key: cord-023705-3q9yr6np
authors: FENNER, FRANK; BACHMANN, PETER A.; GIBBS, E. PAUL J.; MURPHY, FREDERICK A.; STUDDERT, MICHAEL J.; WHITE, DAVID O.
title: Viral Replication
date: 2014-06-27
journal: Veterinary Virology
DOI: 10.1016/b978-0-12-253055-5.50008-6
sha: 
doc_id: 23705
cord_uid: 3q9yr6np

Viral replication is the central focus of much experimental virology and is a significant part of molecular biology. Studies with bacteriophages in their prokaryotic host cells in the 1940s and 1950s provided the first insights into the complexities of viral replication. With the development of mammalian cell culture procedures, the techniques used for the study of bacteriophages were adapted to animal viruses. Progress has been such that the basic mechanisms of transcription, translation, and nucleic acid replication have been characterized for all the major families of animal viruses and the strategy of gene expression and its regulation clarified. Many important biochemical phenomena such as the splicing and other types of posttranscriptional processing of RNA, the posttranslational cleavage and glycosylation of proteins, the replication of RNA, reverse transcription, integration, and the transposition of viral genes and cellular oncogenes were first elucidated by virologists and have general application in cell biology. The chapter provides a general overview on viral replication for understanding pathogenesis, immunity, chemotherapy, and the role of viruses in cancer.

Viral replication is the central focus of much experimental virology and a significant part of molecular biology. Studies with bacteriophages in their prokaryotic host cells in the 1940s and 1950s provided the first insights into the complexities of viral replication. With the development of mammalian cell culture procedures (see Chapter 3), the techniques used for the study of bacteriophages were adapted to animal viruses. Progress has been such that the basic mechanisms of transcription, translation, and nucleic acid replication have been characterized for all the major families of animal viruses and the strategy of gene expression and its regulation clarified. Many important biochemical phenomena, such as splicing and other types of posttranscriptional processing of RNA, posttranslational cleavage and glycosylation of proteins, replica tion of RNA, reverse transcription, integration, and transposition of viral genes and cellular oncogenes, were first elucidated by virologists and have general application in cell biology.

Our knowledge of viral replication is now very detailed and is expand ing rapidly. Every viral family has a different strategy of replication, and for each family several reviews have been published since 1980. It is neither possible nor appropriate to deal comprehensively with the sub ject in this book. This chapter provides a general overview; some addi tional information on particular viral families is provided in Part II. An understanding of viral replication provides a basis for understanding pathogenesis, immunity, chemotherapy, and the role of viruses in cancer.

Following the pattern established in experiments with bacteriophages, studies of the replication of animal viruses began with the onestep growth experiment. In such experiments, all cells in a culture are infected simultaneously, i.e., at high multiplicity of infection. Unadsorbed input virus is removed or neutralized, usually after 1 hour, and the increase in infectious virions over time is followed by titrating cell-free and cell-associated infectivity. Shortly after infection, the inoculated virus "disappears"; infectious particles cannot be demonstrated, even intracellularly. This eclipse period continues until the first pro geny virions become detectable some hours later. Nonenveloped viruses mature within the cell and are detectable for some hours as intracellular virions before they are released by cell lysis. Many enveloped viruses, on the other hand, mature by budding from the plasma membrane and are thus immediately released into the medium. The eclipse period ranges from 5 to 15 hours for the various DNA viruses and from 3 to 10 hours for RNA viruses (see Table 4 -2).

Early studies, relying on quantitative electron microscopy and assay of infectivity, provided a limited amount of information about the early and the late events in the replication cycle (attachment, penetration, intracellular maturation, and budding) but could not tell us anything about what happened during the eclipse period. Investigation of the expression and replication of the viral genome became possible only with the development of molecular methods, and during the last two decades all the sophisticated techniques of molecular biology have been applied to this problem. General features of the viral replication cycle, using a nonenveloped icosahedral DNA virus as a model. No topographical location for any step is implied. One step grades into the next such that, as the cycle progresses, several of these processes are proceeding simultaneously. Release occurs by cell lysis. principally viral structural proteins; some of these are subject to posttranslational modification, such as glycosylation and/or cleavage. As sembly of icosahedral virions occurs in the nucleus or cytoplasm, de pending on the particular family. Enveloped viruses are completed by "budding" through cellular membranes. Each infected cell yields sever al thousand new virions over a period of several hours. 

Many of the processes that can be investigated by morphological stud ies and infectivity assays differ according to viral family (Table 4 -2). There are three principal methods of penetration, and virions may be released by cell lysis or by budding. Some viruses acquire an envelope by budding through plasma membrane, others through nuclear mem brane, and still others in the Golgi complex or the endoplasmic reticulum. Some viruses shut down the synthesis of cellular macromolecules very effectively, whereas others do not. Indeed, some viruses are noncytocidal and others actually induce the cell to divide, or even trans form it to a tumor cell (see Chapter 6).

Even more significant are the differences in the strategy of expression of the viral genome. Under this heading are subsumed the key processes occurring during the eclipse period: transcription and processing of viral mRNA ( Fig. 4 -2, steps 4 and 7), translation and processing of viral pro teins (steps 5 and 8), and replication of the viral nucleic acid (step 6). Before discussing them, we will describe the earlier events: attachment (step 1), penetration (step 2), and uncoating (step 3).

Because virions and cells are both negatively charged at physiological pH, they tend to repel one another, but random collisions do occur and initial (reversible) attachment may be facilitated by cations. Firm binding requires the presence of specific receptors for the virus on the plasma membrane, to which specific molecules on the surface of the virion While there is some specificity about the binding of virions to particular cellular receptors, several different viruses may utilize the same receptor.

