Definition of a Human Herpesvirus-6 Betaherpesvirus-Specific Domain in Glycoprotein gH That Governs Interaction with Glycoprotein gL: Substitution of Human Cytomegalovirus Glycoproteins Permits Group-Specific Complex Formation

Definition of a Human Herpesvirus-6 Betaherpesvirus-Specific Domain in Glycoprotein gH That Governs Interaction with Glycoprotein gL: Substitution of Human Cytomegalovirus Glycoproteins Permits Group-Specific Complex Formation

VIROLOGY 217, 517–526 (1996) 0146 ARTICLE NO. Definition of a Human Herpesvirus-6 Betaherpesvirus-Specific Domain in Glycoprotein gH That Governs I...

211KB Sizes 22 Downloads 30 Views

VIROLOGY

217, 517–526 (1996) 0146

ARTICLE NO.

Definition of a Human Herpesvirus-6 Betaherpesvirus-Specific Domain in Glycoprotein gH That Governs Interaction with Glycoprotein gL: Substitution of Human Cytomegalovirus Glycoproteins Permits Group-Specific Complex Formation R. A. ANDERSON,* D. X. LIU,† and U. A. GOMPELS*,†,1 *Viral Pathogenesis Unit, Department of Clinical Sciences, London School of Hygiene and Tropical Medicine, University of London, London, United Kingdom; and †Department of Medicine, University of Cambridge, Cambridge, United Kingdom Received October 16, 1995; accepted January 9, 1996 Formation of the glycoprotein gH/gL heterooligomer has important implications for understanding the pathology of human herpesvirus-6 (HHV-6)-associated disease because this complex is essential for infectivity and fusogenic cell-to-cell spread. Definition of the HHV-6 gH domain involved in protein–protein interactions was addressed by targeting regions defined by conserved cysteines identified by alignment of gH amino acid sequences representative of all herpesvirus subfamilies. Studies using site-directed mutagenesis and transient cellular expression showed that the N-terminus of HHV-6 gH includes a 230-amino-acid domain required for interaction with HHV-6 gL encompassing residues conserved specifically amongst betaherpesviruses. Interestingly, the human cytomegalovirus (HCMV) homologues, UL75 (gH) or UL115 (gL), can substitute for HHV-6 glycoproteins and participate in heterologous complex formation. Furthermore, the region which governs this heterologous gL binding also maps to the N-terminal portion of HHV-6 gH. Although both proteins can functionally substitute for complex formation there are also specific differences. Surprisingly, further deletion of HHV-6 gH to 145-amino-aciddomain residues abolishes complex formation with HHV-6 gL but allows interaction with HCMV gL. This may be related to requirements in HHV-6 for homodimer formation before complex formation between gH and gL. Under nonreducing conditions HHV-6 gH and gL form multimeric complexes consistent with intra- and intermolecular dimer formation stabilised by disulphide bonds whereas for HCMV there is no evidence for dimer formation for gH and multimeric complexes have only been observed between gH and gL. In summary, both HHV-6 and HCMV glycoproteins can interact and the heterologous complex between HHV-6 gH and HCMV gL is possibly more stable. This may result in important biological consequences in vivo during cellular coinfections by facilitating spread of the viruses, with applications to altered cellular tropisms and effects on reactivation from the latently infected cell. q 1996 Academic Press, Inc.

INTRODUCTION

sented (Challoner et al., 1995). Further, in AIDS patients widely disseminated infections have been observed and there are speculations on its role in AIDS progression (Lusso et al., 1989; Fairfax et al., 1994; Corbellino et al., 1993; Knox and Carrigan, 1994, 1995; Knox et al., 1995b). Moreover, there is evidence that HHV-6 infection can result in immunodeficiency from bone marrow suppression in immunocompetent individuals (Gompels et al., 1994; Knox et al., 1995a). A major cytopathic effect in HHV-6 infections is the formation of cytomegalia, ‘‘giant cells,’’ caused by T-lymphocyte cell membrane fusion (Tedder et al., 1987; Wyatt et al., 1990). This appears as the predominant route of spread from cell to cell during HHV-6 infection. Monoclonal antibodies (MAbs) raised against HHV-6 which can inhibit this cell fusion also inhibit virus cellular spread (Foa-Tomasi et al., 1991a). Similarly, preincubation of infectious virus with these antibodies neutralise virion infectivity and results in the absence of cytopathology. These MAbs are specific for glycoprotein gH (Liu et al., 1993b). Therefore, HHV-6 gH is a glycoprotein with fusogenic properties, involved in both the initiation of infection and spread of infectious virus from cell to cell. All subfamilies of herpesviruses

In most instances, primary HHV-6 infection is asymptomatic or causes exanthem subitum (Yamanishi et al., 1988; Dewhurst et al., 1992) and seroconversion typically occurs early in infancy (Briggs et al., 1988; Okuno et al., 1989) The virus has a preferred cellular tropism in vivo for CD4/ T-lymphocytes and undergoes lytic replication in this cell type (Tedder et al., 1987; Takahashi et al., 1989). Also, there is evidence for a lifelong latent or persistent infection in sites that include salivary gland epithelial cells and monocyte/macrophages (Fox et al., 1990; Kondo et al., 1991). In immunocompromised situations, HHV-6 can reactivate and is associated with secondary infections with life threatening complications in transplant recipients or cancer patients such as pneumonitis, bone marrow suppression, and encephalitis (Drobyski et al., 1993, 1994; Cone et al., 1993; Morris et al., 1989; Carrigan et al., 1991; Knox et al., 1995a,b). A possible association with multiple sclerosis has also been pre1 To whom correspondence and reprint requests should be addressed. Fax: /44 171 637 4314. E-mail: [email protected].

