- Volume 79, Issue 11, 1998

The surface glycoprotein G is the major neutralizing and protective antigen of bovine ephemeral fever rhabdovirus (BEFV). Twelve neutralizing MAbs against BEFV strain BB7721 were used to select 33 neutralization escape mutants. The mutants had been classified previously into three major antigenic sites (G1-G3) based on their cross-neutralization patterns. The nucleotide sequence of the entire extracellular domain of the G protein gene was determined for all mutants. Each contained a single nucleotide change leading to a single amino acid substitution. The 16 mutants assigned to the linear antigenic site G1 mapped to aa 487-503 of the 623 aa G protein. Results of antibody binding to several overlapping octapeptides covering this region mapped the sequence of two common minimal B cell epitopes recognized by the five G1 MAbs to (488)EEDE(491) and (499)NPHE(502). Site G2 mutations mapped either at aa 169 or 187. The 12 mutants representing antigenic site G3 (G3a and G3b) mapped to aa 49, 57, 218, 229 and 265, indicating that this site is likely to combine complex discontinuous epitopes. Comparison of the deduced amino acid sequence from five BEFV field isolates and BB7721 identified aa 218 to be critical for the site G3a neutralization. Alignment of the glycoproteins of rabies virus, vesicular stomatitis Indiana virus, vesicular stomatitis New Jersey virus, infectious haematopoietic necrosis virus and BEFV revealed similarities in the location of the neutralizing epitopes and extensive conservation of cysteine residues, suggesting that basic elements of the folded structure of these glycoproteins are preserved.

H-2d mice are resistant to measles virus-induced encephalitis (MVE) and develop Ld-restricted CD8 T cells which lyse target cells infected with measles virus or with a vaccinia virus recombinant expressing the nucleocapsid protein of measles virus (vvN). In contrast, H-2k mice are susceptible to MVE and generate CD8 T cells which lyse target cells infected with vvN, but not those infected with MV. We were able to demonstrate that this difference is not due to a defect in the antigen processing machinery, but that Kk molecules require 100-fold more peptide to sensitize target cells for lysis by CTL. vvN replicates well in target cells and therefore enhances the level of epitope peptide available for CTL recognition. In contrast, MV infection is abortive in mouse cells and low levels of epitope peptide are produced. As Ld requires 100-fold less peptide than Kk to sensitize target cells for lysis, the low level of epitope peptide is enough to induce lysis by CD8 T cells, whereas for recognition via Kk, increased synthesis of protein is required. We propose that the differences in peptide binding between the two H-2 molecules will have consequences for the kinetics of the generation of CD8 T cells as well as the absolute numbers of CD8 T cells generated.

Natural killer (NK) cells are recruited locally during the initial phases of virus infection and produce cytokines which may affect the subsequent emergence of specific T cells. In this study, cellular responses to primary respiratory syncytial virus (RSV) infection and after vaccination with individual viral proteins were investigated in BALB/c mice using the new NK cell antibody, DX5. Purified DX5 cells caused lysis of YAC-1 cell targets. DX5 cells did not express CD8, CD45R or MHC class II antigens. A small proportion of DX5 cells coexpressed CD4 (10·3%) and CD3 (10·6%). Of the DX5 /CD4 cells, the majority expressed the α/β T cell receptor and less than 1 % expressed the γ/δ T cell receptor. During infection with RSV, lung DX5 /CD3 NK cells peaked on day 4 of primary infection and were the most numerous subset producing IFN-γ, as determined by intracellular staining, at this time-point. Less than 1% of the DX5 cells secreting IFN-γ were CD4 . In the lungs of mice vaccinated with recombinant vaccinia virus expressing individual RSV proteins, increased NK cell cytotoxicity and IFN-γ production correlated with increased numbers of CD8 T cells. Mice with few NK cells subsequently had low CD8 T cells and developed lung eosinophilia. IFN-γ-producing NK cells therefore form a substantial component of the early cellular response to virus infection with important potential influences on the subsequent development of specific immunity.

