Pathol. 17, 72-80. Schetter, C.,Grunemann, B., Holker, I., and Doerfler, W. (1993).Patterns of frog virus 3 DNA methylation and DNA methyltransferase activity in nuclei of infected cells. J. Virol. 67,6973-6978. Schmelz, M., Sodeik, B., Ericsson, M., Wolffe, E. J., Shida, H., Hiller, G., and Griffiths, G. (1994).Assembly of vaccinia virus: The second wrapping cisterna is derived from the trans Golgi network. J. Virol. 68, 130-147. Schmitt, M. P., Tondre, L., Kirn, A., and Aubertin, A. M. (1990).The nucleotide sequence of a delayed early gene (31K)of frog virus 3.Nucleic Acids Res. 18, 4000. Schnitzler, P., and Darai, G. (1993).Identification of the gene encoding the major capsid protein of fish lymphocystis disease virus. J . Gen. Virol. 74, 2143-2150. Schnitzler, P., Soltau, J. B., Fischer, M., Reisner, H., Scholz, J., Delius, H., and Darai, G. (1987).Molecular cloning and physical mapping of the genome of insect iridescent virus type 6;further evidence of for circular permutation of the viral genome. Virology 160,66-74. Schnitzler, P., Rosen-Wolff, A., and Darai, G. (1990).Molecular biology of fish lymphocystis disease virus. In “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 203234.Kluwer, Boston. Schnitzler, P., Handermann, M., Szepe, O., and Darai, G. (1991).The primary structure of the thymidine kinase gene of fish lymphocystis disease virus. Virology 182,835840. Schnitzler, P., Hug, M., Handermann, M., Janssen, W., Koonin, E. V., Delius, H., and Darai, G. (1994a).Identification of genes encoding zinc finger proteins, non-histone chromosomal HMG protein homolgue, and a putative GTP phosphohydrolase in the genome of Chilo iridescent virus. Nucleic Acids Res. 22, 158-166. Schnitzler, P., Sonntag, K. C., Muller, M., Janssen, W., Bugert, J. J., Koonin, E. V., and Darai, G. (1994b).Insect iridescent virus type 6 encodes a polypeptide related to the largest subunit of eukaryotic RNA polymerase 11. J. Gen. Virol. 75, 1557-1567. Scholz, J., Rosen-Wolff, A., Touray, M., Schnitzler, P., and Darai, G. (1988).Identification, mapping, and cloning of the thymidine kinase gene of fish lymphocystis disease virus. Virus Res. 9,63-72.

IRIDOVIRUSES Schultz, G. A., Garthwaite, R. L., and Sassaman, C. (1982).A new family placement for Mauritaniscus littorinus (Miller)N. Comb. from the west coast of North America with ecological notes (Crustacea: Isopoda: Oniscoidea: Bathytropidae). Wasmann J.Biol. 40, 77-89. Schuster, A. M., Burbank, D. E., Meister, B., Skrdla, M. P., Meints, R. H., Hattmann, S., Swinton, D., and Van Etten, J. L. (1986). Characterization of viruses infecting a eukaryotic Chlorella-like green alga. Virology 150, 170-177. Seagull, R., Lee, P. E., and Frosch, M. (1985).Comparison of microtubules and microfilaments in Tipula iridescent virus infected and uninfected cells. Can. J.Biochem. Cell Biol. 63, 543-552. Sieburth, P. J., and Carner, G. R. (1987). Infectivity of an iridescent virus for larvae of Anticarsia gemmatalis (Lepidoptera: Noctuidae). J.Znvertebr. Pathol. 49, 49-53. Sikorowski, P. P., and v s o n , G. E. (1984). Per 0s transmission of iridescent virus of Helothis zea (Lepidoptera: Noctuidae). J. Znuertebr. Pathol. 44,97-102. Silberstein, H., and August, J . T. (1973). Phosphorylation of animal virus proteins by a virion protein kinase. J. Virol. 12, 511-522. Silberstein, H., and August, J. T. (1976). Purification and properties of a virion protein kinase. J.Biol. Chem. 251,3176-3184. Smith, K. M. (1958). The morphology and crystallisation of insect cytoplasmic viruses. Virology 5, 168-175. Smith, W. R., and McAusland, B. R. (1969). Biophysical properties of frog virus 3 and its DNA: Fate of radioactive virus in early stages of infection. J. Virol. 4, 332-347. Soltau, J. B., Fischer, M., Schnitzler, P., Scholz, J., and Darai, G . (1987).Characterization of insect iridescent virus type 6 by physical mapping. J. Gen. Virol. 68, 2717-2722. Sonntag, K. C., and Darai, G. (1992). Characterization of the third origin of DNA replication of the genome of insect iridescent virus type 6. Virus Genes 6, 333-342. Sonntag, K. C., Schnitzler, P., Koonin, E., and Darai, G. (1994).Chilo iridescent virus encodes a putative helicase belonging to a distinct family within the “DEAD/H superfamily: Implications for the evolution of large DNA viruses. Virus Genes 8, 151-158. Speare, R., and Smith, J. R. (1992). An iridovirus-like agent isolated from the ornate burrowing frog Limnodynastes ornatus in northern Australia. Dis.Aquat. Org. 14, 51-57. Stadelbacher, E. A., Adams, J. R., Faust, R. M., and Tompkins, G. J. (1978).An iridescent virus of the bollworm Heliothis zea (Lepidoptera: Noctuidae). J.Znvertebr. Pathol. 32, 71-76. Stebhens, W. E., and Johnston, M. R. L. (1966).The viral nature ofPirhemocyton tarentolae. J. Ultrastruct. Res. 15, 543-554. Stohwasser, R., Raab, K., Schnitzler, P., Janssen, W., and Darai, G. (1993).Identification of the gene encoding the major capsid protein of insect iridescent virus type 6 by polymerase chain reaction. J. Gen. Virol. 74, 873-879. Stoltz, D. B. (1971).The structure of icosahedral cytoplasmic deoxyriboviruses. J. Ultrastruct. Res. 37, 219-239. Stoltz, D. B. (1973). The structure of icosahedral cytoplasmic deoxyriboviruses 11. An alternative model. J. Ultrastruct. Res. 43, 58-74. Stoltz, D. B., and Summers, M. D. (1971). Pathway of infection of mosquito iridescent virus. I. Preliminary observations on the fate of ingested virus. J. Virol. 8,900-909. Stoltz, D. B., Hilsenhoff, W. L., and Stich, H. F. (1968). A virus disease of Chironomus plumosus. J.Znvertebr. Pathol. 12, 118-126. Stuart, D., Upton, C., Higman, M. A., Niles, E. G., and McFadden, G. (1993).A poxvirusencoded uracil DNA glycosylase is essential for virus viability. J. Virol. 67, 25032512.

410

TREVOR WILLIAMS

Tajbakhsh, S., and Seligy, V. L. (1990). Molecular biology of Tipula iridescent virus. In “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 13-46. Kluwer, Boston. Tajbakhsh, S., Dove, M. J.,Lee, P. E., and Seligy, V. L. (1986).DNA components of Tipula iridescent virus. Biochem. Cell Biol. 64, 495-503. Tajbakhsh, S., Lee, P. E., Watson, D. C., and Seligy, V. L. (1990a). Molecular cloning and expression of the Tipula iridescent virus capsid gene. J. Virol. 64, 125-136. Tajbakhsh, S., Kiss, G., Lee, P. E., and Seligy, V. L. (1990b).Semipermissive replication of Tipula iridescent virus in Aedes albopictus C6l36 cells. Virology 174, 264-275. Tanada, Y., and Tanabe, A. M. (1965).Resistance of Galleria mellonella (Linnaeus)to the Tipula iridescent virus at high temperatures. J.Inuertebr. Pathol. 7 , 184-188. Thomas, R. S. (1961).The chemical composition and particle weight of Tipula iridescent virus. Virology 14, 240-252. Thompson, J. P., Granoff, A., and Willis, D. B. (1988).Methylation of the promoter for an immediate-early frog virus 3 gene does not inhibit transcription. J. Virol. 62, 46804685. Tinsley, T. W., and Kelly, D. C. (1970).An interim nomenclature system for the iridescent group of viruses. J. Inuertebr. Pathol. 12, 66-68. Tinsley, T. W., Robertson, J. S., Rivers, C. F., and Service, M. W. (1971). An iridescent virus of Aedes cantans in Great Britain. J.Inuertebr. Pathol. 18, 427-428. Tondre, L., Tham, T. N., Mutin, P. H., and Aubertin, A. M. (1988). Molecular cloning and physical and translational mapping of the frog virus 3 genome. Virology 162, 108117. Tripier-Darcy, F., and Nermut, M. V. (1983). Cryodecapage d’un virus, le FV3 (frog virus 3): Structure fine de la capside et organisation interne. Biol. Cell. 48, 17a. Tripier-Darcy, F., Braunwald, J., and Kirn, A. (1982). Localization of some frog virus 3 structural proteins. Virology 116, 635-640. Tweedell, K., and Granoff, A. (1968). Viruses and renal carcinoma of R a m pipiens. V. Effect of frog virus 3 on developing frog embryos and larvae. J.Natl. Cancer Inst. 40, 407-410. Van Etten, J. L., Meints, R. H., Kuczmarski, D., Burbank, D. E., and Lee, K. (1982). Viruses of symbiotic Chlorella-like algae isolated from Paramecium bursaria and Hydra uiridis. Proc. Natl. Acad. Sci. U S A . 79, 3867-3871. Van Etten, J. L., Lane, L. C., and Meints, R. H. (1991).Unicellular plants also have large dsDNA viruses. Semin. Virol. 2, 71-77. Van Regenmortel, M. H. V. (1990). Virus species, a much overlooked but essential concept in virus classification. Intervirology 31, 241-254. Vilagines, R., and McAuslan, B. R. (1971). Proteins of polyhedral cytoplasmic deoxyribovirus. I1 Nucleotide phosphohydrolase activity associated with frog virus 3. J. Virol. 7, 619-624. Wagner, G. W., and Paschke, J. D. (1977). A comparison of the DNA of R and T strains of mosquito iridescent virus. Virology 81, 298-308. Wagner, G . W., Paschke, J. C., Campbell, W. R., and Webb, S. R. (1973). Biochemical and biophysical properties of two strains of mosquito iridescent virus. Virology 5 2 , 7 2 4 0 . Wagner, G. W., Webb S. R., Paschke, J. D., and Campbell, W. R. (1975). Production and characterization of the cores of the “R’ strain of mosquito iridescent virus. Virology 64,430-437. Wagner, H., Simon, D., Werner, E., Gelderblom, H., Darai, G., and Flugel, R. (1985). Methylation pattern of fish lymphocystis disease virus DNA. J. Virol. 53, 10051007. Walker, R. (1962). Fine structure of a virus tumour of fish. Virology 18,503-511. Ward, V. K., and Kalmakoff, J. (1987). Physical mapping of the DNA genome of insect iridescent virus type 9 from Wiseana spp. larvae. Virology 160, 507-510.

IRIDOVIRUSES

411

Ward, V. K., and Kalmakoff, J. (1991).Invertebrate Zridouiridae.Zn “Viruses of Invertebrates” (E.Kurstak, ed.), pp. 197-226. Dekker, New York. Webb, S. R., Paschke, J. D., Wagner, G. W., and Campbell, W. R. (1973).Infection of Aedes mgypti cells with mosquito iridescent virus. J. Znuertebr. Puthol. 23, 255-258. Webb, S.R., Paschke, J. D., Wagner, G. W., and Campbell, W. R. (1976).Pathology of mosquito iridescent virus of Aedes tmniorhynchus in cell cultures of Aedes aegypti. J. Znvertebr. Puthol. 27, 27-40. Williams, R. C., and Smith, K. M. (1957).A crystallizable insect virus. Nature (London) 179, 119-120. Williams, R. C., and Smith, K. M. (1958).The polyhedral form of the Tcpulu iridescent virus. Biochim. Biophys. Actu 28,464-469. Williams, T.(1993).Covert iridovirus infection of blackfly larvae. Proc. R. SOC.London B 251, 225-230. Williams, T. (1994).Comparative studies of iridoviruses: Further support for a new classification. Virus Res. 33, 99-121. Williams, T.(1995).Patterns of covert infection by invertebrate pathogens: Iridescent viruses of blackflies. Mol. Ecol. 4, 447-457. Williams, T., and Cory, J. S. (1993).DNA restriction fragment polymorphism in iridovirus isolates from individual blackflies (Diptera: Simuliidae). Med. Vet.Entomol. 7, 199-201. Williams, T., and Cory, J. S. (1994).Proposals for a new classification of iridescent viruses. J. Gen. Virol. 75, 1291-1301. Williams, T.,and Thompson, I. P. (1995).Fatty acid profiles of iridescent viruses. Arch. Virol. 140, 975-981. Willis, D. B. (1987).DNA sequences required for trans-activation of an immediate-early frog virus 3 gene. Virology 161, 1-7. Willis, D. B. (1990).Taxonomy of iridoviruses. Zn “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp 1-12. Kluwer, Boston. Willis, D. B., and Granoff, A. (1974).Lipid composition of frog virus 3. Virology 61,256269. Willis, D. B., and Granoff, A. (1976).Macromolecular synthesis in cells infected by frog virus 3. IV. Regulation of virus-specific RNA synthesis. Virology 70, 397-410. Willis, D. B., and Granoff, A. (1978).Macromolecular synthesis in cells infected by frog virus 3. IX. Two temporal classes of early viral RNA. Virology 68,443-453. Willis, D. B., and Granoff, A. (1980).Frog virus 3 is heavily methylated at CpG sequences. Virology 107, 250-257. Willis, D. B., and Granoff, A. (1985).Trans-activation of an immediate-early frog virus 3 gene by a virion protein. J. Virol. 56, 495-501. Willis, D. B., Goorha, R., Miles, M., and Granoff, A. (1977).Macromolecular synthesis in cells infected by frog virus 3.VII. Transcriptional and post-transcriptional regulation of virus gene expression. J . Virol. 24,326-324. Willis, D. B., Goorha, R., and Granoff, A. (1979a).Nongenetic reactivation of frog virus 3 DNA. Virology 98, 476-479. Willis, D. B., Goorha, R., and Granoff, A. (1979b).Macromolecular synthesis in cells infected by frog virus 3. XI. A ts mutant of frog virus 3 that is defective in late transcription. Virology 98, 328-335. Willis, D. B., Goorha, R., and Granoff (1984a).DNA methyltransferase induced by frog virus 3.J. Virol. 49, 86-91. Willis, D. B., Foglesong, D., and Granoff, A. (1984b). Nucleotide sequence of an immediate-early frog virus 3 gene. J. Virol. 52, 905-912. Willis, D. B., Goorha, R., and Chinchar, V. G. (1985).Macromolecular synthesis in cells infected by frog virus 3. Curr. Top. Microbiol. Zmmunol. 116, 77-106.

412

TREVOR WILLIAMS

Willis, D. B., Thompson, J. P., and Beckman, W. (1990). Transcription of frog virus 3. In “Molecular Biology of Iridoviruses” (G. Darai, ed.), pp. 173-186. Kluwer, Boston. Willison, J. H. M., and Cocking, E. C. (1972). Frozen-fractured viruses: A study of virus structure using freeze-etching. J. Microsc. 95, 397-411. Witt, D. J., and Stairs, G. R. (1976). Effects of different temperatures on Tipula iridescent virus infection in Galleria mellonella larvae. J. Znuertebr. Pathol. 28, 151-152. Wrigley, N. G. (1969). An electron microscope study of the structure of Sericesthis iridescent virus. J. Gen. Virol. 5, 123-134. Wrigley, N. G. (1970). An electron microscope study of the structure of Tipula iridescent virus. J. Gen. Virol. 6, 169-173. Wolf, K. (1988). Carp iridovirus. In “Fish Viruses and Fish Viral Diseases,” pp. 313-315. Comstock, London. Wolf, K., Bullock, G. L., Dunbar, C. E., and Quimby, M. C. (1968). Tadpole edema virus: A viscerotropic pathogen for anuran amphibians. J. Infect. Dis. 118, 253-262. Woodard, D. B., and Chapman, H. C. (1968). Laboratory studies with the mosquito iridescent virus (MIV). J. Inuertebr. Pathol. 11, 296-301. Xeros, N. (1954). A second virus disease of the leather jacket, Tipula paludosu. Nature (London) 174,562-563. Xeros, N. (1964). Development of Tipula iridescent virus. J . Insect Pathol. 6, 261-271. Yamamoto, T., Macdonald, R. D., Gillespie, D. C., and Kelly, R. K. (1976).Viruses associated with lymphocystis disease and dermal sarcoma of walleye (Stizostedion vitreum vitreum). J . Fish. Res. Board Can. 33, 2408-2419. Yule, G. B., and Lee, P. E. (1973). A cytological and immunological study of Tipula iridescent virus-infected Galleria mellonella larval hemocytes. Virology 51,409-423. Zwillenberg, L. O., and Wolf, K. (1968).Ultrastructure of lymphocystis virus. J . Virol. 2, 393-399.

ADVANCES IN VIRUS RESEARCH, VOL. 46

MOLECULAR BIOLOGY OF LUTEOVIRUSES

M. A. Mayo and V. Ziegler-GraW Scottish Crop Research Institute Invergowrie, Dundee, Scotland UK and 'Institut de Biologie Moleculaire des Plantes 67084 Strasbourg Cedex, France

I. Introduction 11. Genome Structure A. Arrangement of Open Reading Frames B. Variation among Coding Sequences C. Terminal Structures and Noncoding Regions D. Variation among Strains of Luteoviruses E. Putative Recombination Involving Luteovirus RNA 111. Functions of Gene Products A. Approaches to Determining Gene Function B. Studies on PO C. Studies on P1 and P2 D. Studies on P3 E. Studies on P4 F. Studies on P5 G. Studies on P6 H. Summary of Roles of BWYV Proteins Deduced by Mutagenesis IV. Mechanisms of Gene Expression A. Translational Frameshifting B. Internal Initiation of Translation C. Subgenomic mRNA Synthesis D. Readthrough or Leaky Translational Termination E. Proteolysis and Cap-Independent Translation V. Particle Structure A. Possible Tertiary Structure B. Location of Epitopes C. Presence of Readthrough Protein (P5) D. Heterologous Encapsidation E. Determinants for Particle Assembly VI. Location of Luteovirus Replication A. Limitation to Phloem Tissue B. Cytopathological Effects VII. Phytopathology A. Diagnosis of Luteovirus Infection B. Resistance t o Luteoviruses VIII. Taxonomy A. Speciation B. Structure of Genus Luteovirus C. Problem of Pea Enation Mosaic Virus D. RNA Associated with Luteoviruses IX. Concluding Remarks References 413

Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

414

M. A. MAY0 AND V.ZIEGLERGRAFF

I, INTRODUCTION Luteoviruses were first recognized formally as a group of related viruses by Shepherd et al. (19761, although the effects of luteovirus infection had long been known in the class of “yellowing” diseases (Duffus, 1977). The effects of these diseases can cause significant losses in a variety of crops in many countries of the world (Rochow and Duffus, 1981; Casper, 1988). Luteoviruses were characterized initially by their biological properties; they are aphid-transmitted in a persistent, apparently nonpropagative manner and multiply in the phloem tissues of infected plants (Waterhouse et al., 1988). The characteristic yellowing (hence the prefix luteo- from Latin luteus, yellow) or reddening of the foliage and rolling of the leaves probably reflect the pathological effects of infection on the phloem tissue of the host. The genus Luteouirus, previously known as the luteovirus group (Francki et al., 1991), contains 24 species and 14 tentative species (Randles and Rathjen, 1995), although some viruses in the species list have been regarded by others as being strains of one species (e.g., Waterhouse et al., 1988). Discrimination between different luteoviruses has, in the past, relied mainly on serological differentiation and on biological characters such as the species of vector aphids or the host range of the virus (Casper, 1988). The aphid vectors of different luteoviruses differ; more than 12 aphid genera contain one or more species that are vectors (Rmhow and Duffus, 198l), but individual luteoviruses are often transmitted efficiently by one or a few aphid species (Waterhouse et al., 1988). The breadth of the host ranges of luteoviruses differ markedly. The hosts of beet western yellows virus (BWYV) include plant species from 23 dicotyledonous families, whereas those of barley yellow dwarf virus (BYDV) and soybean dwarf virus (SDV) come from only one family, respectively, Graminae and Leguminosae (Rochow, 1970a; Tamada and Kojima, 1977). However, caution is needed when interpreting reports of host range restriction as these can be a reflection of the host range of the vector aphid rather than, or as well as, that of the virus. Both the variation in host ranges among similar luteoviruses, or strains of one luteovirus, and the variety of aphid species which are efficient vectors would seem to offer useful variation for future molecular studies. Whereas until the mid 1980s it was very difficult to study phloemlimited viruses biochemically because insufficient quantities of virus were available (Rochow and Duffus, 19811, the more recent use of molecular cloning methods has largely overcome this problem. For example, our knowledge of the genome structure of luteoviruses equals that of the genomes of viruses deemed previously to be easy to work

MOLECULAR BIOLOGY OF LUTEOVIRUSES

415

with, such as tymoviruses and tobraviruses. Moreover, in addition to shedding light on relationships among luteoviruses, the use of molecular biological methods has shown that most of the diverse modes of expression of single-stranded RNA (ssRNA) genomes are used during the expression of the 6-kb luteovirus genome. In a previous review, Martin et al. (1990)considered the impact that molecular characterization of luteoviruses has had on thinking about gene expression and the evolution of luteoviruses. The present review is mainly about developments since then in the areas of genome variation, resulting from the accumulation of sequences, the mechanisms by which different genes are expressed [see also a review by Miller et al. (1995)],the determination of the functions of the gene products, mainly by mutagenesis of full-length cDNA copies of genomes, the structure of the particles of luteoviruses, the application of molecular methods in diagnosing luteovirus infection and in obtaining resistance by transformation, and the taxonomy of luteoviruses in the light of increased knowledge of genome structure and gene sequences.