Electron microscopic and other data show that virions can enter cells by at least three different mechanisms: endocytosis, fusion, and translocation. The majority of virions entering a cell fail to initiate infection, many virions taken up by endocytosis being degraded by lysosomal enzymes. However, for some viruses this may be the normal route of penetration, leading to uncoating and productive infection.

The majority of mammalian cells are continuously engaged in receptormediated endocytosis, a specific process for the uptake of essential macromolecules. Viruses may use receptor-mediated endocytosis to initiate infection (Plate 4-1). Following attachment to receptors, virions move down into coated pits. These pits, coated with clathrin, fold inward to produce coated vesicles that enter the cytoplasm and fuse with a lysosome to form a phagolysosome. With enveloped viruses, the envelope of endocytosed virions fuses with the lysosomal membrane, releasing the viral nucleocapsid into the cytoplasm. In this way a virion can be uncoated by a lysosome but escape total degradation by the lysosome's hydrolytic enzymes. Recent studies with influenza virus have identified a pH 5-mediated conformational change in the hemagglutinin molecule which enables fusion to occur between the viral envelope and the mem brane of the phagolysosome.

The F (fusion) glycoprotein of paramyxoviruses, in its cleaved form, enables the envelope of these viruses to fuse directly with the plasma membrane, even at pH 7. This may allow the nucleocapsid to be re leased directly into the cytoplasm. Although a number of other enve loped viruses display a capacity to fuse cells or to lyse erythrocytes, it is not clear whether this is the normal way in which they infect cells.

Some nonenveloped icosahedral viruses appear to be capable of pass ing directly through the plasma membrane. [A, from E. Fries and A. Helenius, Eur. J. Biochem. 8, 213 (1979) ; B, from K. Simons et al., Sci. Am. 246, 46 (1982) , Courtesy Dr. A. Helenius.]

In order that at least the early viral genes may become available for transcription, it is necessary that the virion be at least partially uncoated. With viruses that enter the cell by fusion of their envelope with either the plasma membrane or the membrane of a phagolysosome, the nucleocapsid is discharged directly into the cytoplasm. In the case of vi ruses with helical nucleocapsids, transcription begins from viral RNA while it is still associated with nucleoprotein. In the case of the icosahedral reoviruses only certain capsid proteins are removed and the viral genome expresses all its functions, even though it is never fully released from the core ("subviral particle"; Plate 4-2). Poxviruses are uncoated in two stages: first, to a core, from which half the genome is transcribed; then completely, following the synthesis of a virus-coded uncoating protein. With the picorna viruses, the process of attachment of the virion to the cell leads to a conformational change in the capsid, . Reovirus "cores" that have synthesized mRNA for 8 minutes at 37°C were prepared for electron microscopy by the Kleinschmidt technique, stained with uranyl acetate, and shadowed at a low angle with platinum-palladium, showing the fine fibrils of mRNA being extruded from the cores or occurring free around them. The results of polyacrylamide gel electrophoretic analysis of such mRNA molecules at various times during the replication cycle are illustrated in Fig. 4-6. [From N. M. Bartlett, S. C Gillies, S. Bullivant, and A. R. Bellamy, J. Virol. 14, 315 (1974), courtesy Dr. A. R. Bellamy.] resulting in the loss of capsid proteins VP4 and VP2 and rendering the particle susceptible to proteases; the attachment step itself triggers the process of uncoating. For some viruses that replicate in the nucleus there is evidence that the later stages of uncoating occur there, rather than in the cytoplasm.

The key events in viral replication are the synthesis of viral proteins, the replication of the viral genome, and the assembly of the new compo nents into virions. To synthesize viral proteins, viral mRNAs must be produced in a form capable of being recognized and translated on cel lular ribosomes. Eukaryotic cells synthesize their own mRNA in the nucleus by transcription of the cellular DNA followed by processing of the transcript. They lack the enzymes necessary for synthesizing mRNA off a viral RNA genome and they cannot transcribe viral DNA located in the cytoplasm. Therefore, only those DNA viruses that replicate in the nucleus utilize the cellular machinery for transcription. All other viruses provide their own enzymes to produce mRNAs. Eukaryotic cells have a further constraint, namely, that the protein-synthesizing machinery ap parently cannot recognize internal initiation sites within polycistronic mRNAs. Hence viruses must synthesize a separate (monocistronic) mRNA corresponding to each gene in their genome, or, alternatively, a polycistronic mRNA must be translated into a large precursor "polyprotein" which is then cleaved into individual proteins.

The diverse strategies followed by viruses of different families for transcription and translation are illustrated diagrammatically in Fig. 4 -3 (for DNA viruses) and Fig. 4 -5 (for RNA viruses). Necessarily, the pro cesses summarized in these figures and the descriptions of them involve major oversimplifications. We will describe in turn transcription, trans lation, and replication of the viral nucleic acid.

The viral RNA of (+) sense ssRNA viruses binds directly to ribosomes and is translated in full or in part without the need for any prior transcriptional step. With all other classes of viral genomes, mRNA must be transcribed. In the case of DNA viruses that replicate in the nucleus, the cellular DNA-dependent RNA polymerase II performs this function. All other viruses require unique and specific transcriptases which are virus-coded and are an integral component of the virion. Cytoplasmic dsDNA viruses carry a DNA-dependent RNA polymerase, whereas dsRNA vi ruses have dsRNA-dependent RNA polymerase, and (-) sense ssRNA viruses carry a ssRNA-dependent RNA polymerase (see Tables 2-1 and 2-2).

