Virus Research 117 (2006) 90–104 Review Topics in herpesvirus genomics and evolution Duncan J. McGeoch ∗ , Frazer J. Rixon, Andrew J. Davison Medical Research Council Virology Unit, Institute of Virology, University of Glasgow, Church Street, Glasgow G11 5JR, United Kingdom Available online 21 February 2006 Abstract Herpesviruses comprise an abundant, widely distributed group of large DNA viruses of humans and other vertebrates, and overall are among the most extensively studied large DNA viruses. Many herpesvirus genome sequences have been determined, and interpreted in terms of gene contents to give detailed views of both ubiquitous and lineage-specific functions. Availability of gene sequences has also enabled evaluations of evolutionary relationships. For herpesviruses of mammals, a robust phylogenetic tree has been constructed, which shows many features characteristic of synchronous development of virus and host lineages over large evolutionary timespans. It has also emerged that three distinct groupings of herpesviruses exist: the first containing viruses with mammals, birds and reptiles as natural hosts; the second containing viruses of amphibians and fish; and the third consisting of a single invertebrate herpesvirus. Within each of the first two groups, the genomes show clear evidence of descent from a common ancestor, but relationships between the three groups are extremely remote. Detailed analyses of capsid structures provide the best evidence for a common origin of the three groups. At a finer level, the structure of the capsid shell protein further suggests an element of common origin between herpesviruses and tailed DNA bacteriophages. © 2006 Elsevier B.V. All rights reserved. Keywords: Herpesviridae; Genome annotation; Phylogeny; Coevolution; Capsid structure Contents 1. 2. 3. 4. 5. 6. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Herpesvirus genome sequences and their interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Herpesvirus gene complements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Phylogenetic relationships of mammalian herpesviruses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Phylogenetic relationships among reptilian, avian and mammalian lineages of the Alphaherpesvirinae . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Herpesvirus capsid structures and deep phylogenetic connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 1. Introduction The herpesviruses (HVs) are a group of large DNA viruses with a distinctive virion architecture. Historically, from the 1960s to the 1980s, assignment as a HV was made on the basis of virion morphology. As illustrated in Fig. 1, the HV particle has several distinct components. The genomic DNA is densely packed within an icosahedral (T = 16) capsid, which has 162 surface capsomeres and is of diameter around 115–130 nm. Of the capsomeres, 150 are primarily composed of six molecules of ∗ Corresponding author. Tel.: +44 141 330 4645; fax: +44 141 337 2236. E-mail address: d.mcgeoch@vir.gla.ac.uk (D.J. McGeoch). 0168-1702/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.virusres.2006.01.002 one protein species, and 11 are pentamers of the same protein. The final pentameric position is occupied by the portal complex. Enclosing the capsid is an amorphous layer, the tegument, composed of several protein species. This in turn is bounded by the outermost element, a lipid bilayer envelope with embedded protein molecules (typically glycosylated), giving an overall diameter of about 200 nm. The whole structure presents a characteristic appearance in negatively-stained or thin-sectioned electron microscopic images, and was used to define membership of the taxonomic family Herpesviridae. Biological criteria were then used to make assignments to three subfamilies, the Alpha-, Beta- and Gammaherpesvirinae (Davison et al., 2005a). D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Fig. 1. Electron cryo-microscopic image of an HSV-1 virion. The icosahedral nucleocapsid (c) is embedded in a complex proteinaceous layer called the tegument (t) and enclosed by a lipid envelope containing numerous glycoproteins (e) (bar = 100 nm). DNA sequence data for HV genomes have been accumulating from the start of the 1980s and are now extensive, with many complete genome sequences and large numbers of partial sequences. It became apparent from sequence-based comparisons that mammalian and avian HVs were descended from a common ancestor and that the three taxonomic subfamilies corresponded to major distinct lineages (with only a few virus species reassigned among subfamilies on the basis of sequence relationships). Sequence comparisons then became the primary approach for evaluating phylogenetic and taxonomic relationships among mammalian and avian HVs, and for identifying newly characterized viruses as members of the Herpesviridae and assigning them to subfamilies. Mammalian HVs populate all three subfamilies, while all characterized avian HVs are in the Alphaherpesvirinae. Recently it has become clear that characterized reptilian HVs also belong to the Alphaherpesvirinae (Herbst et al., 2004; Greenblatt et al., 2005; McGeoch and Gatherer, 2005). A completely different picture has emerged for piscine, amphibian and invertebrate HVs: by criteria of their gene contents, piscine and amphibian HVs form a separate group which appears essentially unrelated to the mammalian/avian/reptilian virus group, and the single known invertebrate HV (of bivalve molluscs) belongs to a third group distinct from both of the vertebrate virus groups (Davison, 2002; Davison et al., 2005b). Thus, at this level, the criterion of a common virion morphology appears to indicate the existence of relationships more distant than can be inferred from DNA or protein sequences. The Herpesviridae Study Group of the International Committee on Taxonomy of Viruses has developed proposals to revise higher level taxonomic arrangements for HVs in order to encompass the above findings. At the time of writing this text, these have not been finally passed into accepted taxonomy. However, with the reasonable expectation that this updated taxonomy will 91 enter general use in the near future, we employ it here. The developments proposed are as follows. First, membership of the family Herpesviridae will be restricted to viruses which belong to the Alpha-, Beta- and Gammaherpesvirinae subfamilies on the basis of their gene contents and sequences; these comprise all characterized HVs with mammalian, avian and reptilian hosts. Next, the group of piscine and amphibian HVs will be assigned to a new family, the Alloherpesviridae, and the single known invertebrate HV to another new family, the Malacoherpesviridae. Third, all three families will be grouped in a new higher level taxon, the order Herpesvirales. Additionally, three new genera in the Herpesviridae are proposed. Proboscivirus in the Betaherpesvirinae will contain elephant endothelial HV (EEHV). In the Gammaherpesvirinae, the lineage containing alcelaphine HV 1 (AHV-1) and certain other artiodactyl HVs will form the genus Macavirus, and the lineage containing equine HV 2 (EHV-2) and certain other perissodactyl and carnivore HVs will form the genus Percavirus. Species in both of these last two were formerly assigned to the genus Rhadinovirus, which will become more tightly demarcated. These revised arrangements for genera are clarified by the phylogenetic trees in Fig. 4, as discussed in Section 4. In this review, we first present an evaluation of the current state of HV genomics, and we then describe what can be inferred of the phylogeny of the Herpesvirales from analysis of encoded amino acid sequences. Given that 90% of characterized HVs belong to the revised Herpesviridae, and also by far the greatest part of experimental investigations have concerned members of this family, our treatment inevitably has these viruses in the foreground. Lastly, we move beyond sequence-oriented analyses to outline deep evolutionary connections that are currently emerging from protein structural investigations, first in the common capsid morphology of the three families of the Herpesvirales, and then in similarities that have been observed in the threedimensional (3D) structures of capsid proteins of HVs and large DNA bacteriophages which may suggest an element of common ancestry for these disparate virus groupings. 