- Volume 76, Issue 1, 1995

We have determined the nucleotide sequences of the capsid, E3, E2 and 6K genes of the avirulent Semliki Forest virus variant A774 (SFV A7). The sequence analysis revealed a nucleotide identity of 98% for capsid, 98% for E3, 97% for E2 and 98% for 6K genes, as compared with the prototype SFV strain L10. At the protein level, the capsid and E3 polypeptides of SFV A7 both exhibited two amino acid substitutions, whereas point mutations in the 6K gene did not alter the amino acid sequence. In the E2 gene of SFV A7, seven of the 34 point mutations led to an amino acid difference as compared with the L10 strain. Replacement of the E2 glycoprotein gene of the virulent SFV4 clone with the corresponding region of SFV A7 resulted in a new plasmid construct, pME2, that gave rise to infectious virus CME2. CME2 and SFV4 replicated similarly in an immortalized mouse brain cell line (MBA 13). Intraperitoneal injection of 106 p.f.u. of CME2 into 4- to 6- week-old BALB/c mice caused mild clinical signs in some mice, whereas the majority of the infected animals remained asymptomatic, similar to infection with the avirulent SFV A7. In contrast, infection with the parental SFV4, a derivative of the virulent L10 strain, was lethal in 80% of mice. Virus titres in blood and brain tissue specimens of BALB/c mice were similar after infection with CME2 or A7 viruses. The results suggest that amino acid differences in the E2 glycoprotein individually or in concert cause the attenuation of CME2.

The core protein of hepatitis C virus (HCV) is considered to be cleaved from the N terminus of the large precursor polyprotein by cellular signalase. The HCV cDNA encoding the core protein was expressed (i) in monkey COS cells by a plasmid expression vector driven by the SRα promoter, and (ii) in insect cells by a recombinant baculovirus. The expressed product had an M r of 22000 and was located in the cytoplasm. When the C-terminal hydrophobic domains were deleted, however, the truncated core proteins were translocated into the nucleus. The truncated core proteins were located in the nucleus even when they were expressed as a fusion protein with E. coli β-galactosidase, which is essentially localized in the cytoplasm. Plasmids containing HCV cDNAs with a deletion in one of the regions encoding clusters of basic amino acids were expressed in COS cells and the localization of the core protein was examined. The residues PRRGPR were suggested to play an important role in nuclear localization. HCV is an RNA virus and its life cycle was originally considered to be confined to the cytoplasm; the present study, however, suggests that the HCV core protein can translocate into the nucleus under certain circumstances.

The poliovirus replication complex (RC), the site of genomic 36S RNA synthesis, was previously shown to contain subviral particles of 5S protomer and 14S pentamer antigenicity. The present investigation demonstrates that 5S/14S antigenic subviral particles can be cross-linked to viral RNA by UV irradiation of a subcellular fraction containing the poliovirus RC. Each capsid protein of the subviral particles, i.e. VP0, VP1 and VP3, was cross-linked to viral RNA. SDS–PAGE analysis of the cross-linked capsid proteins revealed a bandshift for VP1, whereas VP0 migrated in several bands, which were interpreted to be multimers of VP0 linked by short stretches of RNA. It was found that 36S RNA rather than replicative intermediate RNA was cross-linked to capsid proteins. Our results indicate that encapsidation of poliovirus RNA starts in the RC and is initiated by 14S pentamers.

In order to derive phylogenetic relationships between rabies virus isolates from different geographical locations and host species in South Africa, two genome regions of the virus, viz. the cytoplasmic domain of the glycoprotein and the G-L intergenic region (pseudogene), were sequenced. A high level of nucleic acid sequence conservation indicated a close phylogenetic relationship between virus isolates from domestic dogs, jackals and bat-eared foxes, i.e. Canidae. These isolates appeared to be distinct from but closely related to European strains of rabies virus. However, a phylogenetically distinguishable and distant group, which contained isolates from mongooses (i.e. Viverridae) was identifiable. The latter group appears to be distantly related to European and vaccine strains of rabies virus and may have evolved uniquely on the central plateau of South Africa. Our data also indicate that spillover from mongooses (or other viverrids) to canid hosts occurs occasionally.