Puumala (PUU) hantavirus is the aetiological agent of nephropathia epidemica (NE), a mild form of haemorrhagic fever with renal syndrome, which occurs in Fennoscandia, central Europe and Russia. In Norway, NE-like disease has been reported since 1946 and about 50 cases are diagnosed annually; however, the causative agent has not been characterized. In this study, a virus originating from bank voles (Clethrionomys glareolus) trapped near the town of Eidsvoll (Akershus county) was isolated and passaged in laboratory-bred bank voles. The bank vole strain was identified as a PUU virus by serological typing and by sequence analysis of the S and M gene segments. For comparison, complete or partial S sequences were determined for wild-type PUU strains from five locations in Sweden, two inhabited by the southern variant of bank vole present in Fennoscandia, and three by the northern variant. Phylogenetic analysis showed that Norwegian PUU strains are clustered together with Swedish strains from the first group forming a well- supported sublineage within the PUU genotype, distinct from other sublineages from northern Sweden, Finland, Russia and France. The results are consistent with the view of a complex evolutionary history of PUU strains in post-glacial Fennoscandia. Analyses of the current collection of nucleotide sequences suggest that PUU is the most variable genotype of the known hantaviruses.

In this study the L segment and the 5′ and 3′ termini of the S, M and L segments of the prototype Tula hantavirus (TUL) were sequenced, thus completing the first determination of the genome sequence of a hantavirus that has not been linked to any human disease. The TUL L segment comprises 6541 nt with one ORF of 6459 nt in the antigenome sense. This ORF potentially encodes a 2153 aa protein with a predicted molecular mass of 247 kDa. The amino acid sequence includes all the motifs conserved in RNA-dependent RNA polymerases. The 5′ termini of all three genome RNAs (vRNAs) had the expected sequences conserved in hantaviruses. The 3′ termini of M vRNAs were also conserved. However, the 3′ termini of S and L vRNAs were heterogeneous as most of the sequenced 3′ termini had either deletions of 1 to 22 nt or an extra 1 to 3 nt. No increase in the level of heterogeneity was seen in vRNAs of virions collected 3, 6, 9 and 12 days postinfection, suggesting that the heterogeneity already exists at the early stages of infection. The S and L vRNAs from infected cells had more truncated 3′ termini than vRNAs from pelleted virus. Heterogeneity of the 3′ termini of genome RNAs could decrease the efficiency of antigenome and mRNA syntheses and contribute to the slow growth observed for TUL and other hantaviruses in cell culture.

Comparative sequence analysis of different isolates of hepatitis G virus (HGV) has demonstrated significant intersubject genetic heterogeneity, but few data on intrasubject genetic evolution have been reported. To further investigate the genetic diversification of the HGV genome, 36 plasma samples from eleven patients chronically infected with HGV serially obtained 2–4 years apart were analysed. We determined the viral nucleotide sequence of the 5′ non-coding (NC) and the NS3 regions by directly sequencing the RT-PCR amplified products obtained from the viral RNAs. Intrasubject sequence variation was found to be 1·3–2·4 × 10 3 base substitutions per genome site per year within the 5′ NC region and 1·3–9·4 × 10 3 base substitutions per genome site per year within the NS3 region. Depending on the genomic region analysed (i.e. 5′ NC or NS3 region), pairwise comparisons and phylogenetic reconstructions showed that intersubject genetic distances were 17·5- to 20·8-fold greater than intrasubject ones. Overall, the evolution rate of HGV in the regions analysed is not significantly different from that found in hepatitis C virus.

The three flavivirus glycoproteins prM, E and NS1 are formed by post-translational cleavage and are glycosylated by the addition of N-linked glycans. NS1 may form homodimers, whereas E may form homodimers, homotrimers or heterodimers (prM-E). Modification of these processes by mutagenesis of the proteins has the potential to generate viruses that are restricted in growth and are possible vaccine candidates. Using an SV40-based expression system, we previously analysed dimerization and secretion of the NS1 protein of dengue virus type 2 (DEN-2) with mutations in the conserved Cys residues, or within hydrophilic or hydrophobic regions, or at glycosylation sites. In this study, mutations which reduce cleavage at the DEN-2 prM/E signalase cleavage site are described. On the basis of earlier and current results with transient expression, six mutations which reduced NS1 dimerization and two mutations which inhibited prM/E cleavage were analysed individually for their effects on virus growth using a genomic length cDNA clone. Two viruses were obtained that showed reduced growth in cell culture and attenuation of neurovirulence when inoculated into 3-day-old mice. One of these viruses encoded NS1 that lacked the second glycosylation site, the other encoded a Ser → Ile change at the -3 position of the prM/E cleavage site. A third virus encoding a mutation in NS1 within a hydrophilic region grew as well as the parental virus. No virus was detected for the remaining five mutations.

