- Volume 75, Issue 4, 1994

The amino acid residues critical for interaction between herpes simplex virus type 1 (HSV-1) glycoprotein C (gC- 1) and cell surface heparan sulphate (HS) were localized to two separate regions within antigenic site II of this glycoprotein. These amino acids were Arg-143, Arg-145, Arg-147 and Thr-150 in one region and Gly-247 in the other. This conclusion is based on the following observations, (i) Monoclonal antibodies defining gC-1 antigenic site II, and not those reactive with antigenic site I, inhibited HSV-1-induced haemagglutination and virus binding to susceptible cells, (ii) A number of HSV- 1 mar mutants, altered at these critical residues, were impaired in attachment to cells, (iii) Synthetic peptides, corresponding to these two regions inhibited virus attachment to cells and infectivity. In addition these peptides were found to agglutinate red blood cells. This agglutination was inhibited by soluble HS, and was prevented by the pretreatment of red blood cells with heparitinase suggesting that cell surface HS was a site of peptide binding. The same was observed with the polycationic substances neomycin and poly-l-lysine. In conclusion, we propose that the regions of gC-1 represented by the HS-binding peptides may form a functional site of a polycationic nature, active in attachment to the polyanionic glycosaminoglycan chain of cell surface HS.

We demonstrate that the immediate early 175K protein (IE175K) of herpes simplex virus type 1 binds to the cap site of the latency-associated promoter (LAP) in an unusual manner. The complex formed on the LAP cap site was significantly larger than that formed on the IE175K cap site and the requirements for binding were qualitatively distinct with respect to both the primary sequence determinants at the site, and the regions of IE175K protein required for binding compared to those for the IE175K cap site. Although purified IE175K was sufficient for this larger complex formed on the LAP cap site, the DNA-binding domain was unable to bind efficiently. This contrasted strikingly with the IE175K cap site where, using precisely analogous probes, the DNA-binding domain exhibited a strong interaction. Surprisingly, from dissociation kinetics we show that binding of the intact protein to the LAP cap site is considerably more stable than the binding of IE175K to its own cap site (half-lives of the complexes 15 min and < 1 min respectively), and this was reflected in more efficient repression of LAP-driven expression than IE175K promoter-driven expression by IE175K. Moreover, primary sequence requirements for IE175K binding to the LAP cap site region differed from previously identified IE175K recognition sequences in that in addition to a partially conserved consensus sequence, neighbouring bases were necessary for binding. Although the LAP cap site exhibits a pseudopalindromic arrangement of core consensus sites, we show that this is not the basis for the higher order, more stable binding to this region. Together these results indicate that IE175K forms an unusual complex at the LAP cap site, broadening the range of previously defined sequences recognized by IE175K.

Two monoclonal antibodies, E3cD8 and E3cA10, were generated to the EBNA 3 c nuclear protein from the B95.8 isolate of Epstein-Barr virus (EBV). Both antibodies efficiently precipitate EBNA 3c from B95.8- transformed lymphoblastoid cell lines, and E3cA10 also detects EBNA 3c on Western blots. Whereas E3cD8 reacts with all 11 Type-1 isolates of EBV tested, and E3cA10 reacts with 14 of 17 Type-1 isolates, neither antibody detects the EBNA 3 c protein encoded by Type-2 isolates. E3cD8 recognizes a peptide sequence (PA/pPQAPYQGY) in a repeat region of the B95.8 EBNA 3c coding sequence which is not present in the prototype Type-2 AG876 sequence. The E3cA10 antibody epitope has been mapped to the minimal five amino acid B95.8 peptide, WAPSV, which has an alanine to valine substitution in the AG876 virus isolate. This substitution was also found in three Type-1 EBV isolates that expressed EBNA 3c proteins not detected by E3cA10. In immunoprecipitation studies E3cA10 additionally coprecipitated the EBNA 2 protein from Type-1 isolates of EBV. The possibility of a direct interaction between EBNA 2 and EBNA 3 c was ruled out by the demonstration that the antibody precipitated EBNA 2 from the Raji cell line which carries a virus with a deleted EBNA 3c gene. Since the WAPSV epitope identified in EBNA 3c is not present in EBNA 2, and no EBNA 2 linear peptide reactivity was detected in ELISA, it seems likely that E3cA10 recognizes a conformational epitope on EBNA 2. However, from the present data we cannot exclude the possibility that the antibody reacts with a cellular protein that physically associates with EBNA 2.

