J . Cell Sci. Suppl. 7, 67-94 (1987)Prin ted in G reat Britain © The Company of Biologists L im ited 1987

67

T H E GENOME OF HERPES SIMPLEX VIRUS:

STRUCTURE, REPLICATION AND EVOLUTION

D U N C A N J. M cG E O C H

M RC Virology Unit, Institute of Virology, University of Glasgow, Church Street, Glasgow G il 5JR, Scotland

S U M M A R Y

The objectives of this paper are to discuss the structure and genetic content of the genome of

herpes simplex virus type 1 (HSV-1), the nature of virus DNA replicative processes, and aspects of the evolution of the virus DNA, in particular those bearing on DNA replication. We are in the late

stages of determining the complete sequence of the DNA of HSV-1, which contains about 155 000 base pairs, and thus the treatment is primarily from a viewpoint of DNA sequence and organization.

The genome possesses around 75 genes, generally densely arranged and without long range

ordering. Introns are present in only a few genes. Protein coding sequences have been predicted,

and the functions of the proteins are being pursued by various means, including use of existing

genetic and biochemical data, computer based analyses, expression of isolated genes and use of

oligopeptide antisera. Many proteins are known to be virion structural components, or to have

regulatory roles, or to function in synthesis of virus DNA. Many, however, still lack an assigned

function.

Two classes of genetic entities necessary for virus DNA replication have been characterized: cis- acting sequences, which include origins of replication and packaging signals, and genes encoding proteins involved in replication. Aside from enzymes of nucleotide metabolism, the latter include

DNA polymerase, DNA binding proteins, and five species detected by genetic assays, but of

presently unknown functions.

Complete genome sequences are now known for the related alphaherpesvirus variceila-zoster

virus and for the very distinct gammaherpesvirus Epstein-Barr virus. Comparisons between the three sequences show various homologies, and also several types of divergence and rearrangement, and so allow models to be proposed for possible events in the evolution of present day herpesvirus

genomes. Another aspect of genome evolution is seen in the wide range of overall base compositions

found in present day herpesvirus DNAs. Finally, certain herpesvirus genes are homologous to non­

herpesvirus genes, giving a glimpse of more remote relationships.

I N T R O D U C T I O N

The genome of herpes simplex virus (HSV) is a DNA molecule of some 155 000

base pairs. By the standards of virology this is a large and complex genetic object. My

colleagues and I, working in the MRC Virology Unit in Glasgow, have undertaken

determination of the whole DNA sequence of HSV type 1 (HSV-1), and we are now

in the late stages of this task. The primary objective of this work, having obtained the

sequence, is to interpret it as fully as possible in terms of genes, transcripts, encoded

proteins, and other functional entities. In addition, other topics are of increasing

interest, as it has become possible to use sequence data in examining relations

between HSV and other herpesviruses, in comparing genes of herpesviruses with

68 L). J. McGeoch

other, non-herpesvirus genes, and in defining more clearly phenomena of herpes­

virus DNA evolution. From this description it can be seen that my interest in the

processes of HSV DNA replication is secondary to my involvement with the

structure and properties of the object to be replicated, the virus genome. Accord­

ingly, this paper sets out to describe our current understanding of the HSV genome,

and of its replication, and to explore implications which analyses of genome sequence

may have for the mechanisms of replication.

HSV is a member of the family Herpesviridae, a large group of viruses responsible

for a variety of diseases of man and animals (Matthews, 1982). All have large,

icosahedral, enveloped virions, whose genomes are linear, double-stranded DNA of

M t in the range 80X106 to 150X106. The group is very diverse in many aspects of

pathology and biology, although one unifying feature is the ability to enter into long­

term, latent infections (Roizman, 1982). Three sub-families have been defined,

based primarily on biological characteristics; these are termed the Alpha-, Beta- and

Gamma-herpesvirinae. The genomes of the herpesviruses vary widely, in their size

and base composition and in the arrangement of large, repeated sequences (reviewed

by Honess, 1984). There are, however, no simple correlations between genome

characteristics and biological properties. Five herpesviruses are recognized as natural

pathogens of Homo sapiens: the two serotypes of HSV (HSV-1 and HSV-2),

varicella-zoster virus (VZV), human cytomegalovirus (HCMV) and Epstein-Barr

virus (EBV). The first three of these belong to the Alphaherpesvirinae, HCMV is

assigned to the Betaherpesvirinae, and EBV to the Gammaherpesvirinae. Although

this paper is primarily concerned with HSV-1, comparative analyses of aspects of

HSV-2, VZV and EBV are also employed. Complete genome sequences have now

been published for EBV (Baer et al. 1984) and for VZV (Davison & Scott, 1986).

S T R U C T U R E A N D O R G A N I Z A T I O N OF THE HSV G E N O M E

Arrangement o f sequence elements in H S V D N A

HSV-1 virion DNA is a linear molecule containing about 155 000 base pairs (Kieff

et al. 1971). As shown in Fig. 1, the DNA is viewed as consisting of two covalently

joined segments, termed Long (L) and Short (S). Each segment contains an unique

sequence flanked by a pair of repeat sequences in opposing orientation. Thus, we

have the long and short unique sequences ( U l and Us), of 110000 and 13 000 base

pairs respectively, the terminal and internal long repeats (TRL and IR l) , each of

9200 base pairs, and the terminal and internal short repeats (TRg and IRs), each of

6600 base pairs. The sequences of R l and R§ are distinct. In addition, the DNA

possesses a direct, terminal repeat of 400 base pairs; for historical reasons this is

termed the a sequence. There is a further copy of the a sequence internally, in

inverted orientation, at the ‘joint’ between the L and S segments. In some molecules

in a virion DNA preparation, the L terminus and the joint may possess multiple

copies of the a sequence (Wagner & Summers, 1978).

Preparations of HSV virion DNA contain equivalent amounts of four sequence-

orientation isomers, which differ in the relative orientations of the two unique

The genome of herpes simplex virus 69

sequences about the joint, as indicated in Fig. 1. These isomers are apparently

functionally equivalent (Davison & Wilkie, 19836; Longnecker & Roizman, 1986).

By convention, one arbitrarily chosen isomer is regarded as the prototype, and this is

used for most purposes in maps of restriction sites and gene organization.

Early sequence analysis of HSV DNA revealed another phenomenon of sequence

repeats, on a finer scale: families of multiple copies of short, directly repeated

sequences were discovered, first in several locations in Rs (Davison & Wilkie, 1981).

Each family had a distinct sequence, although there were marked generic similarities

(high content of G and C, high occurrence of homopolymer runs, marked strand

asymmetry of purine versus pyrimidine content). The copy numbers of several

families were found to vary among individual plaque isolates or molecular clones of

restriction nuclease fragments. Sets of small repeat sequences have now been

observed in Us, Rs and R l, with unit lengths ranging up to 54 residues (McGeoch et al. 1985; Rixon et al. 1984; Perry, 1986).

Organization o f H S V genesOur best current estimate is that there are about 75 genes in the genome of HSV.

In general these are densely arranged, and various forms of gene overlap have been

observed (see for instance: Rixon & McGeoch, 1984; McGeoch et al. 1985; Wagner,

1985). Genes are found in both orientations. Each gene has its own promoter. Thus,

there is no discernible long range order in gene layout (in contrast, for instance, to

0 40 80 120 155 kbp1 i i i t

L S< ->

a a' a

□ ----------------------------- ----- 1— n — □TRl Ul IR l IR s Us TRs

P ----------------------------------->

Is —

II <■

Ils

Fig. 1. Organization of HSV-1 genomic DNA. HSV-1 DNA is regarded as comprising two covalently joined segments, L and S. Each of these consists of an unique sequence (U l and Ug, drawn as solid lines) flanked by a pair of inverted repeat sequences (TRL and IR l , IRs and TRS, drawn as open boxes). The DNA also possesses a 400 base pair directly repeated sequence at each terminus, termed the a sequence. A copy of the a sequence in inverted orientation is present at the ‘joint’ between L and S. A population of HSV DNA contains four equimolar sequence orientation isomers differing in the relative orientations of the L and S segments, as indicated in the lower part of the diagram. These are designated P (an arbitrary prototype), Is (inverted S), I I (inverted L) and Ils (both inverted). See Sheldrick & Berthelot (1974); Roizman (1979).

70 D. J. McGeoch

the case of adenovirus, where many genes are expressed using only a few promoters,

but with complex processing of transcripts). A common feature of transcriptional

organization in HSV is the occurrence of 3'-coterminal families of transcripts. In this

arrangement, several adjacent genes are similarly aligned. For each gene, transcrip­

tion starts 5' to that gene’s coding sequence, but continues through the following

genes of the family until a common, distal polyadenylation site is reached. In each

transcript species, only the 5'-proximal coding region is translated (Rixon &

McGeoch, 1984, 1985). As far as can be judged from our present level of analysis,

introns are an uncommon, but not wholly absent, feature of HSV genes (Rixon &

Clements, 1982; Murchie & McGeoch, 1982; Costa et al. 1985; Perry et al. 1986).

R E P L I C A T I O N OF HSV D N A

Outline o f H S V D N A replicative processes

Our present knowledge of the mechanistic details of HSV DNA replication is quite

limited. In the newly infected cell, incoming virus DNA accumulates in the nucleus.

It is likely that it is quickly converted into a circular form, perhaps by direct ligation

(Davison & Wilkie, 1983«; Poffenberger & Roizman, 1985). Expression of the virus

genes occurs in several stages. The first genes to be transcribed and translated are the

five ‘immediate-early’ (IE) genes (Honess & Roizman, 1974; Watson et al. 1979;

Wagner, 1985). The major recognized function of these is the activation of later

classes of genes, and there is now direct evidence for three of the IE proteins that they

can act as transcriptional activators (Everett, 1984; Gelman & Silverstein, 1986;

O ’Hare & Hayward, 1985; Everett, 1986). A large class of ‘early’ genes is then

expressed. These include many whose products are involved in aspects of HSV DNA

replication, and virus DNA synthesis then ensues. The final classes of HSV genes are

expressed after the onset of DNA replication.

The nature of the replicative forms of HSV DNA is obscure, although electron

microscopic studies have produced some evidence, showing proposed replicative

forks and bubbles (Friedmann et al. 1977). The size and fragility of the virus DNA

present such analyses with severe problems: one paper described replicating

herpesvirus DNA as occurring in ‘large, tangled masses’ (Ben-Porat & Rixon, 1979).

It is known that newly replicated DNA does not possess detectable termini, so that it

is probably circular or in head-to-tail concatemers (Jacob et al. 1979; Jongeneel &

Bachenheimer, 1981). It is not at all clear whether DNA replication first undergoes a

template amplification stage (as circular monomers) before synthesis of DNA for

packaging into progeny virus. Late in infection, replicated DNA is in a very rapidly

sedimentable form, thought to be extensive concatemers (Jacob et al. 1979). It thus

appears likely that replication, at least at this stage, is by a rolling circle mechanism.