2. Herpesvirus genome sequences and their interpretation Genomes isolated from herpesvirions consist of linear, double-stranded DNA with unpaired, complementary nucleotides at each terminus. Genomes of the Herpesviridae range in size from 124 kbp (simian varicella virus (SVV) from the Alphaherpesvirinae; Gray et al., 2001) to 241 kbp (chimpanzee cytomegalovirus from the Betaherpesvirinae; Davison et al., 2003). The genome of channel catfish virus (CCV), the sole sequenced member of the Alloherpesviridae, is 134 kbp in size (Davison, 1992), and viruses in this family that infect carp have the largest known genomes among the Herpesvirales (295 kbp; Hutoran et al., 2005; Waltzek et al., 2005). The single member of the Malacoherpesviridae (ostreid HV 1 (OsHV-1)) has a genome of 207 kbp (Davison et al., 2005b). Nucleotide compositions range widely, from 32 to 75% G + C, even for viruses in the same genus (Honess, 1984). HV DNA molecules characteristically contain regions of unique sequence flanked by direct or inverted repeats. Certain 92 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 genomes (exemplified by herpes simplex virus type 1 (HSV-1)) are particularly complex, containing two unique regions, with one or both flanked by inverted repeats (Sheldrick and Berthelot, 1975; Wadsworth et al., 1975). In such genomes, the unique regions may be present in either orientation in virion DNA (Hayward et al., 1975) presumably as a consequence of recombination between repeats in concatemeric DNA produced during DNA replication. This phenomenon appears to be unconnected to any phenotype of virus functionality (Jenkins and Roizman, 1986). Similar genome arrangements appear to have arisen separately on different occasions during HV evolution (Davison, 1998; Davison et al., 2005b). As a rough approximation, protein-coding genes are distributed with an overall density of one gene per 1.5–2 kbp of HV DNA. They are transcribed by host RNA polymerase II (Alwine et al., 1974). Some of the first genes to be expressed during infection (immediate early genes) encode proteins that regulate the expression of other genes (early and late genes) (Honess and Roizman, 1974). In broad outline, early genes encode enzymes involved in nucleotide metabolism and DNA replication, plus a number of envelope glycoproteins, and late genes encode mainly proteins that make up the virion. With the exception of a few genes that are expressed by splicing from a common 5′ -leader, such as the EBNA genes of Epstein-Barr virus (EBV) (Bodescot et al., 1987), each gene has its own promoter. However, it is common for families of consecutive and similarly oriented genes to share a polyadenylation site downstream from the member at the 3′ -end, giving rise to overlapping, 3′ -coterminal transcripts (Wagner, 1985). Introns are thought to occur only in a minority of HV genes, and to be more common in genes that are specific to a genus or to a subset of species within a genus. Members of the Beta- and Gammaherpesvirinae have more intron-containing genes than those of the Alphaherpesvirinae. Protein-coding regions occupy the great majority of a HV genome. However, sizeable regions also exist that apparently do not specify proteins. Some of these regions express RNAs that appear not to function via translation. Large non-coding RNAs include the latency-associated transcripts in HSV-1 (Stevens et al., 1987) and a spliced RNA of unknown function in human cytomegalovirus (HCMV) (Kulesza and Shenk, 2004). Small RNAs probably transcribed by RNA polymerase III are encoded by members of subfamily Gammaherpesvirinae; examples include the EBER species of EBV (Rosa et al., 1981) and the tRNA-like transcripts of murid HV 4 (MuHV-4) (Bowden et al., 1997). MicroRNAs have been predicted or identified in members of the Alpha-, Beta- and Gammaherpesvirinae (Pfeffer et al., 2004, 2005; Cai et al., 2005; Grey et al., 2005). The functional significance, if any, of these transcripts is a subject of intense interest at present; those among the Gammaherpesvirinae seeming the most promising. Complete sequences are available in the public databases for 56 HV genomes, representing some 37 species (Table 1). Although most sequences have eventually been found to contain errors, the overall level of accuracy is very high. However, the quality of gene description is variable. This is in part because there is no single set of criteria that may be applied with a guarantee of identifying all genes in a sequence. Even though most protein-coding genes can be identified relatively easily, assignments are usually made less confidently (or not at all) for small, spliced, overlapping or poorly conserved protein-coding regions, and in instances where translation initiates in a potential protein-coding region from an internal ATG (methionine) codon or from a codon other than ATG. Also, most criteria are aimed at identifying protein-coding genes, and are not useful for detecting genes that encode functional, non-translated RNAs. In annotating genes, it is important to use as many criteria as possible to develop reliable models of gene content. Most dubious candidates have arisen through over-reliance on a particular analytical tool or a failure to recognize the evolutionary nature of genes. However, given that credible genes are still discovered occasionally even in well characterized HVs, new candidates deserve rigorous examination. A primary step in gene annotation involves identifying potential protein-coding regions lacking in-frame termination codons, that is, open reading frames (ORFs). The number of ORFs may be kept to a manageable number, at least in an initial analysis, by setting a minimum size (e.g. 60 codons). A number of features of HV genes that have been recognized over the years can also help to refine the initial set. For example, the general rarity of splicing or the absence of suitably located splice sites in specific cases may be used to place queries against ORFs in which in-frame ATG codons are lacking or are located so that the putative coding region is very short. Additional indicators may emerge from the perceived rarity of extensive overlap between protein-coding regions and the locations of potential transcriptional signals (a TATA element near the 5′ -end of the ORF and an AATAAA element near the 3′ -end of the ORF or 3′ -coterminal family of ORFs). Experimental data on the expression of an RNA or protein from a putative protein-coding region, or a phenotype arising from mutagenesis studies, are also relevant. These exercises greatly help the derivation of an initial gene map for further analysis. One of the most powerful aids in gene identification is that of comparative genomics, which rests upon the nature of genes as entities shaped by evolution. A very useful approach in this category is that of alignment-based comparisons of nucleic acid or (more usually) amino acid sequence between orthologous genes. This has considerable utility even for small or highly spliced genes (for example, with alleles of human HV 8 (HHV-8) gene K15; Glenn et al., 1999), and may be used even to check regions that do not emerge as protein-coding from initial ORF identification. The power of sequence comparisons depends upon the degree of relatedness. Conservation in very closely related HVs (e.g. strains of the same species) does not usually constitute convincing evidence, since the evolutionary distance may be too small to differentiate coding from non-coding regions. Conservation in more distantly related HVs (e.g. different genera or subfamilies) is very compelling, but is limited by the fact that distantly related viruses usually contain numerous lineage-specific genes. Analyses that offer good discrimination and inclusion of as many genes as possible often arise from comparisons between relatively closely related species in the same genus (e.g. primate members of genus Simplexvirus, Cytomegalovirus or Rhadinovirus; 93 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Table 1 Sequenced HV species Common name of virus Abbreviation Accession Size (bp)a Herpesviridae Alphaherpesvirinae Simplexvirus Herpes simplex virus type 1 Herpes simplex virus type 2 B virus Simian agent 8 HSV-1 HSV-2 HVB SA8 X14112 Z86099 AF533768 AY714813 152261 154746 156789 150715 Varicellovirus Varicella-zoster virus Simian varicella virus Bovine herpesvirus 1 Bovine herpesvirus 5 Pseudorabies virus Equine herpesvirus 1 Equine herpesvirus 4 VZV SVV BHV-1 BHV-5 PRV EHV-1 EHV-4 X04370 AF275348 AJ004801 AY261359 BK001744 AY665713 AF030027 124884 124138 135301 138390 143461 150224 145597 Mardivirus Marek’s disease virus type 1 Marek’s disease virus type 2 Herpesvirus of turkey MDV-1 MDV-2 HVT AF243438 AB049735 AF291866 177874 164270 159160 Iltovirus Infectious laryngotracheitis virus Psittacid herpesvirus 1 ILTV PsHV-1 NC 0066283 AY372243 148687 163025 HCMV CCMV RhCMV AY446894 AF480884 AY186194 235645 241087 221454 Muromegalovirus Murine cytomegalovirus Rat cytomegalovirus MCMV RCMV U68299 AF232689 230278 230138 Roseolovirus Human herpesvirus 6 Human herpesvirus 7 HHV-6 HHV-7 X83413 U43400 159321 144861 Unassigned Tupaiid herpesvirus THV AF281817 195859 EBV RhLCV MarLCV AJ507799 AY037858 AF319782 171823 171096 149696 [137508] 801 133719 112930 1444 108409 1582 108873 2267 119450 Betaherpesvirinae Cytomegalovirus Human cytomegalovirus Chimpanzee cytomegalovirus Rhesus cytomegalovirus Gammaherpesvirinae Lymphocryptovirus Epstein-Barr virus Rhesus lymphocryptovirus Marmoset lymphocryptovirus Rhadinovirus Human herpesvirus 8 HHV-8 Rhesus rhadinovirus RRV Herpesvirus saimiri HVS Herpesvirus ateles HVA Bovine herpesvirus 4 BHV-4 Murid herpesvirus 4 MuHV-4 U75698 U75699 AF083501 X64346 K03361 AF083424 AF126541 AF318573 AF092919 U97553 EHV-2 U20824 184427 AHV-1 AF005370 AF005368 130608 1113 Percavirus Equine herpesvirus 2 Macavirus Alcelaphine herpesvirus 1 94 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Table 1 (Continued ) Size (bp)a Common name of virus Abbreviation Accession Alloherpesviridae Channel catfish virus CCV M75136 134226 Malacoherpesviridae Ostreid herpesvirus 1 OsHV-1 AY509253 207439 Proposed taxonomical revisions have been incorporated. Where several strains of a species have been analysed, the first published sequence is listed, except where a later publication described a genome significantly closer to wild type. a Square brackets indicate a sequence that falls marginally short of full length. Genome sizes for genera Lymphocrypto-, Rhadino- and Macavirus in subfamily Gammaherpesvirinae are larger than those listed owing to the presence of variable copy numbers of terminal repeats at the genome ends. Where a single size is given, this is for either the unique region flanked by partial terminal repeats or for the unique region only. Where two sizes are given, the first is for the unique region and the second for the terminal repeat unit. Overall genome sizes for these genera are approximately 150–180 kbp. Perelygina et al., 2003; Davison et al., 2003; Searles et al., 1999). Nonetheless, the