For all DNA viruses, mRNA must be transcribed by a DNA-depen dent RNA polymerase. Transcription of the viral DNA is programmed such that not all genes are expressed simultaneously or continuously throughout the replication cycle. Particular parts of the genome are tran scribed in sequence, the so-called early genes first, and the late genes later in the cycle. Viruses of different families differ according to whether a cellular or a viral transcriptase is employed, correlating with a nuclear or cytoplasmic site of replication. There are four classes of strategy of ex pression of the viral genome ( Fig. 4 -3A-D), described below.

dsDNA; Cellular Transcriptase ( Fig. 4-3A) . This group comprises the papovaviruses, adenoviruses, and herpesviruses, and has in one respect the most straightforward strategy: the viral DNA is transcribed within the nucleus by a cellular DNA-dependent RNA polymerase. There are at least two temporally separated cycles for adenoviruses and herpesviruses; in each instance the structural proteins of the virion are made from mRNAs produced in the last cycle of transcription. Polycistronic but subgenomic RNA transcripts (corresponding to several genes but less than the whole genome) undergo cleavage and splicing to produce monocistronic mRNAs, introns being removed in the process. dsDNA; Virion Transcriptase ( Fig. 4-3B) . The poxviruses and African swine fever virus, which replicate in the cytoplasm, carry their own transcriptase. It appears that monocistronic mRNAs are transcribed di rectly from the viral DNA. There are at least three cycles of transcription. The transcripts are translated directly into proteins, some of which need to undergo posttranslational cleavage to yield functional molecules. ssDNA; Cellular Transcriptase ( Fig. 4-3C) . The (-) sense ssDNA of the parvoviruses requires the synthesis of a complementary strand to form dsDNA; this is then transcribed in the nucleus and the transcripts are processed to produce mRNAs, before export to the cytoplasm for translation. a virion-associated DNA polymerase, and the DNA then converted into a supercoiled dsDNA. Transcription of mRNA by cellular RNA poly merase II then occurs.

Expression of a DNA Genome. Analysis of the 5224-bp sequence of the circular dsDNA of the papovavirus SV40 and its transcription pro gram have provided insights into these processes ( Fig. 4-4) . The follow ing points should be noted:

1. The early genes and the late genes are transcribed by the host cell RNA polymerase II in opposite directions, from different strands of the DNA.

2. Certain genes overlap and are translated in the same frame, so that their protein products have some amino acid sequences in common.

3. Some regions of the viral DNA are read in overlapping but different reading frames, so that two completely different amino acid sequences are obtained.

4. At least 15% of the viral DNA consists of intervening sequences (introns), which are transcribed but not translated, because they are excised from the primary transcript.

5. Up to three distinct proteins can be produced from mRNAs derived from a primary transcript by different splicing protocols.

Studies with adenoviruses have eluci dated the mechanisms that regulate the expression of viral genomes, which operate principally, but not exclusively, at the level of transcrip tion. Because of the complications arising from posttranscriptional cleav age of mRNA and posttranslational cleavage of precursor proteins in eukaryotic cells, it is no longer adequate to talk of a "gene" and its "geneproduct." More appropriate perhaps is to think in terms of a transcription unit, i.e., that region of the genome beginning with the transcription initiation site, extending right through to the transcription termination site, and including all introns and exons in between. "Simple" transcrip tion units may be defined as those encoding only a single protein, whereas "complex" transcription units code for more than one. There are many adeno virus transcription units. At different stages of the viral replication cycle-"pre-early," "early," "intermediate," and "late"the various transcription units are transcribed in a given temporal se quence. A product of the early-region EIA induces the other early regions including E IB, but following viral DNA replication, there is a 50-fold increase in the rate of transcription from the major late promoter relative to early promoters such as E IB, and a decrease in EIA mRNA levels. A second control operates at the point of termination of transcription.

Transcripts that terminate at a particular point early in infection are read through this termination site later in infection to produce a range of longer transcripts with different polyadenylation sites.

Processing of RNA Transcripts. Primary RNA transcripts from eukaryotic DNA are subject to a series of posttranscriptional alterations in the nucleus, known as processing, prior to export to the cytoplasm as mRNA. First, a cap, consisting of 7-methylguanosine (m 7 Gppp) is added to the 5' terminus. The function of this poly (A) tail is uncertain, but it may act as a recognition signal for processing and for transport of mRNA from the nucleus to the cytoplasm, and it may stabilize mRNA against degradation in the cytoplasm. Third, a methyl group is added at the 6 position to about 1% of the adenylate residues throughout the RNA (methylation). Fourth, introns are removed from the primary tran script and the exons are linked together in a process known as splicing; the precise mechanism is not known but may involve excision of the introns by endonucleases, followed by ligation. Splicing is an important mechanism for regulating gene expression in nuclear DNA viruses. A given RNA transcript can have two or more splicing sites and be spliced in several different ways to produce a variety of mRNA species coding for distinct proteins; both the preferred poly(A) site and the splicing pattern may change in a regulated fashion as infection proceeds. The rate of degradation of mRNA provides another level of regulation. Not only do different mRNA species have different half-lives, but the halflife of a given mRNA species may change as the replication cycle pro gresses.

Transcription is more complicated for the RNA viruses than for rhe DNA viruses, which is perhaps not surprising, since they are the only forms of life that utilize RNA as the repository of genetic information. There are, broadly, three main strategies: (1) the virion RNA of most viruses with (+) sense RNA is itself infectious, because it functions as mRNA, (2) viruses with (-) sense ssRNA, or with dsRNA, carry a virion-associated RNA-dependent RNA polymerase which transcribes mRNA from the viral RNA, and (3) the (+) sense virion RNA of retroviruses is transcribed into DNA, which serves as a template for tran scription of viral mRNAs by a cellular transcriptase. These three general strategies can be further subdivided on the basis of more subtle dif ferences to give seven groups ( Fig. 4 -5A-G).

ssRNA; (+) Sense ( Fig. 4-5A ,B,C). In these groups the (+) sense virion RNA is itself infectious. In the picornaviruses and flaviviruses the ge- nome, acting as a single polycistronic mRNA, is translated into a single polyprotein which is subsequently cleaved to give the individual viral polypeptides (Fig. 4-5A) . Togaviruses of the genus Alphavirus also con tain a single polycistronic (+) sense ssRNA molecule, but only about two-thirds of the viral RNA (the 5' end) is translated; the resulting polyprotein is cleaved into four nonstructural proteins, two of which form the RNA polymerase. This enzyme then copies a full-length (-) sense strand, from which two species of (+) sense strand are copied: full-length virion RNA destined for encapsidation, and a one-third length RNA, which is colinear with the 3' terminus of the viral RNA and is translated into a polyprotein from which three or four structural pro teins are produced by cleavage. The caliciviruses have not been so exten sively studied, but also produce both genome-length and subgenomic mRNA species.