11. GENOMESTRUCTURE A . Arrangement of Open Reading Frames Table I lists the luteoviruses for which some sequence information is available, and Fig. 1 shows the arrangement of open reading frames (ORFs) typical of luteovirus genomes. The genomes are between 5.5 and 6 kb in size, and the coding sequences are in two blocks separated by a region of 100 to 200 nucleotides of noncoding sequence. The more detailed arrangements of the ORFs distinguish two subgroups of luteoviruses (Table I). In the genomes of subgroup I viruses, typified by the PAV strain of BYDV (PAV; Fig. l),the 5'-block of coding sequence consists of 2 ORFs which overlap by 8 t o 13 nucleotides. In contrast, the corresponding block in subgroup I1 virus genomes, typified by potato leafroll virus (PLRV; Fig. l),consists of three ORFs which overlap extensively (Fig. 1).The 5'-most of these ORFs (ORF 0) has no equivalent in subgroup I genomes; ORF 1 and ORF 2 correspond in their encoded polypeptides to ORFs 1 and 2 in subgroup I genomes but overlap by 475 to 628 nucleotides. Despite the different lengths of overlap, ORF 2 of both types of luteovirus genomes are expressed by translational frameshift from ORF 1 (see Section IV,A). The sequences of the proteins encoded by ORF 2, which are thought to have RNA polymerase activity, also distinguish the subgroups; those of subgroup I viruses encode a polymerase with a sequence like that of carnation

TABLE I SOURCES OF GENOME SEQUENCES COMPARED IN TABLE I1

Virus Barley yellow dwarf Strain MAV Strain PAV Strain RPV Bean leaf roll Beet mild yellowing Beet western yellows

Isolate PS1 Victoria P NY

FL1 GB1

Soybean dwarf Pea enation mosaic (RNA-1)

0

6 c

MAV PAV PAV RPV BLRV BMYV

I I I I1 -b

I1

BWYV CABYV GRAV

Cucurbit aphid-borne yellows Groundnut rosette assistor Potato leafroll

Acronym

Genome type

Scottish Dutch

PLRV PLRV

Canadian Australian

PLRV PLRV SDV PEMV

Only values for complete genomes are given, in nucleotides. Not known (-). Size includes nucleotides at 5' end of host origin.

I1 I1 I I1

Genome sizea (nt)

5677 -b

5723 -

Database accession number DO1213 X07653 DO1214 DO1013 X53865

5641 5669

X13062 X13063 X76931

5987= 5882

X14600 YO7496

5883 5882 5861 5705

L24049 LO4573

Reference Ueng et al. (1992) Miller et al. (1988) Ueng et al. (1992) Vincent et al. (1991) Prill et al. (1990) H. Guilley, personal communication Veidt et al. (1988) Veidt et al. (1988) Guilley et al. (1994) K. Scott, personal communication Mayo et al. (1989) Van der Wilk et al. (1989) Keese et al. (1990) Keese et al. (1990) Rathjen et al. (1994) Demler and de Zoeten (1991)

417

MOLECULAR BIOLOGY OF LUTEOVIRUSES Subgroup I (BYDV-PAV)

RNA

3

5’

.

,

ORFs

D ,

m I

P1

1

I

5

-

P3

P6

-

P1 + P2

proteins

3

El

P3 + P5

P4

Subgroup II (PLRV)

RNA ORFs

5

3

rn I

t 1

2

1141

I 1

-

1

5

I

P3

PO

proteins

3

P3 + P5

P1 P1 + P2

P4

FIG.1. Arrangement of the open reading frames (ORFs) in RNA of luteoviruses in subgroup I (e.g., barley yellow dwarf virus, PAV strain; BYDV-PAV) or luteoviruses in subgroup I1 (e.g., potato leafroll virus; PLRV) and the protein products resulting from expression of the ORFs. Boxes represent the ORFs; equivalent ORFs are given the same number. Proteins are represented by horizontal lines, and equivalent proteins are similarly numbered; P1 + P2 and P3 + P5 arise by frameshift or in-frame readthrough, respectively.

mottle carmovirus, whereas the polymerases encoded by subgroup I1 virus genomes resemble that of southern bean mosaic sobemovirus (SBMV) (see Section VII1,B).

418

M. A. MAY0 AND V. ZIEGLERGFUFF

As is the convention, the ORFs of luteovirus genomes have been named numerically commencing with the 5’-most. However, the difference between the genomes of the two subgroups of luteoviruses has meant that different authors have assigned equivalent ORFs different numbers in different luteovirus genomes. Here we follow the system used by Martin et al. (19901, despite the apparent illogicality of naming an ORF as zero, in order to promote consistency. The encoded polypeptides are assigned the same number as the ORF encoding them. Of the three “strains” of BYDV for which nucleotide sequences are known, PAV and the MAV strain (MAV)are in subgroup I (Miller et al., 1988; Ueng et al., 19921, whereas the RPV strain (RPV) is in subgroup I1 (Vincent et al., 1991). Further details of comparisons among the genomes are discussed below. Finally, a small ORF (- 150 nucleotides) is present as the 3’-most gene in the genomes of MAV and PAV, but none has been detected in the genome of the other subgroup I virus SDV, or of subgroup I1 viruses, including RPV.

B . Variation among Coding Sequences Table I1 shows the percentage identities among the amino acid sequences of the proteins encoded by known luteovirus RNA sequences determined by using CLUSTALV (Higgins et al., 1992). Sequences were taken from the GenBank database except those of groundnut rosette assistor virus (GRAV) and beet mild yellowing virus (BMYV), which were personal communications from K. Scott (1994) and H. Guilley (1994), respectively. The P1 and P2 sequences of subgroup I1 viruses are unrelated to the P1 and P2 sequences of subgroup I viruses, and each subgroup is considered separately. PO sequences are the least similar among the proteins of subgroup I1 viruses. In PO, BMYV and cucurbit aphid-borne yellows virus (CABYV)are most alike; both are slightly more similar to BWYV than any of these is to potato leafroll virus (PLRV) or RPV. In P1, CABYV and BWYV are similar, and PLRV and RPV are similar. In P2, CABYV and BWYV are more similar than any other pair. In subgroup I virus genomes, P1 and P2 of MAV and PAV are only 2% different and distinct from P1 and P2 of SDV. In P3, BMYV and BWYV are very similar (90% identical), PAV and MAV are 73% identical, and GRAV and CABYV are 75% identical. Other pairs are 70% (BWYV and GRAV) or less identical. P3 of MAV and PAV are distinct (<47% identical) from P3 of all other viruses. The sequence similarities among luteovirus coat proteins, in particular those of subgroup I1 viruses, are shown in Fig. 2. The regions of sequence similarity explain the ability of polyclonal antisera to indi-

419

MOLECULAR BIOLOGY OF LUTEOVIRUSES TABLE I1 PERCENT IDENTITIES BETWEEN LVTEOVIRUS PROTEINS DEDUCED FROM CLUSTALV ALIGNMENTS BWY

GRA

BMY

CABY

PLR

RPV

25

29 42

18 19 15

22 22 18 27

37

34 29

33 31 37

PAV

33

34 98

61

61 98

65 58 57 55 58 55 57

38 44 42 41 45 41 47 39

39 44 43 47 43 46 46 41 73

38 47 42 47 44 59

52 45 41 44 47 48 42

31 31 26 32 28 29 32 29

28 31 26 32 27 31 33 29 72

65 32 54

61 32 54 64

41 38 44 45 44

38 38 44 44 45 72

Virus

POa

BWY BMY CABY PLR

Pla

BWY CABY PLR

Plb

SD MAV

P2a

BWY CABY PLR

P2b

SD MAV

P3

BLR BWY GRA BMY CABY PLR RPV SD MAV

52

52 70

51 90 66

54 67 75 63

52 62 65 59 61

51 63 59 60 65 63

P4

BLR BWY GRA BMY CABY PLR RPV SD MAV

46

43 50

47 90 47

44 45 58 46

46 49 42 50 48

P5c

BWY CABY PLR RPV SD MAV

32

52 33

65

60 58

SD

MAV

Protein

59 57 58

Subgroup I1 viruses. Subgroup I viruses. c Comparisons among the N-terminal halves of the sequences up to and including a DE dipeptide present in all sequences (see Fig. 4). a

b

420

M. A. MAY0 AND V. ZIEGLERGRAFF

vidual luteoviruses to react with many other luteoviruses (Casper, 1988; Waterhouse et al., 1988). In P4, the similarities between pairs of viruses are like those between P3 although in general less marked. BMYV and BWYV are very similar, but GRAV is less like BWYV or BMYV than it is in P3. P4 of bean leaf roll virus (BLRV) is shorter than that of other luteoviruses because it lacks amino acids equivalent to the N-terminal50 residues (see Fig. 3). Thus, sequence comparisons with BLRV P4 may not be directly comparable to those among other luteoviruses. Alignments of P5 showed that the sequences in the N-terminal half of the proteins were sufficiently similar to be aligned with few gaps, but the sequences to the C-terminal side of a conserved DE dipeptide could not be aligned without many large gaps being added (Fig. 4). To avoid distortion of the results by these gaps, only sequences to the N-terminal side of the DE dipeptide (labeled “proline hinge” and “luteovirus homology” in Fig. 4) were compared. The results were similar to those obtained when P4 sequences were compared except that the P5 sequence of CABYV was distinct from all other sequences including those of PAV and MAV. The results of these sequence comparisons complicate the classification of luteovirus genomes into two subgroups. The distinction becomes increasingly blurred as the proteins being compared are more to the 3’ end of the genomes. The results reinforce those of Rathjen et al. (1994) who pointed out that the genome of SDV was in subgroup I but that the P3, P4, and P5 genes were more similar to those of subgroup I1 viruses than those of subgroup I viruses. Thus SDV was most like RPV in P5, and like BLRV in P3 and P4.

C . Terminal Structures and Noncoding Regions Genomic RNA from PLRV did not bind to oligo(dT)-cellulose(Mayo et al., 1982), and no poly(A) has been detected in luteovirus sequences. RNA extracted from particles of PLRV and RPV was shown to be linked to a protein (virus protein, genome-linked; VPg) with an estimated size of M,7000 for PLRV (Mayo et al., 1982) and 17,000 for RPV (Murphy et al., 1989). There are no reports for VPg being attached to subgroup I virus RNA, and Miller et al. (1995) have suggested that the RNA are capped rather than having a VPg, which would be consistent with their resemblance to RNA from carmovirus-like viruses. The amount of 5’-noncodingsequence ranges between 21 nucleotides for CABYV RNA and 142 nucleotides for SDV RNA (except for the Scottish isolate of PLRV for which an unusually long sequence was reported that proved to be a rare component of RNA populations, most

MOLECULAR BIOLOGY OF LUTEOVIRUSES

42 1

of which were molecules with 5‘ ends like those of other PLRV isolates; see below). The 5’-terminal sequences of RNA of PLRV, BWYV, and CABYV are the same (ACAAAAGA; see Fig. 6). The lengths of the noncoding regions between blocks of coding sequence are 112-113 nucleotides for MAV and PAV, but close to 200 bases for other luteoviruses. Comparisons among these sequences show close similarity (91% identity) between MAV and PAV and slightly more similarity (71% identity) among CABYV, PLRV, and RPV than with, or among, the other viruses (40-63%). Some oligonucleotide sequences are identical in several of the RNAs, and these may have a role in the transcription of the subgenomic mRNA for the coat protein (see Section IV,O. The noncoding sequences at the 3‘ end consist of between 125 (RPV) and 650 nucleotides (SDV). CABYV RNA and BWYV RNA have identical 3’-terminal octanucleotides, but there are no striking similarities in other regions or among RNAs of PLRV, RPV, SDV, and PAV. The termini of subgroup I1 virus RNAs are complementary, 5’-AC . . . with . . . GU-3’ for PLRV, BWYV, and CABYV, or 5‘-CGG . . . with . . . CCG-3’ for RPV.

D . Variation among Strains of Luteoviruses The variations among the amino acid sequences encoded by the RNA of different strains of individual luteoviruses have been found to differ according to which ORFs were being considered. Between the Victoria and P strains of PAV, the sequences of P1, P2, and P3 differed by 1to 3%, whereas those of P4 and P5 differed by 12 and 7%, respectively (Ueng et al., 1992). In contrast, among four strains of PLRV, the genes were equally different (2 to 5%) except for P1 of Australian PLRV which was 11 to 12% different from P1 of the other isolates (Keese et al., 1990). P4 always differed more than P3 both among different strains and among different luteoviruses (Table 11). Because P4 is encoded entirely within the P3 gene, this result suggests that P3 is more constrained in sequence variation than P4, which is unsurprising for a structural protein like the coat protein. Similar conclusions were reached by Veidt et al. (1988) when comparing French and German isolates of BWYV in the P3 and P5 genes. P4 differed more (11 changes) than P3 (8 changes). The P5 sequences differed at 35 positions, and all of these were in the C-terminal part of P5. Comparisons among the P5 sequences of five distinct isolates of PLRV from Scotland also showed this relative conservation in the 5’-half of P5; amino acids differed in one or more of the isolates at 7 positions in the 5’-half but at 19 positions in the 3‘-half (Jolly, 1994).

422

M. A. MAY0 A N D V. ZIEGLERGRAFF

E . Putative Recombination Involving Luteovirus RNA Recombination between blocks of sequence has probably been a major force in the evolution of virus genomes (Gibbs, 1987);the nucleotide sequences of luteovirus genomes have shown that these viruses are prime examples of this process. The discovery that luteoviruses have either of two very different types of polymerase sequences but recognizably similar coat protein sequences strongly suggested that they had evolved by recombination involving ancestors with polymerases like those of carmoviruses (subgroup I) or sobemoviruses (subgroup 11) and one with coat protein and related proteins determining a luteoviruslike mode of transmission (reviewed in Martin et al., 1990). From more detailed sequence comparisons, Gibbs (1995)has argued that subgroup I viruses arose by recombination of the 3’ coding block of an ancestral subgroup I1 virus and the polymerase-coding block of a carmovirus. Rathjen et al. (1994) suggested that the genome of SDV arose from a recombination between viruses belonging to each of the two subgroups. They pointed out that whereas the SDV genome resembles subgroup I genomes in lacking a PO ORF (Fig. 1)and in having a similar sequence for the polymerase (Table 11), the noncoding sequence between the coding blocks is a length (205 nucleotides) characteristic of subgroup I1 virus genomes, and the sequences of the P3, P4, and P5 proteins are more like those encoded by subgroup I1 genomes (BLRV, RPV) than those encoded by subgroup I genomes (Table 11). A different type of recombination was found to occur at low frequency with the genome RNA of three Scottish isolates of PLRV (Mayo and Jolly, 1991). Some RNA molecules were shown to have recombined with a chloroplast mRNA such that the 5’-terminal 21 nucleotides of PLRV RNA were replaced by at least 121 nucleotides of chloroplast RNA. However, the 5’4erminal sequences of most PLRV RNA molecules were similar to that of the Dutch isolate of PLRV (van der Wilk et al., 1989). Some sequence matches were detected between PLRV RNA and the chloroplast mRNA near the site of the putative recombination, and the exact site was only 7 nucleotides from an exon-intron boundary in the transcript RNA from tobacco chloroplast DNA.

111. FUNCTIONS OF GENEPRODUCTS A . Approaches to Determining Gene Function Determination of the functions of the translation products of the different genes encoded by luteoviruses has been approached in several

MOLECULAR BIOLOGY OF LUTEOVIRUSES

423

ways. The biochemical properties of some of the proteins have been analyzed transiently by fusing the polypeptides to a reporter gene such as p-glucuronidase (GUS) to follow their synthesis, and viral genes have been expressed in transgenic plants. Full-length cDNA of PAV and BWYV have been cloned behind bacteriophage RNA polymerase promoters, and transcription of the constructs has yielded RNA transcripts that were infective for protoplasts. The constructs have allowed the effects of site-directed mutagenesis to be assessed for PAV (Young et al., 1991) and BWYV (Veidt et al., 1992; Reutenauer et al., 1993). The study of luteovirus gene products been greatly facilitated by the development of agroinfection technology, first for BWYV (Leiser et al., 1992) and later for PLRV (Commandeur and Martin, 1993). This strategy (see Grimsley, 1990, for a review), which has also been called agroinoculation, uses the Ti plasmid in Agrobacterium tumefaciens to insert cDNA copies of viral genomes into cells such that their transcription can initiate an infection. For luteoviruses this method overcomes the lack of mechanical transmissibility to plants of virus RNA which limits the usefulness of in vitro transcripts of cDNA as sources of infection. For agroinfection experiments with BWYV (Leiser et al., 1992), fulllength cDNA was introduced between the border regions of a binary vector (pBinl9; Bevan, 1984) directly downstream of the cauliflower mosaic virus 35s promoter which is efficiently expressed in most cells, and especially well in vascular tissues (Jefferson et al., 1987). Transcription of this construct resulted in an RNA which had only one nonviral residue at the 5’ end. In the absence of information about the effect of 3’ extensions on transcript activity, three clones were constructed such that transcripts contained (1)a 3’ nonviral extension of 156 nt derived from the nopaline synthase gene cassette plus a 3’ poly(A) tail, (2) an autocatalytic sequence or “ribozyme” just downstream of the viral RNA 3‘ terminus such that the nonviral sequence would be cut off the RNA transcript, or (3) no particular termination sequence downstream of the viral cDNA. In inoculation experiments with progressive dilutions, the three clones were equally infective, as judged by the appearance of viral RNA and proteins 4 weeks after agroinoculation (Leiser et al., 1992, V. Brault, personal communication). About 75% of the plants became infected by inocula containing from 106 to lo9 (the standard amount) bacteria (V. Ziegler-Graff and E. Herrbach, 1994, unpublished results); with inocula containing 105 bacteria, the efficiency dropped to 15 to 35% for each of the clones. These results show that the nonviral extensions present at the 3’ end of the transcripts had no detrimental effect on infectivity. These methods have been used mainly to examine the genes of

424

M. A. MAY0 AND V.ZIEGLERGRAFF

BWYV and BYDV. The results described below for each protein are drawn from this work, and Section II1,H summarizes the results with BWYV, which is the most thoroughly explored system.

B . Studies on PO The PO protein is produced only by subgroup I1 luteoviruses. Thus, its function is either not needed by subgroup I viruses or is performed by one of the other proteins as an additional function. Little significant homology has been found between the PO proteins of the different luteoviruses, except between those of BMYV and CABYV (see Table 11). Thus, little can be deduced as to function from the amino acid sequences of PO. The N-terminal regions are markedly hydrophobic, and that of PLRV has some characteristics of membrane-associated proteins (Mayo et al., 1989). Full-length transcripts of BWYV RNA, in which expression of ORF 0 was prevented by the removal of its initiation codon, were able to multiply in Chenopodium quinoa protoplasts (Veidt et al., 1992).The BWYV RNA synthesized in these protoplasts was unaffected by RNase treatment, presumably because it was encapsidated, which suggests that PO is not required for the assembly of virus particles (Reutenauer et al., 1993).An out-of-frame deletion in the N-terminal part of ORF 0 of a BWYV clone did not alter the efficiency of infection by agroinoculation of, or the symptoms induced in, Nicotiana cleuelandii plants (V. Brault and V. Ziegler-Graff, unpublished results). Moreover, the virus made in these plants was unaltered in aphid transmissibility (E.Herrbach, 1994,personal communication). This indicates that PO of BWYV is not required for infection of tobacco plants, symptom induction, or aphid transmission. Similarly, mutations in PLRV cDNA resulting in the elimination of PO did not affect the ability either of transcripts to infect protoplasts or of cDNA to infect leaf disks following agroinoculation (R. R. Martin, 1994,personal communication). A different approach has been to transform plants with DNA encoding the gene product, although this is an anomalous situation for luteovirus genes as, unlike transgenes which are expressed in virtually all cells in a plant, luteovirus genes are expressed little or not at all in mesophyll cells, which comprise the majority of a plant. Nevertheless, potato plants transformed with cDNA encoding PO of PLRV had disease-like symptoms, which suggests that PO is involved in symptom expression (van der Wilk et al., 1993). The precise function of PO remains unknown; indeed, there have been no reports of detection of this protein in viuo. One speculation has been that it is involved in determining the host range of a particular

MOLECULAR BIOLOGY OF LUTEOVIRUSES

425

virus (Mayo et al.,1989; Veidt et al.,1992). In this role it may act as a n early gene and therefore be ephemeral and difficult to detect.