presence of lineage-specific genes even in closely related species implies that comparative genomics should not be used as an exclusive criterion for gene identification. A second approach to comparative genomics that has long been used in annotation focuses on patterns of nucleotide and codon usage in order to produce a probabilistic evaluation of coding potential. The utility of tools in this category depends upon the source of the input reference pattern. Moreover, the output is often at its least discriminating for the same types of genes that are refractory to other criteria—those containing small, spliced or overlapping protein-coding regions. Injudicious assessment of the output tends to inflate the number of proposed genes by adding numerous small ORFs, often overlapped by larger ORFs whose protein-coding credentials are much more respectable. Particular caution is required for overlapping protein-coding regions, since conservation of a genuine gene can result in indirect conservation of sequences or patterns on the opposing strand and give the misleading impression that an overlapping gene is present. Indeed, the postulation of “cryptic” genes has featured in the genomics of organisms over many years (for an early refutation, see Sharp, 1985). HV genome annotation has not been free from this kind of attention. In summary, a broadly based, rigorous approach is most likely to yield the most reliable gene maps. 3. Herpesvirus gene complements Current conservative estimates for the number of proteincoding genes in the human HVs, as representatives of the three subfamilies of the Herpesviridae, are: HSV-1 and HSV-2, 74 (Dolan et al., 1998); varicella-zoster virus (VZV), 70 (Davison, 2000; Kemble et al., 2000); HCMV, 165 (Dolan et al., 2004); human HV 6 (HHV-6), 86 (Megaw et al., 1998; Zou et al., 1999; French et al., 1999); human HV 7 (HHV-7), 84 (Megaw et al., 1998); EBV, 80 (de Jesus et al., 2003; McGeoch and Davison, unpublished data); and HHV-8, 86 (Neipel et al., 1997). These numbers register duplicates in repeated regions of the genome once only. CCV, the sole sequenced member of the family Alloherpesviridae, has 79 genes (Davison, 1992), and OsHV-1 has 124 (Davison et al., 2005b). Fig. 2 shows the layout of genes in one member of each subfamily of the Herpesviridae, plus one member each of the Alloherpesviridae and Malacoherpesviridae. Within the Herpesviridae, Fig. 2 shows the disposition of the 43 genes that the subfamilies have inherited from a common ancestor (McGeoch and Davison, 1999), and serves to illustrate the relationships of gene content and arrangement among the three subfamilies. These “core” genes (shaded blue in Fig. 2) are largely located in the central regions of the genomes, with many of the more recently evolved “non-core” genes (shaded yellow) located near the termini. The core genes are generally located in the same order in members of the same subfamily, with some exceptions (in the Alphaherpesvirinae, substantial sections in the genomes of pseudorabies virus, infectious laryngotracheitis virus and psittacid HV 1 are rearranged with respect to the norm). In wider comparisons, it is evident that there exist several blocks of conserved genes that are rearranged among subfamilies (Gompels et al., 1995). Table 2 groups the core genes according to function, and shows that most are involved in fundamental aspects of lytic replication of virus. A much greater proportion of core genes than non-core genes are essential for growth of virus in cell culture. Nonetheless, many non-core genes have been shown to have important roles in growth of virus in animal models. This is in accord with the tenet that every gene contributes towards the success of HVs in nature, whether or not a phenotype can be detected in laboratory assays. Where sufficient information is available, it is evident that the genes that HVs employ to establish or reactivate from latency differ between subfamilies and even between genera. Fig. 2 also emphasizes the tenuous nature of sequence-based relationships between the three HV families: only one gene has amino acid sequences detectably conserved among the families in a mode that appears to indicate descent from a common ancestral virus. This gene, shaded red in Fig. 2, has distant relatives in T4 and other double-stranded DNA bacteriophages (Davison, 1992, 2002), which encode the ATPase subunit of a DNA packaging enzyme complex called the terminase (Rao and Black, 1988). For other cases where visibly related genes are present in all three families, such as those encoding DNA polymerase, deoxyuridine triphosphatase (dUTPase) and nucleoside kinases, independent capture events cannot be ruled out. The accumulation of sequence data for the Herpesvirales continues to strengthen our understanding of HV evolution, making D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 95 Fig. 2. Gene layout in HV genomes. HSV-1, HHV-8 and HCMV represent subfamilies Alpha-, Gamma- and Betaherpesvirinae, respectively, in family Herpesviridae. CCV and OsHV-1 represent families Allo- and Malacoherpesviridae, respectively. Substantial direct or inverted repeats are shown in a thicker format than unique regions. In HSV-1, HHV-8 and HCMV, protein-coding regions are shown as arrows shaded blue (core genes) or yellow (non-core genes), and introns as white bars. Blocks of core genes (I–VII) that are arranged differently in the three subfamilies are indicated by open arrows below the genomes. In CCV and OsHV-1, all protein-coding regions but one are shaded yellow. The exception encodes the putative ATPase subunit of the terminase, which is shaded red in all the genomes. This gene is not spliced in OsHV-1, but present as two or three exons in the other genomes. new questions accessible to investigation. The increasing spread of the data has enabled the derivation of ever more extensive and reliable phylogenies, giving clues to the evolution of various lineages in relation to their hosts, as discussed in Section 4 and following. In addition, advances are being gained in understanding how new gene functions arise and to what extent HV species vary in natural host populations. One of the means by which new functions evolve is that of gene duplication. This has been employed in a wide range of organisms, and HVs are no exception. The products of gene duplication – families of related genes – occur in all HV lineages, and are especially prominent among the Betaherpesvirinae (Chee et al., 1990). Gene families continued to be discovered, as exemplified recently by identification of several genes in the Beta- and Gammaherpesvirinae that appear to have evolved from a dUTPase gene. The proteins encoded by this family are predicted to retain the fundamental structural folds of their parent, but have lost motifs conferring the original enzymatic role and have presumably gained novel functions (Davison and Stow, 2005). The part in this example played by structural predictions, over and above sequence conservation, highlights the possibility that other HV genes may have been derived by gene duplication, with divergence having proceeded so far that no evidence remains in sequence comparisons. Future structural studies are anticipated to shed further light on this. A second major means by which HVs have acquired genes is that of capture (or lateral transfer) from the host cell or other viruses. Examples have emerged from all epochs of evolution and in all HV lineages, from ancient events that resulted in genes now common to most HVs to much more recent ones that yielded genes specific to a single lineage or even to a single species. An example of the latter is the ␤-1,6-Nacetylglucosaminyltransferase-mucin gene captured by a relatively recent ancestor of bovine HV 4 (BHV-4) (MarkineGoriaynoff et al., 2003). Many captured genes are involved in manipulation of host defences, with immune evasion and anti-apoptotic functions being particularly productive areas of investigation at present (for recent contributions to HCMV, for example, see Tomasec et al. (2005) and Goldmacher (2005)). Some genes in these categories belong to families whose members presumably have similar or overlapping functions (e.g. the HCMV US6 family; Momburg and Hengel, 2002), thus providing fertile research ground for some time to come. The availability of genome sequences for several strains of some HV species is revealing the extent of variation in virus populations, and adding a growing epidemiological dimension. Thus, intraspecies variation has been examined to a greater or lesser extent for each of the human HVs, and has revealed different modes of variation in the different lineages. HSV-1 and VZV show relatively little variation between strains, although it is possible to detect geographical associations (Norberg et al., 2004; Muir et al., 2002; Bowden et al., 2006). In EBV, four genes associated with the latent phase (EBNA-2, EBNA-3A, EBNA-3B and EBNA-3C) are found in two markedly diverged genotypes (Dambaugh et al., 1984; Sample et al., 1990). A similar property applies to the gene (K1) nearest the left end of the HHV-8 genome, which exists in several genotypes that exhibit strong geographical associations (Kasolo et al., 1998; Zong et al., 1999; Cook et al., 1999). At the other end of the HHV-8 genome, the multiply spliced K15 gene exists in two highly diverged alle- 96 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Table 2 Core genes in family Herpesviridae grouped according to functional class Genea Function DNA replication machinery Catalytic subunit of DNA polymerase complex UL30 UL42 Processivity subunit of DNA polymerase complex UL9 Origin-binding protein; helicase [present in subfamily Alphaherpesvirinae and