- Volume 72, Issue 9, 1991

Primary oligodendrocyte (OL) cultures from three inbred strains of mice with known differences in resistance to herpes simplex virus type 1 (HSV-1) infection in vivo (A/J, susceptible; BALB/cByJ, moderately resistant; C57BL/6J, resistant), also display a similar pattern of resistance in vitro. The nature of the in vitro resistance at the cellular level was investigated. Virus production at different m.o.i.s indicated that the differences in HSV-1 replication are m.o.i.-dependent. Overall, virus yield from the OL cultures infected at a multiplicity of 1 increased 48 h post-infection (p.i.); no additional enhancement occurred 72 h p.i. However, the difference in the replication capacity of the three OL cultures observed at 24 h p.i. persisted at 48 and 72 h p.i. Serial electron microscopy studies on infected OL cultures derived from the different murine strains suggested that the resistance to HSV-1 infection occurs at different stages during the replicative cycle. Virus was detected at the nuclear membrane 5 min p.i. in A/J cells, but was not observed until 120 min p.i. in BALB/cByJ cells, whereas virus could not be detected at the nuclear membrane of C57BL/6J cells, even at 24 h p.i. Virus adsorption, determined by assay of residual non-adsorbed virus infectivity and cell-associated, radiolabelled HSV-1, did not differ in the OL cultures. The cumulative data suggest that A/J cells display the same replication pattern as permissive CV-1 cells, whereas the major replicative blocks in the other two murine strains occur at the level of the cytoplasmic membrane in C57BL/6J OLs, and at the level of the nuclear membrane in BALB/cByJ cells.

We have used the nested polymerase chain reaction (PCR) combined with fluorescence-activated cell sorting to define sites of latency of human cytomegalovirus (HCMV) in the peripheral blood of healthy subjects. Peripheral blood mononuclear (PBM) cells were separated into T cell or non-T cell populations and monocytes, and were then analysed by PCR for the presence of HCMV DNA. In five of six seropositive subjects, HCMV was found predominantly in the non-T cell population. Further analysis suggested that the virus was present in adherent cells and CD14+ cells. In three of nine seronegative subjects we could demonstrate HCMV DNA, which we do not believe was due to contamination, reproducibly by PCR. In one of these seronegative subjects, HCMV DNA was present predominantly in the non-T cell fraction of PBM cells. No HCMV DNA was detectable in the remaining six seronegative subjects. We conclude that, within the PBM cells of normal asymptomatic seropositive and some seronegative subjects, HCMV is present predominantly in the monocyte fraction. In addition, the detection of HCMV sequences in seronegative subjects may indicate that infection with HCMV is more widespread than conventional seroepidemiology suggests.

Hybridomas producing human monoclonal antibodies (HMAbs) against varicella-zoster virus (VZV) were generated by fusing murine myeloma cells with human lymphocytes immunized in vitro. An assay system was developed to select anti-glycoprotein (gp)III HMAbs from the pool of anti-VZV HMAbs. A murine anti-gpIII MAb, 4B7, did not react with a VZV-infected cell homogenate, but did react with a VZV-infected cell monolayer, whereas anti-gpI and anti-gpII MAbs reacted with both antigens. Hybridomas were screened to obtain HMAbs having a reaction profile similar to that of 4B7 and one such clone, V3, stably produces human IgG1 (κ). HMAb V3 immunoprecipitated a VZV antigen of 115K to 120K, which was not immunoabsorbed by an anti-gpII HMAb, implying that V3 recognizes gpIII. V3 neutralized VZV independently of complement, unlike anti-gpI and anti-gpII HMAbs. All five strains of VZV tested were completely neutralized by V3, and the dose of V3 required to reduce the number of virus plaques by 50% ranged from 0.027 to 0.15 µg/ml. V3 was also able to inhibit the spread of virus infection from infected to uninfected cells, whereas anti-gpI and anti-gpII HMAbs could not. In addition, V3 mediated antibody-dependent cellular cytotoxicity but not complement-dependent cytotoxicity of VZV-infected cells. The results suggest that an anti-gpIII HMAb may provide a new means of passive immunoprophylaxis and also help to identify an antigenic epitope appropriate for a subunit vaccine.

