- Volume 80, Issue 4, 1999

A chimeric long terminal repeat (LTR) containing the whole LTR of a human endogenous retrovirus-like element of the H family (HERV-H) inserted downstream of the core enhancer region of the 5′ LTR of a HERV-K retroelement was detected and sequenced in the human 19p12 locus, known to be enriched with genes encoding zinc finger proteins. Similar chimeras were also detected in human chromosomes 10 and Y in human-hamster hybrid cells containing individual human chromosomes. This finding was interpreted as evidence of transpositions of the chimera in the genome. PCR analyses detected the chimera in the genomes of chimpanzee and gorilla, but not in that of orangutan. These data demonstrate that the chimera appeared in the primate germ cells more than 10 million years ago, before divergence of the human/chimpanzee and the gorilla lineages. The combination of the two LTRs forms a new regulatory system that can be involved in nearby gene expression.

Binding sites for transcriptional regulation by the Myb protein have been identified in the long terminal repeats (LTRs) of the H-type human endogenous retrovirus-like elements (HERV-H). Transfection assays using reporter plasmids containing the luciferase gene under the control of a HERV-H LTR disclosed a sevenfold increase in promoter activity in human teratocarcinoma cells when cotransfected with an expression vector for the Myb protein. Binding sites for Myb were unambiguously identified within the LTR by both DNase I footprinting experiments and mobility shift assays using a bacterially expressed purified Myb recombinant protein. Possible roles of these Myb-responsive elements are discussed.

In order to test the hypothesis that macrophages and dendritic cells (DCs) in mucosal tissue play an important role in heterosexual transmission of human immunodeficiency virus, the replication capacities of two simian immunodeficiency viruses (SIVs) were examined in cultured macrophages and DCs as well as in cultured PBMCs in vitro. The virus strains were a T cell-tropic SIV, SIVmac239, and a T cell- and macrophage-tropic (dual-tropic) SIV, SIVmac239/316E. The infectivities of these viruses to cynomolgus macaques by intravaginal inoculation were also compared. Although both virus strains replicated well in cultured PBMCs, SIVmac239 did not replicate in cultured macrophages, whereas SIVmac239/316E did. Both strains showed little replication in cultured DCs, but a high virus yield could be obtained when SIVmac239/316E-infected DCs were co-cultured with uninfected PBMCs. A mixture of these SIVs was inoculated intravaginally to three monkeys and the virus strain that first appeared through the vaginal mucosa was determined. The virus clones detected first in PBMCs, inguinal lymph nodes and vaginal wash cells (VWCs) after the virus inoculation were of SIVmac239 in all cases, except for one clone of SIVmac239/316E in VWCs of one monkey at one time-point. These results show that the infectivity of the virus in intravaginal transmission did not depend on the cell tropism in vitro of the virus.

Application of a highly sensitive PCR-based reverse transcriptase (RT) assay to the analysis of the infection of CD4+ cell lines with human immunodeficiency virus type 1 (HIV-1) demonstrated that virus production can be detected as early as 24 h after infection. Most of the signal at 24 h was due to virus production, as it could be substantially reduced by prior treatment with the RT inhibitor zidovudine. Virus production at 24 and 48 h was unaffected by the protease inhibitor indinavir. Infection of unstimulated peripheral blood mononuclear cells (PBMC) with a macrophage-tropic HIV-1 isolate yielded increasing virus production for 2-3 weeks, while infection with a T-cell line-tropic isolate yielded only low and sporadic virus production. Productive infection of unstimulated PBMC by the macrophage-tropic virus required functional Gag matrix and Vpr proteins; therefore, the monocyte-derived macrophage is probably the virus-producing cell in these cultures.

The evolutionary relationship of retroviruses to the negative-stranded RNA virus superfamily was examined by comparing protein structures. Since protein structures are more conserved over time than primary protein sequences, three-dimensional structural comparisons permit the identification of evolutionary relationships that were previously undetected. Human immunodeficiency virus (HIV) and influenza virus were used as representatives of the virus groups, and proteins with similar functions were compared. Since M1 of influenza virus has membrane- and RNA nucleocapsid-binding activities that are functionally analogous to those of the HIV matrix and capsid proteins, the structural similarities between these proteins were determined. Sequence alignments were based on superimposition of the three-dimensional structures. Helices 2, 2′, 3 and 4 of the HIV matrix protein aligned and superimposed with the four-helix bundle of the membrane-binding N domain of M1 with a root mean square (RMS) of 3.48 Å. Helices A, B and C of the HIV N-terminal capsid protein aligned and superimposed with three helices of the four-helix bundle of the RNA-binding N domain of M1 with an RMS of 2.63 Å. The HIV Gag protein and influenza virus matrix protein may have evolved from a common ancestor protein. The similarities between influenza virus M1 and HIV matrix and capsid proteins may indicate an evolutionary link between retroviruses and negative-sense RNA viruses.