genus Roseolovirus of subfamily Betaherpesvirinae] UL5 Component of DNA helicase-primase complex; helicase Component of DNA helicase-primase complex UL8 UL52 Component of DNA helicase-primase complex; primase UL29 Single-stranded DNA-binding protein Enzymes peripheral to DNA replication UL23 Thymidine (or pyrimidine deoxynucleoside) kinase [present in subfamilies Alpha- and Gammaherpesvirinae] UL39 Large subunit of ribonucleotide reductase [probably not an active enzyme in subfamily Betaherpesvirinae] UL40 Small subunit of ribonucleotide reductase [present in subfamilies Alpha- and Gammaherpesvirinae] UL50 Deoxyuridine triphosphatase [not an active enzyme in subfamily Betaherpesvirinae] UL2 Uracil-DNA glycosylase Processing and packaging of DNA UL12 Deoxyribonuclease; role in DNA maturation and recombination UL15 Putative ATPase subunit of terminase; capsid-associated UL28 Putative subunit of terminase; capsid-associated Portal protein; forms dodecameric ring at capsid vertex; UL6 complexed with terminase Possibly caps the portal after DNA packaging is complete UL25 UL32 Involved in capsid localization in the nucleus UL33 Interacts with terminase Involved in capsid localization in the nucleus UL17 Egress of capsids from nucleus UL31 Nuclear matrix protein; component of capsid docking complex on nuclear lamina Inner nuclear membrane protein; component of capsid UL34 docking complex on nuclear lamina Capsid assembly and structure UL19 Major capsid protein; component of hexons and pentons Component of intercapsomeric triplex between hexons and UL18 pentons UL38 Component of intercapsomeric triplex between hexons and pentons UL35 Small capsid protein located on tips of hexons UL26 Maturational protease; generates mature forms of scaffolding proteins UL26.5 Scaffolding protein removed from capsid during DNA packaging Tegument UL7 UL11 UL14 UL16 UL36 UL37 UL51 Unknown Interacts with UL16 protein; role in virion egress and secondary envelopment in the cytoplasm; myristylated and palmitylated protein Interacts with UL11 protein Interacts with UL11 protein Huge virion protein; interacts with UL37 protein; influences release of DNA from capsids during entry Interacts with UL36 protein Unknown Surface and membrane UL27 Glycoprotein B Table 2 (Continued ) Genea UL1 UL22 UL10 UL49A Function Glycoprotein L; complexed with glycoprotein H Glycoprotein H; complexed with glycoprotein L Glycoprotein M; complexed with glycoprotein N Glycoprotein N; complexed with glycoprotein M; not glycosylated in some HVs Control and modulation UL13 Serine-threonine protein kinase; tegument protein Multifunctional regulator of gene expression UL54 Unknown UL24 Nuclear protein HSV-1 genes are listed, with those essential for growth in cell culture underlined. les, a situation that probably arose by ancient recombination events between HHV-8-like viruses rather than by high divergence rates as envisaged for K1 (Poole et al., 1999; Kakoola et al., 2001). HCMV has an impressive number of genes exhibiting unusual high levels of variation (Pignatelli et al., 2004). The 25 most variable genes identified from a comparison of two genome sequences encode proteins known or predicted to be secreted or located in membranes, though not all membrane proteins are hypervariable (Dolan et al., 2004). This phenomenon probably reflects the evolutionary outcome of a greater overall exposure of HCMV to the host’s immune system than is the case for the members of the Alpha- and Gammaherpesvirinae, perhaps filtered by founder effects in human evolution. Recombination is also a feature of HV evolution, having been documented for HCMV (Pignatelli et al., 2004) and also for the Alpha- and Gammaherpesvirinae (Bowden et al., 2004; Poole et al., 1999). 4. Phylogenetic relationships of mammalian herpesviruses Phylogenetic analysis of HVs of mammals, birds and reptiles is now a well developed field. Utilising primarily concatenated alignments of amino acid sequences from a set of up to eight core genes, these studies have yielded a tree that is overall well resolved and robust, but with demarcated localities where branching patterns remain incompletely revealed (McGeoch, 2001; McGeoch and Gatherer, 2005; McGeoch et al., 1995, 2000, 2005). The tree’s fundamental structure has been stable to addition of data for more species as they became available, and consistent trees can also be obtained based on fewer genes. For this review, we wished to display trees that combined breadth of input gene sequences with the largest accessible number of virus species. The primary tree, shown in Fig. 3, was obtained from analysis of an alignment of amino acid sequences for six genes from 40 virus species and represents an update from versions published earlier. The separation of branches for the three subfamilies is unambiguous, and loci of unresolved multifurcations within the Beta- and Gammaherpesvirinae are depicted. Branch endpoints for most species are distributed closely around the mean, consistent with an overall uniformity in long-term rate of sequence change (within the Alphaherpesvirinae and separately within the Beta- plus Gammaherpesvirinae). There are some striking exceptions: in D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 97 Fig. 3. 40-species phylogenetic tree for the Herpesviridae. A phylogenetic tree was constructed based on an alignment of amino acid sequences for six genes from 40 HV species. Genes used were the orthologues of HSV-1 UL15, UL19, UL27, UL28, UL29 and UL30. The length of the concatenated alignment after removal of unalignable sections and loci with gapping characters was 4555 amino acids. Initial tree evaluation utilised a Bayesian Monte Carlo Markov Chain process (Ronquist and Huelsenbeck, 2003). The output tree from this was modified to include unresolvable multifurcations known from previous work, and then branch lengths recomputed by a maximum likelihood method (Yang, 1997), with discrete gamma distribution of rates and no molecular clock. As presented here the tree’s root was estimated as the midpoint between the mean tip positions of terminal branches in the Alphaherpesvirinae (denoted as ␣) and those in the Beta- plus Gammaherpesvirinae (␤ and ␥). The mean tip position is marked by a dashed line. A divergence scale (substitutions per site) is shown at the foot. Abbreviations for virus nomenclature are explained in Table 1 or in the text, with the addition of GTHV (green turtle HV), SCMV (simian CMV, as two strains), GPCMV (guinea pig CMV) and PLHV-1 (porcine lymphotropic HV 1). the Alphaherpesvirinae the single reptilian HV has a particularly short branch, while in the Gammaherpesvirinae, EBV and its primate HV relatives have uniformly short terminal branches and MuHV-4 has an uniquely extended terminus. These features presumably reflect atypical secular rates of sequence change in these lineages. MuHV-4′ s enhanced rate of change is thought to contribute to difficulties in analysing branching order for the Rhadinovirus clade (McGeoch, 2001; McGeoch et al., 2005). It has become apparent that, within each subfamily, many features of branching and approximate relative scale in the HV tree show congruence with features in the phylogeny of corresponding host organisms, and this equivalence has been interpreted as providing evidence for coevolutionary development of virus and host lineages (McGeoch and Cook, 1994; McGeoch et al., 1995, 2000). As an update of published papers, we describe here the current state of correspondences between host and virus lineages, focusing on mammalian HVs in the first instance and at the level of major lineages rather than individual virus species. Our use of the term “coevolution” is intended to refer to the synchronous development of virus lineages with those of their hosts; we are not primarily concerned here with mutual coadaptation (although we consider self-evident that such processes must have been very extensive). Fig. 4a presents for reference a phylogenetic tree with palaeontological timescale for those major lineages of Eutheria (placental mammals), which include species that are natural hosts for well characterized HV species. The lineages in this tree are equivalent to high level taxa (primarily orders). We have distinguished by colour three deep clades corresponding to the major taxa within Eutheria (Euarchontoglires, Laurasiatheria and Afrotheria; coloured red, green and blue, respectively). It should be noted that only seven of the 18 extant eutherian orders are represented in the figure. The Primates are shown separated into Old World (OW) and New World (NW) lineages. In order to include HVs for which fewer gene sequences were available, further trees were computed based on shorter alignments and loci for species of interest then transferred to the primary tree, as previously described (McGeoch et al., 2000). The construction of such composite trees employed component trees 98 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Fig. 4. Comparison between the phylogeny of orders in the Eutheria and that of subfamilies of the Herpesviridae. (a) Phylogenetic tree for orders in the Eutheria that have well characterized HVs as natural hosts, with timescale as millions of years before the present (Springer et al., 2003) and three deep clades distinguished by colouring, (b); (c); and (d) molecular clock, composite trees for major distinct lineages (approximately equivalent to genera) of the Alpha-, Beta- and Gammaherpesvirinae, with host groupings, viruses in each lineage and genera listed. Clades are coloured to correspond to proposed coevolution with clades in (a), as set out in the text. Divergence scales (substitutions per site) are at the foot. The scales of all panels have been