- Volume 78, Issue 6, 1997

The aphid Myzus persicae (Sulz.) was shown to transmit potato spindle tuber viroid (PSTVd) to potato clone DTO-33 from source plants doubly infected with potato leafroll virus (PLRV) and PSTVd. Transmission was of the persistent type and did not occur when the insects were allowed to feed on singly infected plants. Only low levels of PSTVd were associated with purified PLRV virions, but its resistance to digestion with micrococcal nuclease indicates that the viroid RNA is encapsidated within the PLRV particles. Epidemiological surveys carried out at three locations in China revealed a strong correlation between PSTVd infection and the presence of PLRV, suggesting that PLRV can facilitate PSTVd spread under field conditions.

A 13-amino-acid peptide derived from the V3 loop of human immunodeficiency virus (HIV-1) glycoprotein 120 (gp120) was attached as a C-terminal gene fusion to the coat protein of tomato bushy stunt virus (TBSV). The architecture of this plant virus permitted external display of the foreign sequence 180 times on the surface of the chimaeric virus particle. The chimaera replicated to a level similar to wild-type TBSV and the foreign sequence was retained through six sequential passages in plants. The HIV epitope was detected on the surface of the virus capsid by a V3-specific monoclonal antibody and by human sera from HIV-1-positive patients, demonstrating the potential of using plant-derived chimaeric particles for diagnostic purposes. Chimaeric virus also induced a specific immune response to the foreign HIV epitope when injected into NMRI mice.

Pothos latent virus (PoLV) is an isometric virus with a positive-sense, single-stranded RNA genome of 4415 nt. The genome contains five open reading frames (ORF), coding for five proteins with approximate molecular masses of 25, 84, 40, 27 and 14 kDa, respectively. In vitro synthesized PoLV RNA was infectious to Nicotiana benthamiana plants and protoplasts, but could not support replication of the defective interfering (DI) and satellite RNAs associated with Cymbidium ringspot tombusvirus. No DI RNA related to PoLV was generated after repeated passaging with infected sap. Mutagenesis studiesdefined the role of the ORF 4 product (27 kDa) as a movement protein, and the ORF 5 product (14 kDa) as being responsible for symptom severity. Moreover, it was shown that the coat protein (CP) is important in regulating synthesis of the 14 kDa protein, excess production of which is lethal to infected plants. CP mutants defective in capsid formation infected plants systemically and induced severe necrotic symptoms. Conversely, CP mutants able to form apparently normal virus particles induced symptoms indistinguishable from those elicited by wild-type virus.

Mutational analysis of defective interfering (DI) RNAs of Cymbidium ringspot virus (CymRSV) was used to study the mechanism of DI RNA evolution. It was shown that a highly base-paired structure in the 3′ region of the longer DI RNA directed the formation of smaller DI RNA molecules. Mutations which increased the stability of the computer- predicted, highly structured 3′ region of the longest DI RNA of CymRSV significantly enhanced the generation and accumulation of the smaller derivatives. Sequence analysis of smaller progeny molecules revealed that the highly base-paired region was deleted from the precursor DI RNA. Moreover, sites of recombination were found in other regions of the DI RNA progenies due to transposition of the highly base-paired structure. It is likely that the deletion event was structure- and not sequence- specific, and operated when a foreign sequence containing a 37-nt-long base-paired stem was inserted at the appropriate position of DI RNA.

The complete genome sequence of tobacco necrosis virus strain D (Hungarian isolate, TNV-DH) was determined. The genome (3762 nt) has an organization identical to that reported for TNV-D. Highly infectious synthetic transcripts from a full- length TNV-DH cDNA clone were prepared, the first infectious necrovirus transcript reported. This clone was used for reverse genetic studies to map the viral genes required for replication and movement. Protoplast inoculation with Δ22 and Δ82 mutants revealed that both the 22 kDa and 82 kDa gene products are required for RNA replication. Although the products of three small central genes (p71, p7a and p7b) were not essential for RNA replication in protoplasts, mutations in these ORFs prevented infection of plants. In contrast, viral RNA accumulation and cell-to-cell movement were observed in the inoculated, but not the systemically infected, leaves of Nicotiana benthamiana challenged with RNA lacking the intact coat protein (CP) gene. These results strongly suggest that p71, p7a, p7b and CP are involved in TNV-DHcell-to-cell and longdistance movement, respectively.

Two mutant potato virus X (PVX) movement protein (MP) genes (m12K-Sal and m12K-Kpn) were obtained by inserting specific linkers at the boundary between the N-terminal hydrophobic and putative transmembrane segment, and the central invariant hydrophilic region of the respective 12 kDa, 12K, triple gene block (TGB) protein. Several transgenic potato lines which expressed m12K-Sal or m12K- Kpn to different degrees were resistant to infection by PVX, potato aucuba mosaic potexvirus and the carlaviruses potato virus M and S over a wide range of inoculum concentrations (3-300 μg/ml). However,theywere not resistantto potato virus Y, which lacks a TGB protein. We suggest that the resistance of m12K-Sal and m12K-Kpn transgenic potato lines is MP-derived and not RNA-mediated.