Concatemeric DN A is further processed in the cell nucleus by packaging into nascent,

nucleocapsids, and by scission into genome unit lengths (Vlazny et al. 1982). At

present, the order of the packaging and cutting steps is not clearly resolved.

There are in the HSV genome two classes of genetic entity directly concerned with

genome replication, namely m-acting sequences (involved in initiation of synthesis

The genome of herpes simplex virus 71

and in packaging), and genes encoding proteins active in some aspect of DNA

synthesis. In the following sections we outline what is presently known of these

elements.

cis-acting signals for D N A replication and processing

There was some early evidence, from electron microscopic analyses of HSV DNA

extracted from infected cells, that the virus genome might possess one or more

specific sites at which replicative synthesis of DNA started (Friedmann et al. 1977).

However, study of the existence and location of such origins of replication really only

became productive after analyses of the nature of defective HSV DNA species. As

with many other virus types, defective species are obtained when HSV is repeatedly

passaged at high multiplicity. The HSV defective DNAs were found to consist of

tandem repeats of sequences derived from the standard genome, and were divided

into two classes on the basis of the sequences present. Both classes contained

sequences from the terminus of T RS. Class I molecules contained in addition further

sequences from the S segment, while class I I molecules contained sequences from

near the centre of U l (Kaerner et al. 1979; Vlazny & Frenkel, 1981; Spaete &

Frenkel, 1982). It was also shown that when monomer units from either class of

defective were introduced into cells in the presence of wild-type virus, these could

serve as ‘seeds’ for regeneration of tandemly repeated defective DNAs (Vlazny &

Frenkel, 1981; Spaete & Frenkel, 1982). The ability of such defective molecules to be

replicated and packaged, with the help of wild-type virus, was taken as indicating

that they carried regions of the genome providing necessary cis-acting signals for

DNA synthesis and processing.

With this background, a system was introduced for detailed analysis of m-acting

sequences needed for replication (Stow, 1982). This consisted of introducing into

culture cells a circular plasmid carrying the sequences under test, and using HSV (or

HSV DNA) to provide genes expressing proteins needed for replication. When the

test plasmid contained sequences necessary for its own replication, then amplifi­

cation of the non-HSV sequences in the plasmid could be detected by hybridization

of extracted DNA with labelled parental vector DNA. Thus, the plasmid assumes

the role of an HSV defective DNA, but with the advantage of being amenable to

genetic manipulation. With this system, a small part of the Rs region was identified

as enabling replication of covalently linked sequences, as head-to-tail concatemers,

and was termed ‘on's’, for ‘short region origin of replication’ (Stow & McMonagle,

1983). Much further work on the properties of oris has now been performed, and in

addition a similar element (onL) has been detected in UL. The properties of these

sequences are thoroughly consistent with their being structures where synthesis of a

new DN A strand is initiated, but it should be pointed out that formal proof for this

has yet to be obtained.

Analysis o f oris

Stow & McMonagle (1983) carried out a systematic deletion analysis to localize

orz's activity to a region of 90 base pairs in the Rs sequences of HSV-1, as shown in

I---- 1100 bp

B««< < «« <<<<<<<<<<<>>>>> »> >>>>>>>>>>>>>

GGCCGCCGGGTAAAAGAAGTGAGAACGCGAAGCGTTCGCACTTCGTCCCAATATATATATATTATTAGGGCGAAGTGCGAGCACTGGCGC

t t t t l t l t i t l t l t t t l t

cHSV-1 GGCCGCCGGGTAAAAGAAGTGAGAACGCGAAGCGTTCGCACTTCGTCCC AATATATATATATTATTAGGGCGAAGTGCGAGCACTGGCGC

** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** *

VZV GGGTTTTCATGTTTTGGCATGTG'ICCAACCACCGTTCGCACTTTCTTTCTATATATATATATATATATATATATATATATATAGAGAAAG

Fig. 2. Organization of HSV-1 ori s. (A) The location of theon'g copy in IRs is indicated, together with parts of IE genes 3 and 4. The features numbered are: (1) S' terminus of leftward transcribed IE gene 3 mRNA; (2) location of IE gene3 promoter; (3) location of IE gene 3 upstream control region; (4) location of IE gene 4 upstream control region (ill- defined at right); (5) minimum oris region; (6) location of IE gene 4 promoter; (7) 5' terminus of IE gene 4 mRNA and location of intron; and (8) location of IRg/Us junction. Redrawn from Preston et al. (1984). (B) The minimal oris sequence is shown, as the rightward S' to 3' strand only, of the IRg copy (Stow & McMonagle, 1983). Residues in the left and right arms of the imperfect palindrome are marked “< ’ and “> ’, respectively. The central A + T rich region is overlined, and the sequence protected from DNase by the ori specific binding protein is marked ‘ : : : ’. (C) The minimal HSV-1 oris, as in (B), is compared with the VZV ori sequence (Stow & Davison, 1986) to illustrate conserved features discussed in the text. Positions of pairs of identical residues are marked

Fig. 2. There are thus two copies of orig, one in IRS and one in TRg, but these are

not distinguished in further description in this paper. Ori § lies in an intergenic region

between two divergently transcribed IE genes, and is surrounded by the promoters

and the far-upstream regulatory elements of these genes. As illustrated in Fig. 2, the

orig sequence contains an imperfect palindrome, with each arm consisting of 21

residues. Essential functional sequences are known to lie outside the palindrome

(Stow & McMonagle, 1983). At the centre of the palindrome there is a sequence

(A-T)f,, which is essential for ori function (Stow, 1985).

Two examples from other viruses of sequences corresponding to HSV-1 on's are

now available. First, Whitton & Clements (1984) have shown that in Rs of HSV-2,

strain HG52, there are sequences closely similar to HSV-1 ons. The HSV-2

organization differs in that there are two copies of the ori$ sequences, contained

within almost identical direct repeats of 137 base pairs. Next, Stow & Davison (1986)

have recently identified a functional ori sequence in the Rs regions of the distantly

related alphaherpesvirus VZV. They demonstrated that in a plasmid amplification

assay VZV oris could be activated by HSV-1 replication factors. For comparative

The genome of herpes simplex virus 73

purposes, the VZV example is of particular interest, since in general HSV and VZV

DNA sequences are widely diverged. Like HSV, the VZV oris contains a

palindromic sequence with an (A-T)„ tract at its centre (Fig. 2). The palindromic

sequences flanking this central part are not conserved. However, there is a stretch of

11 residues identical in HSV-1 and VZV, which in HSV-1 lies across the boundary of

the palindrome and in VZV is just outside the palindrome.

Very recently, Elias et al. (1986) have shown that in nuclei from HSV-1 infected

cells there exists a protein which binds specifically to oris DNA. This binding

protects from DNase digestion an 18 base pair region across one end of the

palindrome (Fig. 2). Interestingly, this binding site includes the 11 base pair

sequence conserved between HSV-1 and VZV.

Analysis o f ori/.

The analysis of what is now called oh’l of HSV might have been expected to

proceed systematically from the knowledge of class II defective DNA structure, as

for oris. However, an unexpected obstacle has greatly retarded the progress of this

analysis in the past several years. This was the discovery that DNA containing

functional onL could not be cloned intact into standard bacterial vector systems.

Recombinant clones could indeed be obtained, but these invariably exhibited a

deletion of about 100 base pairs, and were not primarily active in plasmid

amplification assays (Spaete & Frenkel, 1982). When amplification was observed, it

turned out that the deleted region had been restored, presumably through

recombination with DNA of wild-type helper virus. Fine localization of oh'l

functional limits by plasmid manipulation, and also convenient sequence analysis,

were thus frustrated.

The first oril sequence to be determined was for a class I I defective DNA of

HSV-1 strain Angelotti, where the tandem repeat units of the defective served as the

source of suitable DNA for sequence analysis (Gray & Kaerner, 1984). It turned out

that this sequence contained 296 base pairs which became deleted on plasmid

cloning, and that within the deleting region were two long palindromic sequences. It

has since been recognized that this structure is not the simplest form of sequence

associated with onL, so further description of it is postponed until later in this

section. In 1985, sequences were reported for wild-type HSV-1 DNA in the region of

onL, using non-cloned virus DNA fragments as sequencing substrates (Weller et al. 1985; Quinn & MeGeoch, 1985). In addition, molecular cloning of the region in an

undeleted form was reported using a yeast plasmid (Weller et al. 1985).

The oh’l sequences of HSV-1 lie between the divergently transcribed genes for the

major DN A binding protein and the DNA polymerase, as shown in Fig. 3. It was

found that the deleting region contained a long perfect palindrome, with arms each of

72 residues, and that this palindromic sequence showed striking similarity to the oris sequence (Fig. 3). Evidently, this long inverted repeat was the cause of the cloning

instability. Recently, it has been shown that ori^ activity can be obtained with

sequences representing almost the complete palindromic region but lacking any

adjacent non-palindromic sequences (S. K. Weller, personal communication).

74 D. J. McGeoch

As indicated in Fig. 3, the oriL region contains symmetry elements additional to

the 72 residue inverted repeat, in that each arm of the repeat contains short sequences

which together form an ‘inner’ interrupted inverted repeat. As with oris, the onL

region contains an A + T-rich region at the centre of the palindrome. Alignment of

the two oii sequences shows that the whole ons palindrome is closely similar to the

central portion of the onL palindrome (Fig. 3). The sequence similarity extends

beyond the on’s palindrome on one side only. This non-palindromic part of the on’s

dbp OH pal

100 bp

< < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < <

GCGCGTCATCAGCCGGTGGGCGTGGCCGCTATTATAAAAAAAGTGAGAACGCGAAGCGTTCGCACTTTGTCCTAATAATATA

1 2 3 3 * 2 ' 1 *

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> TATATTATTAGGACAAAGTGCGAACGCTTCGCGTTCTCACTTTTTTTATAATAGCGGCCACGCCCACCGGCTACGTCACTCT

1 2 3 3 ' 2 ' 1 *

< < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < < <

O r iL GCGCGTCATCAGCCGGTGGGCGTGGCCGCTATTATAAAAAAAGTGAGAACGCGAAGCGTTCGCACTTTGTCCTAATAATATA**• * • * • * ***** ****** ***** *************************** **** *** *****

O ris GCGGGACCGCCCCAAGGGGGCGGGGCCGCCGGG . TAAAAGAAGTGAGAACGCGAAGCGTTCGCACTTCGTCCCAAT. ATATA<<<<<<<<<< <<<<<< <<<<<

<------------------------------------------s s s s s s s : ) s i : t : t : : :

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>O riL TATATTATTAGGACAAAGTGCGAACGCTTCGCGTTCTCACTTTTTTTATAATAGCGGCCACGCCCACCGGCTACGTCACTCT

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

O ris TATATTATTAGGGCGAAGTGCGAGCACTGGCGCCGTGCCCGACTCCGCGCCGGCCCCGGGGGCGGGCCCGGGCGGCGGGGGG >>>>> >» >>>>>>>>>>>>>

Fig. 3. Organization of HSV-1 onL. (A) The location of theoniy palindrome between the divergent dbp and pol genes is shown, with 5' regions of transcripts marked (Quinn & McGeoch, 1985; Gibbs et al. 1985). (B) The sequence of the ori\. region is shown, as the rightward 5' to 3' strand only. The palindrome is marked with '< ’ and *>’ as for Fig. 2. Within each arm of the palindrome, three regions, 1, 2 and 3, of interrupted, inverted repetition are indicated (Weller et al. 1985). (C) The sequences of the oii\, and oris regions are compared. Two padding characters have been inserted in the oris sequence to optimize alignment. Positions with identical residues are marked V . Palindromic regionsare indicated as before. The mapped limits of on’s are marked ‘< ---------> ’ (Stow &McMonagle, 1983), and the on protein site binding is marked -

The genome of herpes simplex virus 75

sequence is within the mapped functional limits of oris (Stow & McMonagle, 1983),

and it contains part of the oric, protein binding site (Elias et al. 1986).