adjusted to make the green coloured clades of equivalent depth. Abbreviations for virus nomenclature are explained in Table 1 or in the text, with the addition of HVS-1 (saimiriine HV 1), HVA-1 (ateline HV 1) and FHV-1 (feline HV 1). with global molecular clocks imposed, for topologies previously computed; this device enables straightforward interpolation and display, but loses the heterogeneity of terminal branch lengths seen in the original tree. In Fig. 4 parts b, c and d, separate trees with imposed clocks are shown for each of the three HV subfamilies, with the terminal branches aimed at displaying deep distinct lineages rather than virus species. Where we wish to propose a likely coevolutionary pattern in the HV tree, the branches have been coloured to correspond to their counterparts in the host tree. In order to optimise this presentation, the scales of the x-axes of the trees in the four parts of Fig. 4 have been adjusted to give approximately equivalent depths to clades that we consider to represent corresponding features; specifically, this was done by adjusting the depths of the green clades (ungulates plus carnivores, and their HVs) to the same size on the paper. This criterion means that, for the HV tree versions in Fig. 4, the Betaand Gammaherpesvirinae clocks were taken as 1.6 and 1.8 times faster, respectively, than that for the Alphaherpesvirinae. Fig. 4b shows that in the Alphaherpesvirinae there is a convincing coevolutionary clade comprising artiodactyl, perissodactyl and carnivore viruses in the Varicellovirus genus. Next, the Alphaherpesvirinae contain a clade of OW and NW primate viruses in the Simplexvirus genus, which we propose also represents a cospeciational mode. There are, however, within this Simplexvirus clade a wallaby HV clade and a clade containing only bovine HV 2 (BHV-2). Both these are taken to represent instances of interspecies transfer of viruses—with the wallaby viruses a particularly flagrant example. There is a second lineage of OW primate viruses, in the Varicellovirus clade and containing VZV and SVV, which could also be regarded as coevolutionary, and we assigned primacy of this interpretation to the Simplexvirus clade simply because it is better populated and contains both OW and NW primate HVs. In addition, Fig. 4b shows in the Alphaherpesvirinae three deep branching lineages, two of avian viruses and one of reptilian (turtle) viruses, and discussion of these is postponed until Section 5. Fig. 4c presents lineages in the Betaherpesvirinae. Here we have not yet been able to resolve to our satisfaction the detail of branching order among the cytomegaloviruses (CMVs) of primates, murid CMVs and tupaiid HV (THV). As drawn in Fig. 4c, the murid CMV lineage is intended to encompass murine CMV, rat CMV and guinea pig CMV. The branch point for the last of these lies very deep in the clade comprising all the CMVs plus THV, and it may not truly fall into a clade of rodent CMVs only. Such uncertainty notwithstanding, this overall association of Betaherpesvirinae of OW primates, tree shrews (members of the order Scandentia) and rodents strongly suggests a coevolutionary history. Similarly, the branching of the lineage containing EEHV as the earliest event in the Betaherpesvirinae tree corresponds well with the early divergence of the Afrotheria (including Proboscidea) from other placental mammals. This leaves for consideration the intriguing clade consisting of HHV-6 and HHV-7 plus porcine CMV (PCMV). Although the branches here are too sparse to make substantive inferences, we offer the following scenario as consistent with other features of the Betaherpesvirinae and providing a plausible account. First, we can assert that HHV-6 and HHV-7 do not conform to a simple coevolutionary history when viewed in relation to the well populated clade of CMVs plus THV, or considering the substantial depth of the clade containing HHV-6 and HHV-7. However, the location of the PCMV lineage as a sister clade to that of the CMVs plus THV, and also relative to D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 the EEHV lineage, is readily compatible with a coevolutionary origin, and we can then speculate that the HHV-6 and HHV-7 lineages could have arisen by interspecies transfer events from the ungulate lineage of which PCMV is the only known member. The Gammaherpesvirinae present further variations on the same theme (Fig. 4d). We regard the clade of the Lymphocryptovirus genus, containing EBV plus its OW and NW monkey relatives, as having developed by coevolution with primate hosts. Next, there is a grouping of artiodactyl, perissodactyl and carnivore viruses (corresponding to the proposed genera Macavirus and Percavirus), whose internal structure and status as sister group to the lymphocryptoviruses is consistent with a coevolutionary origin. The remaining lineages in the Gammaherpesvirinae all belong to the redefined Rhadinovirus genus. The locus of this clade’s root is not consistent with a coevolutionary origin, although it is plausible that the subsequent development of OW and NW primate lineages within the Rhadinovirus clade has been coevolutionary. The species in the Rhadinovirus clade are notably heterogeneous with respect to their host species, with primate, rodent, artiodactyl and perissodactyl hosts. Additionally, there exists a virus (not shown in Fig. 4d), which is a close relative of MuHV-4 but was isolated, albeit on a single occasion, from a shrew (an insectivore, not a rodent) (Chastel et al., 1994; Davison, unpublished data). Each of the three subfamilies of the Herpesviridae thus shows a substantial proportion of tree-branching features that can be economically rationalised as resulting from coevolution of HV lineages with lineages of placental mammals, as well as other features that require non-coevolutionary explanations. Compatibility with coevolutionary development can also be discerned at a finer level in certain, more populated lineages, notably in the artiodactyl part of the Varicellovirus clade, and in the OW primate simplexviruses and the primate CMVs (not shown). Conversely, there are instances of unpredicted complexities in relationships among closely related viruses. In the Lymphocryptovirus clade, indications are that OW monkey, ape and human viruses are more closely related than expected on the basis of coevolutionary development (Gerner et al., 2004). Another phenomenon is that many examples are emerging of multiple, quite closely similar viruses in a single host species—for example with primate rhadinoviruses and in the Macavirus clade (Schultz et al., 2000; Ehlers et al., 1999). We note that equivalent conclusions regarding both codivergence of HV and host lineages and the duplication of HV species in a single host species have been reached by Jackson (2005) using a formal cophylogeny mapping technique. Accepting an overall correspondence with evolution of the hosts should then allow transfer of the palaeontological timescale for placental mammals to evolution of their HVs. For example, the deepest branchpoint in the Betaherpesvirinae, that for the origin of the EEHV lineage, is taken to correspond to the separation of the Afrotheria, around 105 million years ago (Ma) (Springer et al., 2003). As registered above, several examples of apparent variation in rate of evolutionary change are now becoming visible in the HV phylogenetic tree, although analysis of such aspects has as yet scarcely been attempted. Rate variation increases uncertainty in estimates of dates, particu- 99 larly those involving extrapolation to early events in the tree. Given this caveat, we note that a recent estimate dated the root of the three-subfamily tree (equivalent to the most recent common ancestor of the Herpesviridae) to about 400 Ma (McGeoch and Gatherer, 2005). This figure doubles that obtained a decade earlier (McGeoch et al., 1995), which was based on less abundant data, less developed methodology and the palaeontological estimates of the day. 5. Phylogenetic relationships among reptilian, avian and mammalian lineages of the Alphaherpesvirinae We turn now to considering the avian and reptilian HV lineages in the Alphaherpesvirinae. The two avian HV lineages, genera Mardivirus and Iltovirus, are relatively well studied and each contains several virus species, all with birds as their hosts. The single reptilian HV lineage has been defined only recently and currently contains viruses of turtles (Herbst et al., 2004; Greenblatt et al., 2005; McGeoch and Gatherer, 2005). Limited sequence data for other reptilian HVs, of green iguana and plated lizards, have not proved sufficient to place them precisely in the HV tree, but indicate that they probably belong to the Alphaherpesvirinae, in a lineage originating deep within the tree and possibly the same as that of the turtle HVs. Our core question is: how does the phylogeny of these avian and reptilian HVs relate to the picture of widespread coevolution with hosts that we have described for mammalian HVs? In Fig. 5, we compare the major host and viral lineages involved. The tree in Fig. 5a summarizes current understanding of reptilian lineages and of the evolutionary emergence of birds and mammals from these. The earliest branchpoint (310 Ma) represents the divergence between lineages of diapsid and synapsid reptiles (Benton, 1997). Most modern reptiles are diapsid, while mammals eventually arose from the synapsid lineage. The unresolved branching at 270–285 Ma encompasses divergence events whose order is currently not well resolved, with the resulting lineages giving rise to turtles and birds, as well as scaled lizards and snakes (Cao et al., 2000; Hedges and Poling, 1999; Rest et al., 2003; McGeoch and