- Volume 83, Issue 3, 2002

Herpes simplex virus type 1 (HSV-1)-based amplicon vectors expressing hepatitis C virus (HCV) E1 and E2 glycoproteins were investigated. HSV-1 amplicon vectors carrying the E1E2p7- or E2p7-coding sequences of HCV type 1a under the control of the HSV-1 IE4 (α22/α47) promoter were constructed. Studies of infected HepG2, WRL 68 or Vero cells indicated that HSV-1-based amplicon vectors express high levels of HCV glycoproteins that are processed correctly. Immunofluorescence microscopy combined with immunoprecipitation and endoglycosidase treatment of cells infected with the HSV-1-based vectors expressing E1 and E2 showed that the two glycoproteins were retained in the endoplasmic reticulum and had the expected glycosylation patterns. Furthermore, although most of the E1 and E2 proteins formed disulfide-linked aggregates, significant amounts of monomeric forms of the two proteins were detected by SDS–PAGE under non-reducing conditions, suggesting the presence of non-covalently associated E1 and E2. Similar results were produced by a replication-competent recombinant HSV-1 vector expressing HCV E1 and E2. These results indicated that HSV-1-based amplicon vectors represent a useful expression system for the study of HCV glycoproteins.

A helper-dependent expression system based on transmissible gastroenteritis coronavirus (TGEV) has been developed using a minigenome of 3·9 kb (M39). Expression of the reporter gene β-glucuronidase (GUS) (2–8 μg per 106 cells) and the porcine respiratory and reproductive syndrome virus (PRRSV) ORF5 (1–2 μg per 106 cells) has been shown using a TGEV-derived minigenome. GUS expression levels increased about eightfold with the m.o.i. and were maintained for more than eight passages in cell culture. Nevertheless, instability of the GUS and ORF5 subgenomic mRNAs was observed from passages five and four, respectively. About a quarter of the cells in culture expressing the helper virus also produced the reporter gene as determined by studying GUS mRNA production by in situ hybridization or immunodetection to visualize the protein synthesized. Expression of GUS was detected in the lungs, but not in the gut, of swine immunized with the virus vector. Around a quarter of lung cells showing replication of the helper virus were also positive for the reporter gene. Interestingly, strong humoral immune responses to both GUS and PRRSV ORF5 were induced in swine with this virus vector. The large cloning capacity and the tissue specificity of the TGEV-derived minigenomes suggest that these virus vectors are very promising for vaccine development.

Formation of the coronavirus replication–transcription complex involves the synthesis of large polyprotein precursors that are extensively processed by virus-encoded cysteine proteases. In this study, the coding sequence of the feline infectious peritonitis virus (FIPV) main protease, 3CLpro, was determined. Comparative sequence analyses revealed that FIPV 3CLpro and other coronavirus main proteases are related most closely to the 3C-like proteases of potyviruses. The predicted active centre of the coronavirus enzymes has accepted unique replacements that were probed by extensive mutational analysis. The wild-type FIPV 3CLpro domain and 25 mutants were expressed in Escherichia coli and tested for proteolytic activity in a peptide-based assay. The data strongly suggest that, first, the FIPV 3CLpro catalytic system employs His41 and Cys144 as the principal catalytic residues. Second, the amino acids Tyr160 and His162, which are part of the conserved sequence signature Tyr160–Met161–His162 and are believed to be involved in substrate recognition, were found to be indispensable for proteolytic activity. Third, replacements of Gly83 and Asn64, which were candidates to occupy the position spatially equivalent to that of the catalytic Asp residue of chymotrypsin-like proteases, resulted in proteolytically active proteins. Surprisingly, some of the Asn64 mutants even exhibited strongly increased activities. Similar results were obtained for human coronavirus (HCoV) 3CLpro mutants in which the equivalent Asn residue (HCoV 3CLpro Asn64) was substituted. These data lead us to conclude that both the catalytic systems and substrate-binding pockets of coronavirus main proteases differ from those of other RNA virus 3C and 3C-like proteases.