0042-6822/96 $18.00

517

AID

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

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

vira

AP: Virology

518

ANDERSON, LIU, AND GOMPELS

examined to date have a gH homologue with similar properties (Gompels and Minson, 1986; McGeoch and Davison, 1986; Keller et al., 1987; Oba and Hutt-Fletcher, 1988; Cranage et al., 1988; Gompels et al., 1988a,b; Heineman et al., 1988; Pachl et al., 1989; Gompels et al., 1991, 1992). Sensitive dot matrix analysis of related gH homologues indicates that at the amino acid level conservation is most obvious at the C-terminus whilst N-terminal amino acid sequences are more divergent and only shared within subgroups (Gompels et al., 1988b). HHV-6 gH shares most similarity in both N and C terminal domains with gH from the betaherpesvirus subfamily which includes HCMV (Gompels et al., 1992). We have previously demonstrated that in HHV-6-infected cells gH associates with a conserved virus encoded glycoprotein called gL (Liu et al., 1993a,b). This physical association can be reproduced in vitro by coexpression in rabbit reticulocyte lysates or by transient coexpression in cultured cells. Heterooligomer complex formation has been demonstrated for herpes simplex virus type 1 (HSV-1) (Gompels and Minson, 1989; FoaTomasi et al., 1991b; Hutchinson et al., 1992), HCMV (Kaye et al., 1992) and EBV (Yaswen et al., 1993). Unlike gH, gL does not have significant sequence similarity across herpesvirus subgroups and has been identified as a ‘‘positional’’ homologue which encodes a small secreted glycoprotein with similar properties in complex formation with gH (Kaye et al., 1992). However, like gH, within a subgroup there is sequence similarity throughout the entire sequence based primarily around conserved cysteines and this is observed between both gH and gL homologues of HCMV and HHV-6. Data from HSV-1 and HCMV studies indicate that gH together with gL forms a complex which is an important functional unit required for infection and cellular spread. Expression of gH in the absence of other viral proteins results in an antigenic conformation distinct from those observed in HSV-1-infected cells and wild type (wt) conformation can be restored by the presence of specific virus protein (Gompels and Minson, 1989; Foa-Tomasi et al., 1991b; Gompels et al., 1991). Recently, it has been shown that normal processing can be achieved by coexpression with gL (Hutchinson et al., 1992). Likewise, in the absence of gL, gH is not transported to the plasma membrane and is detected predominantly as perinuclear staining (Gompels and Minson, 1989; Cranage et al., 1988; Roberts et al., 1991; Hutchinson et al., 1992; Kaye et al., 1992; Spaete et al., 1993). Coexpression of gH with gL results in transport to the plasma membrane (Gompels and Minson, 1989; Hutchinson et al., 1992; Kaye et al., 1992). Inappropriate transport of gH through intracellular membranes is reflected by aberrant posttranslational modification (Gompels and Minson, 1989; Roberts et al., 1991) and presumably accounts for at least some of the observed conformational differences. As yet there is only limited evidence that these observations

can be extended to HHV-6 gH and gL in that human sera recognises the gH/gL complex better than gH expressed alone and shows no reaction to gL (Liu et al., 1993a). However, given their similarity, especially to HCMV-encoded products, some of the properties of the complex between gH and gL in mediating processing and transport may also apply to HHV-6. Previous data for the gH and gL homologues indicates there are separate domains for interaction with gL and for mediating cell fusion. All gH homologues have properties in mediating fusogenic cellular spread, and only the C-terminal cysteine domain is conserved in all herpesvirus gH homologues, therefore this domain has been implicated in cell fusion (Gompels et al., 1988b). Consistent with this observation is our previous study showing that at least part of the HHV-6 gH domain involved in fusion is located in the external portion of gH in a 23 amino acid region that includes a conserved N-linked glycosylation motif, although long-range effects that involve the N-terminus of gH cannot be formally excluded as this region was defined by conformation-dependent antibodies (Liu et al., 1993b). In contrast, analysis of antibody-resistant mutants in HSV-1 gH has shown that the nonconserved N-terminal domain may interact with gL. In this study antibody recognition is dependent on gH interaction with another virus protein (subsequently shown to be gL) and maps to a conformational dependant epitope spanning the N-terminal 320 amino acids (Gompels et al., 1991). Thus, in the external domain there is evidence for at least two separate domains, one for gL interaction which is necessary for conformation and the second for mediating fusion or interaction with a cellular fusogen. The object of this work was first to investigate interactions between HHV-6 gH and gL to determine which domain governs these intermolecular interactions within this virus and second, to investigate the nature of the betaherpesvirusspecific sequence conservation across both the N and C terminal domains and whether this is sufficient to permit functional substitution in terms of heterooligomer formation. Finally, as there are conserved cysteines throughout the HHV-6 and HCMV gH and gL homologues our third objective was to determine the role of disulphide bond formation in complex formation. Our results show a HHV-6 gH domain required for interaction between gH and gL that is common to other betaherpesviruses. We also demonstrate betaherpesviruses-specific substitution of gH or gL and heterologous gH/gL complex that is dependent on disulphide bond formation. These results contribute significantly to the understanding the functions of gH and gL in HHV-6 pathogenesis and suggest ways in which related viruses may interact in vivo.

AID

vira

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

MATERIALS AND METHODS Cells, virus, and plasmids Vero cells were grown in Dulbecco’s modified Eagle’s (DME) medium supplemented with 10% fetal calf serum

AP: Virology

HUMAN HERPESVIRUS-6 BETAHERPESVIRUS-SPECIFIC DOMAIN

(FCS) and 1% penicillin–streptomycin (ICN Flow). BHK cells were maintained in Glasgow modified Eagle’s medium supplemented with 10% newborn calf serum (NCS), 10% tryptose phosphate broth and antibiotics as described above. Vaccinia virus (vTF7) was grown in BHK cells and titre determined by plaque assay on Vero cells. Plasmids pH6H2 and pH6CH2, encoding the nucleotide sequence of HHV-6 strain U1102 gH and gL, respectively, and plasmids expressing mutant forms of HHV-6 gH (pH6H4, pH6H5, and pH6H6) have been described elsewhere (Liu et al., 1993a,b). Monoclonal antibodies MAb LP14 (Minson et al., 1986), which recognises a nine-amino-acid linear epitope from HSV-1 gD (Liu et al., 1991), was a gift from Dr. A. C. Minson, Department of Pathology, University of Cambridge, U.K. 12CA5 supplied from BAbCO, U.S.A. is specific for a nine amino acid epitope from influenzavirus haemagglutinin type 1 (Wilson et al., 1984). Site-directed mutagenesis Oligonucleotide directed in vitro mutagenesis of single-stranded DNA was carried out using the Kunkel (1985) method and components supplied by Bio-Rad. Putative mutants were screened, and the entire gene sequence was determined by nucleotide sequencing (Sanger et al., 1977) using a commercially available T7 promoter primer (Promega) and internal gH primers (Gompels et al., 1993). Plasmid constructs Plasmids pH6H7, pH6H8, pH6H9, pH6H10 (based on plasmid pH6H4), and pH6CH4 were constructed by sitespecific mutagenesis (Fig. 1). Deletion mutants pH6H7, pH6H9, and pH6H10, progressively removing portions of the external domain of gH, were constructed by insertion in-frame of a stop codon. Oligonucleotide 1, 5* CCTACCGATTAACGCATGCAG 3*, which contained a single base change (bold type) was used to delete from D 433 giving plasmid pH6H7. Oligonucleotide 2, 5* GACAGATGCAATAGGACAACGTTTCCT 3* was used to construct plasmid pH6H9 which is deleted from Q 145. Oligonucleotide 3, 5* TTGACAGACGATTAACTATTGATAGTC 3*, was utilised to delete from D 230 generating plasmid pH6H10. Oligonucleotide 4, 5* GTTAGCCAAATTTGAGATGTC 3*, was used to make pH6H8, which is deleted for four codons. Plasmid pH6CH4 was constructed by introduction of the DNA sequence encoding the epitope for MAb 12CA5 in-frame in the HHV-6 gL gene after codon 29 using oligonucleotide 5 (the encoded epitope is underlined): 5* CTTTCTATTTTTACCCATACGATGTTCCAGATTACGCTGAAAAACTTGAC 3*. HCMV AD169 UL75 (gH) and UL115 (gL) were subcloned from the vector pRK19