Picornaviruses have been divided into five genera until recently, when a sixth genus, Parechovirus, was defined. Human parechovirus type 1 (HPeV1; formerly echovirus 22) was the first recognized member of this genus and preliminary sequence analysis of echovirus 23 [now renamed human parechovirus type 2 (HPeV2)] suggested that it is also a parechovirus. Here we describe the complete nucleotide and predicted amino acid sequences of HPeV2, which indicate a close relationship to HPeV1 throughout the genome. Sequence covariance in the 5′ untranslated region allows a prediction of the secondary structure, which indicates that these parechoviruses have a type 2 internal ribosome entry site, most closely related to that of cardio-viruses. Overall, HPeV2 has 87·9% amino acid identity with HPeV1, most divergence being seen in regions of the capsid proteins that probably define antigenic sites. The N-terminal sequence extension to VP3, seen only in parechoviruses, is highly basic in both viruses, but has a variable sequence, suggesting that it does not have a sequence-specific role. There is an RGD motif near the C terminus of VP1, in an analogous location to that in HPeV1 which is believed to be functionally significant. The results confirm that both viruses are parechoviruses and give insights into the molecular features of this genus.

The poliovirus 5′ untranslated region (5′ UTR) confers on mRNAs the capacity to be translated by internal initiation. The functionality of this RNA motif has been tested in yeast cells (Saccharomyces cerevisiae) using luciferase (luc) as a reporter gene. Although some luciferase is synthesized from luc mRNA containing the poliovirus 5′ UTR (Leader-luc mRNA), much more luciferase is synthesized in cells that express luc mRNA devoid of the poliovirus 5′ UTR. Since poliovirus 2Apro enhances the translation of Leader-luc mRNAs after eIF-4G cleavage in mammalian cells, yeast cells were produced that synthesize three heterologous proteins, luciferase, poliovirus 2Apro and human eIF-4G. Initially, S. cerevisiae cells constitutively expressing human eIF-4G were isolated. The human eIF-4G gene does not complement yeast cells defective in the initiation factor counterpart, p150, indicating that the human and yeast eIF-4G are not interchangeable. Expression of poliovirus 2Apro in an inducible manner does not affect p150, but led to the efficient cleavage of human eIF-4G in yeast cells. Induction of 2Apro was detrimental to luciferase synthesis either from luc mRNA or Leader-luc mRNA irrespective of the presence or absence of human eIF-4G. 2Apro blocked luciferase expression at the transcriptional level. Finally, the effects of 16 point mutations of poliovirus 2Apro on luciferase expression and human eIF-4G cleavage were analysed. Only those 2Apro variants that generate viable polioviruses actively cleave eIF-4G in yeast.

We examined the antibody responses and protection to a human rotavirus (HRV) in gnotobiotic (Gn) pigs. Pigs were perorally (PO) inoculated with attenuated (group 1), virulent (group 2), or inactivated (group 3) Wa (P1A[8]G1) HRV. Afourth group was inoculated intramuscularly (IM) with inactivated Wa HRV in adjuvant. After PO challenge with virulent Wa HRV at post-inoculation day 21, most group 1,3 and 4 pigs shed virus and developed diarrhoea, whereas this occurred in only a few group 2 pigs. Antibodies to HRV (IgM, IgA or IgG) were detected in serum and intestinal contents of pigs of all groups after virus inoculation or challenge, and the antibody titres in intestinal contents, although lower, showed similar kinetics to the serum responses. There was no correlation between protection and neutralizing antibody titres of serum or intestinal contents, but a positive correlation existed between protection and serum IgA, intestinal IgA and intestinal IgG antibody titres. These findings suggest that serum IgA antibodies to HRV could act as an indicator of IgA antibodies in the intestine after rotavirus infection. The virulent HRV elicited protective immunity and higher levels of serum and intestinal IgA antibodies to HRV compared to attenuated and inactivated HRV. These findings suggest that more efficient mucosal delivery systems and/or adjuvants are needed to enhance the intestinal immune responses to attenuated or inactivated HRV if more successful vaccination is to be achieved in neonates.

Most animal rotaviruses bind to a cell surface molecule that contains sialic acid (SA). We have recently isolated variants from simian rotavirus RRV which show an SA-independent phenotype. The VP4 protein of these variants was shown to have three amino acid changes with respect to the wt protein, one of them being Tyr-267 → Cys. In this work, we have investigated whether the new cysteine could interfere with the disulfide bond (Cys-318/Cys-380) present in the VP5 * subunit of the wt protein. Cysteine residues 318 and 380 were mutagenized in gpr8 and RRV VP4 genes, and the wt and mutant genes were transcribed and translated in vitro. The protein products were analysed by electrophoresis under reducing and non-reducing conditions. This approach showed that, in the VP4 protein synthesized in vitro, Cys- 267 is capable of forming an alternate disulfide bond with Cys-318. This alternate bond also seems to occur in the VP4 protein present in the variant gpr8 virus particles.