The RNA-dependent RNA polymerase (RdRp) of potato virus X (PVX) contains a glycine-lysine- serine (GKS) motif. This motif is present in the replication enzyme of many RNA viruses and is thought to be required for nucleoside triphosphatebinding. Three single amino acid changes, glycine to alanine (AKS), lysine to asparagine (GNS) and lysine to glutamate (GES) within the GKS motif of the PVX RdRp were tested for their effect on PVX accumulation. The GNS and GES mutations rendered the virus unable to accumulate in either tobacco plants or protoplasts, whereas substitution of glycine with alanine had only a minor effect on accumulation of PVX. The glycine to alanine mutation reverted to wild-type after passage on Nicotiana clevelandii plants. These findings suggest that the GKS motif is required for PVX replication and that strong selection pressures are active to maintain necessary sequences of the viral RdRp.

Partial sequences from serologically characterized yam mosaic potyvirus (YMV) isolates were determined in conserved (helper-component proteinase, HC; nuclear inclusion b, NIb) and variable (first protein, P1; third protein, P3; and coat protein, CP) regions of the potyviral genome in order to investigate the intraspecies molecular diversity of YMV. Multiple sequence alignments and pairwise comparisons were used to quantify the sequence polymorphism in these regions. Two levels of diversity were observed among YMV isolates: above 90% nucleotide (nt) sequence identities were found between YMV isolates of the same group (intragroup) regardless of the region considered, whereas identities between isolates from different groups (intergroup) were lower and depended upon the protein chosen. For instance, the average intergroup nt sequence identity between YMV isolates was about 65% in the P1 protein and the N terminus of the CP while there was more than 80% nt identity in the HC, P3 and NIb proteins. Thus P3 appeared to be conserved between YMV isolates even though this region was variable between potyvirus species. Similar analysis of the intraspecies molecular diversity of other potyviruses (potato virus Y, zucchini yellow mosaic virus, plum pox virus, pea seed-borne mosaic virus) led to the same results: (i) two levels of intraspecies molecular diversity were found (intragroup and intergroup); (ii) intraspecies molecular diversity differed from interspecies molecular diversity in the P3, P1 and N- terminal regions.

Maize streak virus (MSV) coat protein (CP) is required for virus movement within the plant. Deletion or mutation of MSV CP does not prevent virus replication in single cells or protoplasts but leads to a loss of infectivity in the inoculated plant. The mechanism by which MSV CP mediates the transfer of MSV DNA from cell to cell and through the vascular bundle is still unknown. Towards understanding the role of MSV CP in virus movement, the interaction of the CP with viral DNA was investigated using the ‘south-western’ assay. Wild-type and truncated MSV CPs were expressed in E. coli and the expressed CPs were used to investigate interactions with single-stranded (ss) and double-stranded (ds) DNA. The results showed that MSV CP bound ss and ds viral and plasmid DNA in a sequence non-specific manner. The binding domain was mapped to within the 104 N-terminal amino acids of the MSV CP. We propose that the binding of CP to MSV DNA is involved in viral DNA nuclear transport as well as encapsidation and thus may have a role in intra- and inter-cellular movement as well as systemic infection.

RNA transcribed from cloned satellite tobacco mosaic virus (STMV) cDNA replicated in Nicotiana benthamiana protoplasts when co-inoculated with tobacco mild green mosaic virus (TMGMV) genomic RNA, but degraded when inoculated alone. STMV genomic RNA extracted from wild-type virions replicated in protoplasts when co-inoculated with TMGMV, tobacco mosaic virus (TMV) or tomato mosaic virus (ToMV). Transcripts from clones of two STMV coat protein (CP) mutants accumulated to the same level as wild-type transcripts in protoplasts when co-inoculated with TMGMV, whereas a third mutant accumulated to detectable levels in some, but not all, experiments. These results confirm that STMV RNA requires helper virus for replication, and that the helper specificity exhibited by cloned STMV reflects a specific requirement for the TMGMV replicase. It also demonstrates that the low accumulation of STMV CP mutants observed previously in whole plants cannot be attributed to inefficient RNA replication.

Isolates of groundnut rosette umbravirus (GRV) contain a satellite RNA(sat-RNA), about 900 nucleotides (nt) in length, different variants of which are responsible for the symptoms of different forms of rosette disease in groundnuts and, in the particular instance of sat-RNA YB3b, for the production of yellow blotch symptoms in Nicotiana benthamiana. Sat-RNA YB3b does not affect the accumulation of GRV genomic or subgenomic RNAs in infected plants. Replication of sat-RNA YB3b and induction of yellow blotch symptoms do not require the production of any sat-RNA-encoded proteins. Experiments with deletion mutants identified three functional untranslated elements in sat-RNA YB3b. One (designated R) comprises nt 47–281, is essential for sat-RNA replication and appears to be cis-acting. The other two (designated A and B) comprise nt 280–470 and 629–849, respectively, are both involved in yellow blotch symptom production and can act in trans. Element A contains the determinant that is unique to sat-RNA YB3b. The process of symptom induction by sat-RNA YB3b apparently involves a novel type of specific interaction of two untranslated RNA elements, which can complement each other, with a host factor or factors.