AID

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

519

(Kaye et al., 1992) into BglII/BamHI cut pIBM6 (Liu et al., 1991). Transient transfection system and labeling of cells The gH and gL genes cloned between the T7 promoter and terminator were expressed transiently in mammalian cells as described by Fuerst et al. (1986). Plasmids were introduced via liposome-mediated transfection using conditions recommended by the manufacturer (Gibco BRL) into subconfluent monolayers of Vero cells already infected with 1 PFU/cell of recombinant vaccinia virus vTF-3 expressing T7 RNA polymerase. Cell culture medium was replaced 4 hr posttransfection with DME medium minus methionine (Gibco BRL) with the growth supplements indicated above. Proteins were radiolabeled with 30 mCi [35S]methionine (ú1000 Ci/mmol; Amersham) for 13 hr at 377. Immunoprecipitation and gel electrophoresis Labeled cells were washed three times in ice-cold PBS, harvested in 0.5 ml RIPA buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1% sodium deoxycholate, 0.1% SDS, 1% Nonidet P-40) and cellular debris pelleted by centrifugation at 13,000 g for 20 min. Cell lysates were precleared by addition of 25 ml of a 1:1 mix of RIPA buffer:protein A Sepharose (Sigma) for 1 hr at 47 followed by centrifugation. Equivalent TCA precipitable counts were diluted to 0.5 ml with RIPA buffer, and 1 ml of the appropriate MAb was added and rotated at 47 for 18 hr. Immune complexes were precipitated using protein A Sepharose, washed three to five times using 1 ml of lysis buffer, and bound proteins disassociated in 15 ml loading buffer. Samples were subjected to electrophoresis in single percentage (12%), or 4–12% gradient precast minipolyacrylamide gels (Novex) under denaturing, reducing conditions unless indicated otherwise. Gels were fixed and treated with Amplify (Amersham) according to the manufacturers instructions prior to exposure to Hyperfilm (Amersham) for fluorography. RESULTS The external portion of gH contains a region required for gL association A schematic representation of HHV-6 gH is shown in Fig. 1. (pH6H2). The amino acid sequence of gH has features typical of N-linked glycosylation (Gavel and von Heijne, 1990) which may occur at 14 possible sites and includes motifs conserved in HHV-6 (N3-N10) and in all herpesvirus subfamilies (N14). There are a total of 14 cysteines residues in gH, 12 of which are found in other representative herpesviruses, suggesting conservation of a functional role. These are abbreviated as C1 for betaherpesviruses only, C2 for beta- and gammaherpesviruses, and C3 for alpha-, beta- and gammaherpesvi-

vira

AP: Virology

520

ANDERSON, LIU, AND GOMPELS

FIG. 1. Schematic view of HHV-6 gH (pH6H2) illustrating the epitope tag position in pH6H4. The putative signal peptide (SP) and transmembrane (TM), external, and cytoplasmic domains are illustrated. Possible sites of N-linked glycosylation are indicated (N1-14). Conserved cysteines are marked C1 (betaherpesvirus specific), C2 (beta and gammaherpesvirus specific), and C3 (alpha, beta, and gammaherpesvirus specific). Mutant forms of gH (pH6H5-10) are pictured underneath. The numbered mutant residues refer to the amino acids in gH excluding the epitope tag. Boxed areas indicate regions conserved amongst gH homolgues determined by dot matrix analysis, amino acid sequence comparisons (Gompels et al., 1988b, 1992), and multiple alignments of HHV-6 gH with all gH homologues identified to date.

ruses (Fig. 1). We have previously defined antibody reagents specific for HHV-6 gH, including human sera and MAbs (Liu et al., 1993a,b). In addition, we have inserted at the beginning of the protein, a linear epitope in gH (pH6H4) which is recognised by LP14 (Liu et al., 1993a). This has provided a useful tool for studying protein– protein interactions without relying on endogenous epitopes which may be the target of antibodies that interfere with binding. The truncated forms of gH employed in this study are also illustrated in Fig. 1. Mutant pH6H5 and pH6H6 have been described previously (Liu et al., 1993b), the former lacking the C-terminal sequences which form the putative cytoplasmic anchor, transmembrane domain, and the N14 conserved glycosylation site within the external domain. The latter, pH6H6, is deleted for a similar region but retains the entire external domain. pH6H7 encodes two-thirds of the gH ectodomain and specifically does not include six cysteine residues conserved amongst all herpesvirus subfamilies (C3). pH6H8 has an in-frame 12-bp deletion that removes the conserved N14 glycosylation site NGSV. pH6H9 retains 145

amino acids from the N-terminus of gH which includes a pair of cysteine residues conserved in betaherpesviruses (C1). Recombinant pH6H10 is deleted for a region that encompasses paired cysteine residues conserved amongst beta and gamma herpesviruses (C2 and C3). The ability of the mutant forms of HHV-6 gH to interact with HHV-6 gL was examined by transient coexpression in Vero cells. The total amount of radiolabeled proteins immunoprecipitated was approximately equal (see Materials and Methods) (Fig. 2, lanes 2–10). As expected, epitope-tagged gH and gL were coimmunoprecipitated by MAb LP14 when expressed in combination (Fig. 2, lane 13), whereas only tagged gH and not gL was immunoprecipitated when both proteins were expressed individually (Fig. 2, lanes 11 and 12, respectively). The gH recombinants were competent for interaction with gL, as demonstrated by coimmunoprecipitation of gL (Fig. 2, lanes 14–17 and 19), with the exception of mutant pH6H9 which failed to immunoprecipitate detectable amounts of gL (Fig. 2, lane 18). This lack of binding was observed even though equal counts had been loaded per lane,