Flaviviruses were recently accorded the status of a family separate from the toga viruses. They do not produce subgenomic mRNAs, and translation of the (+) sense virion RNA initiates with the capsid protein near the 5' end of the genome and proceeds sequentially through the genome to produce one precursor polyprotein. This is rapidly cleaved during the process of translation, so that the complete polyprotein is never seen.

Corona viruses have a unique strategy. Initially, in a step about which little is known, part of the virion RNA acts as mRNA and is translated to produce an RNA polymerase, which then synthesizes a genome-length (-) sense strand. From this, a "nested set" of overlapping subgenomic RNAs is transcribed, of which only the unique (nonoverlapping) se quence in each is translated (see Chapter 28).

ssRNA; (-) Sense; Virion Transcriptase ( Fig. 4-5D,E) . Primary tran scription from the (-) sense ssRNA viruses occurs in the cytoplasm, when the virion RNA is still within the helical nucleocapsid, in associa tion with the nucleoprotein as well as the transcriptase. Particular se quences of 10 to 20 nucleotides, located at or near the termini of each RNA molecule, may serve as recognition signals for transcriptase binding.

The paramyxoviruses and rhabdoviruses have similar transcription strategies, as well as similar consensus sequences at the 3' and 5' termini of their viral RNA, suggestive of a common ancestry. The (-) sense virion RNA is copied in two distinct ways: the replication mode and the transcription mode. Copying in the replication mode produces a fulllength (+) sense strand which is used as a template for the synthesis of new virion RNA. In the transcription mode, five subgenomic (+) sense RNAs are produced; each is capped and polyadenylated and serves as a monocistronic mRNA. It is still not certain what dictates whether the polymerase reads right through from 3' to 5' end of the (-) sense RNA template (replication mode), ignoring internal termination signals which are obeyed in the transcription mode to produce the family of five mono cistronic mRNAs. There is some evidence that the polymerase may have only a certain probability of "falling off" its template as it reaches a termination codon; the five mRNAs are made in decreasing molar amounts, reading from the 3' end of the parental RNA.

The orthomyxoviruses, bunyaviruses, and arenaviruses have seg mented genomes, and each segment is transcribed to yield an mRNA which is translated into one or more proteins (Fig. 4-5E ). In the case of the orthomyxoviruses, most of the segments can be regarded as single genes, for they encode single proteins. Special mention needs to be made of a phenomenon known colloquially as "cap-snatching," which is required by orthomyxoviruses for the initiation of mRNA synthesis. A virion-associated endonuclease enters the nucleus and removes a short segment from the capped 5' terminus of cell mRNA; this is transported back to the cytoplasm, where it binds to the virion RNA and serves as a primer to initiate transcription.

In general, each viral RNA segment of the genomes of the bunya viruses and arenaviruses codes for more than one protein. Furthermore, the S segment, at least, of arenaviruses and the Phlebovirus genus of bunyaviruses is ambisense. The replication strategy of ambisense RNA viruses, like the sense of their genomes, is mixed, with features of both (+) sense and (-) sense ssRNA viruses (see Chapters 29 and 34). Bunyavirus mRNAs also carry nonviral sequences at their 5' termini, presumably derived from cellular mRNA primers. dsRNA; Virion Transcriptase ( Fig. 4-5F) . The two families of viruses with dsRNA (Birnaviridae and Reoviridae) have segmented genomes and each segment is separately transcribed in the cytoplasm by a virionassociated RNA-dependent RNA polymerase. With reoviruses, each of the 10, 11, or 12 dsRNA segments corresponds to a single gene. Monocistronic mRNAs are transcribed from each segment within the partly uncoated subviral particle (see Plate 4-2); these RNAs complex with a protein before each is copied to produce a dsRNA, which serves as the template for further mRNA transcription. ssRNA; (+) Sense; Virion Reverse Transcriptase (Fig. 4-5G ). In the retroviruses the viral RNA is (+) sense, but instead of functioning as mRNA it is transcribed into DNA by a viral RNA-dependent DNA poly merase, and the resulting RNA-DNA hybrid molecule is converted to dsDNA and integrated into the cellular DNA. Transcription of RNA then occurs from the integrated viral DNA via the cellular transcriptase, followed by splicing of the RNA transcript as well as cleavage of the resulting proteins (see Chapters 12 and 31).

Regulation. Transcription from RNA viral genomes is generally not as rigorously regulated as with DNA viruses. In particular, the temporal separation into early genes transcribed before the replication of viral nucleic acid and late genes transcribed thereafter is not nearly so clear. 

K. Joklik, Virology 41, 501 (1970).] increases steadily during the first 6 hours of reovirus infection as more template becomes available, the relative amounts of each of the 10 mRNA species remain unchanged. With some viruses, however, a sub tle form of control can modulate the relative abundance of mRNAs for different proteins. For instance, in the case of the ( -) sense ssRNA rhabdoviruses and paramyxoviruses, where the whole genome is tran scribed into five monocistronic mRNA species, each coding for one of the five structural proteins, the "polarity" of the linear transcription by the viral transcriptase, described earlier, results in favored synthesis of mRNA for the proteins coded by the 3' end of the viral RNA.