The onL sequence in HSV-1 Angelotti class I I defective DNA, described above, is

another example of a ‘double’ origin, consisting of two closely spaced palindromic

sequences (Gray & Kaerner, 1984), and the larger of these is identical to that

described above for HSV-1 strains KOS and 17. The other palindrome is a slightly

shorter version, with 65 residue arms. There is no direct characterization available

for oh'l of wild-type strain Angelotti. However, from size differences in restriction

digests of HSV-1 strain Angelotti, as compared to other HSV-1 strains, it seems

probable that the double origin is present also in non-defective strain Angelotti DNA

(Gray & Kaerner, 1984). Recently, Lockshon & Galloway (1986) examined oh’l in

HSV-2. Here a palindrome, almost perfect, of total length 136 residues is present,

with high sequence homology to the HSV-1 palindrome. In contrast, VZV does not

possess any sequence counterpart of HSV oh'l in the equivalent genome location

(Stow & Davison, 1986).

M echanisms o f ori action

At present we can discern four common elements of alphaherpesvirus ori structure

and surroundings: (1) a palindromic sequence; (2) a repeating (A-T) sequence; (3) a

region conserved between HSV and VZV, which is part of the site for binding an ori specific protein; and (4) an intergenic location, close to transcriptional regulatory

regions of divergently transcribed genes. Palindromic sequences have been found in

other replication origins, for instance in SV40 (Bergsma et al. 1982). In that

example, however, there is no A + T-rich region at the centre of the palindrome, but

an A + T-rich region is located adjacent to the palindrome. SV40 also provides

another example of an origin located between two divergent genes. The perceived

significance of the locations of the HSV ori sequences certainly is enhanced by the

importance of the genes concerned: in the case of ons these are IE genes, one of

which encodes a vital transcriptional activator, and in the case of o h l the genes

encode prominent functional components of the DNA replication machinery. The

arrangement may suggest a functional link between ori activity and transcriptional

activity. Until establishment of a cell-free replication system for HSV and biochemi­

cal analysis of events in initiation of new DNA synthesis, definite roles for the ori elements cannot be assigned. For the moment, it is reasonable to suppose that the

binding of the ori specific protein at the HSV/VZV conserved locus may represent an

early stage in ori recognition and provide a site for assembly of the replicative

complex, while the (A-T)„ region would comprise a site for facile strand separation

prior to initiation of a new DNA strand.

It is quite unclear as to why HSV should possess both ons (in two copies) andon^.

The essential structural difference between them is that the on'L palindrome is

longer. It presumably thus includes two specific protein binding sites. A consequent

attractive possibility is that replication from on'L might be bidirectional, and that

from ong unidirectional. Thus, onL might be active in early, template-amplification

replication of circular monomeric DNA, while on's might act in later rolling-circle

76 D. J. McGeoch

C D -

d r i uh DR2 Uc DR1

- X -- —> <b'

I---------- 1100 bp

Fig. 4. Organization of the HSV-1 a sequence. Sequence elements in the copy of the a sequence at the L/S joint (a') are outlined for HSV-1 strain 17 (Davison & Wilkie, 1981) using the nomenclature of Mocarski & Roizman (1982). ‘DR1’ is a 17 base pair sequence present in two copies in the same orientation. The limits of the a sequence, as defined by positions of the two termini in virion DNA, lie asymmetrically within the DR1 elements. Ub is a locally non-repeated sequence, adjacent to the non-a-sequence part of IRi, (the b' sequence), and Uc is a locally non-repeated sequence adjacent to the non-a-sequence part of IR S (the c' sequence). DR2 is a family of directly repeated 12 base pair sequences.

replication to generate genomes for packaging in progeny virions. These are

speculative and unsupported notions, and they are rather vitiated by the apparent

lack of an ohl equivalent in VZV. There are also other obscure questions concerning

HSV replication origins: (1) is there any significance to the ‘doubling’ of an origin

seen in two cases? (2) what are the evolutionary relations between the highly similar

on's and oh’l sequences?

The a sequence

In early experiments on replication of plasmids carrying oris (Stow, 1982), the

amplified plasmid sequences were not encapsidated. However, Stow et al. (1983)

showed that when the test plasmid also contained a copy of the a sequence, then

processing and encapsidation of progeny plasmid DNA did occur. The sequence

organization of the HSV-1 a sequence is shown in Fig. 4, in the form in which it

occurs at the L/S joint. The a sequence is flanked by a pair of direct repeats of 17

base pairs (‘DR1’). In the genome terminal a sequences, the termini lie within these

repeats. Within the a sequence is a set of tandem reiterations (‘DR2’). Variations in

the sequences and copy number of these reiterations are found in different HSV-1

strains. The events which take place at the a sequence during encapsidation and

cleavage of nascent virion DNA are by no means well resolved. Varmuza & Smiley

(1985) have shown that cleavage can still take place at a partial a sequence, lacking

the Rl proximal copy of DR1 and part of Ub (see Fig. 4), but not at smaller a

The genome of herpes simplex virus 77

sequence fragments. However, no finer definition of recognition sites has yet been

reported. It seems to be the case that a precursor DNA with one a sequence at a given

locus can be cleaved and processed in such a manner that both resulting ends now

possess a copy of the a sequence. Yarmuza & Smiley (1985) presented evidence that

generation of termini involved two distinct cleavages, at each DR1 sequence, and

suggested that the doubling of the a sequence was effected as part of the same

process. In a study of sequences needed for genome isomerization, Chou & Roizman

(1985) ascribed to the DR2 repeat array a role in high frequency, site specific

recombination. A difficulty with this is that the HSV-2 a sequence does not possess a

DR2 repeat set (Davison & Wilkie, 1981).

HSV-encoded enzym es o f nucleotide metabolism

The HSY genome encodes a number of proteins which are involved in virus DNA

synthesis. These include proteins which participate directly in DNA replication, and

also enzymes of nucleotide metabolism. The latter are treated first.

There are two HSV-specified enzymes which catalyse reactions in the biosynthesis

of DNA precursors: thymidine kinase (TK) and ribonucleotide reductase (RR). TK

was first described by Kit & Dubbs (1963) and has been widely studied since.

Genetic analyses have demonstrated that TK is not essential for virus growth in

tissue culture (Jamieson et al. 1974). However, the experimental pathogenicity of

T K deficient mutants is reduced, probably because TK assumes more importance in

the supply of TTP for DNA synthesis in non-dividing cells (Field & Wildy, 1978).

T K is important in the action of nucleoside analogues which inhibit HSV infection

(see Rapp & Wigdahl, 1983). In the infected cell the active form of such compounds

is the 5'-triphosphate equivalent, and the first step in this activation is catalysed by

TK.

The other HSV-encoded enzyme of nucleotide synthesis is ribonucleotide

reductase, which catalyses the reduction of nucleoside diphosphates to deoxynucleo-

side diphosphates (Cohen, 1972; Dutia, 1983). The enzyme is composed of two

subunits, which are encoded by separate but contiguous genes (Preston, Palfreyman

& Dutia, 1984; Bacchetti et al. 1984; Frame eZ,al. 1985; McLauchlan & Clements,

1983). RR from other sources, both prokaryotic and eukaryotic, also consists of two

subunits, and indeed the HSV enzyme shows amino acid homology to these other

enzymes (Caras et al. 1985; Standart et al. 1985). The HSV RR differs from its

counterpart in normal, mammalian cells in that it is resistant to feedback inhibition

by TTP (Cohen, 1972). Unlike TK, RR is essential for virus growth in tissue

culture, presumably because it occupies a more central position in metabolic routes

for the supply of DNA precursors (Preston et al. 1984).

HSV also encodes a deoxyuridine triphosphatase (Wohlrab & Francke, 1980;

Preston & Fisher, 1984), which is more probably of significance in minimizing

incorporation of uracil into nascent DNA than in providing dUMP for TTP

synthesis. It is not essential for growth in culture cells (Fisher & Preston, 1986). It

was recently found, by way of DNA sequence studies, that a thymidylate synthase is

encoded by VZV and also by the gammaherpesvirus, herpesvirus saimiri (HVS)

78 D. J. McGeoch

(Davison & Scott, 1986; Honess et al. 1986). This enzyme catalyses methylation of

dUMP to TMP, but HSV has no corresponding gene.

Proteins active in D N A replication

Apart from the enzymes of nucleotide metabolism just described, a number of

proteins encoded by HSV or found in HSV infected cells have been implicated in

processes of virus DNA replication. The quantity and quality of knowledge

regarding these spans a wide range. Best characterized is the DNA polymerase,

where the protein has been purified, the enzymatic activities have been character­

ized, the gene is known and sequenced, and various mutated forms have been

obtained and studied. At the other end of the spectrum there are poorly characterized

entities; for instance, proteins for which circumstantial evidence, such as copurifi­

cation with a better known protein, suggests an involvement, but for which no

function is known or gene identified.

HSV DNA polymerase was first described by Keir & Gold (1963), and since then

the enzyme has been extensively studied. The protein has been purified as a single

polypeptide chain of M r 136000 (Powell & Purifoy, 1977; Knopf, 1979). This,

however, has been found associated with other species (see below). The polymerase

also exhibits a 3' to 5' exonuclease action, which is thought to have a proofreading

role (Knopf, 1979). As already mentioned, the polymerase gene (po l) lies adjacent to

onL. Several types of mutant have been studied (ts , drug resistance and drug

hypersensitivity; see Coen et al. (1984), and Honess et al. (1984)). The ts mutants

fall into two complementation groups within the one gene (Purifoy & Powell, 1981).