AID

vira

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

AP: Virology

HUMAN HERPESVIRUS-6 BETAHERPESVIRUS-SPECIFIC DOMAIN

521

FIG. 2. Total [35S]methionine-labeled protein extracts from transfected Vero cells (lanes 2 to 10) were immunoprecipitated with LP14 (lanes 11 to 19). Lanes 2 and 11, pH6H4 (LP14 epitope tagged gH); and lanes 3 and 12, pH6CH2 (wt gL). Lanes 3 to 10 and 12 to 19 were all transfected with pH6CH2 and in addition pH6H4 (lanes 4 and 13), pH6H5 (lanes 5 and 14), pH6H6 (lanes 6 and 15), pH6H7 (lanes 7 and 16), pH6H8 (lanes 8 and 17), pH6H9 (lanes 9 and 18), and pH6H10 (lanes 10 and 19), respectively. 14C-labeled molecular weight markers are indicated in kilodaltons (lane 1).

and was highlighted by the observation that while similar intensities were observed for H7 and H9 gH mutant bands, only H7 clearly coprecipitated gL and H9 did not (Fig. 2, lanes 16 and 18). Overexposure from 3 (Fig. 2) to 30 days gave the same result. In the overexposed film H9 shown in lane 18 is darker than any other band shown in Fig. 2, but gL is absent. Therefore, the HHV-6 gH cytoplasmic, transmembrane sequences and part of the external domain which incorporates paired cysteine residues (C3) conserved in all herpesviruses as well as those conserved only in beta and gammaherpesviruses (C2) are not required for complex assembly. These results define the region necessary for association with gL to the N-terminal 230 amino acids of gH which includes paired cysteine residues (C1) only conserved amongst betaherpesviruses (Fig. 1).

ity of the exogenous 12CA5 epitope for gL. 12CA5 coimmunoprecipitated wt gH and epitope-tagged gH (Fig. 3, lanes 23 and 25) with tagged gL. In the reciprocal experiment, LP14 immunoprecipitated tagged gH (Fig. 3, lane 11) and coimmunoprecipitates wt or tagged gL with tagged gH (Fig. 3, lanes 16 and 17). Therefore, addition of the 12CA5 antibody recognition site to gL did not interfere discernibly with gH association and confirmed directly the interaction and identity of HHV-6 gL with gH as observed indirectly previously by coimmunoprecipitation with gH using gH antibodies (Liu et al., 1993a,b). HCMV gH interacts with HHV-6 gL and HCMV gL can substitute for HHV-6 gL to form a heterologous complex with HHV-6 gH

The coding region for a nine-amino-acid linear epitope from influenza virus haemagglutinin type 1 was inserted in-frame at the N-terminus of HHV-6 gL upstream of the putative signal sequence cleavage site. Thus, MAb 12CA5, specific for this epitope, can be used in immunoprecipitations to examine gH/gL interactions through gL. In Fig. 3 wt gH and gL and epitope-tagged gH or gL are expressed individually or in combination. The cell lysate from each transfection was divided into aliquots of equivalent counts (Fig. 3, lanes 2–9) and immunoprecipitated with LP14 or 12CA5 (lanes 10–17 and 18–25, respectively). 12CA5 immunoprecipitated only tagged gL (Fig. 3, lane 21), and not other glycoproteins without the epitope tag (Fig. 3, lanes 18–20), demonstrating the specific-

As the domain defined in HHV-6 gH for interaction with gL contains sequences conserved only in the betaherpesviruses, the possibility of interactions between glycoproteins from different betaherpesviruses was examined. To this end, epitope-tagged HHV-6 gL was coexpressed with HCMV gH and complex formation assayed by immunoprecipitation with the MAb specific for the epitope tag, 12CA5. This MAb coimmunoprecipitated both gL and HCMV gH (Fig. 4, lane 7), whereas HCMV gH was not immunoprecipitated when expressed by itself (Fig. 4, lane 5). This shows that HCMV gH forms a heterologous complex with HHV-6 gL. To further examine interactions between betaherpesviruses, HHV-6-tagged gH (wt and deletions) and HCMV gL were expressed individually or were coexpressed followed by immunoprecipitation with the gH tagging MAb, LP14. Figure 5 shows that HHV-6 gH and HCMV gL form a heterologous complex (lane 13)

AID

vira

Epitope-tagged gL interacts with gH

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

AP: Virology

522

ANDERSON, LIU, AND GOMPELS

FIG. 3. Total [35S]methionine-labeled protein extracts from transfected Vero cells (lanes 2 to 9) were immunoprecipitated with LP14 (lanes 10 to 17) or 12CA5 (lanes 18 to 25). Lanes 2, 10, and 18, pH6H2 (wt gH); lanes 3, 11, and 19, pH6H4 (LP14 epitope-tagged gH); lanes 4, 12, and 20, pH6CH2 (wt gL); lanes 5, 13, and 21, pH6CH4 (12CA5 epitope-tagged gL); lanes 6, 14, and 22, pH6H2 and pH6CH2; lanes 7, 15, and 23, pH6H2 and pH6CH4; lanes 8, 16, and 24, pH6H4 and pH6CH2; lanes 9, 17, and 25, pH6H4 and pH6CH4. 14C-labeled molecular weight markers are indicated in kilodaltons (lane 1).

in transfected cells. Limited evidence derived from in vitro transcription/translation and transient expression studies suggest that the association between HHV-6 gH and HCMV gL is more stable than that between HHV-6 gH and gL (Liu et al., 1993a and unpublished results). HCMV gL was coexpressed with the panel of HHV-6 gH mutants and immunoprecipitated with LP14. HCMV gL was able to interact with all the gH recombinants including pH6H9 which was unable to complex with HHV-6 gL (Fig. 5, lanes 11–19). Therefore, the domain involved in heterologous interaction overlaps the N-ter-

minal one-third of HHV-6 gH described for the homologous system. Since HCMV gL interacts with mutant pH6H9 (Fig. 5, lane 18), this implies that the incapacity of this HHV-6 gH deletion to bind HHV-6 gL is not a consequence of localisation to an inappropriate cellular compartment. Also, this indicates that whilst there are common interaction properties, there are differences between homologous and heterologous complex formation. These features may reflect differences in complex stability and/or a dependence on homodimer formation for quarternary complex assembly by HHV-6 glycoproteins. Disulphide bond formation contributes to gH/gL complex formation

FIG. 4. Total [35S]methionine-labeled protein extracts from transfected Vero cells (lanes 2 to 4) were immunoprecipitated with 12CA5 (lanes 5 to 7). Lanes 2 and 5, pCMVgH (HCMV gH); lanes 3 and 6, pH6CH4 (12CA5 epitope-tagged HHV-6 gL); lanes 4 and 7, pCMVgH and pH6CH4. 14C-labeled molecular weight markers are indicated in kilodaltons (lane 1).