Capped, polyadenylated, and processed monocistronic viral mRNAs bind to ribosomes and are translated into protein in the same fashion as cell mRNAs. The sequence of events has been closely studied for reovirus. Each monocistronic mRNA molecule binds via its capped 5' termi nus to the 40 S ribosomal subunit, which then moves along the mRNA molecule until stopped at the initiation codon. The 60 S ribosomal subunit then binds, together with methionyl tRNA and various initiation factors, after which translation proceeds. Despite the fact that mRNA is transcribed from each of the 10 monocistronic dsRNA reovirus segments in equimolar amounts, there are pronounced differences in the amounts of each protein made, indicating the existence of a regulatory mecha nism at the level of translation.

The proteins translated from the early transcripts of DNA viruses include enzymes and other proteins required for the replication of viral nucleic acid, as well as proteins that suppress host cell RNA and protein synthesis. However, the function of most early viral proteins of the large DNA viruses is still unknown.

The late viral proteins are translated from late mRNA, most of which is transcribed from progeny viral nucleic acid molecules. Most of the late proteins are viral structural proteins, and they are often made in consid erable excess. Some of them also double as regulatory proteins, modu lating the transcription or translation of cellular or early viral genes.

The temporal order and amount of synthesis of particular proteins of DNA viruses is regulated mainly at the level of transcription. With RNA viruses it is also usual for nonstructural proteins to be made early and structural proteins later, but the control is generally not as rigorous as for the DNA viruses and occurs at the level of translation. For instance, in the case of calici viruses, coronaviruses, and toga viruses, only the 5' end of the (+) sense viral RNA, which codes for the nonstructural pro-teins, including the RNA polymerase, is translated early, hence produc tion of complementary (-) sense RNA can commence. This then serves as the template for transcription of subgenomic RNA corresponding to the 3' end of the viral RNA, from which are translated the structural proteins required in abundance later in infection.

In the picornaviruses, the polycistronic viral RNA is translated di rectly into a single polyprotein which carries protease activity. This virus-coded protease cleaves the polyprotein at defined recognition sites into smaller proteins. The first cleavage steps are carried out while the polyprotein is still bound to the polyribosome. Some of the larger inter mediates exist only fleetingly; others are functional but are subsequently cleaved to smaller proteins with alternative functions.

Posttranslational cleavage occurs in several other RNA virus families but is a less prominent feature in the overall production of individual proteins. In the case of the toga viruses and calici viruses, polyproteins corresponding to only part, albeit a large part, of the genome are trans lated from polycistronic mRNA and then cleaved. With viruses of sever al other families, cleavage of particular proteins late in the replication cycle is essential for the production of infectious virions.

Newly synthesized viral proteins must migrate to the various sites in the cell where they are needed, e.g., back into the nucleus in the case of viruses that replicate there. The mechanisms controlling such migration are unknown, but presumably resemble those used for cellular proteins and possibly involve the cytoskeleton. Migration is doubtless intimately dependent on the structural features of particular proteins. In the case of glycoproteins, the polypeptide is translated on membrane-bound ribosomes, i.e., on rough endoplasmic reticulum; various co-and post translational modifications, including acylation, proteolytic cleavage, and addition and subtraction of sugars, occur sequentially as the protein moves in vesicles to the Golgi complex and thence to the plasma mem brane (see below).

Different mechanisms of DNA replication are employed by each fami ly of DNA viruses. We can give only a brief overview here. [From E. D. Sebring et al., J. Virol. 8, 478 (1971).] Papovaviridae. Little is known about the replication of papillomavirus DNA, but the polyomaviruses, especially SV40, have been studied in great detail. The SV40 genome, with its associated cellular histones, morphologically and functionally resembles cellular DNA and utilizes host cell enzymes, including DNA polymerase a, for its replication. An early viral protein, large-T, binds to three sites in the regulatory se quence of the viral DNA, thereby initiating DNA replication. Replication of this circular dsDNA commences from a unique palindromic sequence and proceeds simultaneously in both directions at the same rate ( Fig.  4-7) . As in the replication of mammalian DNA, both continuous and discontinuous DNA synthesis occurs (on leading and lagging strands, respectively) at the two growing forks. The discontinuous synthesis of the lagging strand involves repeated synthesis of short oligoribonucleotide primers, which in turn initiate short nascent strands of DNA {Okazaki fragments), which are then covalently joined to form one of the growing strands.

Adenoviridae. Adenovirus DNA is linear, the 5' terminus of each strand being a mirror image of the other (terminally repeated inverted sequences), and each is covalently linked to a protein. The primer for adenoviral DNA synthesis is not, as is usual, another nucleic acid, but a precursor to this protein, referred to as adenovirus preterminal protein. DNA replication proceeds from both ends, continuously but asynchro nously, in a 5' to 3' direction, using a virus-coded DNA polymerase. It does not require the synthesis of Okazaki fragments.

Herpesviridae. Unlike other DNA viruses that replicate in the nu cleus, herpesviruses specify a large number of enzymes involved in DNA synthesis. Analysis of herpes virus DNA replication is incomplete, but it appears that a rolling-circle mechanism operates, at least in the later stages. The replicating DNA initially consists of circles and linear forked forms, which are later replaced by large bodies of tangled DNA. There are three origins of replication, two on the S component and one on the L component (see Fig. 1-3) , the latter being near the genes that specify the DNA polymerase and the major DNA-binding protein. New ly synthesized viral DNA appears to be cleaved to unit lengths during the process of packaging into newly formed capsids.

