HHV6 Genome Similar and Different

U.A. Gompelsa, F.C. Kasoloa'b'c aDepartment of Infectious and Tropical Diseases, London School of Hygiene & Tropical Medicine, University of London, UK b Virology Unit, University Teaching Hospital, Lusaka, Zambia cAFRO VPD, World Health Organisation, Harare, Zimbabwe

Genome classification and biology

Roseoloviruses, human herpesviruses 6 and 7 (HHV-6, HHV-7) are widespread T lymphotropic and neurotropic viruses causing mostly benign infections. However, particularly for HHV-6, during some primary as well as secondary reactivated infections, which can follow immune aberrations or deficiencies, there can be severe complications which can lead to fatalities. Thus, this is of relevance for immuno-suppressed HIV/AIDS or transplantation patients, as well as increasingly, for those with neurological disease, including encephalitis and a link with multiple sclerosis (primarily HHV-6A). Understanding when and how this virus does or does not cause disease is key to developing effective treatments plus evaluating the impact of HHV-6 infections on worldwide populations. Studies on the virus genome provide a foundation for this exploration and can guide the way towards development of new anti-virals as well as possible novel treatments for immune-related pathologies using this well-adapted virus as a guide.

The general properties of HHV-6 and the closely related HHV-7 and their genomes have been summarized in reviews and the original reports of their complete genomic sequences (Gompels et al., 1995; Nicholas, 1996; Megaw et al., 1998; Dominguez et al., 1999; Isegawa et al., 1999; Gompels, 2004). This chapter reviews overall properties of the genome of HHV-6 with some updates, while further details can be found in the first sequence papers as well as in Genbank nucleotide sequence and genome entries. There are two strain groups for HHV-6 (Ablashi et al., 1993), the prototypes are strain U1102 for HHV-6 variant A (HHV-6A) and strain Z29 for variant B (HHV-6B). HHV-6A strain U1102 was identified and first characterized in the UK (Downing et al., 1987) and HHV-6B strain Z29 in the USA (Lopez et al., 1988); both genomes are sequenced (Gompels et al., 1995; Dominguez et al., 1999). These laboratory strains are from adult HIV/AIDS patients from African countries: U1102 from Uganda and Z29 from Zaire, where the virus has reactivated from the immunosuppression giving a systemic infection. Other laboratory strains studied include HHV-6A strain GS, the first report of HHV-6 infection, from an adult HIV/AIDS patient in USA (Salahuddin et al., 1986), HHV-6A strain AJ, from an adult HIV/AIDS patient in UK (Tedder et al., 1987), and HHV-6B strain HST, from an exanthem subitum pediatric patient in Japan (Yamanishi et al., 1988). There are partial, or fragments of sequence available for strains GS and AJ. While for HHV-6B strain HST, the complete nucleotide sequence was also derived and represents the only primary childhood infection isolate of the genomes analyzed (Isegawa et al., 1999). These are closely related strains differing on average by 5% with increases in variation primarily at the ends of the genomes overlapping repetitive sequences as discussed further below. There is also one hypervariable gene at the centre of the genome which also encodes a variable glycoprotein, U47 or gO, and marks a region of genomic reorganization between herpesvirus subgroups as shown below (Gompels et al., 1995; Kasolo et al., 1997; Dominguez et al., 1999). The two strains of HHV-6B with genomic sequences available, not only show less variation than between the variants, HHV-6A and HHV-6B, strains, but also display increasing variation towards the ends of the genome and overlapping repetitive sequences (Dominguez et al., 1999; Isegawa et al., 1999). The genomes of two strains of HHV-7 have also been sequenced, JI and RK, which show less variation than between the HHV-6 strains (Nicholas, 1996; Megaw et al., 1998).

HHV-6B strains seem more prevalent in primary pediatric infections, where tested, primarily in European countries, USA and Japan, with occasional HHV-6A infection (Hall et al., 1994; van Loon et al., 1995; Ward et al., 2005; Zerr et al., 2005), and some evidence for increased congenital infection with HHV-6A (Hall et al., 1994, 2004; Ward et al., 2005). There also appears to be geographic variation, as in an African country, Zambia, childhood HHV-6A infections appear more frequent (Kasolo et al., 1997). These studies have been performed directly on blood samples taken during acute infections followed by DNA PCR and nucleotide sequencing. However, true distribution by strain-specific serology has yet to be performed, hindered by problems of the antibody cross-reactivity against lysates of whole virus antigen used in serological assays and the lack of defined single antigens with a combination of 100% specificity and immunodominance.