To investigate the role of disulphide bond formation and complex stability the HHV-6 gH and gL were analysed under nonreducing conditions. HHV-6 gH and gL share amino acid homology with their HCMV counterparts. Under nonreducing conditions HCMV gH expressed together with gL forms a multimeric complex that involves disulphide linkages, while HCMV gH expressed individually remains monomeric (Kaye et al., 1992). Given the similarities to HCMV, our current working model is that conserved cysteine residues may play a role in HHV-6 gH/gL complex formation. Our observation above, that the HHV-6 gH domain required for gL interaction includes betaherpesvirus-specific cysteines residues is in agreement with this hypothesis. Therefore, epitope-tagged gH or gL were expressed individually or in combination, immunoprecipitated with the relevant tagging MAbs, LP14 or 12CA5, as indicated and electrophoresed on a denaturing nonreducing 4–12% gradient SDS polyacrylamide gel. Under these conditions a minor species of gH that corresponded to the gH monomer routinely detected in the presence of reducing agents

AID

vira

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

AP: Virology

HUMAN HERPESVIRUS-6 BETAHERPESVIRUS-SPECIFIC DOMAIN

523

FIG. 5. Total [35S]methionine-labeled protein extracts from transfected Vero cells (lanes 2 to 10) were immunoprecipitated with LP14 (lanes 11 to 19). Lanes 2 and 11, pH6H4 (LP14 epitope-tagged HHV-6 gH); lanes 3 and 12, pUL115 (HCMV gL); lanes 4 and 13, pH6H4 and pUL115; lanes 5 and 14, pH6H5 and pUL115; lanes 6 and 15; pH6H6 and pUL115; lanes 7 and 16, pH6H7 and pUL115; lanes 8 and 17, pH6H8 and pUL115; lanes 9 and 18, pH6H9 and pUL115; lanes 10 and 19, pH6H10 and pUL115. 14C-labeled molecular-weight markers are indicated in kilodaltons (lane 1).

was observed (Fig. 6, lane 2). However, most gH was resolved as a high-molecular-weight complex (gH*) that may correspond to a homodimer (Fig. 6, lane 2). Likewise, most gL was discernable as a homodimer (gL*) and there was a small proportion of monomeric gL detected (Fig. 6, lane 5). MAbs directed against gH or gL did not nonspecifically immunoprecipitate other proteins (Fig. 6,

lanes 3 and 4). Thus, homodimer formation mediated by disulphide bonds is observed for either HHV-6 gH or gL. Coexpression of epitope-tagged gH and gL and immunoprecipitation with LP14 resulted in the absence of gL and gL* products and the appearance of a novel species that migrated slower than gH* alone (Fig. 6, lane 6), implying that the gH/gL multimer is complexed via disulphide linkages and is probably a tetramer composed of gH and gL homodimers. In combination with the data from Kaye et al. (1992) these results show that both HHV-6 and HCMV can form multimeric gH/gL complexes mediated by disulphide bonds, but their assembly may differ starting with homodimers for HHV-6 and monomers for HCMV. Therefore, assembly of gH/gL quarternary structures has distinct prerequisites in different betaherpesviruses. DISCUSSION

FIG. 6. Immunoprecipitations with LP14 (lanes 2, 4, and 6) and 12CA5 (lanes 3 and 5) of protein extracts from Vero cells transfected with pH6H4 (LP14 epitope-tagged gH) in lanes 2 and 3, pH6CH4 (12CA5 epitope-tagged gL) in lanes 4 and 5, or pH6H4 and pH6CH4 (lane 6). Products were run on a 4–12% gradient SDS polyacrylamide gel under denaturing, nonreducing conditions. Arrows indicate the positions of gH and gL monomers, gH and gL homodimers (denoted by *), and the position of the high-molecular-weight gH/gL heterooligomer. 14Clabeled molecular-weight markers are indicated in kilodaltons (lane 1).

In this paper we describe a HHV-6 gH region required for interaction with another glycoprotein, gL. The domain consists of 230 amino acids from the N-terminal extracellular portion of gH and contains several amino acids that are conserved in betaherpesviruses including paired cysteines (C1; Figs. 1 and 2). The nature of the betaherpesvirus specificity to this conserved domain suggests that our observations are likely to apply to other betaherpesviruses such as murine cytomegalovirus and human herpesvirus-7 (HHV-7). Also, we present the first demonstration of functional substitution of proteins between different herpesviruses of the betaherpesvirus subgroup (Figs. 4 and 5). The gL interaction domain is formed by the N-terminal