Sub-Saharan Africa is one of the regions of the globe with the highest measles-related morbidity and mortality. Yet only seven virus isolates from this vast region have been phylogenetically characterized on the basis of their nucleoprotein, the last one in 1991. To characterize the prevalent wild-type viruses and to understand their circulation pattern, a large panel (n = 45) of isolates was collected in Ghana and Nigeria in 1997 and 1998. On the basis of their nucleoprotein sequence, the viruses clearly belong to clade B but a reshuffling of the structure of this clade was proposed, tentatively extending the number of genotypes from two to three on the basis of quantitative criteria. The sequences revealed the co-circulation of at least two distinct viruses in the cities of Lagos and Ibadan, suggesting that the number of susceptible individuals seems to be high enough to support endemic circulation of at least two distinct viruses. The endemic co-circulation of several viruses may well be a characteristic of communities with low vaccination rates. One of these viruses was also found in Accra in 1998 as well as in a 1994 case linked to distant Kenya, suggesting that clade B viruses are prevalent in sub-Saharan Africa while non-B viruses seem to dominate the south of Africa.

The non-structural protein 5A (NS5A) of some hepatitis C virus (HCV) isolates has been implicated in the inhibition of the antiviral activity of interferon (IFN). In the present study, the possible inhibitory effects of NS5A from two isolates of HCV subtype 1b, HCV-1bJk and M094AJk, and their chimeric form on the antiviral activity of IFN were examined. HCV-1bJk and M094AJk are categorized as IFN resistant and IFN sensitive, respectively, based on the sequences of the IFN-sensitivity determining region (ISDR). When encephalomyocarditis virus was used as a challenge virus, NS5A was shown to eliminate the antiviral activity of IFN, with inhibition being more prominent with HCV-1bJk NS5A than with M094AJk NS5A. Moreover, the inhibition was significantly weaker in cells expressing a chimeric NS5A that had a short stretch of 49 amino acids (aa 2209-2257), including the ISDR sequence, from M094AJk in the backbone of the HCV-1bJk sequence than in cells expressing the original NS5A from HCV-1bJk. These results suggest an important role for the 49 aa sequence, including the ISDR, in the inhibition of IFN-mediated antiviral activity. On the other hand, only a slight reduction of IFN antiviral activity by HCV-1bJk NS5A was observed when vesicular stomatitis virus was used as a challenge virus, and barely any reduction was observed when Japanese encephalitis virus was used. These results may reflect differential importance of each of the IFN-mediated signalling pathways in conferring resistance against different viruses.

The hepatitis C virus (HCV) genome encodes two membrane-associated envelope glycoproteins (E1 and E2), which are released from the viral polyprotein precursor by host signal peptidase cleavages. These glycoproteins interact to form a noncovalent heterodimeric complex, which is retained in the endoplasmic reticulum. HCV glycoproteins, E1 and E2, are heavily modified by N-linked glycosylation. A recent study has revealed that upon partial deglycosylation with endoglycosidase H only four of the five potential glycosylation sites of HCV glycoprotein E1 are utilized. In this work, the unused glycosylation site on the E1 glycoprotein was identified and the influence of N-linked glycosylation on the formation of the HCV glycoprotein complex was studied by expressing a panel of E1 glycosylation mutants in HepG2 cells. Each of the five potential N-linked glycosylation sites, located at amino acid positions 196, 209, 234, 305 and 325, respectively, on the HCV polyprotein, was mutated separately as well as in combination with the other sites. Expression of the mutated E1 proteins in HepG2 cells indicated that the fifth glycosylation site is not used for the addition of N-linked oligosaccharides and the Pro immediately following the sequon (Asn-Trp-Ser) precludes core glycosylation. The effect of each mutation on the formation of noncovalent E1E2 complexes was also analysed. As determined with the use of a conformation-sensitive monoclonal antibody, mutations at positions N2 and N3 had no, or only minor, effects on the assembly of the E1E2 complex, whereas a mutation at position N1 and predominantly at position N4 dramatically reduced the efficiency of the formation of noncovalent E1E2 complexes.