The key enzyme in coronavirus replicase polyprotein processing is the coronavirus main protease, 3CLpro. The substrate specificities of five coronavirus main proteases, including the prototypic enzymes from the coronavirus groups I, II and III, were characterized. Recombinant main proteases of human coronavirus (HCoV), transmissible gastroenteritis virus (TGEV), feline infectious peritonitis virus, avian infectious bronchitis virus and mouse hepatitis virus (MHV) were tested in peptide-based trans-cleavage assays. The determination of relative rate constants for a set of corresponding HCoV, TGEV and MHV 3CLpro cleavage sites revealed a conserved ranking of these sites. Furthermore, a synthetic peptide representing the N-terminal HCoV 3CLpro cleavage site was shown to be effectively hydrolysed by noncognate main proteases. The data show that the differential cleavage kinetics of sites within pp1a/pp1ab are a conserved feature of coronavirus main proteases and lead us to predict similar processing kinetics for the replicase polyproteins of all coronaviruses.

The haemagglutinin (HA) protein of fowl plague virus A/FPV/Rostock/34 (H7N1) contains three N-linked oligosaccharide side chains in its stem domain. These stem glycans, which are attached to the Asn residues at positions 12, 28 and 478, are highly conserved throughout all HA protein sequences analysed to date. In a previous study, in which mutant HA proteins lacking individual stem glycosylation sites had been expressed from an SV-40 vector, it was shown that these glycans maintain the HA protein in the metastable form required for fusion activity. In the present study, the functional role of the stem N-glycans for virus replication was investigated using recombinant influenza viruses generated by an RNA polymerase I-based system. Studies in Madin–Darby canine kidney cells and embryonated chickens’ eggs revealed that the N-glycan at Asn12 is crucial for virus replication. In both culture systems, growth of virus lacking this glycan (mutant cg1) was completely blocked at 37 °C and inhibited at 33 °C. Loss of the glycan from Asn478 (mutant cg3) caused less striking, but still measurable, effects. Interestingly, it was not possible to generate mutant viruses containing the HA protein lacking the N-glycan at Asn28. It is concluded from this that the N-glycan at Asn28 is indispensable for the formation of replication-competent influenza viruses. When compared to viruses containing wild-type HA protein, mutants cg1 and cg3 showed a significantly decreased pH stability. Taken together, these data show that the HA stem glycans are potent regulators of influenza virus replication.

We have employed immunofluorescence microscopy and transmission electron microscopy to examine the assembly and maturation of respiratory syncytial virus (RSV) in the Vero cell line C1008. RSV matures at the apical cell surface in a filamentous form that extends from the plasma membrane. We observed that inclusion bodies containing viral ribonucleoprotein (RNP) cores predominantly appeared immediately below the plasma membrane, from where RSV filaments form during maturation at the cell surface. A comparison of mock-infected and RSV-infected cells by confocal microscopy revealed a significant change in the pattern of caveolin-1 (cav-1) fluorescence staining. Analysis by immuno-electron microscopy showed that RSV filaments formed in close proximity to cav-1 clusters at the cell surface membrane. In addition, immuno-electron microscopy showed that cav-1 was closely associated with early budding RSV. Further analysis by confocal microscopy showed that cav-1 was subsequently incorporated into the envelope of RSV filaments maturing on the host cell membrane, but was not associated with other virus structures such as the viral RNPs. Although cav-1 was incorporated into the mature virus, it was localized in clusters rather than being uniformly distributed along the length of the viral filaments. Furthermore, when RSV particles in the tissue culture medium from infected cells were examined by immuno-negative staining, the presence of cav-1 on the viral envelope was clearly demonstrated. Collectively, these findings show that cav-1 is incorporated into the envelope of mature RSV particles during egress.

Paramyxoviruses may adopt a similar fusion mechanism to other enveloped viruses, in which an anti-parallel six-helix bundle structure is formed post-fusion in the heptad repeat (HR) regions of the envelope fusion protein. In order to understand the fusion mechanism and identify fusion inhibitors of Newcastle disease virus (NDV), a member of the Paramyxoviridae family, we have developed an E. coli system that separately expresses the F protein HR1 and HR2 regions as GST fusion proteins. The purified cleaved HR1 and HR2 have subsequently been assembled into a stable six-helix bundle heterotrimer complex. Furthermore, both the GST fusion protein and the cleaved HR2 show virus–cell fusion inhibition activity (IC50 of 1·07–2·93 μM). The solubility of the GST–HR2 fusion protein is much higher than that of the corresponding peptide. Hence this provides a plausible method for large-scale production of HR peptides as virus fusion inhibitors.