Gatherer, 2005). Fig. 5b presents the tree of Fig. 4b reduced to lineages corresponding to those in Fig. 5a, with timescale derived by extrapolation of coevolutionary datings for the mammalian HV portion of the Alphaherpesvirinae tree. There is not a global correspondence between branching patterns of the host and Alphaherpesvirinae trees in Fig. 5. However, we can propose two possible scenarios to account economically in a coevolutionary manner for the observed branching pattern of HV lineages in Fig. 5b. Both involve equating the earliest node in the Alphaherpesvirinae tree (the separation of the reptilian HV lineage from those of avian and mammalian HVs, estimated at 240 Ma) with a node in the host tree. First, if this earliest virus node were to correspond to the diapsid-synapsid divergence at 310 Ma in the host tree, then we can see that from this early stage both the reptilian and mammalian alphaherpesvirus lineages could have proceeded in a coevolutionary mode—but that for the two avian virus lineages separate origins by transfer from synapsid reptilian or mammalian HVs would 100 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Fig. 5. Comparison between the phylogeny of reptiles, birds and mammals, and that of the Alphaherpesvirinae. (a) Consensus phylogenetic tree for major lineages of reptiles plus mammals and birds; (b) molecular clock, composite tree for turtle, avian and mammalian lineages of the Alphaherpesvirinae. The timescale at the foot applies to both panels. For (a) the scale is as assembled by McGeoch and Gatherer (2005); and for (b) it was obtained by extrapolation from assumption of coevolution in the mammalian HV portion of the tree. have to be invoked. In the second scenario, if the earliest virus node were to correspond to the divergence of Testudines and Archosauria at 270–285 Ma, then the origin of turtle and avian alphaherpesvirus lineages could have been coevolutionary—but for the mammalian virus lineage an origin by transfer from an avian lineage would be required. Both interpretations place the early history of the Alphaherpesvirinae with early reptiles as hosts. The first interpretation is more parsimonious in that the existence of mammalian Beta- and Gammaherpesvirinae can be similarly accounted for in a consistent way by coevolution rather than by invoking chance events of interspecies transfer, and the absence of known avian viruses in these subfamilies is rationalised. We note that both interpretations also suggest that undetected reptilian HVs may exist in the Beta-and Gammaherpesvirinae. We close this section by noting that phylogenies of the families Alloherpesviridae and Malacoherpesviridae have not yet been explored, although in the case of the Alloherpesviridae we can state that the family contains species whose gene sequences are highly diverged, to an extent comparable to that among the three subfamilies of the Herpesviridae. 6. Herpesvirus capsid structures and deep phylogenetic connections As outlined in Section 1, HVs share a characteristic fourcomponent virion architecture, comprising DNA core, capsid, tegument and envelope. Analyses by cryo-electron microscopy and computer-based image reconstruction have extended our knowledge of HV capsid structures into three dimensions and to much higher resolution, as fine as 8 Å for the most refined work on HSV-1 capsids (Zhou et al., 2000). Three-dimensional reconstructions of capsids for members of all three families of the Herpesvirales have been generated, and comparisons reveal common structural details that were not visible with standard electron microscopic images, as illustrated in Fig. 6. The hexon capsomeres in each case form hollow towers, of closely similar appearance, that extend approximately 10 nm outwards from the capsid floor. In capsids from members of the Herpesviridae, the penton capsomeres are each formed from five molecules of the same protein as the hexons. The size and arrangement of penton subunits for the only member of the Alloherpesviridae yet analysed (CCV) are strongly suggestive of a similar pattern (Booy et al., 1996). However, in the case of OsHV-1, the reconstruction shows large voids at pentagonal vertices, presumably caused by loss of the pentons during preparation, so there are no data bearing on penton structure for the Malacoherpesviridae (Davison et al., 2005b). A particularly distinctive feature of HV capsids is the presence of surface structures termed triplexes, which occupy local three-fold positions between adjacent hexons, and between hexons and pentons. In members of the Herpesviridae, each triplex consists of two molecules of one protein and one of a second protein. It is not known whether the triplexes in capsids from the other two families are organized in the same way, although the protein profile of purified CCV capsids includes suitably sized candidate proteins with appropriate stoichiometries (Davison and Davison, 1995). In summary, although information from the Alloherpesviridae and Malacoherpesviridae is less detailed than that from the Herpesviridae, it is clear that particles of all HVs studied in some detail display a set of structural features that are not found together for any other virus family. Overall, the 3D structures provide a compelling D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 101 Fig. 6. Electron cryo-microscopic reconstructions of HV capsids. (a) HSV-1; (b) CCV; and (c) OsHV-1. Each capsid is shown as viewed along a twofold axis of symmetry. The voids in (c) mark the positions of the pentavalent vertices, which are lost during purification (bar = 50 nm). This figure is a modification of that shown in Davison et al. (2005b). case that the three families have a common evolutionary origin in respect of their capsid architecture, and thus that they possess homologous sets of genes for capsid proteins, although the amino acid sequences of these have diverged beyond detectable similarity between families. In this context, it is noteworthy that the one protein that exhibits features of sequence similarity indicative of common descent for all three families is the ATPase subunit of the terminase—that is, a protein with a role in DNA packaging and thus directly interacting with the nascent capsid. Aspects of capsid structures are now also suggesting a more profound evolutionary link, between HVs and tailed DNA bacteriophages. A possible link between the eukaryote-infecting Herpesvirales and the prokaryote-infecting Caudovirales has long been postulated, based on parallels in their capsid assembly pathways (Casjens and King, 1975; Steven and Spear, 1997), and the likelihood of such a relationship was strengthened by analyses of the portal complexes, through which DNA enters the capsid. These have revealed that the distinctive 12-fold arrangement of subunits seen in several tailed bacteriophages (Valpuesta and Carrascosa, 1994) occurs also in the one HV example studied (Trus et al., 2004). In addition, there has been reported a low level of similarity in amino acid sequences of capsid-scaffold processing proteases from DNA bacteriophages and members of the Herpesviridae, which might be indicative of common ancestry (Cheng et al., 2004; Liu and Mushegian, 2004). However, until very recently there has been no detailed evidence, in the form of protein structural comparisons, that might support a link between these two virus orders. Because the structure of the HV capsid is sufficiently singular for it to be considered as diagnostic for HVs, it might seem an unlikely place to look for evidence of a link between these disparate virus groups, and certainly on superficial examination the overall appearance of HV capsids is very different from that of tailed bacteriophages. However, the improvement in resolution of capsid reconstructions during recent years for both HVs and bacteriophages has progressed to the point that it has enabled comparisons at the level of protein secondary structure between the major capsid protein (VP5) of HSV-1 and analogous bacteriophage proteins, and previously unsuspected similarities have emerged. These may be illustrated by comparing the structure of HSV-1 VP5 (as seen in hexon capsomeres) with the capsid shell protein (gp5) of the lambdoid bacteriophage HK97 (Wikoff et al., 2000). gp5 is a much smaller protein molecule than VP5, but it has very similar dimensions to the floor domain of VP5, which forms the semi-continuous shell from which the capsomere towers rise. When comparison is confined to this domain (excluding the tower component of VP5) both the overall shapes of the two proteins and the dispositions of their secondary structural elements are seen to be strikingly similar, as illustrated in Fig. 7 (Baker et al., 2005). Capsids from several more members of the Caudovirales have now been analysed at sufficient resolution to reveal secondary structural features, and in each case they follow the same fold pattern as the HK97 gp5 protein despite having limited sequence similarity. No HV capsid structure other than that of HSV-1 is known in sufficient detail to carry out an equivalent analysis, but given the general uniformity of their capsid structures, it seems certain that the floor domain fold of VP5 will be maintained throughout the Herpesvirales. The possession of a common structure in a key domain of the capsid protein provides strong support for an evolutionary linkage between Herpesvirales and Caudovirales. This finding, in conjunction with the distinctive arrangement of subunits in the portal and elements of sequence similarity in the terminase subunit, indicates that the entire capsid-packaging machinery is of ancient origin and has been passed down to the extant members of these orders from a