The production of tumour necrosis factor (TNF)-α, interleukin (IL)-l and IL-6, all proinflammatory cytokines, was investigated in radiation-immunosuppressed rats infected with rat cytomegalovirus (RCMV). At day 7 post-infection, when the animals showed disease signs, high TNF-α levels were detected in the serum and in homogenates of various organ tissues. In contrast, IL-1 and IL-6 levels were not significantly elevated. Moreover, replication of RCMV induced TNF-α expression in different types of cells grown in vitro. When frozen tissue sections were examined by immunohistology, TNF-α-producing cells were found in areas with extensive pathology in the lungs, spleen and liver. Both lymphocytes and RCMV-infected cells were identified as the sources of TNF-α. Its abundance in RCMV-infected rats suggests an important role for TNF-α in CMV pathogenesis.

In addition to major proteins of 19K and 55K (176 and 496 residues, 176R and 496R, respectively), early region 1B (E1B) of human adenovirus type 5 (Ad5) is predicted to encode at least three other polypeptides of 156R, 93R and 84R that share 79 amino-terminal residues with 496R. We have used a series of specific antipeptide sera to identify and partially characterize these proteins. 84R was produced in large amounts, 156R somewhat less, and 93R at very low levels. Synthesis of 176R, 496R, as well as the E2A 72K DNA-binding protein commenced shortly after that of E1A proteins in Ad5-infected KB cells. Production of 156R, 93R and 84R began somewhat later, but prior to the synthesis of the late structural protein IX and hexon. 156R, which is composed of the 79 amino-terminal and 77 carboxy-terminal amino acids of 496R, migrated on SDS-PAGE as two species which appeared to differ by their degree of phosphorylation. 156R and 496R yielded identical tryptic phosphopeptides that contained both phosphoserine and phosphothreonine, and one of these was immunoprecipitated by a serum specific for the carboxy terminus. These results suggested that Ser-490 and/or Ser-491 as well as Thr-495 are major sites of phosphorylation in these proteins.

Several H-2 defined cell lines were examined for their ability to support infection and replication of Japanese encephalitis virus (JEV) before their use in in vitro and in vivo stimulation protocols for generating cytotoxic T lymphocytes (CTLs) against JEV. Among 11 different cell lines tested, two H-2d macrophage tumour lines (P388D1, RAW 264.7), an H-2d hybridoma (Sp2/0), an H-2KkDd neuroblastoma (Neuro 2a), and H-2k fibroblast cell line (L929) were found to support JEV infection and replication. These cell lines were used to generate anti-JEV CTLs by using in vivo immunization followed by in vitro stimulation of BALB/c mice. We observed that not only syngeneic and allogeneic infected cells but also JEV-infected xenogeneic cells could prime BALB/c mice for the generation of JEV-specific CTLs upon subsequent in vitro stimulation of splenocytes with JEV- infected syngeneic cells. Although infected xenogeneic cells were used for immunization, the anti-JEV effectors that were generated lysed infected syngeneic targets but not JEV-infected xenogeneic or allogeneic target cells in a 5 h 51Cr release assay. These anti-JEV effectors recognized syngeneic target cells infected with West Nile virus to a lesser extent and were shown to be Lyt-2.2+ T cells. The results of unlabelled cold target competition studies suggested alterations in the cell surface expression of viral antigenic determinants recognized by these CTLs. We further demonstrate that the JEV- specific CTLs generated could virtually block the release of infectious virus particles from infected P388D1 and Neuro 2a cells in vitro.