Three distinct genotypes were identified amongst 50 measles virus (MV) strains characterized in the UK between 1992 and 1995 by direct sequencing of the RT-PCR products amplified from clinical specimens. All three genotypes were related to viruses previously reported in the United States or in continental Europe. Phylogenetic analyses of 255 and 152 nucleotide sequences from the N and M genes, respectively, generated very similar lineages. The degree of divergence between genotypes was 3·5–16% and 2·6–7·9% in the N and the M genes, respectively. MV genotypes which circulated during the 1970s and 1980s in the UK were not detected in this period. Comparison of the C-terminal regions of the N gene sequences of UK strains isolated or detected between 1974 and 1995 suggests that there have been multiple genotypes of MV circulating in the UK over the past 20 years.

The haemagglutinin protein (H) of measles virus (MV) binds to susceptible cells and collaborates with the fusion protein (F) to mediate fusion of the virus with the cell membrane. Binding and fusion activity of the virus can be monitored by haemagglutination and haemolysis, respectively, of monkey erythrocytes. Most monoclonal antibodies (MAbs) with haemolysis inhibiting activity (HLI ) are either MV-F specific and do not inhibit haemagglutination (HI ), or they bind to MV-H and are HI by interfering with virus binding. We describe here a small panel of H- specific MAbs (BH47, BH59, BH103, BH129) which bind to a new linear neutralizing epitope, H244–250 (SELSQLS; NE domain), and which prevent virus-cell fusion (HLI ) but not virus binding (HI ). These antibodies also protect against MV encephalitis in an animal model. They do not compete with an HLI /HI antibody (BH216) which binds to the haemagglutinin noose epitope (HNE). The antibodies described here and the HNE-specific antibodies are functionally distinct and define two topographically non-overlapping interfaces, supposedly with a bias towards the host cell MV- receptor and the fusion protein respectively. The proximity of the CD46 downregulating amino acid Arg-243 may suggest a functional link between the domain described here and the CD46 binding domain. This new protective linear site is also of potential interest for the design of a subunit-based vaccine.

The epitopes recognized by 42 monoclonal antibodies directed against the human parainfluenza virus type 2 (hPIV-2) phosphoprotein (P) were mapped on the primary structure of the P protein by testing their reactivities with deletion mutants. By Western immunoblotting with these monoclonal antibodies and P protein deletion mutants the region essential for P-P interactions was determined. The P protein region encompassing amino acids 211–248 was required for proper folding and oligomerization which is mediated by predicted coiled-coils in this region. The oligomer was shown to be a homotrimer by chemical cross-linking experiments.

A 1622 nucleotide region of the bovine ephemeral fever virus (BEFV) genome, located between the second glycoprotein (GNs) gene and the polymerase (L) gene, has been cloned and sequenced in Australian (BB7721) and Chinese (Beijing-1) isolates of the virus. In the Australian isolate, the region contains five long open reading frames (ORFs) organized into three coding regions (α,β and γ), each of which are bound by a consensus transcription initiation and transcription termination- polyadenylation-like sequences. The α coding region contains three long ORFs (α1, α2 and α3). The a1 ORF encodes a 10 6 kDa polypeptide which contains hydrophobic and highly basic regions characteristic of a viroporin. The α2 ORF encodes a 13 7 kDa polypeptide and overlaps the a3 ORF which encodes a 5·7 kDa polypeptide. The β coding region contains a single long ORF encoding a polypeptide of 12·2 kDa. The y coding region, which does not occur in Adelaide River virus (ARV), contains a single long ORF encoding a polypeptide of 13·4 kDa. The Chinese isolate shares 91% nucleotide sequence identity with the Australian isolate. The organization of the α, β and γ coding regions is preserved and the sequences of the encoded polypeptides are similar to those of BB7721. The major transcription products of the region were identified in BB7721 as polycistronic a (α1-α2-α3) and β-γ mRNAs. Sequence similarities in the BEFV α-β and β-γ gene junctions, and the γ-L and β-L gene junctions of BEFV and ARV, suggest that the γ gene may have evolved from the β-gene by sequence duplication.

RT-PCR methods have been applied to the detection and sequencing of the glycoprotein gene of viral haemorrhagic septicaemia virus (VHSV), the rhab- dovirus which causes viral haemorrhagic septicaemia (VHS) in farmed salmonid fish. Phylogenetic analysis of a 360 nt region of the glycoprotein gene from a range of marine and fresh water VHSV isolates identified three genogroups, I-III. Significantly, two virus isolates recovered from ulcerated North Sea cod caught off the Shetland Islands, and an isolate recovered from diseased turbot farmed on the island of Gigha, Scotland were assigned to the same genogroup. Moreover, a virus isolated from diseased turbot farmed on the Baltic Sea coast shared 99·4% nucleotide sequence similarity with a virus associated with a VHS outbreak in rainbow trout. This is the first time that a genetic link between a VHS outbreak and natural VHSV infections of marine fish species has been demonstrated.