Genomes, cellular tropism and laboratory culture

Both HHV-6 and HHV-7 are T lymphotropic and neurotropic. Furthermore, it is likely that in vivo, the lytic and latent cell types that HHV-6 infects are highly specific subsets. While in vitro, virus is cultured in laboratory-adapted cell lines, and studies suggest that this may influence genomic composition. Such changes have been recently demonstrated for the related betaherpesvirus, human cytomegalovi-rus (HCMV), where serial passage in fibroblast rather than endothelial or leucocyte cell types has resulted in large genomic deletions (Cha et al., 1996; Murphy et al., 2003; Dolan et al., 2004). For HHV-6, some differences upon serial passage of strain Z29 in culture have been observed including expansion of repetitive sequences from the origin of lytic replication, terminal direct repeats, het region, and other rearrangements are possible (Gompels et al., 1995; Stamey et al., 1995; Dominguez et al., 1999; Gompels, unpublished). There may be less deletions observed for HHV-6 than in HCMV, since the reference strains Z29 and U1102 have both been isolated and propagated initially in primary cord blood cells, although passage in various leukaemic cell lines have also been reported. Thus, considerations of the genome of HHV-6 must also address issues of cellular tropism and possible changes arising from in vitro culture.

As described in other chapters, for routine culture both HHV-6 and HHV-7 have been adapted to grow in CD4+ T-leukaemic cell lines, for example, J-JHAN (Jurkat), HSB2, or Molt-3 for HHV-6, and SupT-1 for HHV-6 and HHV-7. In addition, there are some differences reported in growth of HHV-6 strains in leu-kaemic cell lines. However, preferential growth for both HHV-6 and HHV-7 are in activated cord blood lymphocytes or mononuclear cells (CBL, CBMC) or in peripheral blood lymphocytes or mononuclear cells (PBL, PBMC). Screened cord blood is preferred, as infection with laboratory strains can result in reactivation of resident latent virus from adult blood, although there is occasionally a similar risk from cord blood.

Both HHV-6 and HHV-7 have a cellular tropism for T lymphocytes as shown in vivo during viremia from acute infection as well as in vitro. Infection and lytic replication results in a characteristic cytopathic effect of large cells (cytomegalia) and ballooning cells. The cells are completely permissive for replication and virus production, with infection resulting in cell death by necrotic lysis. Although, there is also in vitro and some in vivo evidence that infection also causes cell death by apoptosis in uninfected or non-productively infected bystander cells (Inoue et al., 1997; Secchiero et al., 1997; Yasukawa et al., 1998). Studies show that CD4+, CD8+ and gamma/delta T lymphocytes can be infected, but overall data suggest that activated CD4+ T lymphocytes are the preferential target of fully permissive infection in vivo (Takahashi et al., 1989; Lusso et al., 1991). Antibody to the T-cell-specific marker and signal transduction molecule, CD3 (OKT3) has been shown to augment infection of HHV-6 in both primary (Roffman and Frenkel, 1991) and T-leukaemic cell lines (H.A. Macaulay and U.A. Gompels, unpublished results). This antibody is also often used in transplantation patients and may aid virus replication.

Latent infection has been demonstrated within monocytic/macrophage cells as well as bone marrow progenitor cells similar to that observed for HCMV, and may be a general property of betaherpesvirus infection (Kondo et al., 1991; Gompels et al., 1993, 1994; Kempf et al., 1997; Yasukawa et al., 1997). A strong interaction with monocytic/macrophage cells has been recorded during HHV-6 primary infection and may also include a form of latency with specific restricted transcripts as well as replication within some differentiated subsets (Kondo et al., 2002a,b, 2003). Similarly, latent infections of primary macrophages have also been observed for HHV-7 (Zhang et al., 2001). Higher levels of HHV-7 compared to HHV-6 can be detected by PCR in blood and saliva of asymptomatic adults (Di Luca et al., 1995; Kidd et al., 1996; Gautheret-Dejean et al., 1997). Given the similar prevalence, this suggests that HHV-6 has more stringent regulation of latency. Studies also show that HHV-7 can also act to reactivate latent HHV-6 infections (Katsafanas et al., 1996). Additional HHV-6-specific latency-associated transcripts have been identified from the U94/Rep gene; it is highly conserved between strains, but deleted in HHV-7 (Nicholas, 1996; Megaw et al., 1998; Rapp et al., 2000). This HHV-6 gene has roles in gene expression and replication modulation that may contribute to HHV-6 latency regulation (Rotola et al., 1998; Mori et al., 2000; Rapp et al., 2000; Dhepakson et al., 2002; Turner et al., 2002; Caselli et al., 2005). In rare cases, evidence for viral genome integration has been observed, which can result in high levels of persisting, circulating HHV-6 DNA, not always with concomitant reactivation and gene expression (Tanaka-Taya et al., 2004; Ward et al., 2005). This may be mediated by the numerous repetitive sequences in the genome with similarities to the host genome. The clinical significance, genomic structure and composition are under evaluation.