From the use of pol mutants it is clear that the polymerase participates directly in

virus DNA replication. The HSV DNA polymerase plays a key role in the

mechanism of action of antiviral agents which mimic dNTPs or pyrophosphate, and

the pol gene is a site of mutation to resistance to such compounds (Hay & Subak-

Sharpe, 1976; Chartrand et al. 1979; Crumpacker et al. 1980; Coen et al. 1982;

Honess et al. 1984).

After the polymerase, the best characterized protein involved in HSV DNA

replication is a species termed the major DNA binding protein (DBP) of M r 128 000,

the gene for which (dbp) lies on the other side of onL from pol. Ts mutants in dbp have shown that the protein is essential for HSV DNA replication (Conley et al. 1981; Weller et al. 1983). Its precise role, however, is not clear at present. DBP is not

the same protein as the oris binding protein of Elias et al. (1986). In vitro, DBP has

been shown able to dissociate the strands of duplex DNA (Powell et al. 1981). It has

been suggested that it may act in a complex with DNA polymerase and HSV

exonuclease (Littler et al. 1983).

Like HSV DNA polymerase, HSV exonuclease was first detected by Keir & Gold

(1963), as an increased alkaline nuclease activity in infected cells. Subsequently, the

enzyme was purified and was shown to consist of a single polypeptide chain, with

both 5' and 3' exonuclease activity and also an endonuclease activity (Morrison &

Keir, 1968; Hoffmann & Cheng, 1978, 1979; Stfobel-Fidler & Francke, 1980;

Hoffmann, 1981). The relation of this enzyme to DNA replication remains obscure

The genome of herpes simplex virus 79

and contentious. Studies with ts mutants have come to differing conclusions

regarding requirement of exonuclease for virus DNA synthesis (Moss et al. 1979;

Francke & Garrett, 1982; Moss, 1986). It remains at least possible that this is a

catabolic enzyme, whose function is to degrade host DNA to fuel virus DNA

synthesis.

Biochemical experiments have suggested the involvement of several additional

proteins or enzyme activities in HSV DNA synthesis. Vaughan et al. (1985) have

reported copurification with DNA polymerase of a distinct HSV coded protein of

estimated M r 54 000. Topoisomerase activities have been reported, both copurifying

with DNA polymerase and in virions (Biswal et al. 1983; Muller et al. 1985). As

described above, a protein which binds specifically to part of oris, has been detected

(Elias et al. 1986). Another report has described an a sequence binding protein

(Dalziel & Marsden, 1984). All of these must be regarded as possible components of

the replicative machinery, but more work is needed to authenticate their involvement

and clarify their roles.

Genes required for H S V D N A replicationStudies on the genetics of HSV have assigned ts mutants to about 10 complemen­

tation groups which were considered defective in viral DNA synthesis at non­

permissive temperature (V. G. Preston, personal communication; S. K. Weller,

personal communication). The genes concerned fall into three distinct classes:

(1) genes encoding known proteins, which are involved directly in DNA synthesis;

(2) genes encoding known proteins, which are not thought to be involved in DNA

synthesis, but which influence DNA synthesis as part of a pleiotropic effect; and

(3) genes for proteins of unknown specific function, at least some of which may

participate directly in DNA replication. In the first we include genes for DNA

polymerase, DBP and RR. In the second are several of the IE genes. Until very

recently the membership of the third category has been somewhat ill-defined,

suffering from lack of detailed knowledge of gene organization in the large and

complex genome of this virus, from wide mapping limits for mutants, and from

undoubted background complications of secondary and double mutations. This

situation is now improving, and it seems likely that there are at least five separate

groups, distributed across U l , in our third class, representing unknown proteins

which might be involved in the processes of DNA replication (based on data of V. G.

Preston and of S. K. Weller).

Recently a novel approach has been developed and exploited by M. D. Challberg

to identify HSV genes involved in DNA synthesis (Challberg, 1986; also, personal

communication). It was previously well known that purified HSV DNA could be

introduced into culture cells, where the virus genes were then expressed. Thus, such

transfected DNA could provide in trans the functions to support amplification of an

oris-containing plasmid introduced concurrently. Challberg found that a set of

plasmid clones representing between them most of the HSV genome, in large

fragments, could substitute efficiently for intact virus DNA in supplying replication

functions. Systematic subcloning, deletion analysis and inactivation by restriction

80 D. J. McGeoch

nuclease digestion then identified loci necessary for plasmid amplification. Compari­

son of these loci with our DN A sequence data has indicated clearly the genes

involved. Two classes of loci were found, those essential for plasmid amplification

and those which increased amplification but were not absolutely necessary. The

latter comprised IE genes and also genes for RR, and were thus satisfactorily

explicable. There were seven essential loci, as shown in Fig. 5. Predicted M r values

for the encoded proteins are given in Table 1. As expected, two of the loci represent

thepol and dbp genes. The other five correspond to genes for which no products have

yet been recognized. Each of these either coincides definitively with the position of a

tightly mapped DNA-negative complementation group, or lies within the mapping

limits of a less tightly mapped group (unpublished data of V. G. Preston and of S. K.

Weller). The exonuclease gene is not required.

The plasmid amplification assay is a test for initiation and elongation of DNA

synthesis, in a rolling circle mode. The work just described shows that there are

seven HSV genes which are essential for these functions, including pol and dbp. There are several ways in which possible further HSV genes for proteins involved in

replication could have escaped detection by this assay. First, there may be proteins

whose function could be replaced by a host cell protein, at least in the plasmid assay.

For instance, a virus-specified ligase might fall in this category. Next, it is

conceivable that there might be HSV proteins involved in regulation of replicative

processes, which would not register. Lastly, there are certainly events in the

downstream processing of progeny DNA - packaging and cutting - which cannot be

assayable by the plasmid system. Much the same limitations would apply to the

assignment of ts mutants as ‘DNA negative’. However, the agreement between ts mutant mapping and the results of the plasmid assay system is impressive, and

strongly suggests that the major genes involved primarily in DNA synthesis have

now been located.

' 0 0-2 0-4 0-6 0-8 M• I I I I i

CZD--------------- !------------------ L. IZJ----O

4 44 « ► ► ►UL5 UL9 dbp p o l GC.2L 63.2L

UL8

I f tO riL O ris O ris

Fig. 5. Locations of genes required for HSV-1 DNA synthesis. The prototypic HSV-1

genome is shown, with scale in fractional map units, and the locations and orientations of seven genes necessary for virus DNA synthesis are indicated, together with the positions of onL and on's-

The genome of herpes simplex vim s

Table 1. HSV-1 genes required for D N A replication

81

VZV EBVProtein equivalent equivalent

HSV-1 gene M r genef genej

UL5 99000 55 BBLF4

UL8 80000 52 ?

UL9 94000 51 ?

dbp 128 000 29 BALF2

pol 136000 28 BALF5

‘GC.2L’ 51000 16 j

‘63.2L’ 115 000 6 BSLF1

*The HSV-1 genes identified from the functional assays of M. D. Challberg and from DNA sequence analysis are listed, as in Fig. 5. ‘GC.2L’ and ‘63.2L’ are temporary, working names,

f Obtained by comparison with the sequence of Davison & Scott (1986).I Obtained by comparison with the sequence of Baer et al. (1984). '?’ indicates that no EBV

reading frame showing sequence homology could be found.

There thus exist five proteins encoded by HSV-1 which are clearly implicated as

acting in virus DNA synthesis, but whose roles are at present undefined. In addition,

they have not yet been correlated with known protein species in virus infected cells.

This could well point to their being present only in small amounts. The molecular

weights for these species, as predicted from DNA sequence open reading frames, are

all relatively large, with predicted M r values of 51000, 80 000, 94000, 99000 and

115 000. The predicted amino acid sequences are without particularly extreme or

striking properties - in short, at this informal level of analysis, they look like typical

large, globular proteins. From comparisons with other systems, there is no shortage

of possible functions for these proteins. These could include on recognition and

binding, primer synthesis, primer degradation, unwinding of duplex DNA, main­

tenance of unwound DNA in an extended configuration, DNA phosphorylation and

joining, enhancement of polymerase processivity, and other positive regulatory

activities.

Progress in elucidating the biochemistry of HSV DNA replication ought now to be

substantially facilitated by the enumeration of genes involved, and by the determi­

nation of the DNA sequences and prediction of protein sequences.

ASPECTS OF H E R P E S V I R U S E V O L U T IO N

Comparative analyses o f herpesvirus genomes

This final section examines some aspects of herpesvirus evolution, with emphasis

on topics related to the nature and functioning of the components of the DNA

replicative machinery of the virus. In the last two or three years, the increasing

availability of herpesvirus DNA sequences has greatly enhanced our views of

relationships between herpesvirus DNAs, and of evolution of herpesvirus DNAs,

and this will certainly continue with accumulation of more sequences. Prior to large

scale D N A sequencing, it was clear that many alphaherpesviruses did indeed share a

common ancestry, as evidenced by antigenic cross-reaction between their proteins,

82 D. J. McGeoch

and by hybridization, at some level, between genome sequences (for instance,

Davison & Wilkie, 1983c). There was, however, no comparable evidence for

relations between sub-families. Assignment to the Herpesviridae had been generally

made on the basis of virion morphology, aspects of the biology and gross

characteristics of the virus genome, so that there was in fact no critical, direct

evidence of genomic relatedness at the time of such classification decisions.

Comparisons based on the genome sequences have now provided such evidence of a

common ancestry, in part, for the genomes of alphaherpesviruses (HSV and VZV)

on the one hand, and gammaherpesviruses (as exemplified by EBV) on the other.

They have also illuminated the relations between gene organization in HSV and

VZV, where the viruses, although detectably related by previous methods, were

nonetheless substantially diverged, for instance in gross genome structure and also in

genomic base composition (Davison & McGeoch, 1986).

The genome sequences of these viruses often show little or no meaningful DNA

homology, and comparison of amino acid sequences predicted from the DNA

sequence, using computer methods, has emerged as the appropriate method of

analysis. This has proved useful at several levels. First, it has been used to assign

gene functions by comparison of sequences from a genome poorly studied genetically

with sequences from a genetically better characterized genome. In this way,

homologues in EBV have been detected to HSV genes encoding glycoproteins B and

H, DNA polymerase, ribonucleotide reductase, exonuclease and major DNA

binding protein among others (Pellett et al. 1985; McGeoch & Davison, 19866;

Quinn & McGeoch, 1985; Gibson et al. 1984; McGeoch et al. 19866). Similar, but

more extensive, assignments have been made in VZV by comparison with HSV

sequences (Davison & McGeoch, 1986; Davison & Scott, 1986). The second aspect

of such comparisons concerns the detailed location of conserved residues in a pair of

predicted sequences. In cases where the genomes are extensively diverged, to the

point where little or no DNA homology may be detectable, it is reasonable to suppose

that highly conserved localities within a protein sequence may correspond to

important elements in the protein’s structure or function (see McGeoch et al. 1986a).