The nucleoprotein (NP) of Marburg virus (MBG), a filovirus, is encoded by the gene closest to the 3′ end of the non-segmented negative-strand RNA genome. Sequence comparison has indicated that NP is the functional equivalent to the nucleoproteins of paramyxoviruses and rhabdoviruses. Expression of recombinant NP in two eukaryotic systems using vaccinia virus and baculovirus (vectors pSCll and pAcYMBl, respectively) and analysis of MBG-specific proteins have demonstrated that the NP of MBG is phosphorylated. The NP appeared in two forms differing in Mr by about 2K (94K and 92K respectively). Dephosphorylation clearly demonstrated that the 94K form is phosphorylated whereas the 92K form is unphosphorylated. In virion particles NP was exclusively present in the phosphorylated form. These findings suggest that only the phosphorylated NP can form nucleocapsid complexes and interact with the genomic RNA.

To investigate the effect of persistent measles virus infection on signal transduction in cells of neuronal origin, the mouse neuroblastoma cell line NS20Y/MS, which is persistently infected with measles virus, was used. The results demonstrate an approximate 50% increase in total phosphorylation and a similar increase in protein kinase C (PKC) activity. Western blot analysis with anti-total PKC or anti-PKC-α antibodies revealed a significant increase in the level of an 80K immuno-reactive PKC in NS20Y/MS cells. Following incubation of NS20Y/MS cells with polyclonal anti-measles virus antibodies, which down-regulate the level of measles virus proteins, total and PKC-mediated phosphorylation returned to the basal level of uninfected cells. This effect was reversible and removal of the antibodies resulted in restoration of the high level of total and PKC-mediated phosphorylation. The release of infectious measles virus was strongly inhibited by incubation of NS20Y/MS cells with the PKC inhibitor, 1-(5-isoquinolinylsulphonyl)-2-methylpiperazine (H-7). These results demonstrate that measles virus induces elevation in cellular phosphorylation which is essential for measles virus production.

We have located, cloned, sequenced and characterized the ecdysteroid UDP-glucosyltransferase gene (egt) gene from the baculovirus Lymantria dispar multinucleocapsid nuclear polyhedrosis virus.(LdMNPV), which is specific for the gypsy moth (L. dispar). The egt gene from the related baculovirus Autographa californica multinucleocapsid nuclear polyhedrosis virus (AcMNPV) disrupts the hormonal balance of the host larva by galactosylating ecdysone, which prevents moulting. The location of the LdMNPV egt gene, determined by hybridization analysis using a cloned coding segment of the AcMNPV egt gene, was mapped to between 79·1 and 80·2 map units on the viral genome. This region contains an open reading frame of 1464 nucleotides capable of encoding a 55K polypeptide. This predicted protein exhibits a 42% amino acid identity with the AcMNPV egt polypeptide. Transcripts of the egt gene were analysed by Northern blot and primer extension. The egt gene is transcribed from approximately 12 to 48 h, and maximally at about 16 h post-infection. Transcription occurred in the presence of aphidicolin, a viral DNA synthesis inhibitor, but not in the presence of cycloheximide, a protein synthesis inhibitor. Therefore the LdMNPV egt gene is classified as a delayed early gene. The egt gene is transcribed in a clockwise direction with respect to the circular map, and transcription initiates at a single site. Comparisons between the two baculoviral egt proteins and mammalian UDP-glucuronosyltransferases reveal areas which are conserved between the mammalian and baculoviral genes, as well as areas that are only conserved in the viral egt proteins. The LdMNPV protein sequence appears to include a signal peptide, which would allow the protein to be secreted into the haemolymph.

A secreted form of human immunodeficiency virus type 1 (HIV-1) envelope glycoprotein (gp160s), expressed in HeLa cells from a vaccinia virus recombinant was analysed by velocity-gradient centrifugation and chemical cross-linking. We showed that gp160s existed predominantly as a dimer, but higher forms corresponding to trimers and tetramers were also found. Soluble CD4 (sCD4) and native CD4 expressed by recombinant vaccinia viruses were analysed by sucrose- gradient sedimentation alone or after complexing with gp160s. The sCD4 sedimented in sucrose gradients as a monomer, whereas after solubilization the native CD4 was in a dimeric state. Both forms of CD4 were able to form complexes when incubated with gp160s. In the case of the sCD4, the M r corresponded to a (sCD4)2- (gp160s)2 complex, whereas with CD4 the complexes were of a greater order of magnitude. HIV gpl20 was secreted into the medium in a monomeric state. With sCD4 it gave a one-to-one complex, whereas with the native CD4 high M r complexes were formed. The importance of the oligomeric state of the virus- and cell-receptor proteins are discussed regarding their avidities.