common progenitor. We note that, on the basis of capsid structures known for other groups of viruses, there is at present no indication that the link between Herpesvirales and Caudovirales will be extended more widely. With the estimate for the most recent common ancestor for the family Herpesviridae dated at 400 Ma, in the Devonian Period (McGeoch and Gatherer, 2005), we expect that the common ancestor for the order Herpesvirales with respect to their capsid structure must have a substantially earlier date—to speculate, perhaps close to or predating the emergence of vertebrates in the Cambrian (570–505 Ma; Benton, 1997). And the common elements of capsid architecture in the Herpesvirales and Caudovirales surely must come from a very early stage in the evolution of life. 102 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Fig. 7. Comparison of secondary structure elements in the HSV-1 capsid protein (VP5) and the HK97 capsid protein (gp5). (a) Electron cryo-microscopic reconstruction of the isolated VP5 floor domain. The positions of several ␣-helices (red) and ␤-sheets (green) identified in the structure are indicated; (b) crystal structure of the gp5 protein. The ␣-helices and ␤-sheets are shown in green and blue, respectively; (c) alignment of the core secondary structure elements, demonstrating a clear match between the two structures. This figure is a modification of that shown in Baker et al. (2005). Acknowledgements We thank M. Schleiss and A. Dolan for early sight of sequence data. References Alwine, J.C., Steinhart, W.L., Hill, C.W., 1974. Transcription of herpes simplex type 1 DNA in nuclei isolated from infected HEp-2 and KB cells. Virology 60, 302–307. Baker, M.L., Jiang, W., Rixon, F.J., Chiu, W., 2005. Common ancestry of herpesviruses and tailed DNA bacteriophages. J. Virol. 79, 14967–14970. Benton, M.J., 1997. Vertebrate Palaeontology. Chapman and Hall, London. Bodescot, M., Perricaudet, M., Farrell, P.J., 1987. A promoter for the highly spliced EBNA family of RNAs of Epstein-Barr virus. J. Virol. 61, 3424–3430. Booy, F.P., Trus, B.L., Davison, A.J., Steven, A.C., 1996. The capsid architecture of channel catfish virus, an evolutionarily distant herpesvirus, is largely conserved in the absence of discernible sequence homology with herpes simplex virus. Virology 215, 134–141. Bowden, R., Sakaoka, H., Donnelly, P., Ward, R., 2004. High recombination rate in herpes simplex virus type 1 natural populations suggests significant co-infection. Infect. Genet. Evol. 4, 115–123. Bowden, R., Sakaoka, H., Ward, R., Donnelly, P., 2006. Patterns of Eurasian HSV-1 molecular diversity and inferences of human migrations. Infect. Genet. Evol. 6, 63–74. Bowden, R.J., Simas, J.P., Davis, A.J., Efstathiou, S., 1997. Murine gammaherpesvirus 68 encodes tRNA-like sequences which are expressed during latency. J. Gen. Virol. 78, 1675–1687. Cai, X., Lu, S., Zhang, Z., Gonzalez, C.M., Damania, B., Cullen, B.R., 2005. Kaposi’s sarcoma-associated herpesvirus expresses an array of viral microRNAs in latently infected cells. Proc. Natl. Acad. Sci. U.S.A. 102, 5570–5575. Cao, Y., Sorenson, M.D., Kumasawa, Y., Mindell, D.P., Hasegawa, M., 2000. Phylogenetic position of turtles among amniotes: evidence from mitochondrial and nuclear genes. Gene 259, 139–148. Casjens, S., King, J., 1975. Virus assembly. Annu. Rev. Biochem. 44, 555–611. Chastel, C., Beaucournu, J.-P., Chastel, O., Legrand, M.C., Le Goff, F., 1994. A herpesvirus from an European shrew (Crocidura russula). Acta Virol. 38, 309. Chee, M.S., Bankier, A.T., Beck, S., Bohni, R., Brown, C.M., Cerny, R., Horsnell, T., Hutchison, C.A., Kouzarides, T., Martignetti, J.A., Preddie, E., Satchwell, S.C., Tomlinson, P., Weston, K.M., Barrell, B.G., 1990. Analysis of the protein coding content of the sequence of human cytomegalovirus strain AD169. Curr. Top. Microbiol. Immunol. 154, 125–169. Cheng, H., Shen, N., Pei, J., Grishin, N.V., 2004. Double-stranded DNA bacteriophage prohead protease is homologous to herpesvirus protease. Prot. Sci. 13, 2260–2269. Cook, P.M., Whitby, D., Calabro, M.L., Luppi, M., Kakoola, D.N., Hjalgrim, H., Ariyoshi, K., Ensoli, B., Davison, A.J., Schulz, T.F., 1999. Variability and evolution of Kaposi’s sarcoma-associated herpesvirus in Europe and Africa. AIDS 13, 1165–1176. Dambaugh, T., Hennessy, K., Chamnankit, L., Kieff, E., 1984. U2 region of Epstein-Barr virus DNA may encode Epstein-Barr nuclear antigen 2. Proc. Natl. Acad. Sci. U.S.A. 81, 7632–7636. Davison, A.J., 1992. Channel catfish virus: a new type of herpesvirus. Virology 186, 9–14. Davison, A.J., 1998. The genome of salmonid herpesvirus 1. J. Virol. 72, 1974–1982. Davison, A.J., 2000. Molecular evolution of alphaherpesviruses. In: Arvin, A.M., Gershon, A.A. (Eds.), Varicella-Zoster Virus. Cambridge University Press, Cambridge, pp. 25–50. Davison, A.J., 2002. Evolution of the herpesviruses. Vet. Microbiol. 86, 69–88. Davison, A.J., Davison, M.D., 1995. Identification of structural proteins of channel catfish virus by mass spectrometry. Virology 206, 1035–1043. Davison, A.J., Dolan, A., Akter, P., Addison, C., Dargan, D.J., Alcendor, D.J., McGeoch, D.J., Hayward, G.S., 2003. The human cytomegalovirus genome revisited: comparison with the chimpanzee cytomegalovirus genome. J. Gen. Virol. 84, 17–28. Davison, A., Eberle, R., Hayward, G.S., McGeoch, D.J., Minson, A.C., Pellett, P.E., Roizman, B., Studdert, M.J., Thiry, E., 2005a. Herpesviridae. In: Fauquet, C.M., Mayo, M.A., Maniloff, J., Desselberger, U., Ball, L.A. (Eds.), Virus Taxonomy, VIIIth Report of the ICTV. Elsevier/Academic Press, London, pp. 193–212. Davison, A.J., Stow, N.D., 2005. New genes from old: redeployment of dUTPase by herpesviruses. J. Virol. 79, 12880–12892. Davison, A.J., Trus, B.L., Cheng, N., Steven, A.C., Watson, M.S., Cunningham, C., Le Deuff, R.M., Renault, T., 2005b. A novel class of herpesvirus with bivalve hosts. J. Gen. Virol. 86, 41–53. de Jesus, O., Smith, P.R., Spender, L.C., Elgueta Karstegl, C., Niller, H.H., Huang, D., Farrell, P.J., 2003. Updated Epstein-Barr virus (EBV) DNA D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 sequence and analysis of a promoter for the BART (CST, BARF0) RNAs of EBV. J. Gen. Virol. 84, 1443–1450. Dolan, A., Cunningham, C., Hector, R.D., Hassan-Walker, A.F., Lee, L., Addison, C., Dargan, D.J., McGeoch, D.J., Gatherer, D., Emery, V.C., Griffiths, P.D., Sinzger, C., McSharry, B.P., Wilkinson, G.W.G., Davison, A.J., 2004. Genetic content of wild-type human cytomegalovirus. J. Gen. Virol. 85, 1301–1312. Dolan, A., Jamieson, F.E., Cunningham, C., Barnett, B.C., McGeoch, D.J., 1998. The genome of herpes simplex virus type 2. J. Virol. 72, 2010–2021. Ehlers, B., Ulrich, S., Goltz, M., 1999. Detection of two novel porcine herpesviruses with high similarity to gammaherpesviruses. J. Gen. Virol. 80, 971–978. French, C., Menegazzi, P., Nicholson, L., Macaulay, H., DiLuca, D., Gompels, U.A., 1999. Novel, nonconsensus cellular splicing regulates expression of a gene encoding a chemokine-like protein that shows high variation and is specific for human herpesvirus 6. Virology 262, 139–151. Gerner, C.S., Dolan, A., McGeoch, D.J., 2004. Phylogenetic relationships in the Lymphocryptovirus genus of the Gammaherpesvirinae. Virus Res. 99, 187–192. Glenn, M., Rainbow, L., Aurade, F., Davison, A., Schulz, T.F., 1999. Identification of a spliced gene from Kaposi’s sarcoma-associated herpesvirus encoding a protein with similarities to latent membrane proteins 1 and 2A of Epstein-Barr virus. J. Virol. 73, 6953–6963. Goldmacher, V.S., 2005. Cell death suppression by cytomegaloviruses. Apoptosis 10, 251–265. Gompels, U.A., Nicholas, J., Lawrence, G., Jones, M., Thomson, B.J., Martin, M.E., Efstathiou, S., Craxton, M., Macaulay, H.A., 1995. The DNA sequence of human herpesvirus-6: structure, coding content, and genome evolution. Virology 209, 29–51. Gray, W.L., Starnes, B., White, M.W., Mahalingam, R., 2001. The DNA sequence of the simian varicella virus genome. Virology 284, 123–130. Greenblatt, R.J., Quackenbush, S.L., Casey, R.N., Rovnak, J., Balazs, G.H., Work, T.M., Casey, J.W., Sutton, C.A., 2005. Genomic variation of the fibropapilloma-associated marine turtle herpesvirus across seven geographic areas and three host species. J. Virol. 79, 1125–1132. Grey, F., Antoniewicz, A., Allen, E., Saugstad, J., McShea, A., Carrington, J.C., Nelson, J., 2005. Identification and characterization of human cytomegalovirus-encoded microRNAs. J. Virol. 79, 12095–12099. Hayward, G.S., Jacob, R.J., Wadsworth, S.C., Roizman, B., 1975. Anatomy of herpes simplex virus DNA: evidence for four populations of molecules that differ in the relative orientations of their long and short components. Proc. Natl. Acad. Sci. U.S.A. 72, 4243–4247. Hedges, S.B., Poling, L.L., 1999. A molecular phylogeny of reptiles. Science 283, 998–1001. Herbst, L., Ene, A., Su, M., Desalle, R., Lenz, J., 2004. Tumor outbreaks in marine turtles are not due to recent herpesvirus mutations. Curr. Biol. 14, R697–R699. Honess, R.W., 1984. Herpes simplex and ‘the herpes complex’: diverse observations and a unifying hypothesis. The eighth Fleming lecture. J. Gen. Virol. 65, 2077–2107. Honess, R.W., Roizman, B., 1974. Regulation of herpesvirus macromolecular synthesis. I. Cascade regulation of the synthesis of three groups of viral proteins. J. Virol. 14, 8–19. Hutoran, M., Ronen, A., Perelberg, A., Ilouze, M., Dishon, A., Bejerano, I., Chen, N., Kotler, M., 2005. Description of an as yet unclassified DNA virus from diseased Cyprinus carpio species. J. Virol. 