We have previously shown that a number of isoforms of the non-structural rotavirus protein NSP5 are found in virus-infected cells. These isoforms differ in their level of phosphorylation which, at least in part, appears to occur through autophosphorylation. NSP5 co-localizes with another non-structural protein, NSP2, in the viroplasms of infected cells where virus replication takes place. We now show that NSP5 can be chemically cross-linked in living cells with the viral polymerase VP1 and NSP2. Interaction of NSP5 with NSP2 was also demonstrated by co- immunoprecipitation of NSP2 and NSP5 from extracts of UV-treated rotavirus-infected cells. In addition, in transient transfection assays, NSP5 phosphorylation in vivo was enhanced by coexpression of NSP2. An NSP5 C-terminal domain deletion mutant, was completely unable to be phosphorylated either in the presence or absence of NSP2. However, a 33 aa N-terminal deletion mutant of NSP5 was shown to become hyperphosphoryl- ated in vivo and to be insensitive to NSP2 activation, suggesting a regulatory role for this domain in NSP5 phosphorylation and making it a candidate for the interaction with NSP2. These mutants also allow a preliminary mapping of NSP5 autophosphorylation activity.

Combined androgen and oestrogen treatment of male or female Syrian hamsters results via an unknown mechanism in the formation of leiomyosarcomas in the reproductive tract. We have examined the possibility that retroviral gene expression may play a role in tumorigenesis. Evidence of virus-like particles in epididymis and seminal fluid is shown in electron micrographs. We identified expressed retroviral sequences by using RT-PCRto amplify a conserved retroviral reverse transcriptase coding region in RNA isolated from epididymis, testis, clarified seminal fluid and uterus. Phylogenetic analysis allowed us to classify the sequences into two distinct groups: (1) mammalian type-C viruses, having similarity to Moloney murine leu kaemia virus, feline leukaemia virus and gibbon ape leukaemia virus amongst others; (2) a mixed ABCD group containing, for example, Chinese hamster and murine intracisternal A-particle virus sequences, mouse mammary tumour virus and human and simian retroviral sequences. The presence of putative full-length retrovirus related to mammalian type-C viruses in the epididymis and uterus was confirmed by Northern blot analysis. However, steroid treatment did not alter retroviral RNA levels in the epididymis or in a uterine tumour relative to untreated uterus. In summary, Syrian hamster reproductive tissues were found to express unique retroviral sequences; however, their role, if any, in hormonal carcinogenesis remains unresolved.

To investigate the origin and dissemination of human T-cell lymphotropic virus type I in Latin America, we performed phylogenetic analysis on the LTR and env sequences of 13 HTLV-I isolates from Peruvians of four different ethnic groups: blacks and some mulattos of African origin; Quechuas of Inca origin; Nikkei of Japanese descendance; and Mestizos, a mixed population of white and Indian origin. All Peruvian samples could be situated within the cosmopolitan subtype HTLV-Ia, yet one sample showed an indeterminate Western blot pattern, lacking reactivity towards the HTLV-I type specific MTA1 peptide. Within the LTR, we could confirm the previously reported subdivision into four subgroups - one big transcontinental clade A, a Japanese clade B, a West African/Caribbean clade C and a North African clade D - and we identified a new separate subgroup E of black Peruvian strains. The clustering of the Peruvian samples seemed to depend on the ethnic origin of the host. The largest heterogeneity was observed in the black Peruvian samples. The mitochondrial DNA type of one of these black Peruvian strains of subgroup E was identical to that of West African source populations of the slave trade. Both findings support the idea of multiple post-Columbian introductions of African HTLV-Ia strains into the black Latin American population. Additionally, a tight cluster of Nikkei and Japanese samples implied a separate and rather recent transmission of a Japanese lineage of HTLV-I into Peru. A well-supported cluster of Latin American strains (including Peruvian Quechuas and Colombian Amerindians) could be situated within the transcontinental group. Molecular clock analysis of the Latin American and Japanese clade resulted in an equal evolutionary rate for those strains. Along with the anthropologically documented peopling of the Americas, the analysis was more in favour of a recent (400 to 100 years ago) introduction of HTLV- Ia into the American continent rather than a Palaeolithic introduction.