The African green monkey (AGM) model system for simian immunodeficiency virus (SIVagm) has been used to examine why prolonged infection with the relevant virus does not result in the development of immunodeficiency in its natural host. Blood lymphocyte subset values were determined in uninfected (n=88) and naturally SIVagm-infected AGMs (n=74). A number of blood cell subsets, such as CD8α+CD3+CD28neg, CD8α+CD3neg and CD20+ cells, were expanded significantly in clinically asymptomatic animals carrying a relatively high plasma load of viral RNA (104–107 RNA copies/ml plasma). The expanded CD8α+CD3+CD28neg subpopulation (1094±986 cells/μl blood in infected animals versus 402±364 cells/μl blood, P=0·03) comprised cells that resembled terminally differentiated effector CD8 T cells (CD27neg and CD11a+). In SIVagm-infected animals, the expanded CD8α+CD3neg cell subset shared identity with the CD16+ population (natural killer cells). These results demonstrate for the first time that apathogenic SIVagm infection causes significant changes in the immune system of its natural host. Although previous studies had indicated that noncytotoxic mechanisms might play an important role in the suppression of virus replication in the natural host of SIVagm, this study sheds new light on the possible role of cytotoxic T lymphocytes, the innate immune system and double-positive T helper cells (CD4+CD8α+CD3+) in suppressing virus replication in this animal model of AIDS.

Tax, the trans-activator of human T-lymphotropic virus type I (HTLV-I), is the dominant target antigen for cytotoxic T lymphocytes (CTLs) in the majority of infected individuals, although the reason for this immunodominance is not clear. Tax has been shown to associate physically with the proteasome, a protease that is responsible for the generation of the majority of major histocompatibility complex (MHC) class I ligands recognized by CTLs. This association could lead to the preferential targeting of Tax to the MHC class I pathway and account for its high immunogenicity. Here, the CTL response to Tax was investigated in mice by priming with a Tax expression vector and boosting with a Tax recombinant vaccinia virus (modified vaccinia virus Ankara strain). This approach led to the identification of a new H-2Dk-restricted epitope in Tax, amino acid residues 38–46, sequence ARLHRHALL. Surprisingly, presentation of this epitope was found to be enhanced by the proteasome inhibitor lactacystin, although Tax was shown to associate with proteasomes in murine cells. The difficulties encountered in generating Tax-specific CTL responses and the results of enzyme-linked immunospot (ELISpot) analysis suggested that Tax is only poorly immunogenic for CTLs in mice. Therefore, the immunodominance of Tax in human CTL responses to HTLV-I is probably not due to an intrinsic property of the protein itself, such as an association with the proteasome, but instead may result from the fact that Tax is the predominant protein synthesized early after infection.

The adeno-associated virus type 2 (AAV-2) Rep proteins are essential for AAV DNA replication and regulation of AAV gene expression. We have identified a cellular protein interacting with Rep78 and Rep68 in yeast two-hybrid analysis and in GST pull-down assays. This protein has recently been described as both a p53 (p53BP3) and a topoisomerase I interacting protein (Topors). It contains an arginine/serine-rich domain, a RING finger domain and five PEST sequences. A minimal sequence sufficient for interaction with Rep was mapped to Topors amino acids 871 to 917. We show that the same region is also involved in the interaction with p53. Rep sequences involved in interaction with Topors were mapped to Rep amino acids 172 to 481. Overexpression of Topors stimulated AAV gene expression in the absence of helper virus, suggesting a function of Topors as a transcriptional regulator.

Adenovirus type 5 E1A proteins interact with cellular regulators of transcription to reprogram gene expression in the infected or transformed cell. Although E1A also interacts with DNA directly in vitro, it is not clear how this relates to its function in vivo. The N-terminal conserved regions 1, 2 and 3 and the C-terminal portions of E1A were prepared as purified recombinant proteins and analyses showed that only the C-terminal region bound DNA in vitro. Deletion of E1A amino acids 201–220 inhibited binding and a minimal fragment encompassing amino acids 201–218 of E1A was sufficient for binding single- and double-stranded DNA. This portion of E1A also bound the cation-exchange resins cellulose phosphate and carboxymethyl Sepharose. As this region contains six basic amino acids, in vitro binding of E1A to DNA probably results from an ionic interaction with the phosphodiester backbone of DNA. Studies in Saccharomyces cerevisiae have shown that expression of a strong transcriptional activation domain fused to a DNA-binding domain can inhibit growth. Although fusion of the C-terminal region of E1A to a strong transcriptional activation domain inhibited growth when expressed in yeast, this was not mediated by the DNA-binding domain identified in vitro. These data suggest that E1A does not bind DNA in vivo.