A panel of 14 monoclonal antibodies (MAbs) was used to characterize the high abundance glycoproteins of equine herpesviruses 4 (EHV-4) and 1 (EHV-1), and asinine herpesvirus 3 (AHV-3). The specificities of the MAbs, which had been determined previously for strains of EHV-4 and -1 from the U.S.A., in general were confirmed by ELISA for Australian strains of these viruses. Of the 14 MAbs seven were EHV-4 and -1 type-common and cross-reacted with AHV-3. Of the five MAbs that were EHV-1 type-specific, four cross-reacted with AHV-3, whereas neither of the EHV-4 type-specific MAbs reacted with AHV-3, providing further evidence for a closer evolutionary relationship between EHV-1 and AHV-3 than that between either of these viruses and EHV-4. By Western blot and immunoprecipitation analyses, the identity of the six major glycoproteins, gp2, gp10, gp13, gp14, gp18 and gp21/22a, of an Australian EHV-1 isolate was verified, and it was shown that AHV-3 had cross-reactive glycoproteins of very similar M r to those of EHV-1; five homologous glycoproteins of EHV-4 were also identified. It was determined that the EHV-4 gp13 homologue had a much reduced M r (67K) when the virus was grown in a continuous cell line than when grown in equine foetal kidney cells (95K). It is suggested that altered glycosylation by the cell line is responsible for this change in M r. Those glycoproteins acting as major immunogens in the naturally infected host, at least in their ability to elicit antibody, were identified. It was found that gp2, gp13, gp14, gp18 and a glycoprotein at 120K (EHV-1) or 116K (EHV-4) were all important immunogens in mares following EHV-1-induced abortion, and in a specific pathogen-free foal experimentally infected with EHV-1 and later cross-challenged with EHV-4. Gp2, gp14 and gp18 were the major immunogens in the donkey in response to AHV-3 infection. The type specificity associated with these glycoproteins was also examined and it was found that although most if not all contain type-specific epitopes, gp2 and a glycoprotein at 120K, and to a lesser extent gp13 and gp18, were significantly type-specific in the serum from a mare following natural EHV-1 infection and abortion.

Herpes simplex virus glycoprotein B (HSVgB) and its bovine herpesvirus homologue (BHVgB) share similar primary structures. These glycoproteins are present in the envelope of the virion and are believed to initiate infection by fusing the virus envelope with a host cell membrane. BHVgB, like the membrane-fusing glycoproteins of most enveloped viruses, is normally cleaved and is present as a disulphide-linked complex in the virus envelope and host cell membranes. HSVgB, however, remains uncleaved, presumably because it lacks a similar protease recognition sequence. To determine whether the cleavage of BHVgB is essential for its role in initiating infection, we altered the coding sequence of this glycoprotein by removing the protease cleavage site and making this region similar to that of HSVgB. The mutant BHVgB gene was expressed by an HSV recombinant virus in mouse L cells and produced an uncleaved BHVgB. The uncleaved BHVgB could complement the function of HSVgB which had been neutralized by monoclonal antibody H233. When expressed in mouse L cells, the uncleaved mutant BHVgB retained its ability to fuse membranes.

We have previously classified isolates from a respiratory syncytial (RS) virus epidemic into distinct lineages by restriction mapping and nucleotide sequencing of parts of the nucleocapsid protein and small hydrophobic protein genes, which are areas of the genome not considered to be under immunological pressure. This study has now been extended by the determination of the nucleotide sequences of the attachment (G) protein genes of isolates from each subgroup A lineage. Deduced amino acid identities of the G proteins ranged between 80% and 99%, corresponding closely to the previously determined relatedness of the lineages. The amino acid variability was not evenly distributed; in the extracellular part of the protein there was a sharply defined hypervariable domain which was separated from a more extended variable domain by a highly conserved region. Most nucleotide changes in the variable domains were in the first and second positions of the codon triplets. These results suggest that there may be considerable immunological pressure for change in certain areas of the G protein and this may account for the ability of this virus to reinfect individuals repeatedly. The results presented here reflect the pattern of published data comparing prototype strains of the A and B subgroups.