The genome detailed in this chapter is from the reference HHV-6A strain U1102 with comparisons made to HHV-6B strains HST and Z29 as well as comparisons to HHV-7 and other human herpesviruses. The genomic sequence was derived from plasmid clones from early passage virus propagated in cord blood (Martin et al., 1991; Gompels et al., 1995). Only one region was intractable for plasmid cloning, covering the R3 repeat, and this sequence was determined directly from PCR amplified products. Sequencing from uncultivated virus directly from tissue or blood samples has only been conducted on small fragments to compare strains.

Genome structure and repetitive sequence

HHV-6A strain U1102 has a genome of 159,321 bp, while HHV-6B strains HST and Z29 are slightly larger, the respective sizes are 161,573bp and 162,114bp (Gompels et al., 1995; Dominguez et al., 1999; Isegawa et al., 1999). Much of the differences in length accommodated by variation in the repetitive sequences in the genome include the terminal repeat, which bound both ends of the genome. This direct repeat (DR), or terminal repeat (TR), varies in size: U1102 8087 bp, HST 8231 bp, and Z29 8793 bp. These DR regions are themselves bounded by copies of human telomeric repeats, which have been postulated to play a role in latency possibly by stabilizing the genome as a mini-chromosome (see Fig. 1). The telomeric repeat bounding the left end of the repeats is heterogeneous, het, thus, at

U36 U37 60kb

Hhv6 Treatment

U36 U37 60kb

30kb

90kb

U77 120kb

150kb

Fig. 1 Structure of HHV-6 genome with herpesvirus and betaherpesvirus conserved genes. Herpesvirus conserved genes are indicated in black while betaherpesvirus conserved genes are indicated patterned. Similarities were determined by comparisons of encoded amino acid sequences as described in the text. TR, terminal or direct repeat (DR); t, telomeric repeat; ORI, origin of lytic replication; and R1, R2, R2, repetitive sequences. Conserved or HHV-6-specific genes are indicated for reference: p41 DNA proc-essivity factor also monoclonal antibodies used in diagnostics; pp100, major immunodominant tegument phospho-protein; U94/REP, parvovirus rep homologue and gene/replication regulatory latency gene; IEA and IEB, immediate early regulatory genes; POL, DNA polymerase; gB, gH, gL, gO, gm, gn glycoprotein; dutpase; vccr, viral chemokine receptors; MCP, major capsid protein; GCK, ganciclovir kinase, site resistance mutations; OX-2, homologue of this member of immunoglobulin gene family.

30kb

90kb

U77 120kb

150kb

Fig. 1 Structure of HHV-6 genome with herpesvirus and betaherpesvirus conserved genes. Herpesvirus conserved genes are indicated in black while betaherpesvirus conserved genes are indicated patterned. Similarities were determined by comparisons of encoded amino acid sequences as described in the text. TR, terminal or direct repeat (DR); t, telomeric repeat; ORI, origin of lytic replication; and R1, R2, R2, repetitive sequences. Conserved or HHV-6-specific genes are indicated for reference: p41 DNA proc-essivity factor also monoclonal antibodies used in diagnostics; pp100, major immunodominant tegument phospho-protein; U94/REP, parvovirus rep homologue and gene/replication regulatory latency gene; IEA and IEB, immediate early regulatory genes; POL, DNA polymerase; gB, gH, gL, gO, gm, gn glycoprotein; dutpase; vccr, viral chemokine receptors; MCP, major capsid protein; GCK, ganciclovir kinase, site resistance mutations; OX-2, homologue of this member of immunoglobulin gene family.