79, 1983–1991. Jackson, A.P., 2005. The effect of paralogous lineages on the application of reconciliation analysis by cophylogeny mapping. Syst. Biol. 54, 127–145. Jenkins, F.J., Roizman, B., 1986. Herpes simplex virus 1 recombinants with noninverting genomes frozen in different isomeric arrangements are capable of independent replication. J. Virol. 59, 494–499. Kakoola, D.N., Sheldon, J., Byabazaire, N., Bowden, R.J., KatongoleMbidde, E., Schulz, T.F., Davison, A.J., 2001. Recombination in human herpesvirus-8 strains from Uganda and evolution of the K15 gene. J. Gen. Virol. 82, 2393–2404. Kasolo, F.C., Monze, M., Obel, N., Anderson, R.A., French, C., Gompels, U.A., 1998. Sequence analyses of human herpesvirus-8 strains from both 103 African human immunodeficiency virus-negative and -positive childhood endemic Kaposi’s sarcoma show a close relationship with strains identified in febrile children and high variation in the K1 glycoprotein. J. Gen. Virol. 79, 3055–3065. Kemble, G.W., Annunziato, P., Lungu, O., Winter, R.E., Cha, T.A., Silverstein, S.J., Spaete, R.R., 2000. Open reading frame S/L of varicella-zoster virus encodes a cytoplasmic protein expressed in infected cells. J. Virol. 74, 11311–11321. Kulesza, C.A., Shenk, T., 2004. Human cytomegalovirus 5-kilobase immediate-early RNA is a stable intron. J. Virol. 78, 13182–13189. Liu, J., Mushegian, A., 2004. Displacements of prohead protease genes in the late operons of double-stranded-DNA bacteriophages. J. Bact. 186, 4369–4375. Markine-Goriaynoff, N., Georgin, J.P., Goltz, M., Zimmermann, W., Broll, H., Wamwayi, H.M., Pastoret, P.P., Sharp, P.M., Vanderplasschen, A., 2003. The core 2 ␤-1,6-N-acetylglucosaminyltransferase-mucin encoded by bovine herpesvirus 4 was acquired from an ancestor of the African buffalo. J. Virol. 77, 1784–1792. McGeoch, D.J., 2001. Molecular evolution of the ␥-Herpesvirinae. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356, 421–435. McGeoch, D.J., Cook, S., 1994. Molecular phylogeny of the Alphaherpesvirinae subfamily and a proposed evolutionary timescale. J. Mol. Biol. 238, 9–22. McGeoch, D.J., Gatherer, D., 2005. Integrating reptilian herpesviruses into the family Herpesviridae. J. Virol. 79, 725–731. McGeoch, D.J., Cook, S., Dolan, A., Jamieson, F.E., Telford, E.A.R., 1995. Molecular phylogeny and evolutionary timescale for the family of mammalian herpesviruses. J. Mol. Biol. 247, 443–458. McGeoch, D.J., Davison, A.J., 1999. The molecular evolutionary history of the herpesviruses. In: Domingo, E., Webster, R., Holland, J. (Eds.), Origin and Evolution of Viruses. Academic Press, London, pp. 441–465. McGeoch, D.J., Dolan, A., Ralph, A.C., 2000. Toward a comprehensive phylogeny for mammalian and avian herpesviruses. J. Virol. 74, 10401–10406. McGeoch, D.J., Gatherer, D., Dolan, A., 2005. On phylogenetic relationships among major lineages of the Gammaherpesvirinae. J. Gen. Virol. 86, 307–316. Megaw, A.G., Rapaport, D., Avidor, B., Frenkel, N., Davison, A.J., 1998. The DNA sequence of the RK strain of human herpesvirus 7. Virology 244, 119–132. Momburg, F., Hengel, H., 2002. Corking the bottleneck: the transporter associated with antigen processing as a target for immune subversion by viruses. Curr. Top. Microbiol. Immunol. 269, 57–74. Muir, W.B., Nichols, R., Breuer, J., 2002. Phylogenetic analysis of varicellazoster virus: evidence of intercontinental spread of genotypes and recombination. J. Virol. 76, 1971–1979. Neipel, F., Albrecht, J.C., Fleckenstein, B., 1997. Cell-homologous genes in the Kaposi’s sarcoma-associated rhadinovirus human herpesvirus 8: determinants of its pathogenicity? J. Virol. 71, 4187–4192. Norberg, P., Bergstrom, T., Rekabdar, E., Lindh, M., Liljeqvist, J.A., 2004. Phylogenetic analysis of clinical herpes simplex virus type 1 isolates identified three genetic groups and recombinant viruses. J. Virol. 78, 10755–10764. Perelygina, L., Zhu, L., Zurkuhlen, H., Mills, R., Borodovsky, M., Hilliard, J.K., 2003. Complete sequence and comparative analysis of the genome of herpes B virus (Cercopithecine herpesvirus 1) from a rhesus monkey. J. Virol. 77, 6167–6177. Pfeffer, S., Sewer, A., Lagos-Quintana, M., Sheridan, R., Sander, C., Grasser, F.A., van Dyk, L.F., Ho, C.K., Shuman, S., Chien, M., Russo, J.J., Ju, J., Randall, G., Lindenbach, B.D., Rice, C.M., Simon, V., Ho, D.D., Zavolan, M., Tuschl, T., 2005. Identification of microRNAs of the herpesvirus family. Nat. Methods 2, 269–276. Pfeffer, S., Zavolan, M., Grasser, F.A., Chien, M., Russo, J.J., Ju, J., John, B., Enright, A.J., Marks, D., Sander, C., Tuschl, T., 2004. Identification of virus-encoded microRNAs. Science 304, 734–736. Pignatelli, S., Dal Monte, P., Rossini, G., Landini, M.P., 2004. Genetic polymorphisms among human cytomegalovirus (HCMV) wild-type strains. Rev. Med. Virol. 14, 383–410. 104 D.J. McGeoch et al. / Virus Research 117 (2006) 90–104 Poole, L.J., Zong, J.C., Ciufo, D.M., Alcendor, D.J., Cannon, J.S., Ambinder, R., Orenstein, J.M., Reitz, M.S., Hayward, G.S., 1999. Comparison of genetic variability at multiple loci across the genomes of the major subtypes of Kaposi’s sarcoma-associated herpesvirus reveals evidence for recombination and for two distinct types of open reading frame K15 alleles at the right-hand end. J. Virol. 73, 6646–6660. Rao, V.B., Black, L.W., 1988. Cloning, overexpression and purification of the terminase proteins gp16 and gp17 of bacteriophage T4. Construction of a defined in-vitro DNA packaging system using purified terminase proteins. J. Mol. Biol. 200, 475–488. Rest, J.S., Ast, J.A., Austin, C.C., Waddell, P.J., Tibbetts, E.A., Hay, J.A., Mindell, D.P., 2003. Molecular systematics of primary reptilian lineages and the tuatara mitochondrial genome. Mol. Phylogenet. Evol. 29, 289–297. Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19, 1572–1574. Rosa, M.D., Gottlieb, E., Lerner, M.R., Steitz, J.A., 1981. Striking similarities are exhibited by two small Epstein-Barr virus-encoded ribonucleic acids and the adenovirus-associated ribonucleic acids VAI and VAII. Mol. Cell. Biol. 1, 785–796. Sample, J., Young, L., Martin, B., Chatman, T., Kieff, E., Rickinson, A., Kieff, E., 1990. Epstein-Barr virus types 1 and 2 differ in their EBNA3A, EBNA-3B, and EBNA-3C genes. J. Virol. 64, 4084–4092. Schultz, E.R., Rankin Jr., G.W., Blanc, M.P., Raden, C.C., Tsai, C., Rose, T.M., 2000. Characterization of two divergent lineages of macaque rhadinoviruses related to Kaposi’s sarcoma-associated herpesvirus. J. Virol. 74, 4919–4928. Searles, R.P., Bergquam, E.P., Axthelm, M.K., Wong, S.W., 1999. Sequence and genomic analysis of a rhesus macaque rhadinovirus with similarity to Kaposi’s sarcoma-associated herpesvirus/human herpesvirus 8. J. Virol. 73, 3040–3053. Sharp, P.M., 1985. Does the ‘non-coding’ strand code? Nucleic Acids Res. 13, 1389–1397. Sheldrick, P., Berthelot, N., 1975. Inverted repetitions in the chromosome of herpes simplex virus. Cold Spring Harb. Symp. Quant. Biol. 39, 667–678. Springer, M.S., Murphy, W.J., Eduardo, E., O’Brien, S.J., 2003. Placental mammal diversification and the Cretaceous-Tertiary boundary. Proc. Natl. Acad. Sci. U.S.A. 100, 1056–1061. Steven, A.C., Spear, P.G., 1997. Herpesvirus capsid assembly and envelopment. In: Chiu, W., Burnett, R.M., Garcea, R. (Eds.), Structural Biology of Viruses. Oxford University Press, New York and Oxford, pp. 312–351. Stevens, J.G., Wagner, E.K., Devi-Rao, G.B., Cook, M.L., Feldman, L.T., 1987. RNA complementary to a herpesvirus alpha gene mRNA is prominent in latently infected neurons. Science 235, 1056–1059. Tomasec, P., Wang, E.C., Davison, A.J., Vojtesek, B., Armstrong, M., Griffin, C., McSharry, B.P., Morris, R.J., Llewellyn-Lacey, S., Rickards, C., Nomoto, A., Sinzger, C., Wilkinson, G.W., 2005. Downregulation of natural killer cell-activating ligand CD155 by human cytomegalovirus UL141. Nat. Immunol. 6, 181–188. Trus, B.L., Chen, N.Q., Newcomb, W.W., Homa, F.L., Brown, J.C., Steven, A.C., 2004. Structure and polymorphism of the UL6 portal protein of herpes simplex virus type 1. J. Virol. 78, 12668–12671. Valpuesta, J.M., Carrascosa, J.L., 1994. Structure of viral connectors and their function in bacteriophage assembly and DNA packaging. Q. Rev. Biophys. 27, 107–155. Wadsworth, S., Jacob, R.J., Roizman, B., 1975. Anatomy of herpes simplex virus DNA. II. Size, composition, and arrangement of inverted terminal repetitions. J. Virol. 15, 1487–1497. Wagner, E.K., 1985. Individual HSV transcripts. In: Roizman, B. (Ed.), The Herpesviruses, vol. 3. Plenum Press, New York, pp. 45–104. Waltzek, T.B., Kelley, G.O., Stone, D.M., Way, K., Hanson, L., Fukuda, H., Hirono, I., Aoki, T., Davison, A.J., Hedrick, R.P., 2005. Koi herpesvirus represents a third cyprinid herpesvirus (CyHV-3) in the family Herpesviridae. J. Gen. Virol. 86, 1659–1667. Wikoff, W.R., Liljas, L., Duda, R.L., Tsuruta, H., Hendrix, R.W., Johnson, J.E., 2000. Topologically linked protein rings in the bacteriophage HK97 capsid. Science 289, 2129–2133. Yang, Z., 1997. PAML: a program package for phylogenetic analysis by maximum likelihood. CABIOS 13, 555–556. Zhou, Z.H., Dougherty, M., Jakana, J., He, J., Rixon, F.J., Chiu, W., 2000. Seeing the herpesvirus capsid at 8.5 Å. Science 288, 877–880. Zong, J.C., Ciufo, D.M., Alcendor, D.J., Wan, X., Nicholas, J., Browning, P.J., Rady, P.L., Tyring, S.K., Orenstein, J.M., Rabkin, C.S., Su, I.J., Powell, K.F., Croxson, M., Foreman, K.E., Nickoloff, B.J., Alkan, S., Hayward, G.S., 1999. High-level variability in the ORF-K1 membrane protein gene at the left end of the Kaposi’s sarcoma-associated herpesvirus genome defines four major virus subtypes and multiple variants or clades in different human populations. J. Virol. 73, 4156– 4170. Zou, P., Isegawa, Y., Nakano, K., Haque, M., Horiguchi, Y., Yamanishi, K., 1999. Human herpesvirus 6 open reading frame U83 encodes a functional chemokine. J. Virol. 73, 5926–5933.