C . Studies on P1 and P2 Proteins P1 and P2 are considered together because P2 is expressed by ribosomal frameshifting to make a fusion protein which consists of most of P1 joined to P2 (see Section IV,A). The P1 sequences of all subgroup I1 luteoviruses contain motifs characteristic of chymotrypsin-like serine proteases: H(x_,,)[D/El(x,,-,,)T [R/KlxGxSG (where x is any amino acid) (Gorbalenya et al., 19891, although the basic R/K residue is present only in P1 of RPV and BMYV (Koonin and Dolja, 1993; Guilley et al.,1994; H. Guilley, 1994, personal communication). Koonin and Dolja (1993) have classified these P1 among the poliovirus 3C-like proteases. A similar protease motif has also been identified in the equivalent protein (Pl ; Fig. 8) of pea enation mosaic virus (PEMV) RNA-1 (Demler and de Zoeten, 1991) which resembles luteovirus RNAs in the arrangement of ORFs (see Section VII1,C; Fig. 8). No similarities to proteases have been found in the sequences of proteins of subgroup I luteoviruses. The P2 protein contains the putative RNA-dependent RNA polymerase motif near the C terminus: GxxxTxxxN(x,,~,,)GDD (Kamer and Argos, 1984). Habili and Symons (1989) identified possible helicase motifs in P1 and P2 of several luteoviruses, in particular SDV, but others have contested this, questioning even whether plus-stranded RNA viruses necessarily encode a protein with a helicase motif (Gorbalenya and Koonin, 1989; Dolja and Carrington, 1992; Koonin and Dolja, 1993). From the sequence arrangement of the “core” genes of the replication-associated proteins, subgroup I1 luteoviruses have been classified in the picornavirus-like supergroup of viruses (Dolja and Carrington, 1992; Koonin and Dolja, 1993) which have RNAs containing the sequence domains VPg-protease-polymerase. This suggests that the luteovirus VPg is encoded by ORF 1. It was proposed previously that ORF 4 of PAV encodes the VPg (Miller et al.,19881, but other data have shown that this ORF is not required for BWYV replication (Reutenauer et al.,1993) (see Section 111,E).As there is no evidence of a VPg for any subgroup I luteovirus and no protease consensus has been found, it has been speculated that subgroup I virus RNA do not have a VPg (Miller et al.,1995). In protoplast infection experiments with BWYV transcript RNA, frameshift mutations in ORF 1 and 2 were lethal (Reutenauer et al., 19931,which is direct evidence that P1 and P2 are required for replication. Mutations affecting expression of any other protein encoded by

426

M.A. MAY0 AND V. ZIEGLERGRAFF

BWYV did not destroy transcript infectivity, which suggests that ORF 1 and ORF 2 encode the only viral proteins absolutely necessary for replication in individual cells of C. quinoa.

D . Studies on P3 The P3 protein was shown to be the coat protein for several luteoviruses largely by reaction between antibodies specific to virus particles and P3 expressed, usually as a fusion protein, either in uitro (Veidt et al., 1988), in bacteria (Miller et al., 19881, in HeLa cells (Mayo et al., 19891, or in insect cells (J.Lamb, 1994, personal communication). There is extensive sequence similarity among the sequences of luteovirus coat proteins (Fig. 2). It has been possible, either by aligning the sequences of unrelated viruses of known structure (Dolja and Koonin, 1991) or by using prediction programs (see Section V,A), to predict how luteovirus coat proteins are folded to form a particle and thus which sequences are external. Such surface determinants could play a role in interactions between the virus and its aphid vector during transmission. Van den Heuvel et al. (1993) showed that four monoclonal antibodies specific for readily transmissible PLRV particles reacted in immunoblots with P3 and that these antibodies interfere with PLRV transmission when mixed with particles prior to their being fed to aphids. However, other work with PLRV (Jolly and Mayo, 1994) suggests that a discontinuous or conformation-sensitive epitope absent from PLRV particles of a poorly transmissible isolate is located in P5 (see below). The parts of the virus particle involved in aphid transmission presumably form more than one epitope on more than one particle protein. When C. quinoa protoplasts were inoculated with transcripts of BWYV RNA in which translation of ORF 3 was prevented by mutation, the protoplasts became infected. This showed that coat protein is not strictly required for RNA replication (Reutenauer et al., 19931, although less BWYV RNA accumulated in these protoplasts than in protoplasts inoculated with wild-type RNA. This decrease may have been due either to decreased stability of the RNA progeny in the absence of encapsidation or to feedback inhibition of the replication by the nonsequestered viral RNA; alternatively, it may result from an unknown role of the coat protein in stimulating the replication process. Moreover, in some experiments coat protein synthesis was restored, presumably by the selection of a revertant mutant (Reutenauer et al., 19931, which reinforces the hypothesis that coat protein plays a role in BWYV RNA replication. When N . cleuelandii were agroinoculated with a clone containing the same mutation as that in this tran-

427

MOLECULAR BIOLOGY OF LUTEOVIRUSES

DA

,a,

ELRV MVARGKRVV-------VRQLQTRARRRLPVVLA----TAPVRPQRKRRQRGR-NNKPRGG--NGFAR EMYV MNTVVGRRTINGRRR-------PRRQTRRAQRSQPVVVV----QASRTTQRRPR-RRRRGNNRTRR--TVSTR EWYV UNTVVGRRIINGRRR-------PRRQTRRAQRPQPVVVV----QTSRATQRRPR-RRRRGNNRTGR--TVPTR CASYV MNTVAARNDNAGR-------RRRRNDRPARRDRVVVV----NPIGGPPRGRR-ORRNRRRPNRG---GRAR M N T V V V R R P ~ N G R A-- - - - - - N R R R N ~ R A P R R N P V V V -V - - -QTP- - P Q P N S G - ~ R R R R N R R R A N - - R G S R N GRAV PLRV M ~ G N V N G G V Q Q P R M R R - - - - - - - R Q S L R R R A N R V Q P V V M ~ G Q P R R R R R - R ~ R S R R - - - T G V P R RPV MSTVVLRSNGNGSRRRR-------QRVARRRPAARTQPVVV----VASNGPARRGR-RRRPVGPRRGR--TPRSG MVAVSNVAI-------QRRRSRRAARRAPRVQLMAVPTATSRPQRRGRQRRR-RRNNRGG--SFVSG SDV MNSVCRRNN-------RRRNGPRRARRVSAVRR----MVVVQPNRAGPKRRA-RRRTRGGGANLISG MAV MNSVGRRGPRRANQNGTRRRRRRTVRPVVVVQP----NRAGPRRRNGRRKGR-G----GA--NPVFR PAV

53 59 59 5S 58

64 61 57

55 56

*

De ELRV BUY V EWYV CASYV GRAV PLRV RPV SDV MAV PAV

l

4%

i

125 131 GB~GSSETFVFSKDN~SSSGAITFGPSLSDCPAFSNGMLKAYHEYKISMVILEFVSEASSQMSGSIAYELD 131 R G S P G E T F V F S K D N L T G S S T G S I T F G P S L S E S P A F S S G I L K A Y H E Y K I I M V ~ L E F I S E A S S T S S G S I S Y E L D 128 RGGSGETFVFSKDNLTGSSNGSITFGPSLSDCPAFSSGILKAYHEYKISMVKVEFISEAASTSSGSIAYELD 130 G R G S S E T ~ K D N L V G N T ~ T F G P S L S D C P A F K D G I L K A Y H E Y ~ I L U D F V S E A136 ~S GGSRGETFVFSKDSLAGNSSGSITFGPSLSEYPAFQNGVLKAYHEYKITNCVLQFVSEASSTAAGSISYELD 133 GSGKAHTFVFSKDGINGSSKGSITFGPSLSECKPFSDGILKAYHEYKITSVLLQFITEASSTSSGSIAYELD 129 P A G R T E V F V F S V N D L K A N S S G T I K F G P D L S Q C P A L S G G I L K S Y H L Y K I T N V K V E F K S H A S A S T V G A M F I E L D 127 P T G G T E V F V F S V D N L K A N S S G A I K F G P S L S Q C P A L S D G I L K S Y H R Y K I T S I R V E F K S H A S A N T A G A I F I E L D 128

.

. .

.. . *..

0 . 0 .

' 0 .

' 0 .

....,.. * .

..

l a

OF,

ELRV EMYV EWYV CAEYV GRAV PLRV RPV SDV MAV PAV

DD

a

RSSQVHEFVFSKDNLNGNSKGSITFGPSLSECKPLADGILKAYHEYNITNVELAYITEASSTSSGSIAYELD GTGSSETFVFSKDNLAGSSSGAITFGPSLSDCPRFADGMLKAYHEYKFSMVILEFVCEASSQNSGSIAYELD

~

. . ..*..

*..

DH

DG~-,

...

DI

PHLKNTTIDSKINKFSITKSEKKKFSRKAINGDAWHDTSEDOFRILYEGNGDAK-IAGSFRVTIKVLTONPK 1% P H C K L S A L ~ S T I N K F G I T K P G R R A F A A S YI N G ~ D W H D V A K D Q F R I L Y K G N G S S S - IAGSFRITMKCQF~NPK 202 P H C K L ~ S ~ I T K P G K R A F T A S Y I N G T E W H D V A E D Q F R I L Y K G N G S S S - I A G S F R I T ~ K C Q F H N P K 202 PHCKLSSLQSTINKFGITKSGLRRWTAKQINGMEWHDATEDQFKILYKGNGSSS-VAGSFRITIKCQVQNPK 199 201 PHCKSSSLQSYVNKFGITRGGARSWMGRYINGVEWHDATEDQFRFLYKGNGSSA-IAGSFRFTIKCQVQNPK P H C K V S S L Q S Y V N K F Q I T K G G A K T Y Q A R M I ~ H ~ Q C R I L W K G N G K S S D S A G S F R ~ V A L Q N P K208 PHCKASSLASTINKFTITKTGARSFPAKMINGLEWHPSDEDQFRILYKGNGASS-VAGSFKITLRVQLQNPK 204 PHCKYSEIQSLLNKFSITKSGSKRFPTRAINGLEWHDTSEOQFKIHYKGNGESK-IAGSFKISINVLTQNAK 200 TWCSQSTLGSYINSFTISKSATKTFTAQQIDGKEFRESTVNQFYMLYKANGSTSDTAGQFIITIRVANMTPK 153 TACKQSALGSYINSFTISKTASKTFRSEAINGKEFQESTIDQFWMLYKANGTTTDTAGQFIITMSVSLMTAK 200

.

' : *..

.

0 . 0

.

*

*:

.

*'

..

FIG.2. Alignment of luteovirus coat proteins. Sequences were aligned using MaxHom/HSSP (Rost and Sander, 1993,1994).An asterisk (*I indicates a position at which the amino acids are identical in all sequences, and a dot (J indicates a position at which the amino acids are chemically similar. Regions predicted to have a-helical or P-sheet structure are labeled; lettering of the P sheets is by analogy with other virus coat proteins (Rossmann and Johnson, 1989).Regions probably exposed on the surface of the virus particle are indicated by braces. Shaded amino acids are those thought to be, or contribute to, epitopes.

script, no viral infection could be detected (V. Brault and V. ZieglerGraff, 1994,unpublished results), which strongly suggests that coat protein is absolutely required for infection of whole plants.

E . Studies on P4 The P4 proteins of different luteoviruses are relatively similar (Table 11),except for that of BLRV which lacks amino acid sequence equivalent to the N-terminal approximately 50 residues of the P4 molecules

STELD

428

M.A. MAY0 AND V.ZIEGLERGRAFF

of other luteoviruses (Fig. 3). The alignment shows a sharp distinction between an N-terminal domain with three a-helical regions and a net negative charge and a C-terminal domain that contains some p-sheet regions and has a net positive charge. This polarity of structure and charge was noted first by Tacke et al. (1991) for PLRV P4 and is found in P4 of all luteoviruses. PLRV P4 obtained by expression in a bacterial vector system was shown to bind to single-stranded nucleic acid (Tacke et al., 1991), and by testing deleted forms of the protein, the binding domain was located in the basic C-proximal part of the protein. By mutational analysis, Tacke et al. (1993a) showed that the helix in PLRV P4 corresponding to helix B (Fig. 3) is amphipathic, with negatively charged amino acids located predominantly on one side of the helix, and that it mediates protein-protein interaction, which could

, ELRV EMYV EWYV CABYV CRAV MAV PAV PLRV RPV SDV

ELRV SHY v BWYV CAEYV GRAV MAV PAV PLRV RPV SDV

I

aB

l

aC

MDLPEDQARFTNSYS MEEDOH---VGKHDALSALSQWLWSKPLGQHNADLDDDEEATTGQEELFLPEEQVRAR~SFS MEEDDH---AGKHDALSALSQWLWSKPLGQHNADLDDDEEVTTGQEELFLPEEQVRARHLFS MQGGEGEE----ISALRGATAWLWSTPLGDHRAEDDNEETADALIEEAEL-EEEAQAKHLYF MDE--LTGAV--IGGLQGATQWLWSKPLGNQTAEDDDDETVDALIEEAEI-EE-GLAKHLYF MAQGEQGALAQFGEWLWSNPIEPDQNDELVDAQEE--EGQILYLDQQAGLRYSYS MAQ-EGGAVEQFGQWLWSNPIEQDPDDEMVDAREE--EGQILYLDQQAGLRY5YS MSMVVYNNQECEEGNPFAGALTEFSQWLWSRPLGNPGAED-AEEEAIAAQEELEFPEDEAQARHSCL MAMVRADADR--E-SLGEGLLQERSQWLWSLPTAQPGAED-ADDQLVLGEEELQDLEEEAVARHSFS MSQYNDDAL---VGQQDALQEFSSWLFQRPPADHNAEDDNDDEGEIIEEEALFPEDQARLTHSCF

..

A

t-k

. *

Chpe

-/+ 3/1 15/1 15/4 16/3 14/2 11/1 11/2 15/2 16/4 17/2

58

2/16 3/8 2/9 2/7 3/8 1/6 1/7 2/9 2/9 2/8

82 126 126 124 123 120 119 133 130 129

15 59 59 57 53 52 66 63 62

I

l-4d-i

LRTTSMETPREVSRSGRLYQSASRSQMAYSRPTMSIISRTSSWRTSPRPLPPPQVPSLMNSILTSRT QKTISREVPAEQSRSGRVYQTAROSLMECSRPTMSINSQWSFWSSSARPLPKIPVPSLTSWTHTVNS QKTISREVPAEQSRSGRVYQTARHSLMECSRPTMSIKSQWSFWSSSPKPLPKIPVPSLTSWTHTVNS QRTISRAVPQEVSPSGRLFQRAQHSALEYSRPTMNIRSSWSSWSSSPRPLPPPPVPSLMSWTPTASL QKTTSREVPMEVSRSGRLFQTAQHSVLEYSRPTMNIKSQWLRW5SSPRPLPPLQGRSLTSLIPTANP QSTTLRPTPQGQSSSVPTFRNAQRFQVEYSSPTTFTRSQTSRLSLSHTRPPLQSAQCLLNSTLGAHN QSTTLKPTPPGQSNSAPVYRNAQRFQTEYLSPTTVTRSQVSVLSLSHTRPQLRPALSLLNSTPRANN QRTTSWATPKEVSPSGRVYQTVRHSRMEYSRPTMSIRSQASYFSSSARPLPPPPVPSLMSWTPIAKY QRIHSRATPLEVSPSGRLYQSIRHSRMEYSRPTMNIRSQIVSYSSSARPLPQQPAPSLTSWTPIAKH QRTASHVVPREVSLSGRLYQNASHSLMEYSRPTMNI~SRVSYYS~SPRPLPPRQVPSLMNLTHTAST I

ELRV EMYV EWYV CASYV GRAV MAV PA V PLRV RPV SDV

aA

..

QQSSPKLTNSASPNLRRKSSLGRLSMDRHGTTLQRTNSGFSTKETEMPRLLDRSESLSRY VPFHOPLTSSGSONPAGGHLDRLTSTGRTGMTLPRTNSGSSTKAMVLHR TPFPQLSTSSGSQSPGKGRLQRLTSTERNGTTLPRTNSGSSTKAMVLHR APSNPRLINLESPRVDCDVGPLSRSTGWNGMMQPKTSSRSSIKGMDLPRLRAASESPSSARSRTRNR QVFSPTSINSGSHGHDKEAGWVATSMGLNGTMRRKTNSGSFTRVMDPAQSLVPSGSPSSAKSKTPNR QPWVATLTHSPSQN-QQPKPSPPNRLTGRNSGRVR QPWVATLIPSQSAG-PPQRSSEPKRLTGRNSRNQR HPSSPTSTSSKLRRAAPKLIKRG LHSHQQSISSQS-PKLVRGASQRR PKFNRYSINSVSQRAVRNVSQPELSMASNGMIPVRINSRSTIKGTESPRSQAPSRSRSMS

5/11 -/5

1/6 5/12 3/8 -15

1/5

-/6

-/4 2/16

FIG.3. Alignment of luteovirus P4 proteins. Sequences were aligned using MaxHom/HSSP (Rost and Sander, 1993, 1994).An asterisk (*) indicates a position at which the amino acids are identical in all sequences. Regions predicted to have a-helical or p-sheet structure are labeled. Charge indicates the total negative (D+ E)and total positive (K + R) charges in the sequences on each line of the diagram.

142 175 175 191 190 154 153 i56 154 189

MOLECULAR BIOLOGY OF LUTEOVIRUSES

429

explain the tendency of P4 to dimerize. Indeed, PLRV P4 has been detected in infected potato plants and in transgenic potato plants expressing PLRV ORF 4, mainly as a homodimer (Tacke et al., 1993a). In these plants, P4 was predominantly associated with membraneenriched fractions. P4 has also been found in viuo in a phosphorylated form. The biochemical properties of P4 have led Tacke et al. (1993a) to suggest that it may act as a phloem-specific movement protein. BWYV P4 was detected in infected protoplasts (Reutenauer et al., 1993) and plants (Reutenauer, 1994).However, BWYV RNA replicated and was encapsidated in protoplasts infected with transcripts in which expression of ORF 4 was prevented by a modification of the initiation codon (Reutenauer et al., 1993),showing that P4 plays no essential role in single cell infection. Also, a clone which contained a triple mutation that prevented expression of ORF 4 (but not ORF 3) was infective in agroinoculated N. clevelandii plants, although the infection was delayed compared to that in control plants (V. Ziegler-Graff, 1994, unpublished observations).

F. Studies on P5 Open reading frame 5 is expressed as a fusion protein with the coat protein by translational readthrough of the termination codon of ORF 3 (see Section IV,D). The full-length readthrough proteins of PLRV (Bahner et al., 1990) and BWYV (Reutenauer et al., 1993) have been detected in infected protoplasts and plants by reaction with antibodies specific for P3 or P5. A shortened form of this polypeptide (M,53,000) is present in purified viral particles (Bahner et al., 1990; Reutenauer, 1994); readthrough protein has also been detected in preparations of particles of PAV (Waterhouse et al., 1989) and RPV (Vincent et al., 1991) (see Section V,C). RNA from BWYV containing various frameshift mutations or deletions in ORF 5 was replicated as efficiently as was control RNA in C. quinoa protoplasts. Neither the readthrough protein nor cis-acting sequences present in ORF 5 were required for the infection of single cells (Reutenauer et al., 1993). Miller et al. (1995) reported that PAV P5 was not essential for replication in oat protoplasts, although with Triticurn monococcum protoplasts Young et al. (1991) found that P5 was needed for infectivity. Frameshift or deletion mutants which resulted in the 3’ half of ORF 5 being out-of-frame or deleted did not alter the infectivity of PLRV cDNA for protoplasts or leaf disks (R. R. Martin, 1994, personal communication). In experiments with a variety of BWYV mutants, Reutenauer et al. (1993) showed that P5 was not necessary for the assembly of virus par-

430

M. A. MAY0 AND V. ZIEGLERGRAFF

ticles. Frameshift or deletion mutants in ORF 5 of BWYV were infective when agroinoculated into N. cleuelandii, although the plants accumulated less virus than did plants agroinoculated with the wild-type virus (V. Brault, 1994,personal communication). No symptoms appeared in these plans unless an in uiuo mutation took place, to restore the production of a full-length readthrough protein. These results show that BWYV P5 is not required to initiate infection but is involved in some way in the further development of the infection and also in symptom induction. The results of in situ hybridization experiments on leaves and petiole sections of plants infected with P5 mutants indicated that phloem localization of the mutant virus was the same as that of wild-type virus (V. Ziegler-Graff, 1994,unpublished results). There is considerable homology among the readthrough domain sequences of all luteoviruses (Fig. 4)(see Section 11,B).The region immediately downstream of the coat protein stop codon is very rich in proline residues. This proline hinge may serve as a loose tether joining the coat protein (presumably anchored in the virion capsid) and the rest of the readthrough domain (Bahner et al., 1990;Guilley et al., 1994).It is followed by a region in which there in considerable similarity among all luteoviruses (Table 11) and a C-terminal region in which the sequences are more diverse. However, in this region there is stretch of about 45 residues in which P5 of BWYV, CABYV, and PLRV are similar to one another and distinct from other P5 (Fig. 4). Guilley et al. (1994)have proposed that this region is involved in the specificity of vector transmission, as the three viruses concerned are transmitted by Myzus persicae. Ultrastructural studies of cereal grain aphids transmitting different isolates of BYDV have provided a paradigm for the behavior of virus particles in the aphid vector during transmission (see Gildow, 1991,for a review); particles cross the hindgut epithelium, enter the hemolymph, and then cross the accessory salivary glands to enter the saliva. As no replication occurs in the insect body (Waterhouse et al., 19881, virus particles remain intact, and determinants of the specificity and efficiency of transmission must be part of the P3 and/or P5 sequences, presumably on the surfaces of the particles. Luteovirus-vector speciFIG.4. Alignment of luteovirus P5 proteins. Sequences were aligned using MaxHom/HSSP (Rost and Sander, 1993, 1994). An asterisk (*) indicates a position at which the amino acids are identical in all sequences. Proline hinge indicates a very proline-rich sequence, luteovirus homology indicates the region in which all luteovirus P5 show marked similarities, and BWWICABYVIPLRVhomology indicates the region identified by Guilley et al. (1994) in which many amino acids are common to BWW, CABYV, and PLRV (marked as +).