Monocloral antibodies (MAbs) against Puumala (PUU) virus, the aetiological agent of nephropathia epidemica, were produced by fusing activated spleen cells from a bank vole (Clethrionomys glareolus) with the mouse myeloma cell line SP2/0. This novel approach, utilizing the natural vector of PUU virus for hybridoma production, proved to be highly efficient, and eight stable PUU virus-specific heterohybridomas were isolated and characterized. The bank vole MAbs were all specific for the nucleocapsid protein (N) of PUU virus, as determined by immunoprecipitation. When evaluated by additivity immunoassays, the MAbs were found to recognize several different, distinct or overlapping, epitopes on N. The MAbs were used in immunofluorescence assays to compare eight PUU-related virus isolates, and the prototype Hantaan, Urban rat and Prospect Hill viruses. The reactivity varied among the different MAbs and could be classified into five groups. One MAb reacted exclusively with PUU-related viruses; two MAbs reacted with all PUU-related virus strains tested, as well as Prospect Hill virus, but did not react with Urban rat virus and Hantaan virus; one MAb reacted with all PUU-related virus strains tested and weakly with Hantaan virus, but not with Urban rat and Prospect Hill viruses; two MAbs reacted with all the virus strains tested. Two virus strains, K-27 and CG-1820, isolated in the western U.S.S.R., were distinguished from the other PUU-related virus strains by two MAbs, suggesting that the large group of independently isolated PUU-related viruses may be more heterogeneous than previously believed.

Recently, we reported that hepatitis C virus (HCV) can be classified genetically into two types, HCV-K1 and HCV-K2, which show 67% and 71% identity at the nucleotide and amino acid sequence levels in a 340 bp region which encodes the NS5 gene Gly-Asp-Asp motif. To develop a rapid method to classify the genomes of HCV isolates, we identified restriction fragment length polymorphisms (RFLPs) in reverse transcriptase-polymerase chain reaction products encoding a portion of the NS5 gene. AluI and AccII enabled HCV to be classified into the K1 and K2 types, and Sau96I enabled classification into the K1 type, and the K2a and K2b subtypes. These RFLPs also generally allow Japanese isolates to be distinguished from the prototype (PT, an isolate from the U.S.A.), which is a K1 type. Sequence analysis of the 5′-untranslated regions of Japanese isolates revealed near identity between the K1 type and PT, and 93 to 94% identity between the K1 and K2 types, indicating that there are type K1- and K2-specific RFLPs in this region. Our results suggest that the nucleotide sequences of the K1 and K2 types are different throughout the HCV genome. The incidence of HCV types K1, K2a and K2b, and PT in 50 samples was 74%, 16%, 8% and 2%, respectively.

Using sera from hosts infected with bovine leukaemia virus (BLV), human T cell lymphoma virus types I and II (HTLV-I and -II), or simian T cell lymphoma virus type I (STLV-I), we found that the major gag proteins of these viruses cross-react immunologically. The specificity of this cross-reactivity was demonstrated by absorption using purified viral proteins, virus lysates and extracts of infected cells. The data strongly suggested that the cross-reacting epitope(s), referred to as CE, differs from those responsible for cross-reactions between the major gag proteins of HTLV-I, HTLV-II and STLV-I, and between those of BLV and HTLV-I reported previously. The prevalence of antibodies to CE was low, even amongst infected hosts with high titres to other epitopes present in the major gag proteins of the homologous viruses. CE was not detected in any of the other C- or D-type retroviruses, or lentiviruses examined. Therefore, it is likely that CE can be used to define serologically a subgroup of C-type retroviruses, the genomes of which display unique features and functional activities.

The genomic DNA of a vaccine strain of canine adenovirus type 2 (Vaxitas; ICI Tasman) has been shown to contain two copies of the E1a region, the second being at the far right end of the genome. DNA sequence analysis of the right terminal 2.8 kbp of this vaccine strain showed that numerous point mutations have occurred in the second copy, which would preclude the synthesis of any functional products. However, expression vectors in which the E1a promoter from the right terminus were linked to the chloramphenicol acetyltransferase gene showed that the promoter was fully functional. Furthermore, the activity of the reiterated E1a promoter was considerably greater than that of the normal E4 promoter. This dramatic change in the regulation of E4 expression may be an important factor in determining the altered host cell specificity displayed by this vaccine strain virus.