AID

vira

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

AP: Virology

524

ANDERSON, LIU, AND GOMPELS

third of gH which is variant between subgroups and conserved within subgroups. Similarly, gL itself is only conserved within subgroups. Within the alphaherpesviruses there is also evidence for interaction of HSV-1 gL with the N-terminal third of HSV-1 gH. Here a conformational epitope dependent on interaction with gL has been defined by sequence analyses of antibody-resistant mutants (Gompels et al., 1991). Thus, the N-terminal third of gH homologues may be a general rule for interaction with gL and may explain the diversity of sequence observed here and in gL. Although there may be similarity in domains of gH homologues interacting with gL, there also appear to be differences in strategies for complex assembly. The mechanism for HSV-1 gH and gL complex formation is not known but does not appear to involve disulphide bonds (Hutchinson et al., 1992). In contrast, disulphide linkages are an important feature of gH/gL complex formation in betaherpesviruses examined to date. Under nonreducing conditions HCMV gH and gL form a high-molecular-weight complex that is composed of gH and gL monomers (Kaye et al., 1992). Similarly, interaction between HHV-6 gH and gL results in the formation of a large complex with intermolecular disulphide bridges (Fig. 6). However, there also appear to be distinctions in HHV-6 assembly although they can functionally substitute in complex formation. The HHV-6 complex appears to be formed from gH and gL disulphide-linked homodimers (Fig. 6) whereas this is clearly not observed in studies of HCMV gH (Kaye et al., 1992). This extra requirement in HHV-6 for homodimer formation before gH/gL complex formation may explain the failure of gH mutant pH6H9, 145 residues, to interact with HHV-6 gL whilst retaining the capacity for stable complex formation with HCMV gL. This is also consistent with our results from in vitro transcription–translation reactions where a stable complex is not formed between the homologous HHV-6 gH and gL but is formed between the heterologous HHV-6 gH with HCMV gL. In this regard, it is not clear if mutant pH6H9 lacks homologous gL binding because it does not form conformationally correct dimers or because it fails to form heteroligomers. Recent results show that all the gH mutants can form homodimers, although the ratio between amounts of monomers and dimers is now being assessed for contributions towards conformational competence or stability. The ability of the four conserved cysteine residues to potentiate homodimer formation and oligomerization will be examined by site-directed mutagenesis. We have shown functional substitution of gH and gL by related glycoproteins from the same herpesvirus family. Both HCMV gH and gL were able to form a stable association with their HHV-6 counterparts. Several lines of evidence suggest that these results are not due to nonspecific interactions. First, our expression studies were designed to overproduce glycoproteins in order to facilitate detection but not to the level where it would

AID

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

disturb cellular functions. Thus, a T7-based plasmid vector was used without an internal ribosome binding site (which allows for efficient translation of the abundant uncapped T7 produced transcripts, resulting in abnormally high levels of overexpressed protein) combined with reduced multiplicities of infection with the superinfecting T7 recombinant vaccinia virus. Subsequently, gH or gL expression was not detected above the background total protein profile. Second, heterologous complex formation was not detected between unrelated viral glycoproteins and HHV-6 gH or gL (data not shown). Finally, the interaction of HHV-6 gL with HHV-6 gH could be eliminated by deletion of gH from amino acid 145. We mapped the domain for heterologous interaction to a region similar to the homologous system. It will be important to assess if heterologous binding equates to correct processing, cellular localisation, and cell fusion. We show evidence that the N-terminal domain is involved in interaction with gL and our former studies indicate that the C-terminal domain appears to function in cell fusion (Liu et al., 1993b). This contrasts with the observation in the alphaherpesviruses that amino acids in the HSV-1 gH cytoplasmic tail are important for syncytia formation in a trans complementation assay (Wilson et al., 1994). The reason for this discrepancy is not clear but may be a consequence of mutant forms of HSV-1 gH having altered flexiblity in the plasma membrane, or requirements for rescue of a particular HSV-1 syn mutant strain. HSV-1 is only fusogenic in these in vitro mutants which are derived from mutations in a number of other glycoproteins, whereas both HCMV and HHV-6 wild-type viruses are fusogenic. In this respect, it is notable that HHV-6 and HCMV both have small cytoplasmic domains of 6 amino acids compared to 14 found in HSV-1 gH. Thus, there may be differences in anchorage or membrane fluidity of the glycoproteins from the different subgroups. However, this is clearly not important for gH and gL complex formation as truncated HHV-6 gH molecules containing no transmembrane or cytoplasmic domains interact with either HHV-6 gL or HCMV gL (Figs. 2 and 5). There are precedents in other virus systems for glycoprotein substitution. Herpesvirus glycoprotein gB homologues share a relatively high level of amino acid sequence homology. Trans-complementation experiments using alphaherpesvirus mutants lacking gB have shown that bovine herpesvirus-1 gB can substitute for pseudorabiesvirus (PrV) gB (Kopp and Mettenleiter, 1992) and that PrV gB can compensate for an HSV-1 gB defect (Mettenleiter and Spear, 1994). Interestingly, HSV-1 gB was unable to complement a PrV gB deletion mutant (Mettenleiter and Spear, 1994). Although the sequence similarity between the betaherpesvirus HCMV and HHV-6 gH and gL is less than that for gB homologues, it is sufficient for functional homology in the context of complex formation and probably represents conservation of tertiary structure.

vira

AP: Virology

HUMAN HERPESVIRUS-6 BETAHERPESVIRUS-SPECIFIC DOMAIN

525

Briggs, M., Fox, J., and Tedder, R. S. (1988). Age prevalence of antibody to human herpesvirus 6. Lancet 1, 1058–1059. Carrigan, D. R., Drobyski, W. R., Russler, S. K., Tapper, M. A., Knox, K. K., and Ash, R. C. (1991). Interstitial pneumonitis associated with human herpesvirus-6 infection after marrow transplantation. Lancet 338, 147–149. Challoner, P. B., Smith, K. T., Parker, J. D., MacLeod, D. L., Coulter, S. N., Rose, T. M., Schultz, E. R., Bennett, J. L., Garber, R. L., Chang, M., Schad, P. A., Stewart, P. M., Nowinski, R. C., Brown, J. P., and Burmer, G. C. (1995). Plaque-associated expression of human herpesvirus 6 in multiple sclerosis. Proc. Natl. Acad. Sci. USA 92, 7440– 7444. Cone, R. W., Hackman, R. C., Huang, Mlw, Bowden, R. A., Meyers, J. D., Metcalf, M., Zeh, J., and Ashley, R. (1993). Human herpesvirus 6 in lung tissue from patients with pneumonitis after bone marrow transplantation. New Engl. J. Med. 329, 156–161. Corbellino, M., Lusso, P., Gallo, R. C., Parravicini, C., Galli, M., and Moroni, M. (1993). Disseminated human herpesvirus 6 infection in AIDS. Lancet 342, 1242 Cranage, M. P., Smith, G. L., Bell, S. E., Hart, H., Brown, C., Bankier, A. T., Tomlinson, P., Barrell, B. G., and Minson, A. C. (1988). Identification and expression of a human cytomegalovirus glycoprotein with homology to the Epstein–Barr virus BXLF2 product, Varicella–Zoster virus gpIII and Herpes Simplex virus type 1 glycoprotein H. J. Virol. 62, 1416–1422. Dewhurst, S., Chandran, B., McIntyre, K., Schnabel, K., and Hall, C. B. (1992). Phenotypic and genetic polymorphisms among human herpesvirus-6 isolates from North American infants. Virology 190, 490– 493. Drobyski, W. R., Dunne, W. M., Burd, E. M., Knox, K. K., Ash, R. C., Horowitz, M. M., Flemenbery, N., and Carrigan, D. R. (1993). Human herpesvirus-6 (HHV-6) infection in allogeneic bone marrow transplant recipients. I. Evidence of a marrow suppressive role for HHV-6 in vivo. J. Infect. Dis. 167, 735–739. Drobyski, W. R., Knox, K. K., Majewski, D., and Carrigan, D. R. (1994). Brief report: Fatal encephalitis due to variant B human herpesvirus6 infection in a bone marrow-transplant recipient. New Engl. J. Med. 330, 1356–1360. Fairfax, M. R., Schacker, T., Cone, R. W., Collier, A. C., and Corey, L. (1994). Human herpesvirus 6 DNA in blood cells of human immunodeficiency virus-infected men: Correlation of high levels with high CD4 cell counts. J. Infect. Dis. 169, 1342–1345.