ENYV CABYV PLRV RPV SOV UAV PAV

VOE.....EPGPSPGPSP........SPQPTP Q.KK FIVYTGVPVTRIMAQSTOOAISLYOMP-SQRFRYIEOENMNNTNLOSRNYSQNSLKAIPUIIVPVPQGEN 93 VOGSSP....... PPPSPSPTPPPPPPPQPQPQPCA FNGYEGNPQNKILTAENSRNIOSRPLNFVQYYKI-EOEKNOKVNLQAGYSRNORRCMETYLTIPAOKGKF 100 93 VOS.............GSEPSPSPQPTPTPTP Q.KH FIAYVGIPMLTIQARENOOQIILGSLG-SQRMKYIEOENQNYTKFSSEYYSQSSMQAVPMYYFNVPKGQN VOA.....EPGPSPGPSPDPPPPPSPSPEPAP A.KE F I V Y S G V A H T I I S A Q S T D O S I I V R D I P - O Q R F R Y V E N E N F Y N F Q I A A Q N Y S N T N T K A V P M F V F P V P I G E N 101 99 VOG.....EPGPKPG..POPAPQPTPTPKPTPA.KHERFIAYTGTLSTLISARQSSOSISLYSI R.NQRIRYIEOENSSNTNIOAKNYSQNSVEAIPMFVYPVPEGTN 99 VOSST........PEPSPQPQPEPKPOPQPTPEPRQKRFFEYVGTPYVVIQTRESSOSIAVKAM N.OQSFQYIENETSEQRTVKANNNSNNSVQAQAAFIFP1PAGEY 107 VOSSTSEPQPAPEPTPTPQPTPAPQPAPEPTPAPVPKRFFEYIGTPTGTISTRENSOSISVSKLG-GQSMQYIENEKCETKVIOSFNSTNNNVSAQAAFVYPVPEGSY

... . . . . .

t.

..

Luteovirus homology

SVOISCEGYQPTSSTSOPNRGRSOGMIAYSNAOSOYNNVGEAOGVKISKLRNONTYRQGHP---ELEINSCHFREGQLLEROATISFHVEAPT-OGRFFLVGPAIQKT

198 204 197

SVEISTEGYQATSSTTDPNKGRIOGLIAYDN-SSEGNNIGAGSNVTITNNKAONSNKYGHPO---LEINSCHFNQNQVLEKOGIISFHVKATEKEANFFLVAPPVQKT

205

ENYV CAEYV PLRV RPV SDV YAV PAV

TVEISMEGYQPTSSTTOPNKOKQOGLIAYNOOLSEGNNNGIYNNVEITNNKAONTLKYGHPO---MELNGCHFNQCQCLEROGOLTCHIKTTGONASFFVVGPAVQKQ HVYLEADGEFVVKHIGOELDGSNLGNIAYOVSQRG-NNVGNYKGCKITNYQSNTVFVAGHPOA---TMNGKSFOTARAVEVONFASFELECOOEEGSNAIYPPPIQKO

SVNISCEGFQSVOHIGGNEDGYNIGtIAYSNSSGONNGVGNYKGCSFKNFLATNTWRPGHK---OLKLNOCQFTOGQIVEROAVMSFHVEATGTOACFYLUAPKTMKT

212

EWYV CAEYV PLRV RPV

SKYNYVVSYGANTORMMEYGMIAIALOEQGSSGSVK--TERPKRVGHSMAVSTNETIKLPEKGNSEG---YETSQRQOSKTPPTASGGSOTLOVEEGGLP-LPVEEEI SSYNYTVSYGNYTEKYCENGAISVSIOEONNGNEP----RRIPRRGVMANSTPEPSFSGOOSQRQOFNTPSLEERCSOALESEEKKEEONLLOLEEENIPDVOOOOLN AKVNYTISYGDNTDRDMELGLITVVLDEHLEGTGSANRVRRPPREGHTYMASPHEPEGKPVGNKPROETPIQTQERQPOQTP--SOOVSOAGSVNSGG----PTESLR SKYNYAVSYGAWTDRDMEFGlITVlLDEKRGSGSPT...RKSLRAGHAGVTTTTOLVALPEMENS.G...IETSE..TPSAPVTSSKA...........P.LPT... V AKYNFCVSYGDNTDRDMEFGMVSWVLOEHLEGARSSQYVRKSPRPGHFGVNRSHRL..................Q..OSFTPV......................... D K Y N Y V V S Y G G Y T D K R M E F G T I S V T V O ...S O V E A E R Y S R ........................ HTSTVRRTENRDYGNUNVLPPYNPOQVPEQEOEQP.VV0KE M. D K Y N Y V V S Y G G V T N K R M E F G T I S V T C O ...S O V E A E R I T R ........................ H A E T P I R .....F K H I L V S E Q Y E . Q P L P 1 .......I I O Q C L C

299 289 270 283 280

SDV

YAV PAV

SIEISCEGYQAASSTSOPHRGKCOGMIAYOOOSSKVNNVGQQNNVTIINNKAONONKYG~POPLOLMINGORFOQNQVVEKOGIISFHLVTTGPNASFFLVAPAVKKT 207 S V N I S C E G L Q S V D H I G G N R D G Y N I G ~ I A Y Q S Q S G O Y N G V G N Y V G C O I T N L L G T N T W R P G H E - - - O L E L N S C K F T O G Q I V E R O A V I S F H V K A R G A O P K F Y L U A P K 204 TMKA

...

..

..

PLRV RPV SDV MAV PAY

BWYV CAEYV PLRV RPV SDV YAV PAV

. .

. . . . . . . . ..

+ ENYV CAEYV

. . ..

+ + + *BWW/C,4B~P~?~+L,R,V,IO*mo!og+y* + +

+

300 308

!

PDFVGONPNSDLSTKNSQEEEAMSLESGLRPQLKPPGLPKPQPIRTIRNFOPTPOLVEANRPO-VNPGYSKAOVAAATIIAGGSIKOGRSYIOKRNKAVLOGRKSN-KGISRASEAGTAEDORASTSSRL-----RGNLKPKGLPKGLPKPQPTRTITEFNPGPDLIEVNRPO-LAPGYSKAOVAAATVLAGGSVHEGRDMLERREAKVMOSRKKNGI LEFGVNSDSTYOATVOGTOWPRI------PPPRHPPEPRVSGNSRTVTDFSSKADLLENNDAEHFOPGYSKEOVAAATIIAHGSIQOGRSYLEKREENVKNKTSSNKP SOSESEDDPLSAAPDVGFGGTRLLIOTOIKTI--------POPOVADAFVNSAHVGVOPN-AE-VRA-FKRAQRPP---------------------------RGP--EYVSDDDSSSSSS..--..-.-.I V S N R p S T ........P O N O S O I Q F A N S t K G K L p S ---..-....-Q T K L p p. .......................... KC.-DAGSPIDTASLTSOTEAEKAF---OLKEEELTRAILEYEAATVSIPOAAPOILP---SKSEMSSKPIO-----ROGRSLPKSQTKEVLGTYQGQNI OVQTPEQEQTLVDEEOKQTV-------STEPDIALYEYEAATAEIPOAEEDVLP---SKEQLSSKPVO-----TSGNKIPKPKEPEVLGTYQGQNI

____________ ____________

405 409 401 357 318 368 361

. _ . _ _ . _ _ _ _ _ _G_ SSLA _ S S L_ T G G_ T L K A_ S A K_ S E K_ --LA_ K L T T_ S E R_ A R Y E_ R I K_ R Q Q_ G S T R_ A S E_ F L E -_ - - - - - -_ SLL_ A G E D_ P D S_ RF __ 4 6_ 7 LSSTSSLTSGALKKLSAQSEK----------------------LATLTTGERVQYQRLKNSYGSTVAAEYLE--------KVLADKTS 467 PSLKAVSPAIAKLRSIRKSQPLEGGTLNKDATOGVSSIGSGSLTGGTLKRKATIEERLLQTLTTEQRLNYENFKKTNPPAATQNLFEYQPPPQVORNIAEKPFQGR 507

_______________

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SSVASSSLSGGSLRG----------SLRPKTEDPKDSSKSKSRKNSLGSL--------------------R ............................................... F Q s ..........R L S A R E K E E I s K s K p S N V E R Q V G p L V O .......Ay------Gyps TSDDVPPVIAEKLREVNRAPS--TLLYORQPKQPKNPLTRFVLSNKTSTASPGSQSSTAG-YTREQASEYTRIRKSLGLTAAK---QYKASLODT YPEOVPPIARQKLREAAKTPS--TLLYEKTPKKSNNFLTRFVEANRSPTTPAAPTVSTVSNYTREQLAEYTRIRKSLGLTAAK---EYKAQFQ

398 354 457 449

M. A. MAY0 AND V. ZIEGLER-GRAFF

432

ficity occurs at two levels (Gildow, 1987; 1993: Gildow and Gray, 1993): a recognition that regulates uptake a t the hindgut with little selectivity among luteoviruses, followed by a high-specificity uptake at the salivary gland basal plasmalemma. The surface location of P5 and its ready proteolytic cleavage (see Section V,C) led Bahner et al. (1990) and Guilley et al. (1994) to suggest that P5 plays a role in these recognition events. It is possible that the conserved luteovirus homology domain contains signals involved in the movement across the hindgut epithelium, whereas additional signal sequences, like the M . persicaespecific motif, contribute the high vector specificity required for the movement into the accessory salivary gland cells. These are two lines of evidence concerning the role of P5 in transmission. In the first, the amino acid sequences of highly transmissible and poorly aphid transmissible Scottish isolates of PLRV were compared (Jolly and Mayo, 1994). No consistent amino acid changes were detected between the two types of isolates in either P3 or the luteovirus homology domain of P5. However, two changes were detected in the C-terminal part of P5 that could be responsible for the differences in transmission efficiency. More direct evidence for the role of the P5 protein in vector transmission comes from experiments with frameshift or deletion mutants in ORF 5 of BWYV. Aphids reared on protoplast suspensions or plants infected with these mutants were unable to transmit the progeny (V. Brault, 1994, personal communication).

G . Studies on P6 The genomes of PAV and MAV possess a small ORF located in the 3' end of the genome that encodes proteins of M,. 4300 to 6700 (Miller et al., 1988; Ueng et al., 1992). However, the long 3' terminal sequence of SDV does not contain any coding sequence (Rathjen et al., 1994). That P6 is functional was suggested by the detection of a n abundant subgenomic RNA corresponding to ORF 6 (Miller et aZ., 1988; Ueng et al., 1992; Kelly et al., 1993). Also, a frameshift mutation within this region abolished the infectivity of BYDV-PAV transcripts, suggesting that it plays a role in virus replication (Young et al., 1991).

H . Summary

of

Roles of BWYV Proteins Deduced by Mutagenesis

The most complete picture of the roles of the different luteovirus proteins obtained so far by mutagenesis is for BWYV. As a model for the roles of the gene products, at least of subgroup I1 luteoviruses, Table I11 summarizes the effects of eliminating each of the ORFs on several virus properties. This simple compilation takes no account of

433

MOLECULAR BIOLOGY OF LUTEOVIRUSES TABLE I11 RDLES OF BEETWESTERN YELLOWS VIRUSPROTEINS DEDUCED FROM MUTACENESIS EXPERIMENTS~ Protein Function

PO

P1

P2

P3

P4

P5

Replicationb Particle assembly* Plant infection Symptom formation Transmission

N N N N N

E n.d n.d n.d n.d

E n.d n.d n.d n.d

S E E n.d n.d

N N S S n.d

N N S S E

a N, No effect if deleted; E, Essential for this function; S, Some effect if deleted; n.d., Effect of deletion not determined. b Replication and particle assembly were determined in protoplasts.

possible epistatic effects or of effects mediated by noncoding regions of the genome. IV. MECHANISMS OF GENEEXPRESSION Luteoviruses are remarkable in that the expression of their genomes, which are less than 6 kb in length, involves most of the ways in which genes in RNA virus genomes are expressed. These are (1) translational frameshift between the overlapping ORFs, (2) leaky scanning by ribosomes to translate ORFs downstream of the first AUG in the mRNA, (3) production of subgenomic mRNA to express downstream ORFs, (4) readthrough of termination codons to express downstream ORFs as fusion proteins with the upstream ORF product, and ( 5 ) proteolysis of a precursor protein to produce several proteins from one ORF. The gene products produced by expression of the genomes of luteoviruses are shown diagrammatically in Fig. 1.

A . Translational Frameshifting Frameshift, either as - 1 or + 1, occurs during translation of a variety of RNA species from widely different organisms (Atkins et al., 1990). Relatively few plant viruses have genomes that are expressed by using frameshift, but these include luteoviruses from each subgroup (Brault and Miller, 1992; Prufer et al., 1992) as well as PEMV (Demler et al., 1993) and dianthoviruses (Xiong et al., 1993). In all cases, frameshift results in the translation of the gene for RNAdependent RNA polymerase.

434

M. A. MAY0 AND V. ZIEGLERGRAFF

In luteovirus genomes, ORF 2 appears to be expressed only by -1 frameshift from ORF 1 to produce a fusion protein, P1 + P2. Frameshift was shown to occur in uiuo by inserting the sequence thought to contain the frameshift signal in RNA of PAV or PLRV into the GUS gene (Brault and Miller, 1992; Prufer et al., 1992) and observing transient expression of GUS activity in electroporated protoplasts. For both PAV (Brault and Miller, 1992) and PLRV (Prufer et al., 1992), GUS activity, of around 1%of that observed when the frame was continuous from ORF 1 into the reporter gene, was expressed when a -1 frameshift occurred in the translational reading frame. These experiments showed for the first time that plant viruses, like viruses of vertebrates (Jacks and Varmus, 1985; Moore et al., 1987; Brierley et al., 1987; Bredenbeek et al., 1990; Clare et al., 1988) and yeast (Dinman and Wickner, 19921, use this unusual translational event to produce a protein presumably needed only in small amounts. Frameshift has been shown to occur at a “shifty” heptanucleotide or “slippery site” that, in most cases, is followed by a highly structured sequence thought to cause ribosomes to pause and tRNA to slip into the -1 frame (Jacks et al., 1988a). In the case of PAV, the overlapping sequence is only 13 nucleotides long and contains a GGGUUUU shifty sequence followed immediately by a UAG stop codon. Brault and Miller (1992) showed that a stop codon at the 3‘ end of ORF l is essential for frameshifting; its replacement by UAA or UGA did not affect the frameshift rate (V. Brault, 1994, personal communication). The minimum sequence required for frameshifting in uiuo in PAV RNA is from nucleotides 1132 to 1246 and contains, downstream of the shifty heptanucleotide, a sequence that can form either a double pseudoknot structure called “kissing stem-loops” or a long single stem-loop (Brault and Miller, 1992). Deletions from the 3’ end of this sequence that disrupted the putative secondary structure abolished frameshifting (V. Brault, 1994, personal communication). Secondary structures like these could also be drawn for SDV RNA and PEMV RNA-2 (Miller et al., 1995). The position of the frameshift in PAV RNA was precisely mapped by Di et al. (1993), who engineered the initiation codon of a frameshift fusion protein close to the shifty sequence and, after translation, determined the terminal amino acid sequence of the protein made. The site was located at the codon immediately preceding the UAG or one codon upstream. The ambiguity was because a glycinephenylalanine dipeptide is encoded in both the 0 and - 1reading frames. Sequences involved in ribosomal frameshift in RNA of PLRV and BWYV have been further analyzed by using in uitro transcriptiodtranslation experiments, coupled to site-directed mutagenesis. In contrast to the RNA of subgroup I luteoviruses, the overlapping region

MOLECULAR BIOLOGY OF LUTEOVIRUSES

435

is more extended in subgroup I1 virus RNA (Fig. 1).Prufer et al. (1992) located the domain required for frameshift in PLRV RNA between 458 and 376 nucleotides upstream of the end of ORF 1. Frameshifting efficiency was comparable in this heterologous translation system to that observed in the in uiuo experiments. The shifty sequence was found to be UUUAAAU (Priifer et al., 1992; Kujawa et al., 1993). This sequence represents a new class of frameshift signals ending in AAAU rather than the previously established UUUA (Jacks et al., 1988a,b), AAAC (Hizi et al., 1987), or UUUU (Hatfield and Oroszlan, 1990). By following the effects of mutations that destabilized or restored the stem, Prufer et al. (1992) identified a stem-loop structure downstream the shifty heptanucleotide that strongly influenced the efficiency of frameshifting. However, although using the same in uitro approach and the same translation system (rabbit reticulocyte lysate), Kujawa et al. (1993) concluded that a pseudoknot was required for translational frameshifting. Deletion of the 3’ end of an alternative stem-loop structure (as proposed by Prufer et al., 1992) increased the efficiency of frameshifting by up to 134% rather than causing a decrease. The involvement of a similar pseudoknot structure in BWYV RNA is suggested by the results of experiments in uitro in rabbit reticulocyte lysate o r wheat germ extracts, using a sequence of 50 nucleotides including the heptanucleotide GGGAAAC (Garcia et al., 1993) and sequence that can be folded into a pseudoknot structure similarly to the corresponding sequence in PLRV RNA (Garcia et al., 1993).Similar pseudoknot structures can be made by folding the corresponding overlapping regions of RPV and CABYV RNAs, downstream of the same shifty heptanucleotide GGGAAAC. In the case of BWYV, frameshifting was detected at a frequency of about 1%.Changing the GGGAAAC sequence to UUUAAAC increased the frameshift rate by a factor of 23. Both these signals were poorly recognized in uiuo in E. coli, which suggests that eukaryotic frameshift signals are not recognized by prokaryotes. Evidence that frameshifting is more complex than previously supposed comes from work with BWYV RNA (Veidt, 1991) in which cDNA of the complete overlapping region was flanked by two reporter genes and then transcribed and translated in wheat germ extracts. Both substitution and deletion mutants were made in the 300 nucleotides surrounding the previously proposed shifty site GGGAAAC. Frameshifting was observed both when the shifty sequence was modified to a nonslippery sequence GGAAATT and when the pseudoknot was destabilized (Fig. 5). Also, although removal of the GGGAAAC sequence as part of a BglII fragment did eliminate frameshift, removal in a Sac11 fragment did not (Fig. 5). Thus, although frameshift does occur at the

M.A. MAY0 AND V.ZIEGLERGRAFF

436

Sac

II

Bgl !I

I 1413

"8 0

'

A -__-_- -

sne

GGGAAAC

I1

/

8 8

Bgl I1

I

17U

- --

--. -.-.-. >

/

I I I I

I

I I I

C

I I I

I I I I

Mutation

Frameshift

I I

I

G-C A 4-C-G 4-G-C C-G GGGAAACGGAGUG GAACAAACGGAGAA 3'

~~-LWUGUC

SL

+

H

+

As

+

AB

-

I 1592

FIG.5. Proposed frameshift site in BWYV RNA. The slippery heptanucleotide is underlined, and the pseudoknot is indicated by the joining of the loop structure with the downstream sequence. The diagram shows the deletion mutations AB and A S made by BglII and Sac11 digestion respectively. Point mutations (Hor SL) ae indicated by arrows. The effects of the mutations on frameshift during translation of transcript RNA in wheat germ extracts is shown on the right-hand side.