In this study, overlapping cDNA clones covering the entire S RNA molecule of Junin virus, an arenavirus that causes Argentine haemorrhagic fever, were generated. The complete sequence of this 3400 nucleotide RNA was determined using the dideoxynucleotide chain termination method. The nucleocapsid protein (N) and the glycoprotein precursor (GPC) genes were identified as two non-overlapping open reading frames of opposite polarity, encoding primary translation products of 564 and 481 amino acids, respectively. Intracellular processing of the latter yields the glycoproteins found in the viral envelope. Comparison of the Junin virus N protein with the homologous proteins of other arenaviruses indicated that amino acid sequences are conserved, the identity ranging from 46 to 76%. The N-terminal half of GPC exhibits an even higher degree of conservation (54 to 82%), whereas the C-terminal half is less conserved (21 to 50%). In all comparisons the highest level of amino acid sequence identity was seen when Junin virus and Tacaribe virus sequences were aligned. The nucleotide sequence at the 5′ end of Junin virus S RNA is not identical to that determined of the other sequenced arenaviruses. However, it is complementary to the 3′-terminal sequences and may form a very stable panhandle structure (ΔG -242.7 kJ/mol) involving the complete non-coding regions upstream from both the N and GPC genes. In addition, a distinct secondary structure was identified in the intergenic region, downstream from the coding sequences; Junin virus S RNA shows a potential secondary structure consisting of two hairpin loops (ΔG -163.2 and -239.3 kJ/mol) instead of the single hairpin loop that is usually found in other arenaviruses. The analysis of the arenavirus S RNA nucleotide sequences and their encoded products is discussed in relation to structure and function.

Inhibition of vesicular stomatitis virus (VSV) replication in LB cells by interferon (IFN) resembles the action of IFN on some retroviruses, in that the incorporation of glycoprotein into virions is defective. Primary amines added between 1 and 2 h post-infection significantly enhanced (five- to 1000-fold) the antiviral activity of IFN against VSV, but no enhancement of the antiviral activity of IFN against encephalomyocarditis virus, a virus with no membrane component, by primary amines was seen. SDS-PAGE and immunofluorescence analysis of viral proteins, and Nycodenz gradient fractionation, suggested that both IFN and primary amines inhibited the transport of VSV glycoprotein (G) to the plasma membrane; instead, G accumulated in the trans-Golgi network (TGN). Using sensitive intracellular pH (pHi) indicators, we found that IFN treatment significantly raised the pHi. A further increase in pHi was seen with a combination of IFN and primary amines; the increase in pHi correlated with an enhancement of the antiviral activity of IFN by primary amines. Amiloride inhibited the IFN-induced increase in pHi and a concomitant increase in the concentration of Na+ ions; this observation suggested that IFN induced cytoplasmic alkalinization by activating an Na+/H+ antiporter system. These results indicated that the IFN-induced increase in pHi may be responsible for the accumulation of G in the TGN, thereby producing G-deficient virus particles with reduced infectivity.

The thermal denaturation of poliovirus virions and procapsids in the 42 to 48 °C range was studied using N- and H-specific monoclonal antibodies. The half-life of the N antigen of Mahoney and Sabin 1 virions was extended 50- to 250-fold by either 10 µm WIN 51711 or R 78206. The minimum concentrations required for full stabilization at 46 °C (1.0 µm for WIN 51711, 0.2 µm for R 78206) were independent of the strain or serotype of the virus; 30 to 60 molecules of stabilizer per virion were required for full protection. R 78206 was the most efficient stabilizer of Mahoney procapsids; the half-life of the N-specific epitopes of these particles at 44 °C was extended from less than 1 min to 1 day.

The range of hepatitis A virus (HAV) particles generated during persistent infection of different cell lines was studied. Buoyant density and sedimentation analyses of cell extracts revealed a uniform profile of particles in all cell lines analysed except for BS-C-1 cells. The virion itself usually represented less than 50% of the total mass of virus antigen. A major portion of the antigen was associated with non-infectious, empty particles, which banded at 1.305 g/ml and 1.20 g/ml CsCl, and sedimented in sucrose gradients at 76S and 59S. Empty HAV particles were similar to those of poliovirus with respect to their physical stability and had the characteristic capsid protein content (VP0, VP1 and VP3). An additional RNA-containing particle, probably the provirion, represented only a minor species characterized by a buoyant density of 1.32 g/ml in CsCl and sedimenting at 130S.