The comparisons should therefore be of value in detailed molecular genetic analysis

of the genes. Lastly, over large tracts of a pair of genomes, differences in the pattern

of gene homologies must reflect relative changes in organization which have occurred

since divergence of the two virus species from a common ancestor (McGeoch, 1984;

Davison & McGeoch, 1986).

The overall views which have so far emerged are as follows. The first regions of

HSV-1 and VZV to be compared were the S segments (McGeoch, 1984; Davison &

McGeoch, 1986). The HSV-1 Us region contains ten genes wholly and the major

parts of two more, while each copy of Rs contains one gene. In VZV, Ug contains

two complete genes and the major parts of two more, while each copy of Rs contains

three complete genes. Comparisons of predicted amino acid sequences showed that

each VZV gene had an HSV-1 counterpart, but that six of the HSV-1 genes had no

VZV homologues. The layout of the genes differed in the two cases, and this was

interpreted as showing that the inverted repeat sequences were capable of large scale

The genome of herpes simplex virus 83

expansion or contraction in the course of evolution. By ‘expansion’ is meant that a

region in the unique sequence adjacent to the repeat can be duplicated in inverted

orientation at the other end of the unique sequence and is thus now regarded as part

of the repeat. A model using partially illegitimate recombination was proposed to

account economically for the changes.

Although the genome sequence of HSV-1 is presently unfinished in U l, it is clear

from the available data that the HSV-1 and VZV L regions correspond quite closely

in their content and arrangement of homologous genes, more so than the S segments.

VZV U l is predicted to contain 60 genes, on the basis of DNA sequence

interpretation, and we think at present that four of these have no counterparts in

HSV-1; conversely, we know of three HSV-1 L region genes without VZV

counterparts (unpublished data). Both these numbers may possibly rise a little when

the HSV-1 U l sequence is complete. Next, VZV possesses no large scale equivalent

of HSV R l, but one of the VZV U l genes is equivalent to HSV-1 IE gene 1, in RL

(Perry et al. 1986). These differences are illustrated in Fig. 6. The main conclusions

from these comparisons are that: (1) despite their large difference in base

composition, the genomes of HSV-1 and VZV are closely comparable in gene

arrangement; (2) there are, however, differences in gene content, and many of these

are closely associated with differences in repeat structure; and (3) the major repeats

are clearly highly dynamic structures on an evolutionary time scale.

Comparisons between an alphaherpesvirus genome and that of the gammaherpes­

virus EBV were initiated with HSV-l/EBV examples, but are now most completely

illustrated by the VZV/EBV analysis of Davison & Taylor (1987). The results

obtained in such comparisons are illustrated at two levels of detail in Fig. 7. In

summary, these exercises show the following. EBV and VZV possess twenty nine

pairs of genes which exhibit amino acid sequence homology. These range from

strongly conserved to barely conserved examples. In addition, another fourteen pairs

of genes are regarded as probably homologous, from their positions and from

properties such as distribution of hydrophobic residues, although amino acid

sequence homology was not quantifiable. This, then, accounts for forty three of

VZV’s presently recognized complement of sixty seven unique genes, and it is certain

that each virus possesses a number of genes without counterparts in the other. All of

these related genes are located in the U L regions of the two genomes: the repeat

elements and the short regions appear unrelated. Within U L, conserved genes are

located in three major regions. Within each of these, gene arrangement is generally

conserved, but with some local rearrangements. However, the three large blocks

differ in their relative locations in the two genomes. Thus, as a broad synthesis of

both theHSV/VZV and VZV/EBV analyses, the L segment appears to represent the

core of the general herpesvirus complement of genes, although by no means all of the

genes found there are conserved. The S regions, on the other hand, are characteristic

of sub-groups and also of species within these.

Examining the HSV-1 DNA replication genes in the light of these comparisons,

we see that all seven genes identified by Challberg have counterparts in the U l region

of VZV (Table 1). In some of these the sequence homologies are among the highest

84 D. J. McGeoch

registered in comparing these genomes. In EBV, homologues can be identified to

only four of the genes. This may reflect the roles and relative importances of these

proteins in central processes of DNA replication.

Comparison of replication origins between the genomes is also instructive. As

mentioned above, VZV possesses a counterpart of ons but not, apparently, of onL.

The EBV sequence possesses no element resembling the alphaherpesvirus origin

sequences. A qualification here results from the observed tendency of HSV onL to

delete on cloning into bacterial plasmids: perhaps the EBV ori has deleted, without

detection, in the clones used for sequence analysis. Present knowledge about EBV

DNA replication during the lytic cycle of the virus is limited. However, EBV DNA

exists in transformed, latently infected lymphocytes as a circular, nuclear episome,

TRl UL

HSV-1 L [E

u, IR l

• . :oj j j j

_ i _ i__ i-i

VZV L

IRl UL UL TRl

5 kb

IRs Us TRc

HSV-1 S

i- ii_ioo o

-1 .

5 kb

VZV S

IRs TRS

Fig. 6. Comparisons of gene arrangements of HSV-1 and VZV. In the upper part of the figure, the arrangement of HSV-1 genes adjacent to the two extremities of the L region is compared with the layout of the corresponding parts of VZV’s L region. In order to align the standard VZV map with the prototype orientation of HSV-1 L, the VZV layout is presented with the S region at the left. In the lower part of the figure, the arrangements of genes in the two S regions are compared. Coding regions of genes are shown as arrows. Homologous pairs of genes are joined by dotted lines. For genes in repeat regions, only one homologue is indicated. Genes without homologues are marked ‘O ’. More infor­mation on the gross structure of VZV DNA is given in the legend to Fig. 7. Data are from McGeoch (1984); Davison & McGeoch (1986); Davison & Scott (1986); Chou & Roizman (1986); Perry et al. (1986); Perry (1986); also unpublished data.

The genome of herpes simplex virus 85

and there is a second, ‘maintenance’ mode of EBV DNA replication active in this

situation. It has been found that in this case the only czs-acting element necessary is

located in the U s region (Yates et al. 1984). This is a structure called oriV, which

consists of two separable parts. The first is a tandem array of imperfect repeats of a 21

base pair sequence. The second comprises a palindromic sequence related to the

repeat unit (Reisman et al. 1985). The only essential, trans-acting factor supplied by

EBV is the EBNA-1 protein (which has no counterpart in HSV or VZV) and it is

known that this can bind specifically to the repeat units of on'P (Yates et al. 1985;

Lupton & Levine, 1985; Rawlins et al. 1985). This system thus differs very

significantly from the alphaherpesvirus origins, and can be rationalized as represent­

ing replication in phase with cell division, a capability not known to be required by

HSV or VZV.

Relations o f herpesvirus genes with other genes

All of the differences so far discussed have referred to processes of change within

lines of descent of present day herpesviruses, and the existence of some pre-existing,

ancient herpesvirus genome has been implicit in this. In general, there do not exist

recognizable ‘fossil’ forms of viruses by which we could hope to reconstruct earlier

evolutionary events, and we do not have any real idea as to the nature of a ‘pre­

herpesvirus’ genome, or as to what genetic entities have contributed to herpesvirus

genetic material. However, comparisons of predicted amino acid sequences of

herpesvirus genes with non-herpesvirus sequences have now revealed several

instances of homology between genes of herpesviruses and those from prokaryotic or

eukaryotic cell genomes. The examples found are of particular interest and relevance

to thinking about virus DNA replication. First, it is clear that both subunits of RR

from HSV, VZV and EBV are homologous to RR subunits of eukaryotic and

prokaryotic organisms (Caras et al. 1985; Standart et al. 1985). The same is also true

of the thymidylate synthase of VZV and of the gammaherpesvirus HVS (Davison &

Scott, 1986; Honess et al. 1986). Thus, it is plausible that an earlier herpesvirus or

pre-herpesvirus obtained these genes, directly or indirectly, from a cellular genome.

More instructive for our present purpose is the recent finding that there is an

extended similarity between the DNA polymerases of herpesviruses and of vaccinia

virus (Earl et al. 1986). Vaccinia virus is a member of the Poxviridae, large DNA

viruses which replicate in the cytoplasm. The strategy and details of their replicative

processes are very distinct from those of herpesviruses, and this is the first finding of

gene homology between the groups. Thus, the finding of a similarity in these DNA

polymerases, strongly indicative of a common origin, most simply suggests that both

may have originated from a cellular gene. The best present candidate for this is the

catalytic subunit of DNA polymerase a of eukaryotes. It should be noted that

homologies have been described between other vaccinia virus genes and their cellular

counterparts (see Earl et al. 1986). The homology between polymerases of

herpesviruses and vaccinia virus is considerably more extensive than a previously

found local similarity between herpesvirus polymerases and that of adenovirus,

whose significance was then unclear (Quinn & McGeoch, 1985). However, the

86 D. J. McGeoch

argument on cellular origin also applies, but with less force at present, to adenovirus

DNA polymerase. Apart from these examples, of proteins concerned with synthesis

of D N A and its precursors, the only remaining herpesvirus homology known in this

class is that of the alphaherpesvirus-specific gene proposed to encode a protein kinase

with amino acid sequence homology to members of the eukaryotic protein kinase

family (McGeoch & Davison, 1986a).

The idea that herpesvirus DNA polymerases may be related to the cell’s replicative

polymerase should be illuminating for future work on virus DNA replication. It also

encourages us to speculate that other genes, among those required for virus DNA

synthesis but of presently unidentified function, may bear similarities to factors in

eukaryotic DNA replication, and that this may contribute towards recognition of

functions of the virus genes as the eukaryotic gene sequences become available.

Processes o f herpesvirus genome evolution

The comparisons of herpesvirus genomes which have just been outlined demon­

strate that large changes in herpesvirus DNAs must have occurred during evolution.

Among these, the large scale rearrangements, the qualitative changes in major repeat

elements present, and the alterations in extent of major repeat elements are all

reasonably attributable to the aberrant activities of one or more recombination

systems. At a finer level, the genomes have evidently been subjected to very extensive

processes of point mutation and small addition/deletion events during development

of the present day forms. An obvious candidate for introduction of such alterations is

the genome replicative system. Evidence has been presented in support of the idea

that the DNA polymerase of HSV is relatively error prone, namely that mutants exist

which give rise to a lower rate of mutations in the genome than does the wild type

enzyme (Hall et al. 1985). However, consideration of the wide range of base

compositions found in herpesvirus DNAs shows that processes of generation and

fixation of mutations in these genomes are more complex than just indicated.