In the search for factors that influence the outcome of human polyomavirus JC (JCV) infection, the roles not only of host-related immunological control but also of virus-dependent regulatory steps have to be taken into account. Besides cell-specific control of early expression of the multifunctional virus protein large tumour antigen (T Ag), control mechanisms involve individual steps of the DNA replication process. For the analysis of T Ag DNA binding, the protein was expressed by an adenovirus hybrid vector in the 293 cell line to provide saturating amounts of JCV T Ag. After determination of the size and immunoreactivity, functional activity was analysed by specific DNA binding. To avoid the interference of cellular proteins, T Ag was immunoprecipitated prior to the reaction. Binding to T Ag-binding sites I and II within a 141 bp DNA segment in the control region was analysed using deletion mutants of a JCV subtype from brain tissue of a patient with fatal central nervous system disease. The specificity of the binding was confirmed by recombinant T Ag binding to origin of DNA replication (ori) sequences of wild-type JCV genomes. These data document that recombinant T Ag overexpressed by the adenovirus vector in eukaryotic cells was JCV-specific, had the expected length and exhibited specific ori-binding activity, thus providing the essential tool for future analysis of virus-host interactions at the level of viral DNA replication.

We constructed a non-defective bovine adenovirus type 3 recombinant (BAd3-Luc) containing the firefly luciferase gene inserted in the early region 3 (E3) of the BAd3 genome. Deletion of a 696 bp XhoI–NcoI E3 segment and insertion of the luciferase gene in E3 was confirmed by Southern blot analyses. Luciferase was expressed in Madin-Darby bovine kidney cells infected with BAd3-Luc as measured by enzymic assays and Western blotting. Analyses of luciferase expression in the presence or absence of 1-β-d-arabinofuranosyl-cytosine indicated that approximately 70–75% of luciferase expression was dependent on viral DNA replication, suggesting that transcription of the gene was at least partially under the control of a late promoter. Although yields of infectious virus for BAd3-Luc were approximately 10-fold lower than for wild-type virus, replication of the vector was still relatively efficient. In a Western blot experiment, anti-luciferase antibody reacted with a 62 kDa protein which is of the same molecular mass as the purified firefly luciferase polypeptide. Luciferase was also expressed in the 293 cell line infected with BAd3-Luc for at least 6 days postinfection as monitored by luciferase assays. Based on these observations we suggest that BAd-based expression vectors should have excellent potential for the development of recombinant vaccines for cattle and may also be suitable as vectors for gene transfer into human cells.

The fungal metabolite Brefeldin A was found to inhibit the production of the infectious enveloped form of vaccinia virus, although production of the infectious intracellular form was not affected. Electron microscopic analysis and caesium chloride density centrifugation of progeny virions indicates that the drug block is not due to retention and accumulation of enveloped virions within the cell. Biochemical analysis of the candidate envelopment receptor for vaccinia virus, viral protein P37, shows that the drug has no discernible effect on palmitylation of this protein and does not prevent or alter its association with intracellular membranes. This suggests that P37 may not in fact be the receptor on intracellular membranes for vaccinia virus envelopment, and leaves open the question of what function this molecule performs in the envelopment process.

Stable transfectants were selected from human astrocytoma cells (U373) after transfection with recombinant expression vectors carrying the human cytomegalovirus (HCMV) glycoprotein B (gB; gpUL55) gene with alternative deletions of hydrophobic domain segment 1 (hd1) or segment 2 (hd2) of the carboxy-terminal potential bipartite membrane anchor domain. Comparative analysis of HCMV gB forms from cell lines gB(Mhd1) and gB(Mhd2), expressing mutagenized gB, and those from cells expressing authentic gB showed that deletion of hd1, but not that of hd2, interfered with efficient proteolytic cleavage of the gB precursor. Both mutagenized gB forms exhibited correct transport to the cell surface. Deletion of hd2, but not that of hd1, caused loss of membrane anchoring of the gB molecule, resulting in secretion of the respective gB form into the culture medium. The carboxy-terminal cleavage product of the soluble gB molecule, which migrated more slowly than its authentic counterpart, was modified by complex carbohydrate side chains and formed disulphide-linked complexes. Our observations indicate that hd2 is essential as well as sufficient for membrane anchoring of the HCMV gB molecule. For hd1, a potential fusogenic role is suggested by the conserved positional pattern of glycine residues, which is comparable to that of known fusion peptides of other viruses.