GGGAAAC sequence (Garcia et al., 1993), it also occurs elsewhere between the downstream Sac11 and BglII sites (Fig. 5). This region contains two GGGAAAG sequences and one CCCAAAG. In summary, there is good evidence for in uiuo ribosomal frameshifting, but the mechanism by which it happens is far from being understood. In uitro experiments with mutagenized sequences on and around the slippery site have yielded somewhat discrepant results. Probably it will only be from studies on full-length transcripts that a full analysis of the sequences involved in this mechanism will be possible. Indeed, it has been shown that, in cells infected with full-length PAV cDNA, there is proportionately more of the protein resulting from frameshifting than is found in in vitro translation experiments or when truncated cDNA is expressed in uitro or in uiuo (C. P. Paul and W. A. Miller, 1994, personal communication).

B . Internal Initiation of Translation Most eukaryotic mRNAs are monocistronic, and translation initiation occurs according to the scanning model (Kozak, 1989). The 40s

MOLECULAR BIOLOGY OF LUTEOVIRUSES

437

ribosomal subunit binds to the cap structure at the 5’ end of the mRNA and then scans in the 3’ direction until the first AUG is reached. The 60s ribosomal subunit then binds to it, and protein synthesis begins. However, initiation efficiency can be modulated by the nucleotide context around the AUG, the most favorable being an A at position -3 and a G at +4 (Cavener and Ray, 1991; Kozak, 1989; Lutcke et al., 1987). If the first AUG is in an unfavorable context, some of the scanning ribosomes do not initiate until the second AUG if its context is more favorable. This leaky scanning mechanism (Kozak, 1989) can be utilized by bicistronic mRNAs like those of several animal viruses (see Samuel, 1989, for a review). In uitro translation of PLRV RNA or BWYV RNA in wheat germ extracts resulted in the synthesis of two major products with M, of 28,000 and 70,000 (PLRV; Mayo et al., 1989) or 25,000 and 66,000 (BWYV; Veidt et al., 1992). The apparent M,values correspond to the predicted sizes of the proteins encoded by the respective ORF 0 and ORF 1, indicating that initiation for ORF 1 occurs at an internal AUG. In reticulocyte lysate, the shorter protein encoded by ORF 0 was poorly expressed (Mayo et al., 1982, 1989; Veidt et al., 1992). A third minor polypeptide of apparent M,. 125,000 (PLRV) or 100,000 (BWYV) was also detected, which could represent the fusion protein resulting from frameshift from ORF 1 to ORF 2 (see Section IV, A). The same approach of translating transcript RNA from cloned cDNA has been used to investigate the expression of ORF 3 and ORF 4 of BWYV. The translation products included coat protein (identified by immunoprecipitation) and a protein of apparent M , 22,000 (which was not immunoprecipitated) that could be the translation product of ORF 4 (Veidt et al., 1988). In similar experiments with a transcript of PAV cDNA, Dinesh-Kumar et al. (1992) showed that the coat protein and the M , 17,000 polypeptide encoded by ORF 4 are produced in uitro from a single mRNA, by initiation at the first two AUG codons. The ratio between the amounts of coat protein and P4 produced varied from 1:l to 1:7 depending on the salt concentrations used in the translation. To determine the relative expression of ORF 3 and ORF 4 of PLRV in uiuo, the GUS gene was translationally fused to a cDNA fragment containing one or the other initiation codon. The resulting constructs expressed a GUS protein with an N-terminal part of either 29 PLRV amino acids (translation initiation at the coat protein AUG) or 20 PLRV amino acids (translation initiation at the P4 AUG) (Tacke et al., 1990). Tobacco or potato protoplasts were then electroporated with the construct and GUS activity in the protoplasts measured. Initiation at the internal AUG of ORF 4 was very efficient in this system. The amount of P4 synthesized exceeded by sevenfold that of the coat protein. When a similar approach was used for PAV ORF 3 and ORF 4, the

438

M. A. MAY0 A N D V. ZIEGLERGRAFF

ratio of initiation of P4 synthesis to that of coat protein was about 2 (Dinesh-Kumar and Miller, 1993).Thus, although the ratios differed with both PLRV and PAV, initiation at the second AUG was more common than at the first AUG. This may be because the AUG of ORF 4 is flanked by a more favorable context for initiation than is the coat protein initiation codon. Except for SDV and BMYV, all luteoviruses have a better initiation context around the ORF 4 AUG than around the ORF 3 AUG. The same is true for the initiation codons of ORF 0 and ORF 1. Dinesh-Kumar et al. (1992)reported that a secondary structure could be formed by the leader sequence of PAV subgenomic RNA, which would locate the coat protein AUG in a stem-loop structure, possibly making it less accessible to ribosomes. By site-directed mutagenesis, Dinesh-Kumar and Miller (1993)showed that improving the context of the coat protein AUG resulted in increased expression of ORF 3,compared to that of wild-type ORF 3 both in uitro and in uiuo. Destabilization of the secondary structure increased simultaneous expression of both ORFs, irrespective of the sequence context. Unexpectedly, for a given coat protein AUG context, changes that decreased initiation at the downstream AUG also reduced initiation at the first codon. Therefore, they proposed a new model in which pausing of the ribosomes at the second AUG enhances initiation at the upstream AUG codon. The 80s ribosomes formed at the P4 AUG are proposed to melt some base pairing upstream of this AUG, and therefore cause stacking of the upstream scanning 40s subunit, leaving it time to interact with the coat protein initiation codon and commence translation (Dinesh-Kumar and Miller, 1993).

C . Subgenomic mRNA Synthesis As with many RNA viruses, the ORFs in luteovirus RNA located in the 3’half of the genome are translated from subgenomic messenger RNAs. A major subgenomic species of about 2.6 to 2.9 kb was detected in cells infected with PAV (Dinesh-Kumar et al., 1992),PLRV (Smith and Harris, 1990;Tacke et al., 1990;Miller and Mayo, 1991),BWYV (Falk et al., 1989;Veidt et al., 1992)or CABYV (Guilley et al., 1994). These subgenomic RNAs allow the expression of the gene cluster of ORFs 3,4,and 5 (Tacke et al., 1990;Dinesh-Kumar et al., 1992). A second species of 0.8 kb has been detected in PAV-infected plants (Dinesh-Kumar et al., 1992)which could account for the expression of ORF 6. An extra small RNA of 0.7 kb was found in BWYV-infected plants (Falk et al., 1989), and one of 0.3 kb was detected in plants infected with PAV (Kelly et al., 1993).None of these RNAs was encap-

MOLECULAR BIOLOGY OF LUTEOVIRUSES

439

sidated (Smith and Harris, 1990; Miller and Mayo, 1991; DineshKumar et al., 1992). The positions in the genome RNA of the 5’ ends of the larger subgenomic RNAs of several luteoviruses have been mapped precisely (PLRV, Miller and Mayo, 1991;CABYV, Guilley et al., 1994;BWYV, V. Ziegler-Graff, 1994,unpublished observations). The leader sequences are unusually long (212nucleotides for PLRV and CABYV, 224 nucleotides for BWYV) and the 5‘ ends are located 12,8,or 19 nucleotides upstream of the termination codon of ORF 2. The 5’4erminal sequences of the subgenomic RNA are identical to those of the respective 5’4erminal sequences of the genomic RNA (Fig. 6).Moreover, the first eight nucleotides of the subgenomic and genomic RNA of BWYV, PLRV, and CABYV are identical, which may reflect a conserved replicase recognition signal in the minus-strand RNA. Comparisons among the intergenic regions of different luteovirus RNAs indicate the presence of two conserved regions between nucleotides -102 to -91 and -46 to -24 (Mayo et al., 1989;Vincent et al., 1991).They are particularly A-U-rich and could represent the “core region” of the subgenomic promoter as defined for brome mosaic virus RNA by Marsh et al. (1987).This region may therefore be involved in the synthesis and CABYV

BWYV

PLRV

C A G G A G A A A U U G - N,,

--

FIG.6. The 5’-terminal sequence similarities between the genomic and subgenomic RNAs of CABYV, BWYV, and PLRV. For each virus, the genomic sequence [(l) indicates the 5’-nucleotidel is aligned against the place in the genomic RNA where the subgenomic RNA commences. Vertical lines show the sequence matches between the termini of genomic and subgenomic RNA, the box indicates sequences common to the three viruses. Open boxes indicate the 3’ extremity of ORF 2 (P2) or the 5’ extremities of ORFs 0 or 3 (PO, P3); N, signifies intervening sequence.

440

M. A. MAY0 AND V. ZIEGLERGRAFF

regulation of the larger subgenomic RNA of these luteoviruses. The 5‘ end of the subgenomic RNA of a German isolate of PLRV has been mapped to 40 nucleotides upstream the coat protein AUG codon (Prufer et al., 1992). In contrast to these subgroup I1 viruses, the 5’ end of the subgenomic RNA of PAV is 89 nucleotides upstream of the coat protein initiation codon (Dinesh-Kumar et al., 1992). In this case, the initiation site for the synthesis of the subgenomic RNA is different from the 5’ proximal sequence of the genomic RNA. However, with PAV from Australia, L. Kelly (1994, personal communication) found the subgenomic leader sequence to be 188 nucleotides long, which would mean that, as for subgroup 11 viruses, the 5’ end of the subgenomic RNA is in ORF 2.

D . Readthrough or Leaky Translational Termination Readthrough occurs when a stop codon is suppressed by binding a suppressor tRNA. The action of a suppressor tRNA has been demonstrated in the readthrough of termination codons in RNAs of tobacco mosaic virus (Beier et al., 1984) and tobacco rattle virus (Zerfass and Beier, 1992). The result is a relative abundance of the smaller product and a small proportion of a larger fusion protein that includes the translation product of the next in-frame ORF. In luteoviruses, ORF 5 is expressed as an ORF 3 + ORF 5 fusion protein by translational readthrough of the coat protein UAG termination codon. The sequence context surrounding the leaky stop codon is identical for all luteoviruses: AAAUAGGUAGAC (termination codon in bold type). In the RNA of tobacco mosaic virus (TMV),the two codons downstream of the suppressible UAG codon at the end of the ORF encoding the M, 126,000 protein have been shown to form part of the signal promoting readthrough (Skuzeski et al., 1991). Although the sequence context of the luteovirus leaky stop codon is different from that in TMV RNA (CCAUAGCAAUUA),it is likely that the conserved sequence context in which the UAG is embedded in luteovirus RNAs is similarly important in conferring leakiness. The readthrough protein has been detected in potato plants and tobacco protoplasts infected with PLRV (Bahner et al., 1990) and in C. quinoa protoplasts infected with BWYV (Reutenauer et al., 1993) by using antibodies raised against P5 (see Section II1,F for more details). Readthrough protein has also been detected in preparations of purified particles of PLRV (Bahner et al., 1990),PAV (Waterhouse et al., 19891, RPV (Vincent et al., 1991), and BWYV (see Section V,O. In in uitro translation experiments, the efficiency with which the ORF 3 UAG in PAV RNA was suppressed by readthrough was reported

MOLECULAR BIOLOGY OF LUTEOVIRUSES

441

to vary from 7 to 15% depending on the salt concentration (DineshKumar et al., 1992).The frequency of in uiuo suppression of the UAG stop codon separating the PLRV ORFs 3 and 5 was determined in a transient expression system using a clone in which the 18 nucleotides upstream and 21 nucleotides downstream of the stop codon were translationally fused to the GUS gene. When GUS activity was measured in tobacco and potato protoplasts electroporated with this construct, the efficiency of suppression was about 1% (Tacke et al., 1990). Readthrough was about 5% when the stop codon was in the context of the leaky stop codon in TMV RNA (Skuzeski et al., 1991). Miller et al. (1995)reported that they were unable to observe suppression of the PAV RNA stop codon using the same approach, but they could detect readthrough in uiuo when the construct was a full-length clone of PAV in which the GUS gene was inserted downstream of the coat protein stop codon. Miller et aZ. (1995)suggested therefore that readthrough in PAV RNA requires a viral or virus-induced transacting factorh) or distant cis-acting signals not present in the initial construct. A sequence rich in C residues consisting of 16 uninterrupted CCxxxx repeats located downstream of the coat protein stop codon has been shown to be involved in efficient suppression in protoplasts infected with PAV transcript RNA; 10 such repeats were sufficient for suppression, but elimination of them all resulted in no suppression (S. P. Dinesh-Kumar and W.A. Miller, 1994,personal communication). A similar pattern of from 7 to 16 such repeats is present in the sequences of RNA of all luteoviruses (Miller et al., 1995). The amino acid sequence encoded by this sequence is not involved in the mechanism because the introduction of a frameshifting mutation between the stop codon and the repeats in PAV RNA had no effect on the extent of readthrough (Miller et al., 1995).

E . Proteolysis and Cap-Independent Translation

A protease consensus has been found in P1 proteins encoded by subgroup I1 virus RNA (Koonin and Dolja, 1993;see also Section II1,C). However, no direct evidence has been reported so far for proteolysis of luteovirus proteins. Nonetheless, proteolytic processing of luteovirus proteins, at least for PLRV and RPV, is implied because the VPg does not correspond to the translation product of an ORF. It has been suggested that part of the P1 forms the VPg (Koonin and Dolja, 1993).By analogy with other viruses which have a VPg, it is predicted that the processing to give the VPg is by a virus-coded protease. Miller et al. (1995)reported that a 500-nucleotide sequence located in the 3' end of PAV RNA was required for efficient translation of

442

M. A. MAY0 A N D V. ZIEGLER-GRAFF

uncapped transcripts of PAV cDNA in wheat germ extracts. Deletion of portions of this sequence (from nucleotide 4513 to 5009)resulted in a marked reduction in translation to yield P1.Capping of these deleted transcripts restored full translational activity. These results were not observed in rabbit reticulocyte lysate. Miller et al. (1995)suggested that this cap-independent translation mechanism could be an adaptation of PAV to multiplication in cereals. V. PARTICLE STRUCTURE

A. Possible Tertiary Structure The secondary structure of luteovirus coat proteins was predicted by Dolja and Koonin (1991)on the basis of a multiple sequence alignment with coat proteins of known secondary structure. This model was used by Torrance (1992)to interpret the location of epitopes in PLRV particles (see below). There has been an improvement in the reliability of secondary structure predictions (Rost and Sander, 1993,19941,and the sequences shown in Fig. 2 were submitted to this procedure. The program HSSP/MaxHom was used to align the sequences and then predict the secondary structure of the aligned proteins. The predicted regions of p- sheet and a-helix are shown in Fig. 2.The p sheets are labeled to correspond with the letter coding used for SBMV and tomato bushy stunt virus (TBSV) (Rossmann and Johnson, 1989);the long continuous p sheet near the C terminus is assumed to correspond to p sheets H and I (Fig. 2).As a control, the coat protein of SBMV was assessed, and the program predicted all of the p sheets known to be present (AbadZapatero et al., 1980)except that the edges of the p-sheet regions were all underestimated, suggesting that lengths of p sheets in luteovirus coat proteins shown in Fig. 2 are also underestimates. Assuming that p sheets B to I are arranged in a shell domain, as they are in the coat proteins of many diverse viruses (Rossmann and Johnson, 1989),it is possible to predict the parts of the proteins which would correspond with the parts of SBMV protein known to be exposed on the exterior of the virus particles (Hermodson et al.,1982).These are also indicated on Fig. 2. The N-terminal 58 to 69 amino acids of luteovirus coat proteins to the N-terminal side of p sheet B (Fig. 2) resemble this region (R domain) of many coat proteins of viruses with isometric particles in being highly basic (Rossmann and Johnson, 1989). They contain between 19 and 21 R or K residues separated by relatively nonpolar residues such as G and N, and no acidic residues. By analogy with the

MOLECULAR BIOLOGY OF LUTEOVIRUSES

443

coat proteins of SBMV or TBSV, this region is predicted to be internal, and presumably the basic residues neutralize the negative charge on the virus RNA.

B . Location of Epitopes Inspection of multiple sequence alignments, such as that in Fig. 2, suggest interpretations for the observation that epitopes are often shared between some, but not all, luteoviruses. (e.g., D’Arcy et al., 1989). For example, an epitope common t o BMYV, BWYV, and RPV particles but not on those of PLRV, PAV, MAV, or GRAV could be in the tripeptide LAG or the hexapeptide STINKF. An epitope shared by BLRV, BWYV, and RPV particles but not those of PLRV or SDV could be in INKF. Greater precision is possible with some combinations such as the possibility that an epitope shared by PAV and BMYV but not by BWYV is created by substituting a n M for an I, a T for a n A, or SA for NS or SS. The positions of these possible epitopes are indicated in the BWYV sequence in Fig. 2. The last possibility, like the other possible sites involving INKF, is in the putative E-F loop which should be on the surface of the virus particle. A less conjectural approach is to assess the reactions of the antibodies to each of a set of overlapping peptides that represent the entire protein sequence. By this method, Torrance (1992) has identified 11 epitopes in the coat protein of PLRV and has mapped them to the positions shown in Fig. 2. The major epitope was found to comprise the amino acids at the extreme N terminus. Antibodies to this epitope reacted with intact virus particles, which suggests that the epitope is external although the location of the N-terminal arms of the coat proteins of SBMV and TBSV are known to be internal (Rossmann and Johnson, 1989). Torrance (1992) suggested that the N-terminal amino acids of PLRV coat protein are exposed at the particle surface when particles swell because of changes in pH or ionic conditions.

C . Presence of Readthrough Protein (P5) As discussed in Section III,F, the readthrough protein is expressed as a fusion protein attached to the C terminus of the coat protein, and a few percent of the coat protein molecules produced have this extension. These molecules form part of the virus particles, and P5 has been detected in purified particles of PAV (Waterhouse et al., 19891, PLRV (Bahner et al., 1990), RPV (Vincent et al., 1991), and BWYV (Reutenauer, 1994) (Fig. 7). With both PLRV and RPV there is evidence for P5 being readily degraded after isolation from infected tissue. Bahner

444

M. A. MAY0 AND V. ZEGLERGRAFF

FIG.7. Expression of P5 of PLRV and BWYV. Samples were protein from healthy protoplasts (A, lanes 1 and 4); PLRV-infected protoplasts (A, lanes 2 and 5); purified particles of PLRV (A, lanes 3, 6 and 7); purified particles of BWYV (B, lanes 1 and 2); leaf tissue of Nicotiana cleuelandii infected with BWYV by agroinfection (B, lane 3); or leaves of mock-inoculated N. cleuelandii (B, lane 4). Protein extracts were subjected to electrophoresis and then either staining with Coomassie blue (A, lane 7)or silver nitrate (B, lane 1)or transfer by electroblotting to nitrocellulose. Blots were reacted with mouse monoclonal antibodies to PLRV particles (A, lanes 1, 2, and 3) or rabbit polyclonal antibody to PLRV P5 (A, lanes 4 and 5), to BWYV particles (B, lanes 2, 3, and 4, lower part), or BWYV P5 (B, lanes 2,3, and 4, upper part). Antibody reaction was detected by conjugated anti-mouse or anti-rabbit antibodies. RT indicates the position of the P3 + P5 readthrough protein, RT*indicates the position of the partial degradation product of P3 + P5, and CP indicates the position of the coat protein (P3).

et al. (1990)found that whereas the fusion protein of P3

+ P5 had an

M, of about 80,000 when extracted from infected cells or particles rapidly sedimented from infected protoplasts, the C-terminal half of the P5 part of this fusion protein was rapidly degraded when infected cells were disrupted and its size in particles of purified PLRV was about M, 53,000. The fusion protein in particles of RPV had an apparent M, of 63,000 (theoretical size 66,000) but was often degraded to M, 58,000 and sometimes was undetectable (Vincent et al., 1991).It has been suggested that the readthrough domain is subject to partial proteolytic degradation in the course of virus purification (Bahner et al., 1990;Vincent et al., 1991). Detection of the truncated readthrough protein in purified virus particles of a frameshift mutant of BWYV

MOLECULAR BIOLOGY OF LUTEOVIRUSES

445

indicated that the site of this cleavage was near the C end of the luteovirus conserved domain (see Fig. 4),which suggests that there is a common site for this cleavage in different luteoviruses (V. Brault, 1994,personal communication). The C termini of the coat proteins of SBMV and TBSV are at the surface of the virus particles, and the correspondence in apparent structure of luteoviruses with these viruses (Dolja and Koonin, 1991; Section V,A) makes it likely to be so in luteovirus particles. Thus, the readthrough protein is predicted to protrude on the outside of the particles. This model of luteovirus particles consisting of M, 23,000coat protein molecules and a few much larger proteins of coat plus readthrough protein might explain the occasional protuberances seen on PLRV particles in some electron micrographs (Harrison, 1984). In more recent experiments, it was possible t o demonstrate specific labeling of BWYV particles with antibodies raised against a fusion protein containing part of P5 (J. F. J. M. Van den Heuvel, 1994,personal communication).