Foa-Tomasi, L., Boscaro, A., Di Gaeta, S., and Campadelli-Fiume, G. (1991a). Monoclonal antibodies to gp100 inhibit penetration of human herpesvirus 6 and polykaryocyte formation in susceptible cells. J. Virol. 65, 4124–4129. Foa-Tomasi, L., Avitabile, E., Boscaro, A., Brandimarti, R., Gualandri, R., Manservigi, R., Dall ‘Olio, F., Serafini-Cessi, F., and CampadelliFiume, G. (1991b). Herpes simplex virus (HSV) glycoprotein H is partially processed in a cell line that expresses the glycoprotein and fully processed in cells infected with deletion or ts mutants in the known HSV glycoproteins. Virology 180, 474–482. Fox, J. D., Briggs, M., Ward, P. A., and Tedder, R. S. (1990). Human herpesvirus 6 in salivary glands. Lancet 336, 590–593. Fuerst, T. R., Niles, E. G., Studier, F. W., and Moss, B. (1986). Eukaryotic transient-expression system based on recombinant vaccinia virus that synthesizes bacteriophage T7 RNA polymerase. Proc. Natl. Acad. Sci. USA 83, 8122–8126. Gavel, Y., and von Heijne, G. (1990). Sequence differences between glycosylated and nonglycosylated Asn-X-Thr/Ser acceptor sites: Implications for protein engineering. Protein Eng. 3, 433–442. Gompels, U., and Minson, A. (1986). The properties and sequence of glycoprotein H of herpes simplex virus type 1. Virology 153, 230– 247. Gompels, U. A., Craxton, M. A., and Honess, R. W. (1988a). Conservation of gene organisation in the lymphotropic herpesviruses herpesvirus saimiri and Epstein–Barr virus. J. Virol. 62, 757–767. Gompels, U. A., Craxton, M. A., and Honess, R. W. (1988b). Conservation of glycoprotein H (gH) in herpesviruses: Nucleotide sequence of the gH gene from herpesvirus saimiri. J. Gen. Virol. 69, 2819–2829. Gompels, U. A., and Minson, A. C. (1989). Antigenic properties and cellular localization of herpes simplex virus glycoprotein H synthesized in a mammalian cell expression system. J. Virol. 63, 4744– 4755. Gompels, U. A., Carss, A. L., Saxby, C., Hancock, D. C., Forrester, A., and Minson, A. C. (1991). Characterization and sequence analyses of antibody-selected antigenic variants of herpes simplex virus show a conformationally complex epitope on glycoprotein H. J. Virol. 65, 2393–2401. Gompels, U. A., Carss, A. L., Sun, N., and Arrand, J. R. (1992). Infectivity determinants encoded in a conserved gene block of human herpesvirus-6. J. DNA Sequencing Mapp. 3, 25–39. Gompels, U. A., Carrigan, D. R., Carss, A. L., and Arno, J. (1993). Two groups of human herpesvirus 6 identified by sequence analysis of laboratory strains and variants from Hodgkin’s lymphoma and bone marrow transplant patients. J. Gen. Virol. 74, 613–622. Gompels, U. A., Luxton, J., Knox, K. K., and Carrigan, D. R. (1994). Chronic bone marrow suppression in immunocompetent adult by human herpesvirus 6. Lancet 343, 735–736. Heineman, T., Gong, M., Sample, J., and Kieff, E. (1988). Identification of the Epstein–Barr virus gp85 gene. J. Virol. 62, 1101–1107. Hutchinson, L., Browne, H., Wargent, V., Davis-Poynter, N., Primorac, S., Goldsmith, K., Minson, A. C., and Johnson, D. C. (1992). A novel herpes simplex glycoprotein, gL, forms a complex with with glycoprotein H (gH) and affects normal folding and surface expression of gH. J. Virol. 66, 2240–2250. Kaye, J. F., Gompels, U. A., and Minson, A. C. (1992). Glycoprotein H of human cytomegalovirus (HCMV) forms a stable complex with the HCMV UL115 gene product. J. Gen. Virol. 73, 2693–2698. Keller, P. M., Davison, A. J., Lowe, R. S, Rieman, M. W., and Ellis, R. W. (1987). Identification and sequence of the gene encoding gpIII, a major glycoprotein of varicella-zoster virus. Virology 157, 526–533. Knox, K. K., and Carrigan, D. R. (1994). Disseminated active HHV-6 infections in patients with AIDS. Lancet 343, 577–578. Knox, K. K., and Carrigan, D. R. (1995). Active human herpesvirus (HHV6) infection of the central nervous system in patients with AIDS. J. Acq. Immune Def. Synd. Hum. Retrovirol. 9, 69–73. Knox, K. K., Pietryga, D., Harrington, D. J., Franciosi, R., and Carrigan, D. R. (1995a). Progressive immunodeficiency and fatal pneumonitis

AID

vira

Heterologous complex formation between HHV-6 and HCMV may have biological significance since there is evidence that both viruses share a similar compartment of latent or persistent virus within monocytic cells (Kondo et al., 1991; Taylor-Weideman et al., 1991). Moreover, concurrent reactivation of these viruses has been demonstrated in vivo (Knox and Carrigan, 1994). Glycoproteins gH and gL are produced late in infection; complementation at early stages of the virus life cycle in coinfected cells may overcome limited availablity of these glycoproteins, leading to enhanced recrudescent disease, and possibly facilitating access by cell fusion to new sites of tissue infection affecting the overall pathogenesis. ACKNOWLEDGMENTS We thank the Medical Research Council, U.K., and the Wellcome Trust for support.