An influenza C virus variant, C/AA-cyt, was identified as the agent responsible for highly effective induction of cytopathogenicity in MDCK cells. The cytopathogenic effect was manifested by cell rounding, cell shrinkage and foci of cell destruction leading finally to disruption of the monolayer in a virus dose-dependent manner. Virus-induced cyto- pathogenicity was suppressed by temperatures nonpermissive for virus replication. Maintenance of plasma membrane integrity post-infection, in connection with induction of a DNA fragmentation ladder, revealed the characteristic picture of apoptosis. In support of this, quantitative analysis demonstrated high levels of apoptosis-like oligo- nucleosomal DNA. The results indicate that influenza C viruses can induce programmed cell death, as formerly reported for influenza type A and B viruses.

Hepatitis C virus (HCV) core protein forms the internal viral coat that encapsidates the genomic RNA and is enveloped in a host cell-derived lipid membrane. As the single capsid protein, core should be capable of multimerization but attempts to produce virus-like particles following expression of HCV structural proteins have not been successful. In this study, we have analysed the interaction capacity of full-length and truncated HCV core using the yeast two-hybrid system. Full-length core containing or lacking the translocation signal for the E1 glycoprotein did not interact with full-length or truncated core proteins. Truncation to the N-ter- minal 122 aa revealed an interaction domain which was mapped to the tryptophan-rich sequence from aa 82–102 and was termed the main homotypic interaction domain. The C-terminal hydrophobic fragment of core (aa 122–172) was incapable of interacting with itself but interacted with the main homotypic interaction domain in trans (the weak heterotypic interaction domain). Core proteins truncated at their N and C termini (aa 46-102) were trans-activating when fused to the DNA-binding domain of GAL4. Based on our results, we suggest that the C-terminal segment may interact in cis with the main homotypic interaction domain and thereby prevent multimerization. Core-core interaction was also observed for in wtro-translated proteins bound to truncated immobilized core102. However, interaction was less specific in this system suggesting that protein interaction and possibly conformational alteration of core may be dependent on the experimental system.

Hepatitis C virus (HCV) type 4 is the predominant genotype found throughout the Middle East and parts of Africa, often in association with high population prevalence as in Egypt. To investigate more fully its evolutionary relationship with other genotypes of HCV, and to study its overall genome organization, we have determined the entire sequence encompassing the coding region of the genotype 4a isolate ED43, obtained from an HCV- infected individual from Egypt. The sequence of ED43 contained a single open reading frame encoding a polyprotein of 3008 amino acids (aa), smaller than that reported for other HCV genotypes which vary from 3010 aato 3037 aa.The nucleotide and amino acid sequences were compared with the full-length sequences already reported for genotypes 1a, 1b, 1c, 2a, 2b, 2c, 3a, 3b and those of isolates JK049 and JK046 described as types 10a and 11a. The differences in length of the polyprotein originated in variable regions in the E2 and NS5A genes. The complete sequence of ED43 confirmed the classification of type 4 as a separate major genotype.

We have examined and compared at the molecular level three strains of wild-type yellow fever (YF) virus isolated from Senegal in 1927, 1953 and 1965, termed French viscerotropic virus, Rendu and Dak1279 respectively. Over the structural protein genes, Rendu differed from the other two strains by 8% at the nucleotide level. Rendu also differed antigenically, possessing a ‘vaccine’- specific envelope (E) protein epitope (i.e. an epitope previously shown to be found on 17D and French neurotropic vaccine viruses only and not wild-type strains of YF virus). Consequently, we propose that at least two distinct genotypes of wild-type YF virus have been present in Senegal. Since Rendu virus was isolated from a fatal case of YF, it would indicate that the vaccine-specific epitope on the E protein is not associated with attenuation of the viscerotropism of wild-type YF virus.

Two monoclonal antibody neutralization resistant (MAbR) variants of the yellow fever (YF) 17D-204 vaccine virus strain were selected using YF type- specific MAb B39. These B39R variants were compared with the variant virus selected by Lobigs et al. (Virology 161,474–478, 1987) using a second YF- type specific MAb (2E10) which mapped to amino acid position 71/72 in the envelope (E) protein. Neutralization assays with a panel of MAbs suggested that these two YF-type-specific epitopes are located in two discrete regions of the folded E protein. Each of the B39R variants had a single nucleotide mutation which encoded an amino acid substitution at either position E-155 or E-158. Thus, YF type-specific epitopes map to both domain I (B39) and II (2E10) of the YF virus E protein. The B39 defined epitope represents the first flavivirus neutralizing epitope localized to this region of domain I of the E protein.