Open reading frame (ORF) 71 genes of both equine herpesvirus-1 (EHV-1) and EHV-4 encode a unique glycoprotein, which has been described to vary in molecular mass from 200 to 450 kDa. Using PCR and nucleotide sequence analysis, it was shown that the ORF 71 genes of EHV-1 and EHV-4 are polymorphic due to a variable number of reiterated sequences in two regions, designated regions A and B. Region A was threonine-rich and was located near the N terminus. Region B comprised a 38 amino acid repeat near the C terminus that expanded following cell culture adaptation. Western blot analysis of viruses showed that EHV-4 gp2 was modified by glycosylation and that variation in region A resulted in the marked differences in the molecular mass of EHV-4 gp2.

Ovine herpesvirus-2 (OvHV-2), a member of the gammaherpesviruses (genus Rhadinovirus), asymptomatically infects its natural host, the sheep, but causes malignant catarrhal fever (MCF) in susceptible hosts, such as cattle, deer and pigs. A permissive cell culture system for virus replication has not been identified but viral DNA is present within lymphoblastoid cell lines (LCLs) established from cases of MCF. During this study, a cDNA expression library generated from LCLs was screened with sheep sera and two cDNAs were isolated. One cDNA contained two open reading frames (ORFs) that show similarity to ORFs 58 and 59 of alcelaphine herpesvirus-1 (AlHV-1), a closely related gammaherpesvirus that also causes MCF. Both ORFs 58 and 59 are conserved throughout the gammaherpesviruses. ORF 58 is predicted to be a membrane protein, while ORF 59 has been shown to be an early lytic gene that functions as a DNA polymerase processivity factor. The second cDNA clone contained a partial ORF showing limited similarity to AlHV-1 ORF 73, a homologue of the latency-associated nuclear antigen of human herpesvirus-8, which is associated with latent infections. The full-length OvHV-2 ORF 73 was cloned subsequently by PCR. The ORFs isolated from the library were cloned into a bacterial expression vector and the recombinant proteins tested for their reactivity to sera from OvHV-2-infected animals. An ORF 59 fusion protein was recognized specifically by sera from OvHV-2-infected cattle and will be used to develop a sero-diagnostic test.

The vaccinia virus (VV) interferon (IFN)-γ receptor (IFN-γR) is a 43 kDa soluble glycoprotein that is secreted from infected cells early during infection. Here we demonstrate that the IFN-γR from VV, cowpox virus and camelpox virus exists naturally as a homodimer, whereas the cellular IFN-γR dimerizes only upon binding the homodimeric IFN-γ. The existence of the virus protein as a dimer in the absence of ligand may provide an advantage to the virus in efficient binding and inhibition of IFN-γ in solution.

During virus infection of insect cells, the Autographa californica nucleopolyhedrovirus chitinase is localized primarily within the endoplasmic reticulum (ER), which is consistent with the presence of a carboxy-terminal ER retention motif (KDEL). Release of chitinase into the extracellular medium appears to be concomitant with terminal cell lysis, rather than by active secretion. In this study, we have shown that mutation of the KDEL motif induces a partial redistribution of the chitinase at both early and late times post-infection. Deletion of the KDEL motif or substitution with glycine residues allowed chitinase to move through the secretory pathway, accumulating to detectable levels in the extracellular medium by 24 h post-infection; more than 48 h prior to cell lysis. Deletion of the KDEL motif did not compromise enzyme activity, with the modified enzyme exhibiting characteristic endo- and exo-chitinolytic activity. Trichoplusia ni larvae infected with the modified virus were found to liquefy approximately 24 h earlier than larvae infected with a control virus in which the chitinase KDEL motif had not been deleted.