Malignant catarrhal fever virus (MCFV), a gamma-herpesvirus, causes a severe inflammatory and lymphoproliferative disease of cattle and other susceptible ruminants. Polyclonal antisera and monoclonal antibodies (MAbs) to the Minnesota isolate of MCFV were produced and used to examine the characteristics of the viral proteins. Immunoprecipitation of antigens of the Minnesota isolate of MCFV with polyclonal antisera revealed at least 11 proteins with molecular masses ranging from 17 kDa to 145 kDa. Among 279 candidate anti-MCFV hybridomas, 14 were selected and clustered into six groups on the basis of the patterns of reactivity to viral proteins in immunoprecipitation and immunoblot. The group I MAbs exhibited strong neutralizing activity and recognized a glycosylation-dependent conformational epitope on a 110 kDa protein. The MAbs in group II bound a non-neutralizing conformational epitope on a 130 kDa non-glycosylated protein. A glycosylated protein complex of 115/110/105/78/45 kDa moieties was identified by the MAbs in group III. The MAbs in groups IV, V and VI reacted with nonglycosylated proteins of 36/34 kDa, 24 kDa and 17 kDa, respectively. Comparison of three MCFV isolates [the Minnesota isolate, the Austrian isolate (Au-732) and the African prototype isolate (WC-11)] revealed no apparent differences in immunoprecipitation patterns with the single exception that the 110 kDa protein of WC-11 was slightly smaller than its counterpart in the Minnesota isolate.

The presence of transcripts of the Epstein–Barr virus genes for Epstein—Barr nuclear antigen (EBNA)-1 and EBNA-2 and for latent membrane protein (LMP)-1, LMP-2A and LMP-2B was investigated in 24 nasopharyngeal carcinoma (NPC) biopsies of Chinese origin and two NPC-derived solid tumour lines, CAO and C15, of Chinese and north African origin respectively, propagated by serial transplantation in nude mice. Transcripts were detected by PCR amplification of cDNA. EBNA-1 transcripts were present in all biopsies tested and originated exclusively from the FQ promoter while the C and W promoters were inactive. Using nested primers, LMP-1 and LMP-2B RNAs were found to be co-ordinately expressed in 22 of the 24 biopsies, while the two remaining tumours were negative for both. LMP-2A transcription was detected in 17 of the 22 LMP-1-positive biopsies. In summary, the following patterns of viral gene expression were observed in the tumour biopsies: (i) LMP-1-LMP-2B and LMP-2A positive (17 biopsies); (ii) LMP-1-LMP-2B positive but LMP-2A negative (five biopsies); (iii) no viral gene other than EBNA-1 expressed (two biopsies).

Latent membrane protein 2A (LMP-2A) is expressed in Epstein—Barr virus transformed B lymphocytes in vitro and has been detected in various types of EBV-associated malignancies. LMP-2A interferes with membrane signal transduction through phosphorylation of its hydrophilic N-terminal domain and binding of the cellular tyrosine kinases encoded by fyn and lyn. In vitro, the domain can block calcium influx and participate in signal transduction inducing cytokine production. These two activities are differently affected by site-directed mutagenesis of potentially phosphorylated amino acid residues. Several potential tyrosine protein kinase recognition motifs have been identified including an antigen recognition motif (ARAM). ARAMs are activated by tyrosine phosphorylation that enables binding of tyrosine protein kinases such as lyn and fyn. To assess the importance of potential sequence variation in natural EBV infection and in tumourigenesis, the sequence of the LMP-2A N-terminal domain was determined in 28 EBV isolates, including 14 fresh tumour isolates. Comparison of the corresponding sequences with the prototype B95 strain indicates that LMP-2 is generally conserved with a few base pair changes resulting in conservative amino acid changes in an occasional isolate. However, five single-base loci were frequently mutated, resulting in three patterns of sequence polymorphism in exon 1 of LMP-2A. The patterns did not segregate with EBV Type 1 or Type 2 and were detected in both lymphoid and epithelial tissues. Four of the most frequent mutations at loci 166627, 166750, 166796 and 166805 (codons 23, 63, 79 and 82) could potentially affect tyrosine protein kinase binding motifs. The pivotal tyrosines (codons 74 and 85) and leucines (codons 77 and 88) of the LMP-2 ARAM were not affected in any of the isolates, suggesting that ARAM function is important for EBV infection in vivo. However, the interspacing positions 79 and 82 were distinct in more than 50% of the isolates. These prevalent polymorphisms could influence interaction of the LMP-2 cytoplasmic domain with specific cellular ligand proteins.