Infections with pestiviruses occur in cattle, sheep, pigs and also in numerous other ungulate species. In the present study, pestiviruses from goat, buffalo, deer and giraffe were analysed at the molecular level; unusual strains from cattle and pigs were also included. A phylogenetic analysis of the respective pestiviruses was undertaken on the basis of a fragment from the 5′ noncoding region as well as the gene encoding autoprotease Npro. Statistical analyses of the respective phylogenetic trees based on the 5′ NCR revealed low confidence levels for most of the branches, while the structure of the tree based on the Npro gene was supported by high bootstrap values. Accordingly, the isolates from goat, buffalo and deer can be grouped together with bovine viral diarrhoea virus (pestivirus type 1); within this genotype three subgroups and one disparate virus have been identified. One isolate from pig and one from cattle belong to the group of ‘true’ border disease virus (pestivirus type 3), which can be further subdivided into two major subgroups. Interestingly, the giraffe isolate does not belong to one of the four established pestivirus genotypes. The phylogenetic analysis strongly suggests that genotype 1 pestiviruses occur world-wide in many ruminant species. Furthermore, phylogenetic trees based on the Npro gene nucleotide sequences show that the respective sequences do not segregate into discrete lineages based on host-species origin.

It has been previously shown that rotavirus triplelayered particles induce permeabilization of liposomes and membrane vesicles. These effects were mediated by one or both of the solubilized outer-capsid proteins, VP4 and VP7. Permeabilization was dependent on trypsin treatment of the viral particles, suggesting that VP4 was involved. To analyse the respective roles of the outer-capsid proteins in this permeabilization process, we have used membrane vesicles loaded with carboxyfluorescein and virus-like particles derived from insect cells co-expressing various sets of capsid proteins. Virus-like particles containing VP2, VP6 and VP7 (VLP2/6/7) are as efficient in permeabilizing vesicles as triple-layered particles. As with double-layered particles, virus-like particles made of VP2 and VP6 had no effect on vesicle permeabilization. Permeabilization of membrane vesicles required trypsinization of the VP7 solubilized from VLP2/6/7. These results show that solubilized and trypsinized VP7 is able to induce membrane permeabilization, independently of the presence of VP4.

Rotavirus serotype G5 isolates were recently recovered from children with diarrhoea in Brazil. Like most human strains, they exhibited long electro- pherotypes and subgroup II and Wa-like VP4 specificity. We report the successful propagation and the molecular and antigenic characterization of one of these isolates (IAL-28). Cross-neutralization of IAL-28 and a single gene reassortant, UK × IAL- 28, which contains the gene encoding the IAL-28 VP7 in the UK genomic background, with prototype G1 to G14 rotaviruses demonstrated that IAL-28 has antigenic determinants specific for both G5 and G11 serotypes. Sequence analysis of the gene encoding VP7 suggested that one or two amino acid substitutions at positions 96 and 100 on IAL-28 VP7 were possibly responsible for the additional G11 specificity. G5 rotaviruses are found in horses and predominate in piglets, whereas G11 has been identified exclusively as a swine pathogen.

The nucleotide and deduced amino acid sequences of genome segment 10 of aquareovirus strain SBR, encoding the major outer capsid protein (VP7), have been determined. Genome segment 10 of SBR virus is 986 nucleotides long and encodes a polypeptide of 298 amino acids with a predicted molecular mass of 32430 Da. There are 26 non- translated nucleotides at the 5′ end and 66 non- translated nucleotides at the 3′ end. Using a recombinant baculovirus system, the VP7 protein of SBR virus was expressed to a high level. The baculovirus-produced VP7 protein was similar both in its size and antigenic properties to the authentic aquareovirus VP7 protein. Antiserum from a rabbit immunized with the baculovirus-produced VP7 protein failed to neutralize the homologous aquareovirus strain. As determined by Western blotting, this antiserum reacted with aquareovirus strains belonging to the same genogroup as SBR virus, but did not react with aquareovirus strains belonging to the other genogroups.

The relative mobilities of the 11 dsRNA genomic segments of 22 aquareovirus isolates from fish and shellfish obtained from different geographical areas of the world were compared by PAGE. Using reciprocal RNA-RNA dot blot hybridization, a new sixth genetic group of aquareovirus (genogroup F) was identified. Genogroup A was represented by eight and genogroup B by 12 isolates. The remaining two isolates represented the new sixth genogroup (genogroup F). The genetic relationship of these aquareoviruses with mammalian rotavirus group A (SA11) was also examined by reciprocal RNA-RNA blot hybridization but none was found under any of the stringency conditions used.