Poxviridae. The special features of poxvirus DNA replication are that it occurs in the cytoplasm and depends entirely on virus-coded proteins; it can occur in enucleated cells. Replication appears to begin at each end of the genome and involves a strand displacement mechanism, with the formation of small DNA fragments covalently linked to RNA primers. The discovery of the loop structure at the ends of the vaccinia virus genome (see Fig. 1 -3) suggested a model whereby nicks near the ends of the genome allow self-priming by the 3' ends thus generated.

Parvoviridae. In the autonomous parvoviruses (genus Parvovirus), DNA replication occurs in close association with cellular chromatin and is dependent on cellular functions provided in the S phase of the cell cycle, i.e., when cellular DNA synthesis is occurring, a feature that is correlated with the pathogenic potential of these viruses (see Chapter 22). The virion (-) sense DNA is copied to give a dsDNA replicative form. Further DNA synthesis requires the binding of a virus-coded pro tein to the 5' termini. Production of viral ssDNA appears to occur after nicks at the 5' end and repeated rounds of synthesis.

Hepadnaviridae. Replication occurs in the nucleus by a unique pro cess. The viral DNA polymerase converts the viral ss/dsDNA into a complete circular dsDNA. The (-) sense strand of this molecule is then transcribed by the cellular RNA polymerase to produce a full-length "pregenome" RNA. This (+) sense RNA is then encapsidated in viral cores together with newly synthesized DNA polymerase, which also carries reverse transcriptase activity. Minus-strand DNA is then synthe sized by reverse transcription of the pregenome RNA; the template is degraded to leave a full-length (-) sense DNA strand. A small RNA fragment from the 5' end of the pregenome is then used to prime the synthesis of the (+) sense DNA strand. Complete synthesis of this strand is not necessary for maturation of the virus, hence infectious particles contain dsDNA with a single-stranded region.

The replication of RNA is a phenomenon restricted to viruses. Tran scription of RNA from an RNA template requires an RNA-dependent RNA polymerase, a virus-coded enzyme not normally found in cells. It is not known whether the polymerase required to transcribe (+) sense RNA from (-) sense RNA is different from that needed to transcribe (-) sense RNA from (+) sense RNA. Both processes are essential because the replication of virion RNA requires first the synthesis of complemen tary RNA, which then serves as a template for making more virion RNA.

Where virion RNA is of (-) sense the complementary RNA is of (+) sense and the RNA polymerase is the virion-associated transcriptase used for transcription of subgenomic RNAs. However, whereas the pri mary transcripts from such (-) sense virion RNA are subsequently cleaved (in most cases) to produce mRNAs, some must remain uncleaved to serve as a full-length template for virion RNA synthesis.

In the case of (+) sense virion RNA, the complementary RNA is of (-) sense. Several RNA molecules can be transcribed simultaneously from a single complementary RNA template, each RNA transcript being the product of a separately bound polymerase molecule. The resulting struc ture, known as the replicative intermediate, is therefore partially doublestranded, with single-stranded tails ( Fig. 4-8) . Initiation of replication of picornavirus and calicivirus RNA, like that of adenovirus DNA, requires a protein, rather than a ribonucleoside triphosphate, as primer. This small protein, VPg, is covalently bound to the 5' terminus of nascent (+) and (-) RNA strands, as well as virion RNA, but not to mRNA.

Retroviruses have a genome consist ing of (+) sense ssRNA. Unlike other RNA viruses, they replicate via a DNA intermediate. A virion-associated RNA-dependent DNA poly merase (reverse transcriptase), using a tRNA molecule as a primer, makes a ssDNA copy. The reverse transcriptase, functioning as a ribonuclease, then removes the parental RNA molecule from the DNA-RNA hybrid. The free (-) sense ssDNA strand is then converted into linear dsDNA, which contains an additional sequence known as the long terminal repeat at each end. This linear dsDNA then migrates to the nucleus and becomes integrated into cellular DNA. Transcription of the viral RNA can then occur from this integrated (proviral) DNA (see Chap ter 12). 

Structural proteins of nonenveloped icosahedral viruses associate spontaneously to form capsomers, which self-assemble to form empty procapsids, into which viral nucleic acid is packaged. Completion of the virion often involves proteolytic cleavage of one or more species of capsid protein. The best-studied examples among animal viruses are the picornaviruses (Fig. 4-9) . The capsomer precursor protein (noncapsid viral protein, NCVPla) aggregates to form pentamers; each of the 5 NCVPla molecules is then cleaved by virus-specific proteases into VPO, VP1, and VP3. Twelve such pentamers aggregate to form a procapsid. A final proteolytic event, which cleaves the VPO molecule into VP2 and VP4, is required for RNA incorporation. The mature virion is a dodeca hedron with 60 capsomers, each of which is made up of one molecule each of VP1, 2, 3, and 4. There are also one or two uncleaved molecules of VPO in the virion. X-Ray crystallography shows that the assembling units are not just rigid preformed "bricks"; they have extensions that reach across adjacent units to form second-and third-nearest neighbor relationships. Such studies have also shown that there is no fixed way in which RNA interacts with ordered parts of the protein.

The mechanism of packaging viral nucleic acid into a preassembled empty procapsid has been elucidated for adeno virus. One terminus of the viral DNA is characterized by a nucleotide sequence referred to as the packaging sequence, which enables the DNA to enter the procapsid bound to basic core proteins, after which some of the capsid proteins are cleaved to make the mature virion.

All mammalian viruses with helical nucleocapsids, as well as some with icosahedral nucleocapsids, acquire an envelope by budding through cellular membranes. Since such envelopes always contain viral glycoproteins, we begin by discussing the mechanism of glycosylation of these proteins.

Glycosylation of Envelope Proteins. Much of our understanding of the glycosylation of viral proteins comes from studies with vesicular stomatitis virus (a rhabdovirus), Semliki Forest virus (a togavirus), and the orthomyxoviruses and paramyxoviruses. The essential steps appear much the same for all enveloped viruses, hence a general overview is presented (Fig. 4-10) . Viruses exploit existing cellular pathways nor mally used for the synthesis of membrane-inserted and exported se cretory glycoproteins.