The nucleotide sequence of 3134 bp from the genome of human herpesvirus 6 (HHV-6) strain U1102 was determined. The sequence overlaps and is contiguous with the 21858 bp nucleotide sequence published by us previously. The sequence reported here encodes two open reading frames, named 18L and 19R. The protein encoded by 18L shares amino acid sequence similarity with the multiply hydrophobic glycoprotein M conserved in the genomes of all herpesviruses sequenced to date. ORF 19R encodes a protein which shares a significant degree of amino acid sequence conservation with the origin-binding protein homologues encoded by members of the alphaherpesvirus subgroup, but does not share detectable amino acid sequence homology with positionally analogous open reading frames present in the genomes of other betaherpesviruses or in the genomes of gammaherpesviruses.

The potential of selected proteins of human cytomegalovirus (HCMV) to induce a helper T (Th) cell immune response was investigated in healthy HCMV-seropositive donors. Recombinant derived glycoproteins B (gpUL55), H (gpUL75), integral membrane protein (pUL100), the US6-US11 glycoprotein family (pUS6-US11), the matrix proteins pp65 (ppUL83), pp28 (ppUL99) and the immediate early proteins IE1 (pUL123), IE2 (pUL122) and UL69 (pUL69) were used as stimulating antigens in a lymphocyte proliferation assay. The antigen-specific proliferative response was measured in HCMV-specific T cell lines (phenotype CD4+ CD8-) generated from five donors by stimulation of peripheral blood mononuclear cells with purified HCMV or HCMV-infected fibroblasts. A proliferative T cell response was induced by pp65, gB, gH, IE1, IE2 and UL69, with a dominant response to pp65 in all donors. Three T cell lines responded to gB and gH, respectively. For IE1, IE2 and UL69 a T cell stimulation could be demonstrated in single cell lines generated with lysate of HCMV-infected fibroblasts.

Budgerigar fledgling disease virus 1 (BFDV-1) is the first avian polyomavirus to be identified, and it possesses uncommon structural and biological properties. Here we present an analysis of the processed viral RNAs in infected chicken embryo fibroblast cells. Two early and 18 late BFDV-1 mRNAs were defined according to their 5′ ends and internal splice patterns. In the early region of the genome an incomplete splice reaction covering 195 nt is responsible for creating two mRNAs that could encode small t and large T antigens, which would be initiated from a hypothetical early promoter, PE. The late mRNA 5′ ends define two putative promoter regions (PL1 and PL2), 111 nt apart in the BFDV-1 genome noncoding region. The overall splicing pattern of the late mRNAs is further complicated by an alternative splice reaction of intron 2 (deletion of either 64 nt in intron 2a or of 256 nt in intron 2b) and a splice removing intron 3 (870 nt), resulting in deletion of most of the VP2-VP3 coding region. The positions of the late mRNA 5′ ends and the splicing pattern indicate the existence of two open reading frames, putatively encoding two pairs of agnoproteins, in the 5′ region of several late mRNAs. These mRNAs appear to be bicistronic and to encode one of the agnoproteins together with one of the viral coat proteins.