Degenerate and specific PCR assays were developed for bovine leukaemia virus (BLV) and/or primate T cell leukaemia/lymphoma viruses (PTLV). The degenerate assays detected all major variants of the BLV/PTLV genus at a sensitivity of 10-100 copies of input DNA; the specific systems detected 1–10 copies of input target. Sensitivity was 100% in specific DNA-PCR assays done on peripheral blood from seropositive BLV-infected cattle and HTLV-I- or HTLV-II-infected humans, and 62% in RNA/DNA- PCR assays on sera from BLV seropositive cattle. The pol fragments from 21 different BLV strains, isolated from cattle in North and Central America, were cloned and sequenced, and compared to other published BLV and PTLV pol sequences. BLV and PTLV sequences differed by 42%. Sequence divergence was up to 6% among the BLV strains, and up to 36% among the PTLV strains (with PTLV-I and PTLV-II differing among themselves by 15% and 8%, respectively). Some cows were infected with several BLV strains. Among retroviruses, BLV and PTLV sequences formed a distinct clade. The data support the interpretation that BLV and PTLV evolved from a common ancestor many millennia ago, and some considerable time before the PTLV- I and PTLV-II strains diverged from each other. The dissemination of the BLV strains studied probably resulted from the export of European cattle throughout the world over the last 500 years. The relatively similar mutation rates of BLV and PTLV, after their various points of divergence, suggest that there could be a much wider genetic range of BLV than has currently been defined.

To identify the products of the bovine herpesvirus- 1 (BHV-1) gE and gI genes, we constructed baculo- virus recombinants containing the putative gE or gI genes. These recombinant viruses synthesized BHV-1 gE and gI with apparent molecular masses of 84 and 41 kDa, respectively. Polyclonal antibodies against these recombinant gE or gI proteins were produced and by using these antibodies, we showed the presence of gE and gI with apparent molecular masses of 94 and 45 kDa, respectively, in purified BHV-1 virions. We also demonstrated that like their herpes simplex virus-1 and pseudorabies virus counterparts BHV-1 gE and gI form a complex. A gI BHV-1 mutant failed to express gI but gE was found in the virions. On the other hand, neither gE nor gI was found in the virions of a gE BHV-1 mutant. In th gE BHV-1 mutant, gI was produced but released into the medium without being integrated in the virions.

Herpesvirus saimiri strain C488 transforms human CD4 T-lymphocytes to continuous interleukin- 2-dependent growth. Unlike human T-cell lines derived from tumours or those transformed by human T-lymphotropic virus 1, herpesvirus saimiri- immortalized T-cells (HVS T-cells) retain many functions of primary activated T-lymphocytes. We have characterized the course of human immunodeficiency virus types 1 and 2 (HIV-1/-2) infection in three HVS T-cell lines. Our results confirm that HVS T-cells are highly permissive to both HIV-1/-2 prototype viruses and to poorly replicating HIV-2 strains of restricted cell tropism. However, the infection was persistently productive for up to 5 months. The down-regulation of surface CD4 molecules was delayed and virus yields significantly exceeded those obtained in T-cell lines.

We have previously demonstrated that two distinct transcripts are produced from ORF 50, the major transcriptional activating gene of herpesvirus saimiri. The products of these transcripts trans- activate the delayed-early ORF 6 promoter, though to different degrees. Deletion analysis demonstrated that the ORF 50 responsive elements are contained in a 132 bp fragment situated 127–259 bp from the transcription initiation site within the ORF 6 promoter. This fragment conferred ORF 50-responsiveness on an enhancerless simian virus 40. Gel retardation analysis further mapped the responsive elements to a 38 bp fragment.

Epstein-Barr virus (EBV) infection in animal model systems has been studied previously in marmosets and tamarins using serology and PCR of saliva. Here we directly demonstrated long-term persistence of EBV in the peripheral blood of marmosets by assaying EBER RNA expression. A new reverse transcription-PCR assay, able to distinguish a naturally occurring strain polymorphism in EBER 2 that may be useful as a strain marker for monitoring persistence and interactions between multiple strains in the same animal or person, has been developed. In situ hybridization and immunohisto-chemistry have also been used to search for EBV-infected cells in the animals. The carrier state in the common marmoset is similar to that of humans in that it is asymptomatic, long-lived and displays a very low level of circulating virus-infected cells. It differs from the human in lacking the characteristic antibody response to EBNA 1.

Murine gammaherpesvirus 68 (MHV-68) is a relatively recently discovered pathogen of wild rodents and provides a unique opportunity to explore in detail the interactions of a gammaherpesvirus with its natural host. It may also provide a much needed small animal model for human gammaherpes-viruses. As a step in the detailed analysis of virus gene structure and expression we have sequenced over 20 kb of the MHV-68 genome and mapped gene transcripts by Northern blot hybridization. The region we chose to analyse contains several conserved gene blocks as well as some less well conserved genes and allowed us to estimate the relationship of this virus to other herpesvirus family members. Of particular interest is the fact that none of the characteristic Epstein-Barr virus (EBV) genes is present at this genomic locus although MHV-68 does have one gene encoding a membrane glycoprotein, gp150, which shows similarities to the major membrane glycoprotein of EBV. Our results further confirm that MHV-68 is a gammaherpesvirus marginally more closely related to a cluster of gammaherpesviruses including herpesvirus saimiri than to EBV. Northern analysis shows that the temporal regulation of expression is broadly similar to that of other herpesviruses in this region of the genome. We also show that like other gammaherpesviruses, MHV-68 splices its homologue of the EBV transcriptional activator gene BMRF1.