At the whole genome level, herpesvirus DNA base compositions lie in the range of

32 to 75 % G + C (Honess, 1984). For the DNAs of viruses discussed in this paper,

HSV-1 is 67 %, VZV is 46 % and EBV is 60 % G + C. For HSV-1 and VZV we thus

Fig. 7. Comparison of gene arrangements in the genomes of VZV and EBV. (A) The

gross structures of the VZV and EBV genomes are shown. At this level, the genome of VZV resembles that of HSV, except that the Ri, element is only eighty eight base pairs, there is no counterpart of the a sequence, and proportions of the four genome isomers are not all equally abundant (Davison, 1*984). EBV DNA possesses at each terminus, and in the same orientation, a set of tandem repeat sequences, shown jointly as ‘TR ’. Internally there is another set of direct repeats, marked ‘IR ’. There is only one orientation isomer (see Baer et al. 1984). The locations in the three genomes of the three major blocks (A, B and C) of corresponding genes are indicated. In EBV DNA the A block is relatively inverted and is marked as ‘A”. (B) The arrangements of genes in the U l regions of the VZV and EBV genomes are shown, with the EBV genome rearranged to align blocks A, B and C with the continuous VZV sequence. Lengths are marked in kilobase pairs, and for EBV refer to the non-rearranged sequence. Coding regions and orientations of all predicted genes are indicated, and pairs of genes which show amino acid sequence homology are joined by dotted lines. Data are from Baer et al. (1984); Davison & Scott (1986); Davison & Taylor (1987).

A

A B Cl------- li-------- Il------------------- 1

VZV

The genome of herpes simplex virus 87

I--------------------------------t r l u l i r l i r s t r s

Us

C A' BI--------------------- II-------il--------- 1

EBV

□---------------------------------------------------------r u l l i l i i t t i----- □

TR UL IR Us TR

l I20 kb

VZV

<0H

A10

><20 30 40

*

EBV H| ■■■■"I . "■■■ | ........................................ | -I I r90 80 70 100 110 120< A > < B >

VZV50

—1_60

_ l_70 80

_ k _90

—L.100

—I— ■ C l

>

T”10

—r-20

“ I—40

—r—6030 50

Fig. 7

88 D. J. McGeoch

have the remarkable result that two genomes which are clearly related, with very

similar genetic organization over most of their length, differ by 21 percentage points

in gross base composition, the difference being distributed rather evenly over the

whole of the comparable sequences. The phenomenon of base compositional

variation also occurs at an intragenomic level: for example, the Rs element of HSV-1

DNA is 79-5% G + C, while the adjacent Ug is 64-3% (McGeoch et al. 1985,

19866). The nature of evolutionary forces responsible for such extensive effects are at

present quite obscure. It is evident from detailed examination of the sequences that

requirements of polypeptide coding are not in general primarily responsible, but

that, in contrast, evolution to an extreme of G + C content has affected amino acid

composition. This is most clearly shown by comparing the HSV-1 and VZV versions

of the transactivator gene in Rs (McGeoch et al. 1986a).

Although the ‘why’ of this phenomenon remains unknown, we can discuss ‘how’:

that is, the nature of the biochemical machinery responsible. Possession of an error

prone DNA polymerase cannot in itself be sufficient, since it would be difficult to

account for large differences in base composition between extensive regions within

one genome by this means. We have previously classified elements of the machinery

responsible under three functional headings, as a mutation producer, a biasing

mechanism and a disseminating mechanism. DNA polymerase certainly may be

important in the production of mutations. The biasing mechanism is required to

impart directionality. Whatever its nature, this must be metastable over evolutionary

time periods. It is possible that DNA polymerase could be involved here also.

Another suggestion is that virus coded enzymes for nucleotide synthesis could

influence frequency and direction of substitution mutations by altering dNTP pool

sizes (Honess et al. 1986). We have proposed a model whereby directionality is

imparted by recombination in the DNA molecules in an infected cell, by a form of

biased gene conversion in which survival of alleles is affected by their base

composition (McGeoch et al. 1986a). This notion had its roots in ideas on the

concerted evolution of multigene families of eukaryotes (Dover, 1982). Whether or

not directionality is mediated by recombination, we wish to invoke recombination as

a disseminator of mutations through an intracellular population of DNA molecules,

in order to explain differences in base composition between repeat and unique

sequences. The abundance of repeat sequences is, by definition, higher than that of

unique elements. More potentially fixable mutations therefore occur at a given

sequence position over the whole repeat population, and the occurrence of

recombination for dissemination of these mutations should also be higher, resulting

in a more rapid evolution of repeat sequences. We conclude by re-emphasizing that

these speculations have been concerned only with how base composition changes

might have been implemented, but do not at all approach the question of why they

should represent favourable or even survivable operations.

Thanks are due to M. D. Challberg, A. J. Davison, V. G. Preston and S. K. Weller for access to unpublished data, and to N. D. Stow for discussion. References to unpublished HSV-1 sequences determined in the author’s laboratory included work by M. A. Dalrymple, A. Dolan, M. C. Frame, D. McNab and L. J. Perry.

The genome of herpes simplex virus 89

R E F E R E N C E S

B a c c h et t i, S ., E v e l e g h , M. J., M u ir h e a d , B . & S pector , T. (1984). Immunological characterization of herpes simplex virus type 1 and type 2 polypeptide(s) involved in viral ribonucleotide reductase activities. J ’. Virol. 49, 591-593.

B a e r , R., B a n k ie r , A . T ., B ig g in , M. D . , D e in in g e r , P. L., F a r r e l l , P. J., G ib s o n , T . J., H a t f u l l , G., H u d s o n , G. S ., Sa t c h w e l l , S. C., Se g u in , C., T u f f n e l l , P. S. & B a r r e l l ,

B . G. (1984). D N A sequence and expression of the B95-8 Epstein-Barr virus genome. Nature, Lond. 310, 207-211.

B e n -Po r a t , T. & R ix o n , F. J. (1979). Replication of herpesvirus DNA. IV. Analysis of concatemers. Virology 94, 61-70.

B e r g s m a , D. J., O l iv e , D. M., H a r t z e l l , S . W. & S u b r a m a n ia n , K. N. (1982). Territorial limits and functional anatomy of the simian virus 40 replication origin. Proc. natn. Acad. Sci. U.S.A. 79, 381-385.

B is w a l , N., F e l d a n , P. & L e v y , C. C. (1983). A DNA topoisomerase activity copurifies with the DNA polymerase induced by herpes simplex virus. Biochim. biophys. Acta 740, 379-389.

C a r a s , I. W ., L e v in s o n , B. B., F a r b r y , M., W il l ia m s , S . R . & M a r t in , D. W . (1985). Cloned mouse ribonucleotide reductase subunit M l cDNA reveals amino acid sequence homology with Escherichia coli and herpesvirus ribonucleotide reductases. J . Biol. Chem. 260, 7015-7022.

C h a l l b e r g , M. D. (1986). A method for identifying the viral genes required for herpesvirus DNA replication. Proc. natn. Acad. Sci. U.S.A. 83 , 9094-9098.

C h a r t r a n d , P., St o w , N. D., T im b u r y , M. C . & W il k ie , N. M. (1979). Physical mapping of paaT mutations of herpes simplex virus type 1 and type 2 by intertypic marker rescue. J. Virol. 31, 265-276.

C h o u , J. & R o iz m a n , B . (1985). Isomerization of herpes simplex virus 1 genome: Identification of the cw-acting and recombination sites within the domain of the a sequence. Cell 41, 803-811.

C h o u , J . & R o iz m a n , B . (1986). The terminal a sequence of the herpes simplex virus genome contains the promoter of a gene located in the repeat sequences of the L component. J'. Virol. 57, 629-637.

C o e n , D . M., A s c h m a n , D . P., G e l e p , P. T ., R e t o n d o , M. J., W e l l e r , S. K . & Sc h a f f e r ,

P. A . (1984). Fine mapping and molecular cloning of mutations in the herpes simplex virus D N A polymerase locus. J . Virol. 49, 236-247.

C o e n , D. M., F u r m a n , P. A., G e l e p , P. T. & Sc h a f f e r , P. A. (1982). Mutations in the herpes

sim plex virus DNA polymerase gene can confer resistance to 9-/3-D-arabinofuranosyladenine.

J . Virol. 41, 909-918.

C o h e n , G. H. (1972). Ribonucleotide reductase activity of synchronized KB cells infected with herpes simplex virus. J . Virol. 9, 408—418.

C o n l e y , A. J., K n ip e , D. M., I o n e s , P. C . & R o iz m a n , B . (1981). Molecular genetics of herpes simplex virus. V II. Characterization of a temperature-sensitive mutant produced by in vitro mutagenesis and defective in DNA synthesis and accumulation of y polypeptides. J . Virol. 37, 191-206.

C o st a , R. H ., D r a p e r , K . G., K e l l y , T. J. & W a g n e r , E . K . (1985). A n unusua l spliced herpes

sim plex virus type 1 transcrip t w ith sequence hom ology to Epstein-Barr virus D N A . J . Virol. 54, 317-328.

C r u m p a c k e r , C . S ., C h a r t r a n d , P., S u b a k -Sh a r p e , J. H. & W il k ie , N. M . (1980). Resistance of herpes simplex virus to acycloguanosine - genetic and physical analysis. Virology 105, 171-184.

D a l z ie l , R. G. & M a r s d e n , H. S. (1984). Identification of two herpes simplex virus type 1- induced proteins (21K and 22K) which interact specifically with the a sequence of herpes simplex virus D N A ._7 . gen. Virol. 65, 1467-1475.

DAVISON, A. J. (1984). Structure of the genome termini of varicella-zoster virus, jf. gen. Virol. 65, 1969-1977.

D a v is o n , A. J. & M c G e o c h , D . J. (1986). Evolutionary comparisons of the S segments in the genomes of herpes simplex virus type 1 and varicella-zoster virus.,7. gen. Virol. 67, 597-611.

D a v is o n , A . J. & S cott , J. E . (1986). T he complete D N A sequence of varicella-zoster virus.

J. gen. Virol. 67, 1759-1816.

90 D. J. McGeoch

D a v is o n , A. J . & T a y l o r , P. (1987). Genetic relations between varicella-zoster virus and Epstein- Barr virus. J . gen. Virol, (in press).

D a v is o n , A. J. & W il k ie , N. M. (1981). Nucleotide sequences of the joint between the L and S

segments of herpes simplex virus types 1 and 2. J. gen. Virol. 55, 315-331.

D a v is o n , A . J . & W il k ie , N . M . (1983a). Inversion of the two segments o f the herpes simplex

v irus genom e in intertyp ic recom binants . J . gen. Virol. 64, 1-18.

D a v is o n , A. J . & W il k ie , N. M. (19836). Either orientation of the L segment of the herpes simplex virus genome may participate in the production of viable intertypic recombinants. J . gen. Virol. 64, 247-250.

D a v is o n , A. J . & W il k ie , N. M. (1983c). Location and orientation of homologous sequences in the genomes of five herpesviruses. J'. gen. Virol. 64, 1927-1942.

D o v e r , G. (1982). Molecular drive: a cohesive mode of species evolution. Nature, Lond. 299, 111-117.