Amino acids 731–752 (731PRGPDRPEGIEEEGGERD- RDRS752) of the transmembrane glycoprotein gp41 of human immunodeficiency virus type 1 (HIV-1)are conserved in most virus isolates and are controversially reported to be implicated in virus neutralization. The humoral response in infected patients against this region is poor and humans immunized with gp160 show high levels of antibodies against the peptide but poor neutralization titres. Nonetheless, several groups have succeeded in obtaining neutralizing antibodies against this sequence using different antigen-presenting systems. In order to identify the sequence(s) against which the neutralizing response was generated, we used the flock house virus (FHV) antigen-presenting system to analyse neutralizing antisera from mice immunized with a cowpea mosaic virus (CPMV) chimera expressing the 731–752 sequence. Data show that the neutralizing response is uniquely directed against a conformational epitope mapping to the ERDRD portion of this sequence, although the major antibody response, which is non-linear, and is not neutralizing, is against an IEEE epitope. These results provide an explanation for the controversy regarding the immunogenicity of this region of gp41 and suggest that this conformational epitope, in the absence of the non-neutralizing epitope, should be considered for a subunit vaccine. In addition, this study highlights the usefulness of antigen- presenting systems that preserve epitope conformation in the investigation of immune responses.

The Nef protein of simian immunodeficiency virus (SIV) is dispensable for replication in established T- cell lines but essential for high level virus replication in the adult host, though the mechanism by which Nef contributes to this has remained unclear. We demonstrate here that SIV Nef binds to the ζ chain of the T-cell receptor (TCR). SIV Nef proteins that interact with TCR ζ in a yeast two-hybrid system also reduce T-cell surface expression levels of TCR αβ, CD3 and CD4. These findings are the first demonstration that Nef can bind directly to a component of the TCR-CD3 complex and modulate its surface expression.

Feline immunodeficiency virus (FIV) is more readily isolated from CD8 T cell-depleted peripheral blood mononuclear cells (PBMC) of FIV-infected cats than from unfractionated PBMC cultures. However, it is not known whether feline CD8 T cells down- regulate FIV expression by direct interaction with FIV-infected cells or via a soluble mediator. Furthermore, it is not known whether this anti-FIV activity involves a lytic or non-lytic mechanism. In the present study, we demonstrated that autologous and allogeneic CD8 T cells from asympto-matic FIV-infected cats inhibited the replication of FIV in CD8 T cell-depleted PBMC cultures in a dose-dependent manner. The inhibitory effect was mediated by a non-lytic mechanism, and was not dependent on direct cell-to-cell contact: an inhibitory effect was exerted by CD8 T cells across a semi-permeable membrane, and an inhibitory activity was also present in cell-free supernatants from CD8 T cells. These results suggest that this suppressive effect is mediated, at least in part, by soluble factors produced by CD8 T cells.

The hepatitis B virus × protein (HBx) is suggested to regulate transcription by stimulation of intracellular signalling pathways. We have analysed the effects of HBx on activation of the MAP kinase (Erk) and JNK/SAPK signalling pathways and confirm a stimulation of the Erk/MAP kinase in quiescent cells. However, a substantial Erk-independent activation of AP-1, and phosphorylation of c-Jun (serine-63), but not Erk-2, was induced by HBx in dividing, serum-maintained cells. These data suggest that HBx promiscuously activates Erk and JNK responsive pathways and that its overall effect on signalling may be influenced by external mitogenic stimuli.

A fraction of the large surface protein (L) of duck hepatitis B virus (DHBV) is phosphorylated at serine or threonine residues (E. Grgacic & D. Anderson, Journal of Virology 68,7344–7350, 1994). We now report the identification of phosphorylation sites in DHBV L protein. Using site-directed mutagenesis, we have identified serine-118 (S118) as the major phosphorylation site, accepting approximately 64% of the total phosphate groups incorporated in L, and resulting in retarded migration of phos- phorylated L in SDS-PAGE. Proline-119 is indispensable for S118 phosphorylation. Mutation of other serine/threonine residues which are followed by prolines (T79, T89, S117 and T155) together with S118 further reduced phosphorylation to around 19% of wild-type. Non-equilibrium pH gel electrophoresis (NEPHGE) and SDS-PAGE of 33P- labelled L protein revealed two phosphorylated L species, while protein with the S118 to alanine mutation was detected as only one labelled species, consistent with multiple phosphorylations in wild- type L. Together, these results demonstrate that serine 118 is the major phosphorylation site for a proline-directed kinase, and that a proportion of L molecules are additionally phosphorylated at one of a number of secondary sites. DHBV mutants encoding L proteins with minimal phosphorylation (alanine mutants) or mimicking constitutive phosphorylation (aspartic acid mutants) remained infectious both in cell culture and in ducks, demonstrating that L phosphorylation may play only a minor role in DHBV replication.