The major capsid protein of human polyomavirus JC virus, VP1, has been cloned into a baculovirus genome and expressed in insect cells. The VP1 protein was expressed in the cytoplasm and transported into the nucleus. It was then purified by a sucrose cushion and CsCl density gradient centrifugation to near homogeneity. Electron microscopy showed that isolated recombinant VP1 protein self-assembled into a capsid-like structure similar to the natural empty capsid. Both chelator (EDTA) and reducing agent (DTT) are required to disrupt the capsid structure into the pentameric capsomeres, as demonstrated by haemagglutination assay and electron microscopy. These results suggest that JC virus VP1 can be transported into the nucleus and self-assembled to form capsid-like particles without the involvement of the viral minor capsid proteins, VP2 and VP3. In addition, metal ions and disulphide bonds appear to be important in maintaining the integrity of the viral capsid structure.

Recently, we hypothesized that the tumour-suppressive, human helper-virus-dependent, adeno- associated parvoviruses (AAV) may interfere with transforming functions of human papillomaviruses (HPV) in the development of cervical carcinoma. Here, we demonstrate that in cervical epithelium containing papillomavirus DNA, AAV DNA can be detected in a replication-competent form and that AAV proteins are expressed. In cultured cells con-taining integrated AAV-2 DNA, transfection of HPV- 16 DNA induced rescue of infectious AAV-2, revealing helper functions of HPV-16. Similarly, cotransfection of HPV-16 and AAV-2 DNAs into human epithelial cell lines led to replication of AAV-2, and, in keratinocytes, to a cytopathic effect. These data suggest an interaction of the two viruses, possibly influencing the development of HPV-related lesions.

Capsid proteins VP1, VP2 and VP3 of adeno- associated virus type 2 (AAV-2) were separately expressed by recombinant baculoviruses, purified under denaturing conditions and renatured in the presence of 0·5 M arginine, followed by dialysis against buffers of physiological ionic strength. At a protein concentration of 0·05 mg/ml, the three capsid proteins predominantly formed monomers and, to a lesser extent, oligomers, as determined by sedimentation analysis. Oligomerization increased at higher protein concentrations. The capsid protein oligomers consisted of globular, non-capsid-like structures, as detected by electron microscopy. Addition of a HeLa cell extract significantly stimulated oligomerization of the capsid proteins, probably due to interactions with HeLa cell proteins. Characterization of structures sedimenting around 60S by immunoprecipitation and electron microscopy showed that, in addition to other aggregates, empty capsid-like structures were formed in vitro. The identity of these structures as empty AAV capsidswasverified by immunoelectron microscopy. Analysis of capsid formation in HeLa cells by transfection of VP expression constructs allowing separate expression of VP1, VP2 and VP3 showed that they were able to form capsids, although with a reduced efficiency as compared to VP proteins expressed from the wt cap gene. This finding suggests that the mutations introduced to allow separate capsid protein expression reduced the efficiency of capsid assembly in vivo and might also explain the reduced recovery of empty capsids employing the in vitro assembly procedure.

Covalently closed circular DNA (cccDNA) is the first hepatitis B virus (HBV) DNA replicative intermediate formed from the genomic DNA and serves as the template for synthesis of viral mRNA and pregenomic RNA. It also appears to be produced by intracellular recycling of relaxed circular DNA intermediates. Here, we report that none of the forms of HBV surface antigen affect this intracellular recycling of HBV DNA. Elimination of the initiation codons for the large and middle surface proteins did not affect the detection of surface antigen in culture supernatants. In contrast, detection of surface antigen was eliminated by the removal of, or the introduction of two stop codons downstream of, the initiation codon of the small (major) surface antigen. None of the mutations affected the production, in transfected HepG2 cells, of HBV DNA replicative intermediates, including cccDNA.

A translational stop in the hepatitis B virus (HBV) precore codon 28 and specific changes in the core promoter region of the X gene have been suggested to influence the level of circulating HBeAg in patients. We analysed the core promoter region and precore sequences from 59 HBV strains (including 14 from the databank) of different genotypes and from patients with different HBeAg/anti-HBe patterns. The initiator and TATA elements for transcription of precore and pregenomic RNA were highly conserved. The majority of X gene deletions in the core promoter region would lead to translational frame-shifts and stops, truncating the C- terminal end of the X protein. We found significant associations between specific changes in core promoter positions 1762 to 1764, or in precore codon 28, and absence of circulating HBeAg. For the core promoter mutations alone, this association was related to the apparent degree of liver damage (as estimated by alanine aminotransferase levels) at the time of sampling. Mutations at nucleotides 1762 and/or 1764 were often accompanied by point mutations at positions 1751 to 1755. Since mutations at nucleotide positions 1762 and 1764 have recently been shown by in vitro studies to suppress HBeAg production with a concomitant enhancement of virus production, disappearance of the HBeAg- positive phenotype associated with 1762 to 1764 mutations may thus have at least as much significance for the course of infection as HBeAg absence associated with precore codon 28 stop mutations. These observations are considered against a secondary structural model for the 3′ end of HBV pregenomic RNA which also predicts enhancement of virus replication after mutation at positions 1762 and 1764.