D u t ia , B . M . (1983). Ribonucleotide reductase induced by herpes simplex virus has a virus specified constituent.^, gen. Virol. 65, 513-521.

E a r l , P. L., J o n e s , E . V. & Moss, B. (1986). Homology between DNA polymerases of poxviruses, herpesviruses, and adenoviruses: nucleotide sequence of the vaccinia virus DNA polymerase gene. Proc. natn. Acad. Sci. U.S.A. 83, 3659-3663.

E l ia s , P ., O ’D o n n e l l , M . E . , M o c a r s k i, E . S. & L e h m a n , I . R . (1986). A D N A - b in d in g prote in

specific for an orig in of replication of herpes sim plex virus type 1. Proc. natn. Acad. Sci. U.S.A. 83, 6322-6326.

E verett , R. D. (1984). Transactivation of transcription of herpes virus products; requirements for

two HSV-1 immediate early polypeptides for maximum activity. EM BOJ. 3, 3135-3141.

E verett , R. D . (1986). The products of herpes simplex virus type 1 (HSV-1) immediate early genes 1, 2 and 3 can activate HSV gene expression in trails. J . gen. Virol. 67, 2507-2513.

F i e l d , H. J. & W i l d y , P. (1978). The pathogenicity of thymidine kinase-deficient mutants of herpes simplex virus in mice. J . Hygiene 81, 261-277.

F is h e r , F . B. & P r e s t o n , V. G. (1986). Isolation and characterisation of herpes simplex virus type1 mutants which fail to induce dUTPase activity. Virology 148, 190-197.

F r a m e , M . C., M a r s d e n , H . S. & D u t ia , B. M . (1985). The ribonucleotide reductase induced by herpes simplex virus type 1 involves minimally a complex of two polypeptides (136K and 38K). J. gen. Virol. 66, 1581-1587.

F r a n c k e , B. & G arret t , B. (1982). The effect of a temperature-sensitive lesion in the alkaline DNase of herpes simplex virus type 2 on the synthesis of viral DNA. Virology 116, 116-127.

F r ie d m a n n , A., Sh l o m a i , J . & Be c k e r , Y. (1977). Electron microscopy of herpes simplex virus DNA molecules isolated from infected cells by centrifugation in CsCl density grandients. J . gen. Virol. 34, 507-522.

G e l m a n , I. H . & S il v e r s t e in , S. (1985). Identification of immediate early genes from herpes simplex virus that transactivate the virus thymidine kinase. Proc. natn. Acad. Sci. U.S.A. 82, 5265-5269.

G ib b s , J. S ., C h io u , H . C ., H a l l , J. D ., M o u n t , D. W ., R e t o n d o , M . J., W e l l e r , S. K. & C o e n , D. M . (1985). Sequence and m app ing analyses of the herpes s im plex virus DNA polymerase gene predict a C-term inal substrate b in d in g dom ain . Proc. natn. Acad. Sci. U.S.A. 82, 7969-7973.

G ib s o n , T., St o c k w e l l , P., G in s b u r g , M. & B a r r e l l , B. (1984). Homology between two EBV early genes and HSV ribonucleotide reductase and 38K genes. Nucl. Acids Res. 12, 5087-5099.

G r a y , C. P. & K a e r n e r , H . C. (1984). Sequence of the putative origin of replication in the region of herpes simplex virus type 1 ANG DNA. J . gen. Virol. 65, 2109—2119.

H a l l , J. D ., F u r m a n , P . A., St C l a ir , M . H . & K n o p f , C . W . (1985). Reduced in vivo mutagenesis by m u tan t herpes sim plex DNA polymerase involves im proved nucleotide

selection. Proc. natn. Acad. Sci. U.S.A. 82, 3889-3893.

H a y , J. & Su b a k -Sh a r p e , J. H . (1976). Mutants of herpes simplex virus types 1 and 2 that are resistant to phosphonoacetic acid induce altered DNA polymerase activities in infected cells. J. gen. Virol. 31, 145-148.

H o f f m a n n , P. J. (1981). Mechanism of degradation of duplex DNA by the DNase induced by herpes simplex virus. J . Virol. 38, 1005-1014. .

The genome of herpes simplex virus 91

H o f f m a n n , P. J. & C h e n g , Y.-C. (1978). The deoxyribonuclease induced after infection of KB cells by herpes simplex virus type 1 or type 2. J . Biol. Chem. 253, 3557-3562.

H o f f m a n n , P. J. & C h e n g , Y.-C. (1979). DNase induced after infection of KB cells by herpes simplex virus type 1 or type 2. II. Characterization of an associated endonuclease activity. jf. Virol. 32, 449-457. "

H o n e s s , R. W. (1984). Herpes simplex and ‘the herpes complex’: diverse observations and a unifying hypothesis. J . gen. Virol. 65, 2077-2107.

H o n e s s , R . W. & R o iz m a n , B. (1974). Regulation of herpesvirus macromolecular synthesis. I. Cascade regulation of the synthesis of three groups of viral proteins. J . Virol. 14, 8-19.

H o n e s s , R. W., B o d e m e r , W., C a m e r o n , K. R., N il l e r , H .- H . & F l e c k e n s t e in , B . (1986). The A + T-rich genome of herpesvirus saimiri contains a highly conserved gene for thymidylate synthase. Proc. natn. Acad. Sei. U.S.A. 83, 3604-3608.

H o n e s s , R. W., P u r if o y , D. J. M., Y o u n g , D., G o p a l , R., C a m m a c k , N. & O ’H a r e , P . (1984). Single mutations at many sites within the DNA polymerase locus of herpes simplex viruses can confer hypersensitivity to aphidicolin and resistance to phosphonoacetic acid. J. gen. Virol. 65, 1-17.

J a c o b , R . J . , M o r s e , L . S. & R o iz m a n , B . (1979). Anatomy of herpes simplex virus DNA. XII. Accumulation of head-to-tail concatemers in nuclei of infected cells and their role in the

generation of the four isomeric arrangements of viral DNA. J. Virol. 29, 448-457.

J a m ie s o n , A. T., G e n t ry , G . A. & Su b a k -Sh a r p e , J . H. (1974). Induction of both thymidine and deoxycytidine kinase activity by herpes viruses. J. gen. Virol. 24, 465-480.

J o n g e n e e l , C. V. & B a c h e n h e im e r , S. L . (1981). Structure of replicating herpes simplex virus DNA. J . Virol. 39, 656-660.

K a e r n e r , H. C., M a ic h l e , I. B., O tt, A. & S c h r ö d e r , C. H. (1979). Origin of two different classes of defective HSV-1 Angelotti DNA. Nucl. Acids Res. 6, 1467-1478.

K e ir , H . M . & G o l d , E . (1963). D eoxyribonucle ic acid nucleotidyltransferase and deoxyribonu ­

clease from cu ltured cells infected w ith herpes sim plex virus. Biochim. biophys. Acta 72, 263-276.

K ie f f , E. D . , B a c h e n h e im e r , S. L . & R o iz m a n , B. (1971). Size, composition and structure of the deoxyribonucleic acid of herpes simplex virus subtypes 1 and 2. J. Virol. 8, 125-132.

K it , S. & D u b b s , D . R . (1963). A cqu is ition of thym id ine kinase activity by herpes simplex

infected m ouse fibroblast cells. Biochem. Biophys. Res. Comm. 11, 55-59.

K n o p f , K .- W . (1979). Properties of herpes simplex virus DNA polymerase and characterization of its associated exonuclease activity. E ur.J . Biochem. 98, 231-244.

L it t ler , E., Pu r if o y , D., M in s o n , A. & Po w e l l , K. L . (1983). Herpes simplex virus non- structural proteins. III. Function of the major DNA-binding protein..?, gen. Virol. 64, 983-995.

L o c k s h o n , D. & G a l l o w a y , D. A. (1986). Cloning and characterization of oh'l2 , a large palindromic DNA replication origin of herpes simplex virus type 2 .J . Virol. 58, 513-521.

L o n g n e c k e r , R . & R o iz m a n , B. (1986). Generation of an inverting herpes simplex virus 1 mutant lacking the L-S junction a sequences, an origin of DNA synthesis, and several genes including those specifying glycoprotein E and the alpha 47 gene. J. Virol. 58, 583-591.

L u p t o n , S. & L e v in e , A. J. (1985). Mapping genetic elements of Epstein-Barr virus that facilitate extrachromosomal persistence of Epstein-Barr virus-derived plasmids in human cells. Mol. cell. Biol. 5, 2533-2542.

M a t t h e w s , R. E. F. (1982). Classification and nomenclature of viruses. Intervirology 17, 1-200.M c G e o c h , D. J . (1984). The nature of animal virus genetic material. In The Microbe 1984, part 1,

Viruses (ed. B. W. J. Mahy & J. R. Pattison), pp. 75-107. Cambridge: Cambridge University

Press.M c G e o c h , D . J. & D a v is o n , A. J. (1986a). Alphaherpesviruses possess a gene homologous to the

protein kinase gene family of eukaryotes and retroviruses. Nucl. Acids Res. 14, 1765-1777.

M c G e o c h , D . J. & D a v is o n , A. J. (1986£>). DNA sequence of the herpes simplex virus type 1 gene encoding glycoprotein gH, and identification of homologues in the genome of varicella-zoster virus and Epstein-Barr virus. Nucl. Acids Res. 14, 4281-4292.

M c G e o c h , D . J., D o l a n , A ., D o n a l d , S. & Br a u e r , D . H. K. (1986a). Complete D N A

sequence of the short repeat region in the genome of herpes simplex virus type 1. Nucl. Acids Res. 14, 1727-1745.

92 D. J. McGeoch

M c G e o c h , D . J., D o l a n , A ., D o n a l d , S. & R i x o n , F. J. (1985). Sequence determination and genetic content of the short unique region in the genome of herpes simplex virus type 1. J . mol. Biol. 181, 1-13.

M c G e o c h , D . J., D o l a n , A . & F r a m e , M . C . (19866). D N A sequence of the region in the genom e

of herpes s im plex virus type 1 conta in ing the exonuclease gene and ne ighbouring genes. Nucl. Acids Res. 14, 3435-3447.

M c L a u c h l a n , J. & C l e m e n t s , J. B. (1983). Organization of the herpes simplex virus type 1 transcription unit encoding two early proteins with molecular weights of 140000 and 40 000. J . gen. Virol. 64, 997-1006.

M o c a r s k i, E. S. & R o iz m a n , B . (1982). Structure and role of the herpes simplex virus DNA termini in inversion, circularization and generation of virion DNA. Cell 31, 89-97.

M o r r is o n , J. M . & K e ir , H. M . (1968). A new DNA-exonuclease in cells infected with herpes virus: partial purification and properties of the enzyme. J. gen. Virol. 3, 337-347.

Moss, H. (1986). The herpes simplex virus type 2 alkaline DNase activity is essential for replication and growth. J . gen. Virol. 67, 1173—1178.