D . Heterologous Encapsidation It has been known for some time that aphids can transmit a luteovirus which they do not normally transmit if the virus is acquired from plants also infected with a transmissible luteovirus. For example, Myzus persicae was able to transmit carrot red leaf virus (CRLV)when feeding on plants doubly infected by CRLV and PLRV (Waterhouse and Murant, 1983).Similar results were obtained with mixtures of strains of BYDV (Rochow, 1970b;H u et al., 1988).Work in which virus particles were trapped by reaction with specific antibodies and the RNA molecules contained in the particles were characterized by reaction with specific nucleotide sequence probes has demonstrated two types of heterologous encapsidations (Creamer and Falk, 1990;Wen and Lister, 1991). These are transcapsidation or genomic masking, where particles are formed from the protein of one virus and the RNA of another virus, and phenotypic mixing, where the coat protein shell contains proteins from more than one virus. Transcapsidation was demonstrated between RPV and either PAV or MAV and between MAV and the RMV strain of BYDV (RMV) (Creamer and Falk, 1990), whereas phenotypic mixing occurred with PAV and MAV and with RPV and RMV (Wen and Lister, 1991).Thus, although PAV and MAV coat proteins differ in 54 amino acids, of which 31 are in the shell domain (Section V,A), they can coassemble. Transcapsidation also suggests a lack of specificity in the interaction of virus RNA with coat protein. A more extreme example of this is

446

M. A. MAY0 AND V.ZIEGLERGRAFF

the association between luteoviruses and several other viruses to form transmission complexes (Waterhouse et al., 1988; Murant, 1993). The dependent viruses lack coat proteins and thus depend on the luteovirus for RNA protection and a means of being transmitted by vector aphids. They have been classified in the genus Umbravirus (Murant et al., 1995). Other examples are the symbiotic association of the two RNAs comprising the genome of PEMV and the dependency of ST9-associated RNA on the ST9 strain of BWYV (see Section VII1,D).

E . Determinants for Particle Assembly Purified preparations of some viruses with isometric particles characteristically contain RNA-free shells, often called top component. Particle assembly of such viruses presumably does not require RNA. Only with one isolate of BWYV has a top component been found in preparations of luteovirus particles (Hewings and D’Arcy, 1986). Some slowly sedimenting PLRV protein structures were found in PLRVinfected tobacco protoplasts 2 days after inoculation, but no virus-like particles could be detected (Miller, 1992). Work with mutants of BWYV has shown that the readthrough protein contributes little or nothing to the assembly of virus particle (Reutenauer et al., 1993).Virus particles were made in protoplasts infected with mutants lacking almost all of the readthrough protein (see Section 111,F). More recent work in which PLRV coat protein has been expressed in heterologous systems may lead to a system for determining the factors needed for PLRV particle assembly (J. Lamb, 1994, personal communication). When Spodoptera frugiperda cells were infected with a recombinant baculovirus that contained cDNA encoding PLRV coat protein under the control of the polyhedrin promoter, they expressed large amounts of P3. The PLRV protein accumulated in the nuclei of the insect cells apparently in amorphous aggregates. When the inserted cDNA also contained sequence encoding a histidine tag at the N terminus of the coat protein, the resultant modified coat protein also accumulated in nuclei, but it formed viruslike particles that aggregated into crystallike structures. The cDNA did not encode P5, and the result therefore suggests that P5 is not needed for the assembly of particles. However, it is not known what effect the amino acid extension of the N terminus has; it may be mimicking the RNA component of PLRV particles, or it may mimic the role played by P5, even though P5 is a C-terminal extension of P3 rather than an N-terminal extension.

MOLECULAR BIOLOGY OF LUTEOVIRUSES

447

VI. LOCATION OF LUTEOVIRUS REPLICATION

A . Limitation to Phloem Tissue Microscopic examination of tissues infected with luteoviruses shows that virus accumulates almost exclusively in the sieve elements and companion cells of the phloem and occasionally in phloem parenchyma. Typically, not all phloem cells contain virus particles, and movement vertically in infected plants is probably relatively rapid whereas horizontal movement is slow and inefficient (Waterhouse et al., 1988). Restriction of luteovirus multiplication to these cells is not because luteoviruses cannot replicate in other cell types, as luteoviruses have been shown to multiply in isolated mesophyll protoplasts and in a few cells outside the vascular tissue. Replication has been demonstrated in mesophyll protoplasts inoculated with PAV (Barnett et al., 1981; Dinesh-Kumar et al., 1992),PLRV (Kubo and Takanami, 19791,tobacco necrotic dwarf virus (TNDV) (Kubo, 1981), BWYV (Veidt et al., 19921, and RPV (Silver et al., 1994). TNDV was shown to multiply in inoculated epidermal cells (Imaizumi and Kubo, 1980), BYDV was sometimes found in xylem parenchyma cells (Gill and Chong, 1981), and in N . cleuelandii plants infected with PLRV a few mesophyll cells became infected (Barker, 1987). Moreover, double infection of N . cleuelandii with PLRV and potato virus Y (PVY) resulted in about a sevenfold increase in the proportion of the mesophyll cells that contained PLRV. This effect was also found in plants infected with PLRV and potexviruses, tobraviruses, or carrot mottle virus (CMoV) (type species of the genus Umbravirus) but not in plants infected with a variety of other viruses including PEMV (Barker, 1989). A similar effect was detected in plants infected with BWYV (Barker, 1989). An interesting parallel occurs with plants infected with PEMV. The luteovirus-like RNA-1 of PEMV (see Section VII1,C) multiplies in infected protoplasts but does not multiply when inoculated into plants unless these are also inoculated with RNA-2. In plants inoculated with RNA-2 alone it can spread systemically, but no virus particles are formed (Demler et al., 1994). In this way, PEMV RNA-2 resembles umbraviruses such as CMoV. It seems likely that the CMoV or PVY in the plants doubly infected with PLRV provide a factor which moderately enhances the very restricted movement of the PLRV in nonphloem cells; the RNA 2 of PEMV provides a fully functional movement protein that assists or permits PEMV to spread systemically. A different coinfection occurs with the ST9 strain of BWYV. When plants are infected with BWYV-ST9, the presence of a smaller, ST9-

448

M. A. MAY0 A N D V. ZIEGLERGRAFF

associated RNA (see Section VII1,D) enhances the amount of BWYV accumulating in infected tissue (Passmore et al., 1993). However, despite this extra accumulation, BWYV could not be detected outside the phloem tissue (Sanger et al., 1994). A distinction between BWYV-ST9 and the PLRV-CMoV system is that CMoV spreads systemically in infected plants, whereas, although the ST9-associated RNA can infect inoculated leaves, it does not spread systemically (Passmore et al., 1993).

B . Cytopathological Effects The effects that infection has on cell ultrastructure differ according to which subgroup the virus belongs (Gill and Chong, 1979). Infection of oat plants with MAV or PAV (subgroup I) resulted in the formation of single membrane-bound vesicles and dense filaments that seemed to accumulate in the nuclei. Virus particles accumulated in the cytoplasm, and late in infection the nuclei became distorted, and deteriorated. In contrast, infection of the same host species with RPV (subgroup 11)resulted in the formation of double membrane-bound vesicles and tubules. Virus particles accumulated in the nuclei, which remained intact (Gill and Chong, 1979). The effects of other subgroup I1 genome viruses were similar, although no PLRV particles were detected in nuclei (Shepardson et al., 1980). Infection of isolated protoplasts with PEMV RNA-1 induced the formation of vesicles, and virus particles were detected within the nuclei as well as in the cytoplasm (Demler et al., 1994). As discussed above, the expression in insect cells of the coat protein gene of PLRV in a recombinant baculovirus resulted in the accumulation of the coat protein in the nuclei (J.W. Lamb, G. H. Duncan, M. A. Mayo, and R. T. Hay, 1994, unpublished results). When the gene was modified to produce a tagged protein, the modified protein formed into viruslike particles. The coat protein of PLRV and most other luteoviruses contain highly basic sequences near the N termini that resemble nuclear location signals (Garcia-Bustos et al., 1991), but the sequence is probably on the inside of intact particles where it would normally interact with virus RNA (see Section V,A). These observations reinforce the suggestion made by Esau and Hoefert (1972) for BWYV that virus particles assemble in the nucleus. VII. PHYTOPATHOLOGY

A. Diagnosis of Luteovirus Infection Diagnosis of virus infection can be of two types. Either a broad specificity detection is required so that even viruses only marginally

MOLECULAR BIOLOGY OF LUTEOVIRUSES

449

related to the reference standard are detected (e.g., as with polyclonal antiserum), or a highly specific detection is required in order to discriminate the virus from close relatives (e.g., as with monoclonal antibodies). The accumulation of cloned cDNA to luteovirus RNA and knowledge of the nucleotide sequences of several luteoviruses and their strains has led to the development of nucleic acid hybridizationbased probes for each of the types of diagnostic tests. Not surprisingly in the light of the data in Table 11, probes made from cDNA to different parts of the genome were either highly specific or reacted to some extent with RNA from other luteoviruses (e.g., Waterhouse et ul., 1986). Probes complementary to the coat protein gene of PLRV (Robinson and Romero, 1991) or BWYV (Herrbach et al., 1991) have proved to be the least specific. Probes to the coat protein gene of PLRV detected BWYV and RPV readily, GRAV and CRLV weakly, but did not detect PAV or MAV (Robinson and Romero, 1991); probes to the coat protein gene of BWYV detected BMYV and PLRV readily, RPV and PAV weakly, but did not detect MAV (Herrbach et al., 1991). These results do not correlate well with the amount of amino acid sequence identity between the coat proteins of the various luteoviruses (Table 11). A potentially powerful method for diagnosing luteovirus infection is the use of reverse transcription followed by polymerase chain reaction (RT/PCR) using primers designed to hybridize to all luteovirus RNAs. Robertson et al. (1991) described such a pair of primers, one of which is partially degenerate, that hybridized with RNA and cDNA of PLRV, BWYV, and PAV. The amplified product represents the 3’-terminal464 nucleotides of the coat protein gene. Detection by this method is potentially highly discriminatory, as sequence analysis of the PCR product would show how similar the detected luteovirus was to known viruses. However, the greater number of sequences available now show that, at least for some luteoviruses, the primers are unlikely to work because of mismatching. The downstream primer, which includes the termination codon, hybridizes well, but the upstream primer is a poor match to SDV RNA and also may not hybridize with BLRV RNA. A possible new upstream “universal luteovirus primer” was derived from an alignment of the coat protein genes of BLRV, BWYV, CABYV, MAV, PAV, PLRV, RPV, and SDV and should yield a PCR product corresponding to the 3’-terminal 319 nucleotides of the coat protein gene (M. A. Mayo and C. A. Jolly, 1994, unpublished results). Detection can be made more sensitive and more specific, without recourse to sequencing, by using immune-capture PCR in which virus particles are bound to a plate coated with a specific antibody and then RNA is extracted from the bound virus and subjected to RT/PCR. The method has been used for detecting potyvirus particles (Wetzel et al.,

450

M. A. MAY0 AND V. ZIEGLERGRAFF

1992) and is also effective for PLRV; the sensitivity obtained was sufficient to detect the PLRV particles circulating in a single aphid (C. A. Jolly, 1994, personal communication).

B . Resistance to Luteoviruses For many viruses, it has been shown that plants are more resistant to virus infection and/or multiplication when they have been transformed such that they express the coat protein gene of the target virus (Reavy and Mayo, 1992). Application of transformation methods to produce plants resistant to luteoviruses has been done with PLRV. DNA encoding PLRV coat protein (and thereby also P4) has been inserted into the genomes of potato plants (Kawchuk et al., 1990,1991; Van der Wilk et al., 1991; Barker et al., 1992) and tobacco plants (Barker et al., 1993). The transformants were often resistant to infection by the feeding activity of viruliferous aphids (Kawchuk et al., 1990) and/or showed a restriction in the amount of virus accumulating after the primary infection of potato (Kawchuk et al., 1991) or tobacco (Barker et al., 1993). Transgenic potato plants were also resistant to secondary infection (Barker et al., 1992). The resistance seemed to be independent of the production of coat protein, as little or no protein could be detected in many transgenic lines that were nonetheless resistant to PLRV. Resistance to infection was not always particularly strong, but multiplication resistance resulting in virus accumulating to about 10 to 20% of the amount in control plants is comparable to the resistance being used in current breeding programs. Antisense constructs were also effective (Kawchuk et al., 1991; Van der Wilk et al., 19911, which reinforces the view that the resistance is caused by the production of PLRV RNA sequences (or their complement) rather than coat protein. Multiplication resistance was shown t o be related to the amount of transcript synthesized, but not in a strict correlation (Barker et al., 1992, 1993). Derrick and Barker (1992) showed that in potato resistant to PLRV, either because of transformation with the PLRV coat protein gene or because of the action of a host gene(s), only the adaxial phloem was infected and fewer cells were infected than in control plants. In experiments with the same transgene, it was shown that the effects of the transgene and that of the host resistance gene were additive; transformation of relatively resistant genotypes further enhanced their resistance to levels approaching that of extremely resistant wild Solanurn species (Barker et al., 1994). The similarity in phenotypic effect of the PLRV coat protein gene and host resistance genes may be coincidence, but if not, it raises the intriguing prospect that the two types of resis-

MOLECULAR BIOLOGY OF LUTEOVIRUSES

451

tance are induced by the same or similar mechanisms (H. Barker, 1994, personal communication). Plants expressing a mutated form of P4 were found to be resistant to virus spread, and it was suggested that the transgene product had interferred with the transport role of the P4 made during the course of an infection (Tacke et al., 1993b). VIII. TAXONOMY

A . Speciation The available sequences provide sufficient data to pose the question of when strains of a luteovirus should be considered different viruses. BYDV isolates have been placed in two subgroups on the basis of serological relationships, cytopathological effects, and the doublestranded RNA formed in infected tissue (Rochow 1970a; Waterhouse et al., 1988; Ueng et al.,1992). Strains MAV and PAV were placed in one subgroup, and RPV was placed in the other subgroup. The molecular evidence therefore reinforces the biological evidence in arguing for considering BYDV to consist of two species; the corollary is that one should be renamed. However, the situation is more complex when sequences of individual genes are compared. If P1 or P2 are considered, then PAV and MAV should be considered as strains because they are 98% identical; if genes in the 3’ coding block are compared then values of 73% (P3), 72% (P4), and 72% (P5) identity (Table 11) suggest dissimilarity approaching that thought of for other virus groups such as potyviruses (Shukla and Ward, 1989) to indicate that the viruses belong to different species. Although MAV and PAV could perhaps be considered as strains of one virus because of their similar host ranges and ability to cross-protect against one another (Wen et al., 1991), this seems less reasonable for CABYV and GRAV which, although they have coat proteins that are 75% identical (Table II), are transmitted by different aphid species and infect different hosts. In contrast, with the four strains of PLRV sequenced the minimum similarity was 88%between P1 of Australian PLRV and that of Canadian or Dutch PLRV. In comparisons among isolates obtained in Scotland, the sequences of P3 and P5 were 96 to 99% identical (Jolly, 1994). Thus the variation among isolates from diverse countries (Table 11) was similar to that obtained in Scotland. PLRV appears to vary very little. In a larger study, J. De Miranda and R. Hull (1994, personal communication) compared 37 isolates of BWYV-like viruses from Europe and

452

M. A. MAY0 AND V. ZIEGLERGRAFF

Iran by sequencing RNA consisting of about 400 nucleotides of ORF 2, the intergenic region and ORF 3. The isolates could be clustered in different ways according to which regions of the sequences were considered. The ORF 3 of all isolates were 93% or more identical, the ORF 2 sequences fell into two groups with only 63% identity between the groups, but the intergenic sequences in one of these groups fell into two subgroups with only 68% identity between them. There have been a number of attempts to rationalize the classification of luteoviruses with some serological relationship to BWYV. BMYV, RPV, CRLV, and the RGV strain of BYDV have all been considered to be strains of BWYV (Casper, 1988). This lumping approach does not seem very useful for pathologists seeking to discriminate between pathogens and is not supported by the differences between the sequences of the coat proteins of BWYV and RPV (Table 11).Moreover, a hybridization probe corresponding in sequence to ORF 0 of BMYV, which is pathogenic for beet, did not cross-react with RNA of BWYV, which is not pathogenic for beet (0.Lemaire, 1994, personal communication).

B . Structure of Genus Luteovirus As discussed above, viruses in the genus Luteovirus have one of two genome arrangements, each typified by distinctive polymerase sequences (Fig. 1).The polymerases resemble those of either carmoviruses or sobemoviruses, which are considered to be very different types of RNA viruses (Dolja and Carrington, 1992; Gibbs, 1995). Thus, classification on the basis of the polymerase sequence would probably place viruses in each luteovirus subgroup in different families. But the biology of the viruses (phloem restriction, persistent aphid transmission) and the molecular features (suppression of the amber termination codon of the coat protein gene to yield a readthrough protein, encoding of P4 inside the P3 gene) are consistent with luteoviruses belonging to one genus. The pragmatic solution to this difficulty would seem to be to retain the grouping known at present as the genus Luteovirus but to consider each of the subgroups as genera within the larger grouping. This arrangement would keep biologically similar viruses together. The implication would be that building a phylogenetic tree of viruses from the sequences of polymerase domains does not necessarily lead to workable taxa. Indeed, the extent of detectable recombination within the genomes of luteoviruses should be a strong disincentive to the use of the polymerase gene to typify a virus.

MOLECULAR BIOLOGY OF LUTEOVIRUSES

453

C . Problem of Pea Enation Mosaic Virus Pea enation mosaic virus resembles luteoviruses in being persistently transmitted by aphids but is distinct in being mechanically transmissible, being capable of invading mesophyll tissues in infected plants, and having a bipartite genome (Hull and Salquero, 1991). Molecular analysis of the PEMV genome has provided an explanation for these properties and has posed a taxonomic challenge (Demler et al., 1994).Figure 8 shows the arrangement of the ORFs in the two genome RNAs. The larger RNA of the genome (RNA-1)resembles that of subgroup I1 luteoviruses. It contains ORFs which are equivalent in position and in sequences of the encoded proteins to ORFs 0, 1 , 2 , 3 , and 5 of luteoviruses. Moreover, in the overlap between ORF 1 and ORF 2 there are slippery sequences, and ORF 2 is probably expressed by translational frameshift from the ORF 1 frame, just as for luteovirus subgroup I1 genomes. There is no ORF equivalent to ORF 4 of luteoviruses in PEMV RNA-1, and the ORF corresponding to ORF 5 is relatively short (Fig. 8).The ORF 2 region of RNA-1 encodes a putative polymerase that is similar to subgroup I1 luteovirus polymerases and the RNA can multiply in inoculated protoplasts independently of the smaller genome RNA (RNA-2)(Demler et al., 1994).ORF 3 encodes the coat protein, which is about 30 to 35%identical to those of luteoviruses, and virus particles accumulate in protoplasts inoculated with RNA-1 alone. However, RNA-1 is incapable of systemic movement in plants. RNA-2 also resembles the genome of other viruses, in this case carmoviruses (Fig. 8). RNA-2 encodes a polymerase and is capable of multiplying in both inoculated protoplasts and inoculated plants (Demler et al., 1993) independently of RNA-1. In plants, RNA-2 moves systemically. The combination of the two somewhat defective RNA RNA 1

FIG.8. Genome organization of pea enation mosaic virus. Boxes indicate the ORFs. Those in RNA-1 are labeled to correspond with the ORFs of PLRV in Fig. 1.

454

M. A. MAY0 AND V. ZIEGLERGRAFF

molecules, each resembling a different type of virus, produces an agent which behaves like a virus and which is stable in nature. A major biological feature of PEMV is its persistent transmission by aphids, and a possible solution to the problem of how to classify PEMV is to regard it as a luteovirus, albeit one with anomalous properties. A further complication is that the polymerase of PEMV RNA-2 has some sequence relatedness with the polymerases of the subgroup I genomes of luteoviruses. It is conceivable that the recombinatorial origin of the polymerase part of the subgroup I genomes was a virus like PEMV RNA-2. Whatever the eventual taxonomic outcome, the elucidation of the genome structure of PEMV is a vivid example of the contribution of molecular biology to taxonomic thinking.