the left end of the genome and the junction of the right end of unique sequence with the right DR. Redundant copies of this repeat are also present in opposite orientation in either side of the origin of lytic replication and suggest a role of RNA copies in replication possibly through priming the lagging strand (Gompels and Macaulay, 1995; Gompels et al., 1995; Mrazek and Karlin, 1998; Dominguez et al., 1999). These repeats are adjacent to the pac repeats, directing DNA packaging, at the ends of the genomes and the junction between the DR and the unique sequences. These have been functionally defined and described in detail for both HHV-6A and HHV-6B (Thomson et al., 1994; Gompels et al., 1995; Dominguez et al., 1999; Turner et al., 2002; Borenstein et al., 2004). The origin was defined as a minimal 400 bp, but further study suggests that the functioning ori is larger, 1.3 kb containing both repeats for binding the origin-binding protein (OBP) followed by imperfect direct repeats, IDR1, 2, 3, which vary between the strains (Dykes et al., 1997; Turner et al., 2002). The ORI, is mutagenic and multiple copies have been identified in tissue culture passaged Z29 strains (Stamey et al., 1995). Pac sequences together with the origin of replication can form 'amplicons' which could be used as artificial vectors for gene delivery as described with some applications for stem cell delivery (Deng and Dewhurst, 1998; Turner et al., 2002; Borenstein et al., 2004). A series of repetitive sequence regions are found at the heterogeneous right end of the genome, R1, R2, and R3. R1 has reiterations encoding an SR domain in the IE2 homologue in HHV-6 U86 (Gompels et al., 1995). R2 contains simple TG repeats, resulting in a large reading frame open in six frames, U88, and unlikely to be coding. In HHV-6B it has been further sub-divided to R2A and R2B regions (Dominguez et al., 1999). R3 contains tandem repeats of an approximately 110 bp sequence which includes a KpnI restriction endonuclease site, and has been shown to have a role as an enhancer for U95 expression (Takemoto et al., 2001). This region in HHV-6A was refractory to cloning, possibly due to secondary structure formation from the repeated sequences there (Martin et al., 1991; Gompels et al., 1995). Not surprisingly, this right end of the genome which contains most of the repetitive sequence of the genome is heterogeneous between all strains and includes the R1, R2, R3 repeats followed by the het region and the DR repeats.

The overall composition of the genome is 43% G + C with 58% in the DR and 41% in the unique sequences. Lower G + C composition is also found at the origin for lytic replication, ORI. There is compositional polarity around the ORI as found in other organisms and particularly betaherpesviruses, this could be driven by the telomeric repeats as described above, or could be due to errors accumulated during copy repair synthesis of the lagging strand from priming with Okazaki fragments (Gompels et al., 1995; Gompels and Macaulay, 1995; Mrazek and Karlin, 1998; Dominguez et al., 1999). Other compositional biases are present across the IEA region which has marked localized CpG suppression, suggesting mutagenic effects of methylation possibly during regulation of gene expression during latent infections as described previously (Gompels et al., 1995). This localized suppression has only been observed in betaherpesviruses, while in alphaherpesviruses there is none and in gammaherpesviruses there is global suppression suggested related to distinct forms of latency (Honess et al., 1989). Thus, localized CpG suppression is a feature in betaherpesviruses, which may reflect similarities in latency control in monocytic/ macrophages or bone marrow progenitor cells for this herpesvirus subgroup. The overall structural features of the HHV-6 genome show some similarities to all betaherpesviruses, but in roseoloviruses, in particular HHV-6, there are unique features as well, which are the residues of this virus specific evolution and frame its biology.

Genome rearrangements and relationship to other herpesviruses

HHV-6 together with HHV-7 form the Roseolovirus grouping of the betaherpes-virus subgroup of herpesviridae. This has also been termed as beta-2 herpesviruses, compared to the Cytomegalovirus (CMV) grouping of beta-1 herpesviruses. Recent studies, particularly on gB and polymerase sequences of extensive sets of animal herpesviruses suggest that this classification may be broadened. Examples include the roseolovirus, porcine CMV-significant in xeno-transplantation, the fatal elephant roseolovirus, and recently identified chimpanzee HHV-6 (Ehlers et al., 2001; Chmielewicz et al., 2003; Lacoste et al., 2005). Similar to other betaherpesvirus, infection is species specific, thus animal models with human viruses are restricted, although as with some of the CMV group, there may be some utility using these related (albeit distant) animal viruses for study of in vivo models for human infection.