The amino-terminus of viral envelope proteins initially contains a se quence of 15 to 30 hydrophobic amino acids, known as the signal sequence, which characterizes the protein as one destined for insertion into membrane and/or export from the cell. The hydrophobicity of the signal sequence facilitates binding of the growing polypeptide chain to a recep tor site on the cytoplasmic side of the rough endoplasmic reticulum and its passage through the lipid bilayer to the luminal side. A signal peptidase then removes the signal sequence. Oligosaccharides are added to asparagine residues of the nascent polypeptide in the lumen of the rough endoplasmic reticulum by en bloc transfer of a mannose-rich core of preformed oligosaccharides from a lipid-linked intermediate, an oligosaccharide pyrophosphoryldolichol. Glucose residues are then re moved by glycosidases, a process known as "trimming" of the core. The viral glycoprotein is then transported from the rough endoplasmic re ticulum to the Golgi complex, probably within a coated vesicle. Here the core carbohydrate is further modified by the removal of several mannose residues and the addition of further N-acetylglucosamine, galactose, and the terminal sugars, sialic acid and fucose. The completed side chains are a mixture of simple ("high-mannose") and complex oligosac charides. While in the Golgi complex the glycoprotein may become acylated, by the covalent attachment of fatty acids such as methyl palmitate to the hydrophobic membrane attachment end of the molecule. Another coated vesicle then transports the completed glycoprotein to the cellular membrane from which the particular virus buds.

Transport of Glycoproteins. Different viruses bud from different sites in the plasma membrane (orthomyxoviruses, paramyxoviruses, rhabdoviruses, arenaviruses, togaviruses, and retroviruses), some from the apical and others from the basolateral surface. Others bud from intracytoplasmic smooth endoplasmic reticulum (flaviviruses, bunyaviruses, corona viruses) or from the nuclear membrane (herpesviruses). Presum ably some structural feature of the glycoprotein serves as the "zip code" ensuring delivery to the correct location in the cell.

Cleavage of Envelope Proteins. With the orthomyxoviruses and para myxoviruses, which bud through the plasma membrane, a cellular pro tease cleaves the envelope protein at the time of its insertion into the membrane into two polypeptide chains, which remain covalently linked by disulfide bonds. Cleavage is not required for viral release and does not occur in certain types of host cells, but it is essential for the produc tion of infectious virions in the orthomyxoviruses (cleavage of the hemagglutinin) and paramyxoviruses (cleavage of both the hemagglutinin-neuraminidase and the fusion protein). Following fusion of the coated vesicle with the plasma membrane, the hydrophilic N-terminus of the glycoprotein projects from the external surface of the membrane, while the hydrophobic domain, which is near the C-terminus, remains anchored in the lipid bilayer. Budding. Budding may be regarded as a nonphysiological form of exocytosis ( Fig. 4-11) . The process begins with the insertion of the com pleted viral glycoprotein into the appropriate cellular membrane. Be cause proteins are free to move laterally in the "sea of lipid" that con stitutes the lipid bilayer of the plasma membrane, cellular proteins are displaced from the patch of membrane into which viral glycoproteins are inserted. It is not known whether there is selection of particular lipids for incorporation into the viral envelope, but the ratio of phospholipids to glycolipids and cholesterol is essentially the same as that of the mem brane of the particular host cell.

The monomeric, cleaved viral glycoprotein molecules associate into oligomers, to form the typical rod-shaped peplomer with a prominent hydrophilic domain projecting from the external surface of the mem brane; the hydrophobic domain near the C-terminus spans the mem brane and a short hydrophilic domain at the C-terminus projects slightly into the cytoplasm. In the icosahedral togaviruses (Plate 4-3A), each protein molecule of the nucleocapsid binds directly to the C-terminus of a glycoprotein oligomer of the envelope. Multi valent attachment of nu merous peplomers, each to an underlying molecule on the surface of the icosahedron, molds the envelope around the nucleocapsid, forcing it to bulge progressively outward until finally the nucleocapsid is completely enclosed in a tightly fitting envelope and the new virion buds off. The capsid proteins of most enveloped viruses with helical nucleocapsids do not bind directly to envelope glycoprotein but to a matrix protein which is bound to the cytoplasmic side of the plasma membrane beneath patches of viral glycoprotein (Fig. 4-11) .

Coronaviruses and bunyaviruses bud from rough endoplasmic reticulum and the Golgi complex; orthopoxviruses may acquire an enve lope in the Golgi, but enveloped forms are released from the plasma membrane. The envelope of the herpesviruses is acquired by budding through the inner lamella of the nuclear membrane; the enveloped vir ions then pass directly from the space between the two lamellae of the nuclear membrane to the exterior of the cell via the cisternae of the endoplasmic reticulum.

There are basically two mechanisms for the release of mature virions from the infected cell. With most nonenveloped viruses that accumulate within the cytoplasm or nucleus, release occurs only when the cell lyses. This may occur shortly after the completion of viral replication; e.g., cells infected with picornaviruses lyse as soon as assembly of virions is completed, with immediate release of the progeny virions. On the other hand, parvoviruses accumulate within the cell nucleus and are not re leased until the cell slowly degenerates and dies. Most enveloped vi ruses, on the other hand, are released by budding, a process which can occur over a prolonged period without much damage to the cell, hence many such viruses (e.g., arenaviruses, retroviruses) are noncytopathogenic and are associated with persistent infections. However, some en veloped viruses that are released by budding are cytolytic, e.g., the alphaherpesviruses. Orthopoxviruses may be released as enveloped forms by budding from the plasma membrane or as nonenveloped forms, by cell lysis; both forms are infectious.