Hepatitis B virus (HBV) genomes with deletions in the precore-core (preC-C) promoter have been detected in HBV infections without serological markers. To address whether the mutations are responsible for the reduced production of virus antigens, either an 8 bp (8d, position 1763 to 1770) or a 20 bp (20d, 1753 to 1772) deletion was created in a wild-type (wt) HBV clone. Both mutations cause premature termination of the overlapping X ORF. When introduced into HepG2 cells, both mutants produced reduced amounts of HBsAg, HBcAg and HBeAg, but released the same or more virion- associated DNA compared with the wt. A co-transfection of the 20d mutant with a small amount of intact X gene resulted in a 3-fold increase of HBcAg production compared to transfection with either the 20d or wt alone. When the promoter region was cloned into CAT plasmids, the 8d preC promoter showed weak activity and its initiation site was shifted 6 to 10 bp downstream. The preC promoter activity of 20d was not detectable by CAT ELISA and 5′ RACE. The levels of C transcripts of both mutants were higher than that of the wt, and their start sites were not altered. Therefore, the deletions cause the reduction of HBsAg, HBcAg and HBeAg although the mutant viruses can still replicate in cultured cells. The reduction of HBeAg is due to both the reduced preC promoter activity and the defect in HBx. The reduction of HBcAg is due to the disrupted X gene, despite augmented C promoter activity.

C-terminally truncated surface proteins of hepatitis B virus (HBV) are frequently translated from genomically integrated viral sequences. They may be relevant for hepatocarcinogenesis by stimulating gene expression. First, we examined the transactivating potential of middle hepatitis B surface protein truncated at amino acid (aa) position 167 (MHBst167) on the HBV regulatory element. In transient cotransfection assays using Chang liver or HepG2 cell lines and chloramphenicol acetyl- transferase (CAT) reporter constructs only the HBV enhancer I, but no other HBV regulatory elements like the X promoter, the S1 or S2 promoter or the enhancer II/core promoter could be stimulated by MHBst167. Since there is no evidence for a direct interaction of MHBst167 with DNA, we subsequently analysed whether cellular transcription factors were involved in mediating transactivation. This was tested both with isolated transcription-factor-binding sites and in the natural context of viral and cellular promoter elements. Deletion analysis and electrophoretic mobility shift assays revealed that Sp1,AP1 and NF-kB can mediate transactivation by MHBst167. No involvement of CREB, NF1 or the liver- specific factor C/EBP was found. These data indicate that MHBst167 is a pleiotropic, non-liver-specific transactivator which exerts its effect via ubiquitous cellular transcription factors that are also involved in the regulation of expression of cellular genes relevant for proliferation and inflammation.

The genome of Spodoptera exigua multicapsid nucleopolyhedrovirus (SeMNPV) was screened for the presence of putative origins of DNA replication (oris). Using a transient DNA replication assay, several fragments were identified that underwent SeMNPV-dependent DNA replication in Spodoptera frugiperda cells (Sf-AE-21). Preliminary sequence data revealed the presence of multiple copies of homologous repeats (hrs). Restriction fragment XbaI-F2 showed a distinct sequence reminiscent of Autograpfia californica and Orgyia pseudotsugata MNPV (AcMNPV and OpMNPV) non-fir oris. Deletion analysis of this fragment indicated that the essential sequences of this putative non-hr ori mapped within a region of 800 bp. Sequence analysis of this region showed a unique distribution of six different (im)-perfect palindromes, several polyadenylation motifs and the occurrence of multiple direct repeats. No sequence homology or similarities to other reported baculovirus oris were detected. The spatial and modular distribution of these motifs are similar to those of the non-hr oris of AcMNPV and OpMNPV. Comparison of baculovirus non-fir and consensus eukaryotic oris revealed no consensus ori but indicated that each of the non-hrs studied so far is unique. From the structural similarity, however, it was concluded that the SeMNPV XbaI-F2 ori represents a baculovirus non-fir type ori. In addition, evidence is provided that SeMNPV renders more specificity to baculovirus DNA replication than AcMNPV.

IE1 of Autographa californica multicapsid nuclear polyhedrosis virus acts as a transactivator of several viral promoters in insect cells. Transient expression assays indicate that IE1 is involved in the activation of the early promoter he65 in both insect TN-368 and mammalian BHK-21 cell lines. IE1 activation of the he65 promoter was compared with IE1 activation of the early 39K promoter. In contrast to the he65 promoter, the 39K promoter was not inducible by IE1 in BHK-21 cells.