A recombinant vaccinia virus containing most of the P1 region of hepatitis A virus (HAV) was constructed. Cell lysates of cultures infected with the virus contained HAV proteins detectable by radioimmunoassay. Western blot analysis revealed the presence of a single protein of M r 60K to 62K, bearing epitopes from structural polypeptides VP4, -3 and -2, and the N terminus of VP1. The size of the protein suggests that at least some of the vaccinia virus thymidine kinase is also expressed. Inoculation of tamarin monkeys with the recombinant virus resulted in the development of a specific anti-HAV immune response which was protective against challenge with a virulent strain of HAV. Recombinant viruses expressing the above region of HAV or the proteins expressed by such viruses may be useful in the development of a vaccine suitable for use in man.

Borna disease virus (BDV) is an infectious agent that causes profound disturbances in motor function and behaviour in a wide range of animal species and possibly humans. The infectious nature of BDV has long been established, but the aetiological agent has not been isolated or classified. Recently, we have reported the isolation of BDV-specific cDNA clones using subtractive libraries constructed from mRNA from infected material. Here we describe studies on one of these cDNA clones, B8, and confirm its specificity by in situ hybridization on sections of BDV-infected brain. The complete nucleotide sequence of BDV-specific clone B8 was determined. Oligonucleotides of positive and negative polarity synthesized from sequences from the 5′ and 3′ ends, as well as the central part, of clone B8 identified both positive- and negative-strand BDV-specific RNAs in infected rat brain. All B8 sequences used as oligonucleotide probes were found to be contained in the larger positive- and negative-strand RNAs. Thus, the structure of the BDV-specific RNAs appears to be a nested set of multiple, overlapping subgenomic positive- and negative-strand RNA transcripts.

This study outlines the effects of a modification of the actin-based cytoskeleton on the maturation of rabies virus in human neuroblastoma cell and primary rat cortical neuron cultures. In a Ca2+-depleted or an EGTA-containing medium, disruption of microfilaments did not affect intracellular viral nucleoprotein synthesis, as demonstrated by dual-immunofluorescence microscopy, and caused no change in the extracellular titre of rabies virus. Furthermore, the continuous presence of the anti-calmodulin drugs trifluoperazine (1 to 20 µm) and chlorpromazine (1 to 30 µm), or the l-type Ca2+ channel antagonist nifedepine (1 to 10 µm) or the Ca2+-specific ionophore A23187 (0.05 to 1.0 µm), did not modify the extracellular titre of rabies virus significantly over a 48 h period. The inference from these studies is that the maturation of rabies virus is independent of the integrity of the microfilament structures and calmodulin-dependent processes of neuronal cells.

The localization of the blue tongue virus (BTV) non-structural proteins NS3 and NS3a has been identified using immunoelectron microscopical techniques. NS3 and NS3a have been observed in the plasma membrane of BTV- and recombinant vaccinia virus (expressing NS3)-infected cells. The NS3 protein was associated with areas of membrane perturbation. There was a good correlation between the presence of NS3 and NS3a and BTV release. The NS3 protein was associated with membrane fragments and the inability to detect it on the extracellular aspect of intact cells suggested that the protein was not exposed extracellularly. Electron microscopical and biochemical evidence suggested that fragments of plasma membrane containing NS3 and NS3a were released from infected cells. Collectively, the data indicate that NS3 and NS3a may be involved in the final stages of BTV morphogenesis, i.e. the release of BTV from infected cells.

The E4 gene of several human papillomavirus types is expressed in association with vegetative viral DNA synthesis in differentiated epidermal cells. To develop reagents to study expression of the bovine papillomavirus type 1 (BPV-1) E4 gene in warts and in virus-transformed cell lines, rabbit polyclonal antiserum was raised to the BPV-1 E4 antigen produced as a fusion polypeptide in Escherichia coli. By immunoblotting analysis of productively infected bovine fibropapilloma tissue, E4-related proteins of 16K, 21K, 30K and 42K were detected. In some but not all C127 cell lines transformed by BPV-1 or by a replication-competent BPV-1 deletion mutant, cytoplasmic E4 antigen with a predominantly perinuclear localization was detected by immunofluorescence analysis in a subpopulation of cells in stationary-phase cultures. The E4-expressing cells were identified by their grossly enlarged size to represent the same cell subpopulation shown earlier to support BPV-1 DNA amplification. The observation of synthesis of the E4 protein in association with viral DNA amplification in this system provides further evidence that there is a switch in viral early region gene expression in a subpopulation of division-arrested cells, which may accurately reflect events occurring during the vegetative phase of BPV-1 replication in terminally differentiated cells in vivo.