REFERENCES

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

AP: Virology

526

ANDERSON, LIU, AND GOMPELS

associated with human herpesvirus 6 infection in an infant. Clin. Infect. Dis. 20, 406–413. Knox, K. K., Harrington, D. P., and Carrigan, D. R. (1995b). Fulminant human herpesvirus six encephalitis in a human immunodeficiency virus-infected infant. J. Med. Virol. 45, 288–292. Kondo, K., Kondo, T., Okuno, T., Takahashi, M., and Yamanishi, K. (1991). Latent human herpesvirus 6 infection of human monocytes/macrophages. J. Gen. Virol. 72, 1401–1408. Kopp, A., and Mettenleiter, T. C. (1992). Stable rescue of a glycoprotein gII deletion mutant of pseudorabies virus by glycoprotein gI of bovine herpesvirus 1. J. Virol. 66, 2754–2762. Kunkel, T. A. (1985). Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA 82, 488–492. Liu, D. X., Cavanagh, D., Green, P., and Inglis, S. C. (1991). A polycistronic mRNA specified by the coronavirus infectious bronchitis virus. Virology 184, 531–544. Liu, D. X., Gompels, U. A., Nicholas, J., and Lelliott, C. (1993a). Identification and expression of the human herpesvirus 6 glycoprotein H and interaction with an accessory 40K glycoprotein. J. Gen. Virol. 74, 1847–1857. Liu, D. X., Gompels, U. A., Foa-Tomasi, L., and Campadelli-Fiume, G. (1993b). Human herpesvirus-6 glycoprotein H and L homologues are components of the gp100 complex and the gH external domain is the target for neutralizing monoclonal antibodies. Virology 197, 12–22. Lusso, P., Ensoli, B., Markham, P. D., Ablashi, D. V., Salahuddin, S. Z., Tschachler, E., Wong-Staal, F., and Gallo, R. C. (1989). Productive dual infection of human CD4/ T lymphocytes by HIV-1 and HHV-6. Nature (London) 337, 370–373. McGeoch, D. J., and Davison, A. J. (1986). DNA sequence of the herpes simplex virus type 1 glycoprotein gH, and identification of homologues in the genomes of varicella–zoster virus and Epstein–Barr virus. Nucleic Acids Res. 14, 4281–4292. Mettenleiter, T. C., and Spear, P. G. (1994). Glycoprotein gB (gII) of pseudorabies virus can functionally substitute for glycoprotein gB in herpes simplex virus type 1. J. Virol. 68, 500–504. Minson, A. C., Hodgman, T. C., Digard, P., Hancock, D. C., Bell, S. E., and Buckmaster, E. A. (1986). An analysis of the biological properties of monoclonal antibodies against glycoprotein D of herpes simplex virus and identification of amino acid substitutions that confer resistance to neutralization. J. Gen. Virol. 67, 1001–1013. Morris, D. J., Littler, E., Arrand, J. R., Jordan, D., Mallick, N. P., and Johnson, R. W. G. (1989). Human herpesvirus 6 infection in renal transplant recipients. N. Engl. J. Med. 320, 1560–1561. Oba, D. E., and Hutt-Fletcher, L. M. (1988). Induction of antibodies to the Epstein–Barr virus glycoprotein gp85 with a synthetic peptide

corresponding to a sequence in the BXLF2 open reading frame. J. Virol. 62, 1108–1114. Okuno, T., Takahashi, K., Balachandra, K., Shiraki, K., Yamanishi, K., Takahashi, M., and Baba, K. (1989). Seroepidemiology of human herpesvirus 6 infection in normal children and adults. J. Clin. Microbiol. 27, 651–653. Pachl, C., Probert, W. S., Hermsen, K. M., Masiarz, F. R., Rasmussen, L., Merigan, T. C., and Spaete, R. R. (1989). The human cytomegalovirus strain towne glycoprotein H gene encodes glycoprotein p86. Virology 169, 418–426. Roberts, S. R., Ponce de Leon, M., Cohen, G. H., and Eisenberg, R. J. (1991). Analysis of the intracellular maturation of the herpes simplex type 1 glycoprotein H in infected and transfected cells. Virology 184, 609–624. Sanger, F., Nicklen, S., and Coulson, A. R. (1977). DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463– 5467. Spaete, R. R., Perot, K., Scott, P. I., Nelson, J. A., Stinski, M. F., and Pachl, C. (1993). Coexpression of truncated human cytomegalovirus gH with the UL115 gene product or the truncated human fibroblast growth factor receptor results in transport of gH to the cell surface. Virology 193, 853–861. Takahashi, K., Sonoda, S., Higashi, K., Kondo, T., Takahashi, H., Takahashi, M., and Yamanashi, K. (1989). Predominant CD4 T-lymphocyte tropism of human herpesvirus 6-related virus. J. Virol. 63, 3161–3163. Taylor-Wiedeman, J., Sissons, J. G. P., Borysiewicz, L. K., and Sinclair, J. H. (1991). Monocytes are a major site of persistence of human cytomegalovirus in peripheral blood mononuclear cells. J. Gen. Virol. 72, 2059–2064 Tedder, R. S., Briggs, M., Cameron, C. H., Honess, R., Robertson, D., and Whittle, H. (1987). A novel lymphotropic herpesvirus. Lancet 2, 390–392 Wilson, D. W., Davis-Poynter, N., and Minson, A. C. (1994). Mutations in the cytoplasmic tail of herpes simplex virus glycoprotein H suppress cell fusion by a syncytial strain. J. Virol. 68, 6985–6993. Wilson, I. A., Niman, H. L., Houghten, R. A., Cherenson, A. R., Connolly, M. L., and Lerner, R. A. (1984). The structure of an antigenic determinant in a protein. Cell 37, 767–778. Wyatt, L. S., Balachandran, N., and Frenkel, N. (1990). Variations in the replication and antigenic properties of human herpesvirus 6 strains. J. Infect. Dis. 162, 852–857. Yamanishi, K., Okuno, T., Shiraki, K., Takahashi, M., Kondo, T., Asano, Y., and Kurata, T. (1988). Identification of human herpesvirus-6 as a causal agent for exanthem subitum. Lancet 1, 1065–1067. Yaswen, L. R., Stephens, E. B., Davenport, L. C., and Hutt-Fletcher, L. M. (1993). Epstein–Barr virus glycoprotein gp85 associates with the BKRF2 gene product and is incompletely processed as a recombinant protein. Virology 195, 387–396.

AID

vira

VY 7775

/

6a12$$$$81

02-13-96 15:15:56

AP: Virology