Macrophage activation resulting from phagocytosis has the potential to modulate human immunodeficiency virus (HIV) replication. We have determined the effects of phagocytosis of particulate stimuli on transcription and release of HIV. Using THP-1 and Mono Mac 6 human monocytic cell lines transfected with HIV long terminal repeat sequence chloramphenicol acetytransferase (LTR CAT) constructs we have demonstrated that phagocytosis of Mycobacterium tuberculosis enhanced HIV-1 and -2 LTR CAT expression. However phagocytosis of zymosan or inert latex beads had little or no effect on CAT expression. Enhancement of HIV LTR CAT expression was dependent upon intact NF-κB binding sites and was independent of tumour necrosis factor alpha secretion. M. tuberculosis strains of different degrees of virulence induced similar levels of enhanced CAT expression. In contrast, phagocytosis of M. tuberculosis by HIV-1-infected THP-1 cells reduced supernatant reverse transcriptase (RT) activity without suppression of p24 antigen release. Phagocytosis of zymosan granules or latex particles did not alter released RT activity. However, phagocytosis of either M. tuberculosis, zymosan granules or latex particles by HIV-1-infected human peripheral blood monocyte-derived macrophages reduced supernatant RT activity. These data indicate that phagocytosis of M. tuberculosis may enhance HIV transcription in monocytic cells although it may reduce release of intact HIV.

We have compared cytoplasmic CD4 mRNA accumulation, CD4 biosynthesis and steady-state levels of both CD4 protein and mRNA in a variety of clonal derivatives of U-937 cells, chronically infected with human immunodeficiency virus type 1IIIB (HIV-1), that express various cellular and viral phenotypes. These phenotypes included defective processing of either gpl60 or Gag-Pol, viruses with severely limited host-range, and inability to generate viral products. All clones, with the exception of the one that failed to generate viral mRNA and proteins, did not express cell surface CD4. Furthermore, each of these clones had steady-state levels of CD4 mRNA which were either equivalent to or higher than those of the parental U-937 cell line. Patterns of cytoplasmic CD4 mRNA levels resembled those of total RNA, suggesting that CD4 mRNA transport from the nucleus to the cytoplasm was unaffected by HIV-1 infection. Profiles of steady-state levels of the CD4 protein resembled those of CD4 mRNA in the UHC clones, but CD4 biosynthesis was reduced in all clones with the exception of that which failed to express viral products. This report is the first demonstration that steady-state CD4 biosynthesis is reduced in HIV-l-infected cells. In general, there was a good correlation between high levels of expression of gpl60 and reduced CD4 biosynthesis. These results suggest that HIV-1 env gene products may contribute to the observed reduction in levels of CD4 biosynthesis.

Length polymorphisms and partial nucleotide sequences were determined for the hypervariable regions, V1 to V5, of the human immunodeficiency virus type 1 (HIV-1) env gene obtained from proviral DNA of sequential peripheral blood samples, from viral RNA in plasma, and from proviral DNA obtained from different organs of individuals at autopsy. The lengths of several env regions of HIV-1 proviral DNA differed markedly when obtained from different organs of an individual. Nucleotide sequences of the hypervariable V3 region of HIV-1 obtained from different organs of one patient demonstrated distinct viral variants. Most proviral DNA sequences found in organs were also present in viral RNA obtained from plasma. The majority of HIV-1 V3 variants present in the lymph tissue could be found in the plasma viral population obtained at autopsy and in the sequential blood samples obtained before death, but were absent from the cardiac blood provirus population obtained at autopsy. However, sequence variants found in the brain proviral DNA were not detected in either plasma or the sequential blood samples. Sequence differences were observed at the apex of the V3 loop between HIV-1 variants present in sequential blood samples and in blood lymphocytes and nervous tissue, lymph tissue and plasma obtained post-mortem. The potential effect of lymph tissue on the long-term persistence of different viral variants is discussed. Virus obtained from the two sequential blood samples produced syncytia in primary cultures and was easily transmitted to the continuous JM cell line. Consensus (majority) V3 loop sequences determined for the adapted viruses demonstrated that some, but not all, sequences were represented within the in vivo viral population.