D . RNA Associated with Luteoviruses 1 . ST9-Associated Satellite-Like RNA Another example of a “symbiotic” association between viral RNAs is the ST9 strain of BWYV. Particles of BWYV-ST9 contain two RNA species (Falk et al., 1989). The larger resembles the genome of other isolates of BWYV, and it multiplies when inoculated into protoplasts. The smaller is a 2.8-kb RNA which encodes a polymerase (Chin et al., 1993) and is capable of replicating in inoculated plants independently of BWYV RNA (Passmore et al., 1993); however, it does not encode a coat protein (Chin et al., 1993). When plants are infected by BWYVST9, symptoms are more severe, and virus yields are about 10 times greater, than in plants infected with other isolates of BWYV (Falk et al., 1989). Aphids transmit both RNA species, each encapsidated in a different particle (Sanger et al., 1994). This association resembles that between the RNAs of PEMV but is different in that BWYV-ST9 does not invade mesophyll tissue and is not mechanically transmissible. According to the current definition of satellites (Mayo, 19911, BWYVST9-associated RNA is not a satellite as it is capable of replication independent of its associated virus (BWYV-ST9).It, and indeed PEMV RNA-2, resembles umbraviruses in being dependent on a luteovirus for encapsidation and therefore transmission (Murant et al., 1995). However, the 2.8-kb ST9-associated RNA is much smaller than the approximately 4.5-kb RNA of umbraviruses. 2 . Satellite RNA of Barley Yellow Dwarf Virus (RPV Strain) Satellite RNA has been detected in RPV cultures (Miller et al., 1991; Silver et al., 1994).The satellite is a D-type (Mayo, 1991)that occurs as circular molecules of 322 nucleotides. The RNA replicates by a rolling

MOLECULAR BIOLOGY OF LUTEOVIRUSES

455

circle mechanism during which the RNA undergoes self-cleavage of both (+I and (-) sense RNA to release monomers from the polymeric products of replication. The satellite replicates in protoplasts coinfected with RPV but not in those coinfected with PAV (Silver et al., 1994).

IX. CONCLUDING REMARKS Molecular analysis has shown that the genomes of luteoviruses combine most of the strategies used to express the monopartite genomes of (+) sense ssRNA viruses (Morch and Haenni, 1987). Moreover, luteovirus genomes have clearly evolved by recombination between blocks of coding sequence derived from distinct ancestral viruses. Thus, despite the difficulties of studying these viruses, many interesting molecular biology features have been demonstrated and would repay more intensive study. The biological features of luteoviruses are also unusual in that, except in peculiar circumstances, luteoviruses are confined to the phloem of their hosts (Section VI,A) and during transmission luteovirus particles interact with surfaces in their insect vectors to cross several cell boundaries (Section 111,F).Both features can be explained in a general way, but the molecular bases for the properties are as yet poorly understood and are thus excellent candidates for more penetrating molecular study. In this review we have attempted to show progress beyond the first phase of luteovirus molecular biology in which sequences have been accumulated. It seems clear that much more progress can be anticipated on several fronts in the near future and that in many cases the knowledge gained should convey lessons applicable in the wider field of RNA virus molecular biology.

ACKNOWLEDGMENTS We thank Allen Miller for sharing a review with us prior to publication and to H. Barker, V. Brault, S. P. Dinesh-Kumar, H. Guilley, E. Herrbach, R. Hull, C. A. Jolly, L. Kelly, J. W. Lamb, 0. Lemaire, R. R. Martin, J. de Miranda, C. P. Paul, K. Scott, J. F. J. M. Van den Heuvel, and I. Veidt for allowing us to refer to unpublished results andlor personal opinions. Financial support from the Scottish Office Agriculture and Fisheries Department is also acknowledged.

REFERENCES Abad-Zapatero, C., Abdel-Meguid, S. S., Johnson, J. E., Leslie, A. G. W., Rayment, I., Rossmann, M. G., Suck, D., and Tsukihara, T. (1980). Nature (London) 286, 33-39.

456

M. A. MAY0 AND V. ZIEGLERGRAFF

Atkins, J. F., Weiss, R. B., and Gesteland, R. F. (1990).Cell (Cambridge, Mass.) 62,413423. Bahner, I., Lamb, J., Mayo, M. A., and Hay, R. T. (1990). J. Gen. Virol. 71, 2251-2256. Barker, H. (1987).J. Gen. Virol. 68, 1223-1227. Barker, H. (1989). Ann. Appl. Biol. 115, 71-78. Barker, H., Reavy, B., Kumar, A., Webster, K. D., and Mayo, M. A. (1992). Ann. Appl. Biol. 120, 55-64. Barker, H., Reavy, B., Webster, K. D., Jolly, C. A., Kumar, A., and Mayo, M. A. (1993). Plant Cell Rep. 13, 54-58. Barker, H., Webster, K. D., Jolly, C. A., Reavy, B., Kumar, A., and Mayo, M. A. (1994). Mol. Plant-Microbe Interact. 7, 528-530. Barnett, A., Hammond, J., and Lister, R. M. (1981). J . Gen. Virol. 57, 397-401. Beier, H., Barciszewska, M., Krupp, G., Mitnacht, R., and Gross, H. J. (1984).EMBO J. 3, 351-356. Bevan, M. (1984). Nucleic Acids Res. 12, 8711-8721. Brault, V., and Miller, W. A. (1992). Proc. Natl. Acad. Sci. U.S.A.89, 2262-2266. Bredenbeek, P. J., Pachuk, C. J., Noten, A. F. H., Charit6, J., Lutyjes, W., Weiss, S. R., and Spaan, W. J. M. (1990).Nucleic Acids Res. 18, 1825-1832. Brierley, I., Boursnell, M. E. G., Binns, M. M., Bilimoria, B., Block, V. C., Brown, T.D. K., and Inglis, S. C. (1987). EMBO J. 6, 3779-3785. Casper, R. (1988). In “The Plant Viruses: Polyhedral Virions with Monopartite RNA Genomes” (R. Koenig, ed.), pp. 235-258. Plenum, New York. Cavener, D. R., and Ray, S. C. (1991). Nucleic Acids Res. 19, 3185-3192. Chin, L. S., Foster, J. L., and Falk, B. W. (1993). Virology 192,473-482. Clare, J. J., Belcourt, M., and Farabaugh, P. J. (1988).Proc. Natl. Acad. Sci. U S A . 86, 6816-6820. Commandeur, U., and Martin, R. (1993). Phytopathology 83, 1426. Creamer, R., and Falk, B. W. (1990). J. Gen. Virol. 71, 211-217. D’Arcy, C. J., Torrance, L., and Martin, R. R. (1989).Phytopathology 79, 869-873. Demler, S. A., and de Zoeten, G. A. (1991). J. Gen. Virol. 72, 1819-1834. Demler, S. A., Rucker, D. G., and de Zoeten, G. A. (1993).J. Gen. Virol. 74, 1-14. Demler, S. A. Borkhsenious, 0. N., Rucker, D. G., and De Zoeten, G. A. (1994). J. Gen. Virol. 76, 997-1007. Derrick, P. M., and Barker, H. (1992). Ann. Appl. Biol. 120, 451-457. Di, R., Dinesh-Kumar, S. P., and Miller, W. A. (1993). Mol. Plant-Microbe Interact. 6, 444-452. Dinesh-Kumar, S. P., and Miller, W. A. (1993). Plant Cell 5, 679-692. Dinesh-Kumar, S. P., Brault, V., and Miller, W. A. (1992). Virology 187, 711-722. Dinman, J. D., and Wickner, R. B. (1992). J. Virol. 66, 3669-3676. Dolja, V. V., and Carrington, J. C. (1992). Semin. Virol. 3, 315-326. Dolja, V. V., and Koonin, E. V. (1991). J . Gen. Virol. 72, 1481-1486. Duffus, J. E. (1977). In “Aphids as Virus Vectors” (K. F. Harris and K. Maramorosch, eds.), pp. 361-383. Academic Press, New York and London. Esau, K., and Hoefert, L. L. (1972). Virology 48, 724-738. Falk, B. W., Chin, I.-S., and Duffus, J. E. (1989). J. Gen. Virol. 70, 1301-1309. Francki, R. I. B., Fauquet, C. M., Knudson, D. L., and Brown, F. (1991). Classification and nomenclature of viruses. Fifth report of the International Committee on Taxonomy of Viruses. Arch. Virol. Suppl. 2, 450. Garcia, A., Vanduin, J., and Pleij, C. W. A. (1993).Nucleic Acids Res. 21, 401-406. Garcia-Bustos, J., Heitman, J., and Hall, M. N. (1991).Biochim. Biophys. Actu 1071,83101. Gibbs, A. J. (1987). J. Cell Sci. Suppl. 7, 319-317.

MOLECULAR BIOLOGY OF LUTEOVIRUSES

457

Gibbs, M. J. (1995).In “Molecular Basis of Viral Evolution” (A. J. Gibbs, C. H. Calisher, and F. Garcia-Arenal, eds.), Cambridge Univ. Press, Cambridge. In press. Gildow, F. E. (1987).Curr. Top. Vector Res. 4,93-120. Gildow, F. E. (1991).In “Aphid-Plant Interactions: Populations to Molecules” (D.C. Peters, J. A. Webster, and C. S. Chlouber, eds.), pp. 165-177.USDAIARS, Oklahoma State University, Stillwater, Oklahoma. Gildow, F. E. (1993).Phytopathology 83,270-277. Gildow, F. E., and Gray, S. M. (1993).Phytopathology 83, 1293-1302. Gill, C.C.,and Chong, J. (1979).Virology 95, 59-69. Gill, C. C., and Chong, J. (1981).Virology 114,405-413. Gorbalenya, A. E.,and Koonin, E. V. (1989).Nucleic Acids Res. 21,8413-8438. Gorbalenya, A. E.,Donchenko, A. P., Blinov, V. M., and Koonin, E. V. (1989).FEBS Lett. 243, 103-114. Grimsley, N. (1990).Physiol. Plant. 79, 147-153. Guilley, H., Wipf-Scheibel, C., Richards, K., Lecoq, H., and Jonard, G. (1994).Virology 202, 1012-1017. Habili, N., and Symons, R. H. (1989).Nucleic Acids Res. 17, 9543-9555. Harrison, B. D. (1984).CMI/AAB Descriptions of Plant Viruses. No. 291. Hatfield, D., and Oroszlan, S. (1990).Trends Biochen. Sci. 15, 186-190. Hermodson, M. A., Abad-Zapatero, C., Abdel-Meguid, S. S., Pundak, S., Rossmann, M. G., and Tremaine, J. H. (1982).Virology 119, 133-149. Herrbach, E., Lemaire, O., Zeigler-Graff, V., Lot, H., Rabenstein, F., and Bouchery, Y. (1991).Ann. Appl. Biol. 118, 127-138. Hewings, A. D., and DArcy, C. J. (1986).Phytopathology 76, 1270-1274. Higgins, D. G., Bleasby, A. J., and Fuchs, R. (1992).CABZOS 8, 189-191. Hizi, A., Henderson, L. E., Copeland, T.D., Sowder, R. C., Hixson, C. V., and Oroszlan, S. (1987).Proc. Natl. Acad. Sci. U.S.A. 84,7041-7045. Hu, J. S., Rochow, W. F., Palukaitis, P., and Dietert, R. R. (1988).Phytopathology 78, 1326-1330. Hull, R., and Salquero, S. (1991).In Classification and nomenclature of viruses. Fifth report of the International Committee on Taxonomy of Viruses Arch. Virol. Suppl. 2, 375-376. Imaizumi, S., and Kubo, S. (1980).Ann. Phytoputhol. SOC.Jpn. 46,54-56. Jacks, T.,and Varmus, H. E. (1985).Science 230, 1237-1242. Jacks, T.,Madhani, H. D., Masiarz, F. R., and Varmus, H. E. (1988a).Cell (Cambridge, Mass.) 55, 447-458. Jacks, T., Power, M. D., Marsiarz, F. R., Luciw, P. A., Barr, P. J., and Varmus, H. E. (1988b).Nature (London) 331, 280-283. Jefferson, R. A,, Kavanagh, T.A,, and Bevan, M. W. (1987).EMBO J. 6,3901-3907. Jolly, C. A. (1994).M.Sci. Thesis, University of Dundee, Scotland. Jolly, C. A., and Mayo, M. A. (1994).Virology 201, 182-185. Kamer, G., and Argos, P. (1984).Nucleic Acids Res. 12,7269-7282. Kawchuk, L.M.,Martin, R. R., and McPherson, J. (1990).Mol. Plant-MicrobeZnteract. 3, 301-307. Kawchuk, L. M., Martin, R. R., and McPherson, J. (1991).Mol. Plant-MicrobeZnteract. 4, 247-253. Keese, P., Martin, R. R., Kawchuk, L. M., Waterhouse, P. M., and Gerlach, W. L. (1990). J. Gen, Virol. 71, 719-724. Kelly, L.,Gerlach, W. L., and Gibbs, A. (1993).Abstr. ZXthZnt. Cong. Virol.,Glasgow, 329. Koonin, E. V., and Dolja, V. V. (1993).Crit. Rev.Biochem. Mol. Biol. 28,375-430. Kozak, M.(1989).J. Cell Biol. 108,229-241. Kubo, S. (1981).CMUAAB Descriptions of Plant Viruses. No. 234.

458

M. A. MAY0 AND V. ZIEGLERGRAFF

Kubo, S., and Takanami, Y. (1979).J. Gen. Virol. 42,387-398. Kujawa, A. B., Drugeon, G., Hulanicka, D., and Haenni, A. L. (1993).Nucleic Acids Res. 21,2165-2171. Leiser, R. M., Ziegler-Graff, V., Reutenauer, A., Herrbach, E., Lemaire, O., Guilley, H., Richards, K., and Jonard, G. (1992).Proc. Nutl. Acud. Sci. U.S.A.89,9136-9140. Lutcke, H. A., Chow, K. C., Mickel, F. S., Moss, K. A., Kern, H. F., and Scheele, G. A. (1987).EMEO J . 6,43-48. Marsh, L. E., Dreher, T. W., and Hall, T. C. (1987).Nucleic Acids Res. 16, 981-995. Martin, R. R., Keese, P. K., Young, M. J., Waterhouse, P. M., and Gerlach, W. L. (1990). Annu. Rev. Phytoputhol. 28, 341-363 Mayo, M. A. (1991).In Classification and nomenclature of viruses. Fifth report of the International Committee on Taxonomy of Viruses Arch. Virol. Suppl. 2, 400-402. Mayo, M. A., and Jolly, C. A. (1991).J. Gen. Virol. 72,2591-2595. Mayo, M. A., Barker, H., Robinson, D. J., Tamada, T., and Harrison, B. D. (1982).J. Gen. Virol. 59, 163-167. Mayo, M. A., Robinson, D. J., Jolly, C. A., and Hyman, L. (1989).J. Gen. Virol. 70,10371051. Miller, J. S. (1992).Ph.D. Thesis, University of Dundee, Scotland. Miller, J. S., and Mayo, M. A. (1991).J. Gen. Virol. 72, 2633-2638. Miller, W. A,, Waterhouse, P. M., and Gerlach, W. L. (1988).Nucleic Acids Res. 16,60976111. Miller, W. A., Hercus, T., Waterhouse, P. M., and Gerlach, W. L. (1991).Virology 183, 711-720. Miller, W. A., Dinesh-Kumar, S. P., and Paul, C. P. (1995).Crit. Rev. Plant Sci. 14, 179211. Morch, M. D., and Haenni, A. L. (1987).In “The Molecular Biology of the Positive Strand RNA Viruses” (D. J. Rowlands, M. A. Mayo, and B. W. J. Mahy, eds.), pp. 153-175. Academic Press, London. Moore, R., Dixon, M., Smith, R., Peters, G., and Dickson, C. (1987). J . Virol. 61,480-490. Murant, A. F. (1993).In “Diagnosis of Plant Virus Diseases” (R. E. F. Matthews, ed.), pp. 333-357. CRC Press, Boca Raton, Florida. Murant, A. F., Robinson, D. J., and Gibbs, M. J. (1995).In “Virus Taxonomy-The Classification and Nomenclature of Viruses: Sixth Report of the International Committee on Taxonomy of Viruses” (F. A. Murphy, C. M. Fauquet, D. H. L. Bishop, S. A. Ghabrial, A. W. Jarvis, G. P. Martelli, M. A. Mayo, and M. D. Summers, eds.), pp. 388391.Springer-Verlag, Vienna. Murphy, J. F., D’Arcy, C. J., and Clark, J. M. (1989).J . Gen. Virol. 70,2253-2256. Passmore, B. K., Sanger, M., Chin, L. S., Falk, B. W., and Bruening, G. (1993).Proc. Nutl. Acud. Sci. U.S.A.90, 10168-10172. Prill, B., Maiss, E., Katul, L., and Casper, R. (1990).Nucleic Acids Res. 18, 5544. Prufer, D., Tacke, E., Schmitz, J., Kull, B., Kaufmann, A., and Rohde, W. (1992).EMEO J . 11, 1111-1117. Randles, J., and Rathjen, J. P. (1995).In “Virus Taxonomy-The Classification and Nomenclature of Viruses: Sixth Report of the International Committee on Taxonomy of Viruses” (F. A. Murphy, C. M. Fauquet, D. H. L. Bishop, S. A. Ghabrial, A. W. Jarvis, G. P. Martelli, M. A. Mayo, and M. D. Summers, eds.), pp. 379-383. SpringerVerlag, Vienna. Rathjen, J. P., Karageorgos, L. E., Habili, N., Waterhouse, P. M., and Symons, R. H. (1994).Virology 198,671-679. Reavy, B., and Mayo, M. A. (1992).In “Plant Genetic Manipulation for Crop Protection” (A. M. R. Gatehouse, V. A. Hilder, and D. Boulter, eds.), pp. 183-214. CAB International, Wallingford, U.K.

MOLECULAR BIOLOGY OF LUTEOVIRUSES

459

Reutenauer, A. (1994).Etude par mutagenbse et agroinfection des fonctions des proteines codees par le virus de la jaunisse occidentale de la betterave. Ph.D. Thesis. Universite Louis Pasteur, Strasbourg. Reutenauer, A., Ziegler-Graff, V., Lot, H., Scheidecker, D., Guilley, H., Richards, K., and Jonard, G. (1993).Virology 105,692-699. Robertson, N. L., French, R., and Gray, S. M. (1991).J.Gen. Virol. 72, 1473-1477. Robinson, D. J., and Romero, J. (1991).J. Virol. Methods 34, 209-219. Rochow, W. F. (1970a).CMI/AAB Descriptions of Plant Viruses, No. 32. Rochow, W. F. (1970b).Science 167,875-878. Rochow, W. F., and Duffus, J. E. (1981).Zn “Handbook of Plant Virus Infections and Comparative Diagnosis” (E. Kurstak, ed.), pp. 147-170. Elsevier/North-Holland,Amsterdam. Rossman, M. G., and Johnson, J. E. (1989).Annu. Rev. Biochem. 58,533-573. Rost, B.,and Sander, C. (1993).J . Mol. Biol. 232, 584-599. Rost, B., and Sander, C. (1994).Proeins Struct. Funct. Genet. 19, 55-72. Samuel, C. E. (1989).Prog. Nucleic Acid Res. Mol. Biol. 37, 127-153. Sanger, M., Passmore, B., Falk, B. W., Bruening, G., Ding, B., and Lucas, W. J. (1994). Virology 200,48-55. Shepardson, S., Esau, K., and McCrum, R. (1980).Virology 105,379-392. Shepherd, R. J., Francki, R. I. B., Hirth, L., Hollings, M., Inouye, T., Macleod, R., Purcifull, D. E., Sinha, R. C., Tremaine, J. H., Valenta, V., and Wetter, C. (1976).Zntervirology 6, 181-184. Shukla, D. D., and Ward, C. W. (1989).Arch. Virol. 106, 171-200. Silver, S. L., Rasochova, L., Dinesh-Kumar, S. P., and Miller, W. A. (1994).Virology 198, 331-335. Skuzeski, J. M., Nichols, L. M., Gesteland, R. F., and Atkins, J. F. (1991).J. Mol. Biol. 218,365-373. Smith, 0. P., and Harris, K. F. (1990).Phytopathology 80,609-614. Tacke, E., Priifer, D., Salamini, F.,and Rohde, W. (1990).J.Gen. Virol. 71,2265-2272. Tacke, E., Priifer, D., Schmitz, J., and Rohde, W. (1991).J. Gen. Virol. 72, 2035-2038. Tacke, E., Schmitz, J., Priifer, D., and Rohde, W. (1993a).Virology 197, 274-282. Tacke, E., Priifer, D., Schmitz, J., and Rohde, W. (1993b).Abstr. 9th Int. Cong. Virol., Glusgow, 100. Tamada, T., and Kojima, M. (1977).CMI/AAB Descriptions of Plant Viruses. No. 179. Torrance, L. (1992).Virology 191,485-489. Ueng, P. P., Vincent, J. R., Kawata, E. E., Lei, C. H., Lister, R. M., and Larkins, B. A. (1992).J. Gen. Virol. 73, 487-492. Van den Heuvel, J. F. J. M., Verbeek, M., and Peters, D. (1993).Phytopathology 83,11251129. Van der Wilk, F., Huismann, M. J., Cornelissen, B. J. C., Huttinga, H., and Goldbach, R. \ (1989).FEBS Lett. 245, 51-56. Van der Wilk, F., Posthumus-Lutke Willink, D., Huisman, M. J., Huttinga, H., and Goldbach, R. (1991).Plant Mol. Biol. 17,431-439. Van der Wilk, F., Molthoff, J., van den Heuvel, J. F. J. M., Dekker, B., Goldbach, R., and Huttinga, H. (1993).Abstr. 9th Znt. Cong. Virol., Glasgow, 341. Veidt, I. (1991).Construction d‘un clone ADNc complet du virus des jaunisses occidentales de la betterave: Obtention d’un transcrit infectieux et etudes de ses propri6tRs messagbres. Ph.D. Thesis, Universite Louis Pasteur, Strasbourg. Veidt, I., Lot, H., Leiser, M., Scheidecker, D., Guilley, H., Richards, K., and Jonard, G. (1988).Nucleic Acids Res. 16, 9917-9932. Veidt, I., Bouzoubaa, S. E., Leiser, R. M., Zeigler-Graff, V., Guilley, H., Richards, K., and Jonard, G. (1992).Virology 186, 192-200.