If this had been a book about bacterial diseases of domestic animals, there would have been at least one chapter on antibacterial chemother apy. However, of the hundreds of antibiotics and other antibacterial compounds now available, not one has the slightest effect on any virus, and there are no specifically antiviral chemotherapeutic agents in com mon use. The reason is that viruses are absolutely dependent on the metabolic pathways of the host cell for their replication, hence most agents that interfere with viral replication are toxic to the cell. Increased knowledge of the biochemistry of viral replication has led to a more rational approach to the search for antiviral chemotherapeutic agents.

Several steps in the viral replication cycle represent potential targets for selective attack. Theoretically, all virus-coded enzymes are vulnera ble, as are all processes (enzymatic or nonenzymatic) that are more essential to the replication of the virus than to the survival of the cell. Obvious examples include: (1) transcription of viral mRNA (or copy DNA, in the case of the retroviruses) by the viral transcriptase, (2) rep lication of viral DNA or RNA by the virus-coded DNA polymerase or RNA-dependent RNA polymerase, (3) posttranslational cleavage of protein(s) by (virus-coded) protease(s). Less obvious at first sight, but proven points of action of currently known antiviral agents are: (4) penetration/uncoating, (5) polyadenylation, methylation, or capping of viral mRNA, (6) translation of viral mRNA into protein, and (7) assem bly/maturation of the virion.

A logical approach to the discovery of new antiviral chemotherapeutic agents is to isolate or synthesize substances that might be predicted to serve as an inhibitor of a known virus-coded enzyme. Analogs (con geners) of this prototype are then synthesized with a view to enhancing activity and/or selectivity. The discovery of a class of nucleoside analogs which selectively inhibit herpesvirus DNA synthesis has led to a realiza tion that virus-coded enzymes with a broader (or different) substrate specificity than their cellular counterparts can be exploited to convert an inactive precursor ("prodrug") to an active antiviral agent. Because the viral enzyme occurs only in infected cells, such drugs are nontoxic for uninfected cells. Exploitation of this principle may revolutionize anti viral chemotherapy.

Acycloguanosine, now commonly known as acyclovir, is a guanine derivative with an acyclic side chain, the full chemical name being 9-(2hydroxyethoxymethyl)guanine ( Fig. 4-12 ). Its unique advantage over earlier nucleoside derivatives is that it requires the herpesvirus-specified enzyme, deoxythymidine-deoxycytidine kinase, to phosphorylate it intracellularly to acycloguanosine monophosphate; a cellular GMP kinase then completes the phosphorylation to the active agent, acycloguano sine triphosphate (Fig. 4-12) . Acycloguanosine triphosphate inhibits the herpesvirus-specified DNA polymerase. Since activation of the prodrug needs the viral thymidine kinase, acyclovir is essentially nontoxic to uninfected cells but is powerfully inhibitory to viral DNA synthesis in infected cells.

Acyclovir and various derivatives, as well as other nucleoside analogs dependent on viral enzymes for conversion to the active form, are begin ning to be used in human medicine for the treatment of herpesvirus infections. It is a small start, but it does demonstrate that antiviral che motherapy may have a future. Such drugs find limited use in veterinary medicine, e.g., for treatment of feline herpesvirus 1 corneal ulcers.

A few other antiviral agents are in use in human medicine. For exam ple, rimantadine and amantadine can prevent the uncoating of influenza virus, and several compounds known to inhibit reverse transcriptase are undergoing clinical trials against AIDS.

In theory at least, interferons are the ideal antiviral antibiotics. They are naturally occurring, relatively nontoxic, and display a broad spec trum of activity against essentially all viruses (see Chapters 6 and 8). However, clinical trials in humans have been disappointing. Currently, it appears that they have a demonstrable effect on infections with papilloma viruses, herpesviruses, and rhino viruses. It is now possible to produce large amounts of various human and other interferons using cloned interferon genes, but it is still uncertain whether they will be of clinical value in humans. Their use for therapy in viral diseases of do mestic animals is even further away.

Overall, in spite of decades of effort and massive expenditure by the pharmaceutical industry, the yield of useful antiviral drugs has been meager. Only a handful of marginally effective agents have found a place in human medicine, and very few are used in veterinary practice. Nevertheless, it is important to be aware of the continuing research in this field, for antiviral chemotherapy may one day come to constitute an integral part of veterinary medicine.

A) Togavirus. (B) Accumulation of paramyxovirus SV5 nucleocapsids. (C, D) Budding of SV5 from the plasma membrane, with some filamentous forms (bars = 100 nm)

Expression of animal virus genomes

Segmented Negative Strand Viruses: Arena viruses, Bunya viruses, and Orthomyxo viruses

Cotranslational and posttranslational processing of viral glycoproteins

Comparison of initiation of protein synthesis in procaryotes, eucaryotes and organelles

Animal Virus Receptors. Series B: Receptors and Recognition

Structure of a human common cold virus and functional rela tionship to other picornaviruses

Viral Replication

Initiation Signals in Viral Gene Expression

How an animal virus gets into and out of its host cell

Replication strategies of the single-stranded RNA viruses of eukaryotes

Antiviral Chemotherapy, Interferons and Vaccines

Replication of arenaviruses and bunyaviruses

Rhabdoviruses. In "Virology

Structure and assembly of alphaviruses

The Adenoviruses

Autonomous parvo virus DNA structure and replication

The Reoviridae

Orthomyxo-and paramyxoviruses and their replication

The gene structure and replication of influenza virus

Replication of poxviruses

Nucleotide sequence of yellow fever virus; implications for flavivirus gene expression and evolution

The Herpesviruses

Picornaviruses and their replication

The molecular biology of corona viruses

The hepatitis B virus

DNA Tumor Viruses

RNA Tumor Viruses," and "RNA Tumor Viruses 2 Supplements

African swine fever