Moss, H ., C h a r t r a n d , P ., T im b u r y , M. C . & H a y , J. (1979). Mutant of herpes simplex virus type 2 with temperature-sensitive lesions affecting virion thermostability and DNase activity: identification of the lethal mutation and physical mapping of the nuc~ lesion. J . Virol. 32, 140-146.

M u l l e r , M . T., B o l l e s , C. S. & P a r r is , D. S. (1985). Association of type I DNA topoisomerase with herpes simplex virus, jf. gen. Virol. 66, 1565-1574.

M u r c h ie , M.-J. & M c G e o c h , D. J. (1982). DNA sequence analysis of an immediate-early gene region of the herpes simplex virus type 1 genome (map coordinates 0.950 to 0.978).^. gen. Virol. 62, 1-15.

O ’H a r e , P . & H a y w a r d , G. S. (1985). Evidence for a direct role for both the 175,000 and 110,000 molecular weight immediate-early proteins of herpes simplex virus in the trans activation of delayed-early promoters. J'. Virol. 53, 751-760.

P ell e t t , P . E., B ig g in , M. D ., Ba r r e l l , B. & R o iz m a n , B. (1985). Epstein-Barr virus genome may encode a protein showing significant amino acid and predicted secondary structure homology with glycoprotein B of herpes simplex virus 1. Jf. Virol. 56, 807-813.

Pe r r y , L . J. (1986). DNA sequence analysis of the repeat and adjoining unique region of the longsegment of herpes simplex virus type 1. P h D Thesis, University of Glasgow.

P e r r y , L. J., R ix o n , F. J., E verett , R . D., F r a m e , M . C. & M c G e o c h , D. J. (1986). Characterization of the IEllOgene of herpes simplex virus type 1 .J .gen . Virol. 67, 2365-2380.

Po f f e n b e r g e r , K. L. & R o iz m a n , B. (1985). A noninverting genome of a viable herpes simplex virus 1: Presence of head-to-tail linkages in packaged genomes and requirements for circulariz­ation after infection. J. Virol. 53, 587-595.

Po w e l l , K. L . & P u r if o y , D. J. M. (1977). Nonstructural proteins of herpes simplex virus.I. Purification of the induced DNA polymerase. J'. Virol. 24, 618-626.

Po w e l l , K . L . , L it t ler , E . & PURIFOY, D.J.M .(1981). Nonstructural proteins of herpes simplex virus. II. Major virus-specific DNA-binding protein J ’. Virol. 39, 894-902.

Pr e s t o n , C . M., C o r d in g l e y , M. G. & St o w , N. D. (1984). Analysis of DNA sequences which regulate the transcription of a herpes simplex virus immediate early gene. J . Virol. 50, 708-716.

P r e s t o n , V. G. & F is h e r , F . B . (1984). Identification of the herpes simplex virus type 1 gene encoding the duTPase. Virology 138, 58-68.

Pr e s t o n , V. G., P a l f r e y m a n , J. W. & D u t ia , B. M. (1984). Identification of a herpes simplex virus type 1 polypeptide which is a component of the virus-induced ribonucleotide reductase. J. gen. Virol. 65, 1457-1466.

P u r if o y , D. J. M. & Po w e l l , K. L . (1981). Temperature-sensitive mutants in two distinct complementation groups of herpes simplex virus type 1 specify thermolabile DNA polymerase. J . gen. Virol. 54, 219-222.

Q u in n , J . P . & M cG e o c h , D. J. (1985). DNA sequence of the region in the genom e of herpes

s im p lex virus type 1 con ta in ing the genes for DNA polymerase and the m ajor DNA b in d in g

p ro te in . Nucl. Acids Res. 13, 8143-8163.R a p p , F . & W ig d a h l , B. (1983). Herpesvirus target considerations for the design of antiviral

agents. In Targets for the design o f antiviral agents (ed. E. De Clercq & R . T . Walker), pp. 29-60. New York: Plenum Press.

The genome of herpes simplex virus 93

R a w l in s , D. R ., M il m a n , G., H a y w a r d , S. D. & H a y w a r d , G. S. (1985). Sequence-specific DNA binding of the Epstein-Barr virus nuclear antigen (EBNA-1) to clustered sites in the plasmid maintenance region. Cell 42, 859-868.

R e is m a n , D., Y a t es , J. & S u g d e n , B. (1985). A putative origin of replication of plasmids derived from Epstein-Barr virus is composed of two as-acting components. Mol. cell. Biol. 5, 1822-1832.

R i x o n , F. J. & C l e m e n t s , J. B. (1982). Deta iled structural analysis of two spliced HSV-1 immediate-early m R N A s . Nucl. Acids Res. 10, 2241-2256.

R ix o n , F. J. & M c G e o c h , D. J. (1984). A 3' co-terminal fam ily of m R N A s from the herpes

sim plex virus type 1 short region: T w o overlapp ing reading frames encode unrelated

polypeptides one of w h ich has a h igh ly reiterated am ino acid sequence. Nucl. Acids Res. 12, 2473-2487. '

R ix o n , F . J . & M c G e o c h , D . J . (1985). Deta iled analysis of the m R N A s m app ing in the short

u n ique region of herpes sim plex virus type 1. Nucl. Acids Res. 13, 953-973.

R ix o n , F. J., C a m p b e l l , M. E. & C l e m e n t s , J. B. (1984). A tandemly reiterated DNA sequence in the long repeat region of herpes simplex virus type 1 found in close proximity to immediate- early mRNA 1 .J . Virol. 52, 715-718.

R o iz m a n , B. (1979). The structure and isomerization of herpes simplex virus genomes. Cell 16, 481-494.

R o iz m a n , B. (1982). The family herpesviridae: General description, taxonomy, and classification. In The herpesviruses, vol. 1 (ed. B. Roizman), pp. 1-23. New York and London: Plenum Press.

S h e l d r ic k , P. & B e r t h e l o t , N. (1974). Inverted repetitions in the chromosome of herpes simplex virus. Cold Spring Harbor Symp. Quant. Biol. 39, 667-678.

S pa et e , R. R. & F r e n k e l , N. (1982). The herpes simplex virus amplicon: a new eukaryotic defective-virus cloning-amplifying vector. Cell 30, 295-304.

St a n d a r t , N . , B r a y , S . J., G e o r g e , E. L., H u n t , T . & R u d e r m a n , J. V. (1985). T he small

s ubun it of ribonuc leotide reductase is encoded by one of the m ost abundan t translationally

regulated m aternal R N A s in clam and sea urch in eggs. J . Cell Biol. 100, 1968—1976.

Stow, N. D. (1982). Localization of an origin of DNA replication within the TRg/lRg repeated region of the herpes simplex virus type 1 genome. E M BO J. 1, 863-867.

St o w , N. D. (1985). Mutagenesis of a herpes simplex virus origin of DNA replication and its-effect on viral interference. .7- gen. Virol. 66, 31-42.

St o w , N . D . & D a v is o n , A . J. (1986). Identification of a varicella-zoster virus origin of D N A

replication and its activation by herpes simplex virus type 1 gene products. J . gen. Virol. 67, 1613-1623.

S to w , N. D. & M c M o n a g le , E . C. (1983). Characterization of the T R g / lR g origin of DNA replication of herpes simplex virus type 1. Virology 130, 427-438.

S t o w , N . D . , M cM o n a g l e , E. C. & D a v is o n , A . J. (1983). Fragments from both termini of the herpes simplex virus type 1 genome contain signals required for the encapsidation of viral D N A .

Nucl. Acids Res. 11, 8205-8220.

S t ro b e l-Fid l e r , M. & F r a n c k e , B. (1980). Alkaline deoxyribonuclease induced by herpes simplex virus type 1: Composition and properties of the purified enzyme. Virology 103, 493-501.

V a r m u z a , S . & S m il e y , J. R. (1985). Signals for site-specific cleavage of HSV DNA: Maturation involves two separate cleavage events at sites distal to the recognition sequences. Cell 41, 793-802.

V a u g h a n , P . J., P u r if o y , D. J. & Po w e l l , K. L. (1985). DNA-binding protein associated with herpes simplex virus DNA polymerase. J. Virol. 53, 501-508.

V l a z n y , D. A. & F r e n k e l , N. (1981). Replication of herpes simplex virus DNA: localization of replication recognition signals within defective virus genomes. Proc. natn. Acad. Sci. U.S.A. 78, 742-746.

V l a z n y , D. A., K w o n g , A. & F r e n k e l , N. (1982). Site-specific cleavage/packaging of herpes simplex virus DNA and the selective maturation of nucleocapsids containing full-length DNA. Proc. natn. Acad. Sci. U.S.A. 79, 1423-1427.

W a g n e r , E. K. (1985). Individual HSV transcripts: Characterization of specific genes. In The herpesviruses, vol. 3 (ed. B. Roizman), pp. 45-104. New York and London: Plenum Press.

W a g n e r , M. J. & S u m m e r s , W . C. (1978). Structure of the joint region and the termini of the DNA of herpes simplex virus type 1. J'. Virol. 27, 374-387.

94 D. J. McGeoch

W a t s o n , R. J., Pr e s t o n , C . M. & C le m e n t s , J. B. (1979). Separation and characterization of herpes simplex virus type 1 immediate-early mRNAs. J . Virol. 31, 42-52.

W e l l e r , S . K., L e e , K. J., Sa b o u r in , D. J. & Sc h a f f e r , P. A. (1983). Genetic analysis of temperature-sensitive mutants which define the gene for the major herpes simplex virus type 1 DNA-binding protein. J . Virol. 45, 354-366.

W e l l e r , S . K., S p a d a r o , A., Sc h a f f e r , J. E., M u r r a y , A. E., M a x a m , A. M . & Sc h a f f e r ,

P. A. (1985). Cloning, sequencing, and functional analysis of onL, a herpes simplex virus type 1 origin of DNA synthesis. Mol. cell. Biol. 5, 930-942.

W h it t o n , J. L. & C le m e n t s , J. B. (1984). Replication origins and a sequence involved in coordinate induction of the immediate early gene family are conserved in an intergenic region of herpes simplex virus. Nucl. Acids Res. 12, 2061-2079.

WOHLRAB, F . & FRANCKE, B. (1980). Deoxyribopyrimidine triphosphatase activity specific for cells infected with herpes simplex virus type 1. Proc. natn. Acad. Sci. U.S.A. 77, 1872-1876.

Y a t es , J., W a r r e n , N., R e is m a n , D. & Su g d e n , B. (1984). A cis acting element from the Epstein-Barr viral genome that permits stable replication of recombinant plasmids in latently infected cells. Proc. natn. Acad. Sci. U.S.A. 81, 3806-3810.

Y a t es , J. L., W a r r e n , N. & S u g d e n , B. (1985). Stable replication of plasmids derived from Epstein-Barr virus in various mammalian cells. Nature, Lond. 313, 812-815.