We investigated permissiveness of the malignantly transformed trophoblast (choriocarcinoma) cell lines JAR, BeWo and JEG-3 to the human T cell lymphotropic virus type I (HTLV-I). After co-culture with the productively infected cell line MT-2 the choriocarcinoma cell lines were analysed for infection over a period of three months. The presence of HTLV-I viral DNA was examined by PCR using primers targeting the gag, pol, env and pX specific sequences. All amplified segments were found consistently in the cell cultures throughout the period of study. Further analysis that aimed to characterize the size variation of the integrated proviral DNA by Southern blotting revealed the presence of integrated proviral sequences which consisted, for the most part, of incomplete genomes. The presence of the full-length HTLV-I genome, however, was unequivocally confirmed in clonally expanded cell cultures derived from the originally infected parental cells. In order to analyse virus expression at the transcriptional level, we used reverse transcriptase (RT)-mediated PCR that was targeted at intraexon regions (gag, pol, env and pX), and the splicing sites of the env and pX-tax/rex mRNAs. When compared with MT-2 cells, substantially lower levels of all transcripts were found in all the cell lines analysed. We were unsuccessful in attempts to detect viral protein expression using polyvalent or Tax- and Gag-specific monoclonal antibodies by Western blot analysis or immunoprecipitation, and we could not detect any RT activity released into the supernatant of the infected cells either. Collectively, these data suggest that the trophoblastic cells may become persistently but essentially non-productively infected with HTLV-I.

Paired serum and cerebrospinal fluid (CSF) samples from 10 AIDS patients with and 10 without AIDS dementia complex (ADC) were studied, in an attempt to uncover ADC-associated variation in V3 sequences. Sequences were obtained from four of the patients with and eight of those without ADC. Comparison of the sequences using a resampling technique revealed a significant ADC-associated difference occurring at several amino acid positions. Results from serum and CSF sequences were comparable. These differences may indicate that the virus found in ADC and that in non-ADC patients have different biological properties. Comparison of serum versus CSF sequences within samples from both ADC and non-ADC patients, using the same resampling technique, revealed no clear distinctions. In some patients, the sequence populations in serum and CSF were completely distinct, while in others, there was no difference in distribution. These patterns were not associated with ADC.

A genomic DNA library was constructed, in a bacteriophage λ vector, from line 0 chick embryo fibroblasts (CEFs) infected with HPRS-103, an exogenous avian leukosis virus (ALV; envelope subgroup J) recently isolated from meat-type chickens. The library was screened at high stringency using a full length RAV-1 (subgroup A) proviral probe. From 106 plaques, two clones which hybridized strongly to the RAV-1 probe were isolated; one contained a full-length copy of the proviral genome of HPRS-103 and the other contained a copy lacking the 5′-long terminal repeat (LTR) and part of gag. The relative strength of hybridization of RAV-1 and HPRS-103 clones, to RAV-1 probes representing different parts of the proviral genome, indicated that the gag and pol genes of HPRS-103 share a high level of identity with those of RAV-1 but that the env gene and the LTRs are considerably less well conserved. Infectious virus was recovered from CEFs transfected with the full-length clone, as detected by ELISA. The recovered virus appeared to be identical to HPRS-103 by electron microscopy and by Southern blotting of proviral DNA. The recovered virus was shown to be of the same subgroup as HPRS-103 by serum neutralization and receptor interference assays. Sequence analysis of the env gene of HPRS-103 shows that it differs considerably from the env genes of other ALV subgroups, particularly in the host range determinants, consistent with the finding that HPRS-103 represents a new subgroup (designated J).

Jembrana disease virus, the cause of an acute, severe disease in Bali (Bos javanicus) cattle in Indonesia was recently identified as a retrovirus, and possibly a lentivirus. We have produced sequence data representing 598 bp of the pol gene, amplified by PCR from viral cDNA using broadly reactive universal primers for retroviruses and more specific genus-reactive primers for lentiviruses. When the sequence data were compared with that of known lentiviruses and other bovine retroviruses, the closest alignment was with bovine immunodeficiency-like lentivirus (BIV), showing 74% nucleotide sequence identity. This confirmed that JDV is a lentivirus and that it is distinguishable from BIV. The pathogenesis of Jembrana disease is most unusual for a lentivirus infection and differs markedly from that reported for BIV infection.