The HHV-6 genome organization of conserved genes is a betaherpesvirus-specific arrangement, as first described for HHV-6A strain U1102. Both the genome organization and the encoded protein sequences share a closer relation than to the other herpesvirus lineages (Gompels and Macaulay, 1995; Dominguez et al., 1999) (Fig. 1). Seven discrete blocks of conserved genes can be identified as compared to other human herpesviruses, and these are rearranged in the alpha and gammaherpesvirus lineages of bird and mammalian herpesviruses (Gompels et al., 1995) (Fig. 2). The genome rearrangements shown by these lineage comparisons suggest that recombination has played a major role in their evolution. For example, the central conserved gene block III, appears to have recombined into repetitive sequences at the right end (relative) of the genome in the other two lineages (Gompels et al., 1995). This ancient recombination is supported by recent sequencing studies on hypervariable genes adjacent to this region in the centre of the genome or in the repetitive sequences towards the right end of the genome in HCMV where switching between phylogeny defined groups between adjacent genes are observed (Dolan et al., 2004; Mattick et al., 2004). In HHV-6, study at this same central locus, HCMV UL74/HHV-6U47, also provides evidence for strain recombination as described (Kasolo et al., 1997) and discussed below.

Genome composition and general molecular biology

HHV-6 encodes approximately 100 open reading frames, almost all of these appear coding, a few have been deleted or annotated in the updated gene list shown in

Fig. 2 Relationship between human herpesvirus genome organizations. HHV-6 shares with HCMV a betaherpesvirus specific arrangement of core-conserved genes, here collected into seven gene blocks, I-VII. The core genes are identified in Fig. 1. The numbers refer to the HHV-6 homologues of these core conserved genes. HSV and EBV represent the alpha and gammaherpesvirus lineages, respectively.

Fig. 2 Relationship between human herpesvirus genome organizations. HHV-6 shares with HCMV a betaherpesvirus specific arrangement of core-conserved genes, here collected into seven gene blocks, I-VII. The core genes are identified in Fig. 1. The numbers refer to the HHV-6 homologues of these core conserved genes. HSV and EBV represent the alpha and gammaherpesvirus lineages, respectively.

Table 1, owing to comparisons between HHV-7 and strains, spliced products identified, as well as using gene prediction software. In this chapter to highlight HHV-6-specific genes versus, Roseolovirus, betaherpesvirus-specific or herpesvirus conserved genes selected genome comparisons will be made, in particular to the other human roseolovirus, HHV-7. References are also to the prototypic beta-1 herpesvirus, HCMV, plus the initial prototypic herpesvirus, HSV, where many gene functions were first defined. HHV-7 is more compact at 144 kb or 153 kb in size for strains JI and RK, respectively, with similar long unique regions of 133 kb and variation in the DR of 5.8 and 10 kb each (Nicholas, 1996; Megaw et al., 1998). With a few notable exceptions described below, it shares all genes with HHV-6 as indicated with varying degrees of conservation of encoded amino acid sequences (Table 1, Fig. 1). In HHV-6, the open reading frames are designated from U1 to U100 with those in the direct repeats from DR1 to DR7, although recent analyses show evidence for splicing as well as expression for two of these, DR2 and DR6, which include previous DR designations as indicated (Fig. 1, Table 1) (Gompels et al., 1995). In HHV-7 similar nomenclature is used with homologous genes U2-U100 (Nicholas, 1996; Megaw et al., 1998). A few genes are lacking in HHV-7 compared to HHV-6 as discussed further below, but there are also a few HHV-7-specific genes noted H1-H7 (Gompels et al., 1995; Nicholas, 1996; Megaw et al., 1998). There is limited splicing observed in HHV-6, approximately 10% of the genes, these are primarily in the immediate early genes and selected early/late genes as noted experimentally or predicted from sequence motifs for DR1, DR6, U12, U15, U7, U66, U79, U83, U90 and U100 (Figs. 1, 2, Table 1). These are likely to be

Table 1

Features of HHV-6 strain U1102 genes

Table 1

Features of HHV-6 strain U1102 genes

Genea

Was this article helpful?

0 0

Post a comment