The murine AIDS (MAIDS) virus has a unique sequence in the gag p12 region, which could be responsible for MAIDS development. RNA preparations from the spleens of normal uninfected C57BL/6 mice contain a transcript hybridizing with this sequence. Levels of the transcript in the kidney of C57BL/6 mice were higher than in the spleen, liver or thymus. Although BALB/c, NFS, DBA/2 and SL murine strains also contained genomic sequences hybridizing with the MAIDS virus- specific probe, no transcript hybridizing with the probe was detected in these strains of mice. The cDNAs carrying the transcript expressed in C57BL/6 mice were molecularly cloned. The complete nucleotide sequence of the clone indicates that the transcript is one of the endogenous murine leukaemia virus-related sequences containing large deletions from the R and U5 regions of the 5′ long terminal repeat (LTR) to gagpl5, from the C- terminal region of pol p40 (integrase) to the N-terminal region of env pl5E, and many short deletions in the 3′ LTR U3 region. The nucleotide sequence in the gag pi 2 region of the transcript was closely similar to that of the MAIDS virus, but the amino acid sequence was less similar because of frameshifting, even when translated. As the MAIDS virus was isolated from C57BL/6 mice with radiation-induced leukaemia, this transcript may be the progenitor of the MAIDS virus. To determine whether the gag p12 region of the transcript contains a functional sequence, a recombinant virus was generated by replacing the gag p12 region of a replication- competent BM5eco virus with that of the endogenous transcript. The recombinant virus was replication- competent, and the p12 region of the transcript retained the functional sequence present in the BM5eco virus.

Hybridomas secreting monoclonal antibodies (MAbs) specific for the surface protein (SU) of feline immunodeficiency virus were generated. Four MAbs were obtained which could be assigned to two groups based on their neutralization and competition behaviour. Using SU protein fragments expressed in Escherichia coli the antigenic site recognized by one of the MAbs (2H11) could be mapped to the c terminus. The neutralizing MAb 1E1 did not bind to any of the SU protein fragments and was directed to a conformational epitope. Binding of the MAb 1E1 to native SU protein could be blocked with a rabbit serum raised against the SU3 fragment (amino acids 361 to 445). These data indicate that at least part of the epitope is located on this SU3 domain. In competition experiments most sera of naturally infected cats were able to inhibit binding of the MAbs. This shows the conserved and immunodominant nature of the epitopes involved.

Cells infected with vaccinia viruses expressing the equine infectious anaemia virus (EIAV) gag gene (VGag) or gag plus the 5′ pol encoding protease (VGag/PR) were evaluated with monoclonal antibody to a p26 capsid protein linear epitope (QEISKFLTD). Both recombinant viruses expressed Gag precursor protein (55K) whereas only VGag/PR expressed a detectable Gag-Pol fusion protein (82K) with a functional protease, shown by subviral particles containing processed p26. Horses inoculated with VGag/PR produced antibodies reactive with EIAV Gag proteins.

Bovine herpesvirus 1 (BHV-1) glycoprotein gill functions both as a major virus attachment protein and haemagglutinating protein. Here we constructed recombinant baculovirus incorporating the BHV-1 gill coding sequence to characterize the expression, function and immunogenicity of the glycoprotein in insect cells. The recombinant gill had an Mr of 72K and seemed to form homodimers. The gill was expressed on the surface of insect cells and a rosette formation assay demonstrated haemadsorbing activity of the glycoprotein. Antigenic authenticity of the recombinant gill was confirmed by a panel of monoclonal antibodies specific for the glycoprotein produced in mammalian cells. Antisera raised to recombinant gill neutralized the infectivity of BHV-1. These data suggest that recombinant gill produced in insect cells may be a useful immunogen in a BHV-1 vaccine.