M. A. MAY0 AND V. ZIEGLERGRAFF Vincent, J. R., Lister, R. M., and Larkins, B. A. (1991). J. Gen. Virol. 72, 2347-2355. Waterhouse, P. M., and Murant, A. F. (1983). Ann. Appl. Biol. 103,455-464. Waterhouse,P. M., Gerlach, W. L., and Miller, W. A. (1986).J. Gen. Virol. 67,1273-1281. Waterhouse, P. M., Gildow, F. E., and Johnstone, G. R. (1988). CMUAAB Descriptions of Plant Viruses. No. 339. 3 Waterhouse, P. M., Martin, R. R., and Gerlach, W. L. (1989). Phytopathology 79, 1215. Wen, F., and Lister, R. M. (1991). J. Gen. Virol. 72, 2217-2223. Wen, F., Lister, R. M., and Fattouh, F. A. (1991).J. Gen. Virol. 72, 791-799. Wezel, T.,Candresse, T., Macquaire, G., Ravelonandro, M., and Dunez, J. (1992).J.Virol. Methods 39, 27-37. Xiong, Z. Kim, K. H., Kendall, T. L., and Lommel, S. A. (1993). Virology 193, 213-221. Young, M. J., Kelly, L., Larkins, P. J., Waterhouse, P. M., and Gerlach, W. L. (1991). Virology 180, 372-379. Zerfass, K., and Beier, H. (1992). EMBO J . 11, 4167-4173.

INDEX

A

Antigen e, mediating host immune responses, 174-175

B Barley yellows dwarf virus heterologous encapsidation, 447 P5 protein, 432 RNA translation, 439-440 Beet western yellows virus agroinfection experiments, 425 heterologous encapsidation, 447 infection diagnosis, 451 P5 protein, 431-434 protein roles, deduction by mutagenesis, 434-435 replication, 449-450 ribosomal frameshift, 436-437 RNA translation, 439 subgenomic mRNA synthesis, 440-442 Borna disease virus, 333, 335-336 Bragg reflections, iridoviruses, 371

C

Capsid, see also Complementation system empty, poliovirus, 25-29, 55-56 iridoviruses, 367-369 mutations affecting RNA encapsidation, poliovirus, 47-52 myristylation, poliovirus, functional significance, 37-38 precursor, poliovirus, proteolytic cleavage, 44-47 Carrot red leaf virus, 447 Central nervous system involvement, cytomegalovirus, 217218 viral persistence, 335 Chickenpox, see Varicella-zoster virus

46 1

Chloriridovirus classification, 351, 363 hybridization complexes, 357, 359 Cleavage maturation, using complementation system, poliovirus, 52-53 proteolytic, capsid precursor, poliovirus, 44-47 Complementation system poliovirus encapsidation study, 39-53 assembly phenotypes, 42 capsid mutations, 47-52 affecting RNA encapsidation, 47-52 capsid precursor, proteolytic cleavage, 44-47 capsid proteins, 43-44 maturation cleavage, 52-53 nucleating role of RNA genome, 43 P1 precursors, 39, 42-43 Cucurbit aphid-borne yellows virus, subgenomic mRNA synthesis, 440-442 Cytomegalovirus, 197-241 human infection determinants, 198216 cell culture systems, 200-201 cell differentiation, 213 gene expression early phase, 210 late, 211 permissive culture cells, 199-212 IE2 protein functions, 204-205 immediate early gene location, 202204 initial events, 214-215 monoclonal antibodies, 206-207 noninfectious enveloped particles, 211-212 nonpermissive, 212-214 polypeptides, immunoblot analyses, 207-209 replication cycle, 199, 202 permissive culture cells, 199-212 strain variabilities, 216

462

INDEX

tegument proteins, 215 virion protein role in initiating infection, 215 infection hematopoietic system and circulating cells, 228, 231-232 latent, 236-241 cell culture models, 239-241 latency site, 238-239 murine, as model, 237-238 tissue cells, 219-220, 224-230 organ tropism, 216-219 spread and pathogenesis, cell types involved, 234-236 transmission modes, 232-234 Cytopathology, luteoviruses, 450 Cytoskeleton, viral interactions affecting cell function, 337-338 Cytotoxicity assays, varicella-zoster virus, 282-283

repetitive, iridoviruses, 390-391 replication, iridoviruses, 375-376 torsion-induced bends, HIV-1 reverse transcription, 115

E Encapsidation, see also Complementation system heterologous, luteoviruses, 447-448 RNA, 30-34 capsid mutations affecting, poliovirus, 47-52 hepadnaviruses, 176-180 requirements for, poliovirus, 31 signals, 32-33 subcellular location, 33-34 Enzymatic activities, iridoviruses, 385386 Epitopes, location, luteoviruses, 445

D

F Dazaifu IV genes, 388-389 repetitive DNA, 390 Defective interfering particles, poliovirus, 31-32, 40 Deoxynucleoside 5"-triphosphate, HIV-1 reverse transcriptase binding site, 116-1 19 DNA complementary, frameshifting, luteoviruses, 437-438 covalently closed circular formation, 191-192 hepadnaviruses, 170-172 invertebrate iridoviruses, restriction endonuclease profile, 354-357 methylation iridoviruses, 376-377 transcription, 380 polymerization, RNA- and DNAdependent, 107, 131-133 priming and synthesis, hepadnaviruses minus strand, 184-189 plus strand, 189-191 proviral, double-stranded, HIV-1, 101, 104

Failure to thrive, 333 Feline leukemia virus, 333 Frameshift, translational, 435-438 G

Genes, iridoviruses, 388-390 Glutamine-glycine dipeptide sites, cleavage, poliovirus, 12-13 Growth hormone deficiency syndrome, caused by lymphocytic choriomeningitis virus persistent infection, 324-333 Guinea pig model, varicella-zoster virus, cell-mediated immunity, 303-306

H Hepadnaviral polymerase, 180-183 experimental approaches, 180-181 minus strand priming, 182-183 mutational analyses, 181-182 sequence similarities, 181

Hepadnaviruses, 167-192, see also Hepatitis B virus DNA, covalently closed circular, 170172 life cycle, 170-172 protein products, 174-176 reverse transcription, 172, 184-192 covalently closed circular DNA formation, 191-192 minus strand DNA priming and synthesis, 184-189 plus strand DNA priming and synthesis, 189-191 pregenomic RNA organization, 184 RNA packaging, 186 RNA, encapsidation cis-acting signals on pregenomic RNA, 177-180 core particle assembly, 176-177 missense mutations, 177 surface antigen forms, 175-176 transcripts, 173-174 virion, structure, 170-171 Hepatitis B virus, 167-168, see also Hepadnaviruses genome organization, 173 historical background, 168-170 minus strand priming, 182-183 mutational analyses, 181-182 open reading frames, 174-176 P protein expression, 187 transfections, 186 Herpesvirus, see Varicella-zoster virus Human foamy virus, 100 Human immunodeficiency virus type 1, 99-146 DNA double-stranded proviral, 101, 104 significance in infectiousness, 123124 LTR, 104 proteolytic processing, 106 replication, 101-106 initial events, 101, 104-105 reverse transcription first template switch, 137-146 efficiency, 142 genomic RNA dimerization, 143 interstrand and intrastand, 145 mode, 145

oligoribonucleotide size, 141 RNA-dependent DNA polymerization, 140 RNase H digestion, 141 origin, 122-125 RNA genomic, interaction with tRNAitLys3, 127-130 templates, 129 trans-activation response element, 104 virus assembly and maturation, 102103, 105-106 virus-host cell membrane fusion, 101 Human immunodeficiency virus type 1 reverse transcriptase, 107-119 association constants, DNA primerDNA template, 131 DNA polymerization, RNA- and DNAdependent, 131-133 template-DNA primer duplex, 114115 torsion-induced bends, 115 160itgag-pol precursor, 124 heterodimer, 108-109 hydroxyl-radical footprinting, 113-114 interaction with primer and template, 113-116 polymerase active site, 113 specific binding, 115-116 nuclease footprinting, 114 polymerase active site and deoxynucleoside 5-triphosphate binding site, 116-119 polymerization by, fidelity, 133-134 recombinant, 108 ribonucleases, 134-137 structure, 109-113 dimerization sites, 112 functional domains, 110-111 p51 and p66 domains, 109, 112 Human T-cell lymphotropic virus, 100 Hybridization, RNA, rotaviruses, 72-73

I Immunocompromised hosts, cytomegalovirus infection, 218, 220223

464

INDEX

Iridoviruses, 347-401 capsid, 367-369 classification, 350-366 alternative approaches, 360-361 comparative studies, 352-359 current system, 350-352 invertebrate, 362-364 new nomenclature, 359-360 new scheme, 365-366 suggested changes to current system, 362-366 vertebrate, 364-365 description, 348-349 DNA-DNA dot-blot hybridization values, 356-358 ecology, 391-399 future directions, 399-401 genes, 388-390 hybridization complexes, 357, 359 infectious particles per host, 392 invertebrate classification, 362-364 DNA restriction endonuclease profile, 354-357 iridescence phenomenon, 371-372 lipid membrane, 361, 368-370 models of host-iridovirus population dynamics, 391 molecular biology, 386-391 particle core, 370-371 persistence alternative hosts, 398-399 in host populations, 396-397 physical, 395 physicochemical properties, 366-367 repetitive DNA, 390-391 replication cell penetration and uncoating, 373 cytoskeletal manipulation, 383-385 DNA, 375-376 enzymatic activities, 385-386 host macromolecular synthesis shutdown, 373-375 methylated DNA transcription, 380 methylation of DNA, 376-377 mRNA stability and methylation, 380-381 nongenetic reactivation, 377-380 transcription, 377-381 translation, 381-382 virion packaging, 382-383

transmission, 393-395 vertebrate, 348 classification. 364-365

L Lentiviruses, 100 Lipid membrane, iridoviruses, 361, 368370 Luteoviruses, 415-457 gene expression, 435-444 internal initiation of translation, 438-440 proteolysis and cap-independent translation, 443-444 readthrough, 442-443 subgenomic mRNA synthesis, 440442 translational frameshifting, 435-438 gene function determination, 424-425 genome structure, 417-424 open reading frame arrangement, 417-420 putative recombination, 424 terminal structures and noncoding regions, 422-423 variation among coding sequences, 420-422 variation among strains, 423 infection diagnosis, 450-452 open reading frames 5 and 6, 431-434 particle structure, 444-448 epitope location, 445 heterologous encapsidation, 447-448 particle assembly determinants, 448 readthrough protein, 445-447 tertiary, 444-445 PO protein, 426-427 P1 and P2 proteins, 427-428 P3 protein, 428-429 P4 proteins, 429-431 P5 proteins, 431-434, 445-447 replication, 449-450 resistance to, 452-453 taxonomy barley yellows dwarf virus, satellite RNA, 456-457 pea enation mosaic virus problem, 455-456 speciation, 453-454

INDEX ST9-associated satellite-like RNA, 456 structure, 454 Lyrnphocystiuirus,classification, 351 Lymphocytic choriomeningitis virus, persistent infection, growth hormone deficiency syndrome, 324-333

M Memory T-lymphocyte responses, varicella-zoster virus, 285-286, 290300 natural infection, 291-296 varicella vaccine, 296-300 Methylation DNA iridoviruses, 376-377 transcription, 380 mRNA, 380-381 Microfilaments, cytoskeletal manipulation, iridoviruses, 384 Microtubules, cytoskeletal manipulation, 383-384 Moloney murine leukemia virus, 333 Mosquitoes, iridoviruses transmission, 393 Murine cytomegalovirus, as model system, 237-238 Myristylation, poliovirus capsid proteins, 18-19 functional significance, 37-38

N Neuroendocrine dysfunctions, virusinduced, in absence of cytolysis and inflammation, 333-337 Nongenetic reactivation, iridoviruses, 377-380

P Particle assembly, determinants, luteoviruses, 448 Pathogenesis cytomegalovirus, 234-236 new perspective, 314

465

Pea enation mosaic virus genome organization, 455 replication, 449 taxonomy, 455-456 Pentamer, 14S, 22-24 poliovirus, 55 Phloem tissue, luteovirus tissue, 449450 Picornuviridae,2, see also Poliovirus Poliovirus, 1-56 arginine residues, 50-51 assembly pathways, 20-21 assembly phenotypes, 53-54 assembly process, 34-38, 53-56 capsid myristylation functional significance, 37-38 P1 and 3CD expression using recombinant vaccinia virus vectors, 35-37 using recombinant vaccinia viruses, 34-35 f3-barrel, 16-17 f3 strands, 16-17 capsid cavity associated with, 50 empty, 25-29, 55-56 protein-RNA binding, 51 cascade of polyprotein processing, 3-4 coding portion, 5 genomic organization, 3-6 life cycle, 6-14 events, 8 protease 2Apr0 release, 11-12 protease 3Cpr0 release, 12-13 RNA replication, 13-14 translation, 10-11 virus entry and uncoding, 7, 9-10 morphogenesis, 19-30 empty capsid, 25-29 55 protomer, 20-22 provirion, 29-30 1 4 s protomer, 22-24 5”-NTR, 4-5 1 4 s pentamers, 55 proteolytic cleavages, 5-6 recombination, 85 RNA, encapsidation, 30-34 defective interfering particles, 3 1-32 RNA requirements, 31 signals, 32-33

studies, complementation system, 39-53 subcellular location, 33-34 RNA-protein interactions, 49-50 serological types, 2 subgenomic replicon, 52 virion, 14-19 capsid protein myristylation, 18-19 properties, 15-16 structure, 16-18 VP1, amino-terminal portion, 47-49 Polymerization DNA,RNA- andDNA-dependent,131133 by HIV-1 reverse transcriptase, fidelity, 133-134 Potato leafroll virus infection diagnosis, 451 particle assembly determinants, 448 P4 protein, 430-431 replication, 449-450 resistance to, 452 RNA translation, 439-440 subgenomic mRNA synthesis, 440-442 Procapsid, poliovirus, 25-29 Protease 2Apr0, release, poliovirus, 11-12 3CD, P1 precursor cleavage, 35-36 30'0, release, poliovirus, 12-13 Proteins, see also Luteoviruses capsid interaction with poliovirus RNA, 51-52 poliovirus myristylation, 18-19 targeted, 43 VP4, poliovirus, 9 RNA interactions, poliovirus, 49-50 TATA-binding, cytomegalovirus, 205 varicella-zoster virus, putative functions, 267 Proteolysis, cap-independent translation and, luteoviruses, 443-444 Proteolytic processing, HIV-1, 106 Protomer, 5S, poliovirus, 20-22 Provirion, poliovirus, 29-30

R Rabies virus, 333, 335 Ranavirus, classification, 351, 364-365

Readthrough, protein presence, luteoviruses, 445-447 Recombinant vaccinia virus, poliovirus assembly process studies, 34-35 vectors, P1 and 3CD expression, 35-37 Reouiridae, see also Rotaviruses genome rearrangements, 92 Replication, see also Iridoviruses HIV-2 initial events, 101, 104-105 virus assembly and maturation, 102-103, 105-106 luteoviruses, 449-450 nonlytic strategy, viral persistence, 321-322 RNA, poliovirus, 7, 13-14 Retrovirus reverse transcription scheme, 119-122 species, 100 type D, human immunodeficiency virus type 105-106 Reverse transcriptase, 100, see also Human immunodeficiency virus type 1 reverse transcriptase Reverse transcription hepadnaviruses, see Hepadnaviruses HIV-1 first template switch, 137-146 host tRNAitLys3 primer, 125-131 origin, 122-125 sites of hypermutability and pausing, 134 scheme, retroviruses, 119-122 Rhinovirus, provirion, 29 Ribonucleases, HIV-1 reverse transcriptase, 134-137 RNA encapsidation, 30-34 capsid mutations affecting, poliovirus, 47-52 hepadnaviruses, 176-180 requirements, poliovirus, 31 signals, 32-33 subcellular location, 33-34 frameshift, luteoviruses, 436-437 hybridization, rotaviruses, 72-73 luteovirus, putative recombination, 424 messenger genomic and subgenomic, hepadnaviral transcripts, 173-174

467

INDEX methylation and stability, 380-381 subgenomic synthesis, luteoviruses, 440-442 translation, poliovirus, 6 poliovirus, interaction with capsid proteins, 51-52 pregenomic cis-acting signals, hepadnaviruses, 177-180 organization, hepadnaviruses, 184 profiles, rotaviruses, 72-74 protein interactions, poliovirus, 49-50 replication, poliovirus, 7, 13-14 satellite, barley yellow dwarf virus, 456-457 ST9-associated satellite-like, luteoviruses, 456 transfer isoacceptor species, 125-127, 129 primer, retroviruses, 121 tRNAitLys3 primer, HIV-1 reverse transcription, 125-131 translation, poliovirus, 10-1 1 RNA polymerase, RNA-dependent, poliovirus, 5 Rotaviruses, 71-93 biophysical data, 86-87 genome rearrangements discovery, 71-74 duplication, 83 evolution, 91-92 extent, 75 in vitro in cultured cells, 79-82 mechanisms, 82-86 other genera of Reouiridue, 92 RNA, segments 5 and 6,89-90 sequence data, 75-79 groups, 71 in uitro growth properties, 89 rearranged genes, function, 86-91 RNA hybridization, 72-73 profiles, 72-74 plaque-purified, 79-80 second generation genome rearrangements, 82-83 segment 5, normal and rearranged forms, 77-78 standard gene 5, 77-78 3 UTR, 75,83 variants, 77

S

Sedimentation coefficient, poliovirus virion, 15-16 Severe combined immunodeficiency, rotaviruses infections, 71

T Tipula oleracae, iridoviruses transmission, 394 T-lymphocyte proliferation antigen-specific, varicella-zoster virus, 281 assay tests, varicella-zoster virus, 282 T lymphocytes, varicella-zoster virus tropism, severe combined immunodeficient hu mouse, 276-280 Trans-activation response element, HIV-1,104 Transcription, iridoviruses, 377-381 methylated DNA, 380 Translation iridoviruses, 381-382 luteoviruses cap-independent, proteolysis and, 443-444 internal initiation, 438-440 leaky termination, 442-443 poliovirus mRNA, 6 RNA, 10-11 Transmission, iridoviruses, 393-395

V Varicella vaccine maintaining cell-mediated immunity, 302-303 memory T-lymphocyte responses, 296300 primary cell-mediated immune response, 285-290 Varicella-zoster virus, 265-306 cell-mediated immune response, 280306 assessment methods, 281-283 guinea pig model, 303-306 long-term immunity, 293

468

INDEX

maintenance mechanisms, 300-303 memory T-lymphocyte, 285-286, 290-298 primary, 283-290 natural infection, 284-285 varicella vaccine, 285-290 disease correlations with cell-mediated immunity, 272-273 gene sequences, 268-269 infection peripheral blood cells in uitro, 273275 primary, viremia, 267-271 medical significance of related disease, 266 pathogenesis, 271 proteins putative functions, 267 recognized by T-lymphocytes, 291292 reactivation maintaining cell-mediated immunity, 300-302 viremia, 271-273 reexposure, maintaining cell-mediated immunity, 300 subclinical viremia, relationship between episodes, 305 tropism, T lymphocytes, severe combined immunodeficient hu mouse, 276-280 Venezuelan encephalitis virus, 334 Viral persistence, 313-339 iridoviruses alternative hosts, 398-399 in host populations, 396-397 physical, 395

requirements avoidance of recognition of infected cells by specific immune response, 316-318 block of action of nonspecific antiviral defense mechanisms on infected cells, 318-319 essential, 315 induction of suppression of host immune response, 319-321 nonlytic strategy of replication, 321322 virus-induced changes in cells, 316321 virus-induced alterations, host cellular differentiated functions in absence of cytolysis, 323-338 experimental evidence, 323-324 growth hormone deficiency syndrome, 324-333 neuroendocrine dysfunctions, 333337 viral interactions with cytoskeleton, 337-338 Viremia during primary varicella-zoster virus infection, 267-271 varicella-zoster virus reactivation, 271-273 Virions, iridoviruses, packaging, 382383

W Wasting syndrome, 333

This Page Intentionally Left Blank

I S B N O-L2-037846-X