A single genetic locus, flavivirus resistance (Flv), controls virus titres and severity of flavivirus infection in mouse brain. It has been mapped to mouse chromosome 5 and shown to include different allelic forms. While the majority of laboratory mouse strains are susceptible to flaviviruses and carry the Flv(s) allele, wild mice and laboratory mouse strains recently derived from them are resistant and carry flavivirus-resistance alleles including Flv r-like and Flv mr alleles. Although there is a mouse model of flavivirus resistance conferred by the Flv r allele, other resistance alleles have not been adequately studied due to a lack of appropriate animal models. In this paper we describe the development of new flavivirus-resistant mouse strains, C3H.M.domesticus-Flv r and C3H.MOLD-Flv mr, which carry the novel resistance alleles Flv r-like and Flv mr on the genetic background of flavivirus susceptible C3H/HeJ mice. The new strains were created by 10 to 11 generations of backcrossing followed by brother-sister matings resulting in a generation of homozygous founder stocks. Genome analysis of the newly developed mouse strains has revealed chromosomal regions of approximately 9 and 11 cM, respectively, encompassing Flv on chromosome 5, which are derived from resistant donor mice. These segments are much smaller than the segment of approximately 31 cM described in the congenic resistant mouse strain C3H.PRI-Flv r (also known as C3H/RV). The new congenic mouse strains, which were created to carry the Flv r-like and Flv mr alleles on the standardized genetic background of susceptible mice, represent new animal models of flavivirus resistance conferred by these novel resistance alleles.

Poliovirus strains derived from the oral poliovirus vaccine (Sabin) can be differentiated from wild-type poliovirus by tests based on either immunological or genetic properties of the strains. The characterization of a recently identified poliovirus type 1 isolate with exceptional properties is described. Initial phenotypic analysis of the virus by use of polyclonal absorbed antisera suggested a wild-type character. However, the different genomic analyses all confirmed the Sabin-derived character of the virus. All 17 plaques isolated from the strain shared these properties, thus excluding the possibility of a mixture of a wild-type and a Sabin-derived strain. To elucidate the properties of this virus further, the nucleotide sequences of the P1 region and most of the 5′ non-coding region were established. Although the nucleotide identity with Sabin 1 was more than 99.4%, mutations were observed in regions encoding three major antigenic sites; the deduced amino acid substitutions confirmed the aberrant results of micro-neutralization assays with site-specific monoclonal antibodies. The most striking feature was the existence of a hexanucleotide deletion in the VP1 gene, which gave rise to a two amino acid deletion in the BC loop. In spite of these antigenic changes, the strain was readily serotyped as poliovirus type 1 under standard conditions. Likewise, replication of the virus under cell culture conditions was not affected by these mutations or by the deletion. Standard polio vaccination protects against this aberrant virus, and its epidemiological significance remains open.

Several mutants of the Mahoney strain of poliovirus type 1 have been generated by introducing mutations into the stem-loop II (SLII) structure within the internal ribosomal entry site (IRES). Four of these mutants (SLII-1, -4, -5 and -6 mutants) have been characterized previously and are host-range mutants that replicate well in human HeLa cells but not in mouse cells. Two deletion mutants, SLII-2 and SLII-3, were non-viable, even in HeLa cells. It is now reported that SLII-2 was defective in genome RNA synthesis and viral protein synthesis, while SLII-3 was defective only in viral protein synthesis. These results indicate that the SLII region contains a cis-element for RNA replication as well as for IRES-dependent translation and that these two functions lie at the same sites within the SLII region. The host cellular factors that interacted with wild-type SLII and mutant SLII-2 and SLII-3 RNAs were different, suggesting that different host-factor binding regulates expression of mutant phenotypes.

Although enteroviruses cause a great variety of diseases including meningitis, paralysis and myocarditis, the life-cycle of these viruses in man is still quite poorly understood. The role of human leukocytes as a target for enterovirus infections was studied in this report. Despite great similarity in the structure and replication of coxsackievirus B3 (CBV3), echovirus 1 (EV1), and poliovirus 1 (PV1), the ability of these viruses to infect human peripheral blood mononuclear cells (PBMC), and B (Raji), T (Molt-4) and monocytic (U-937) cell lines differed markedly. CBV3 attached both to PBMC and the three haematopoietic cell lines studied, whereas EV1 bound only to PBMC. Generally, the attachment of PV1 was very poor. The binding assays mostly correlated with the expression of CD55 (decay accelerating factor, DAF) and alpha2beta1-integrin (VLA-2), which are known to act as receptors for CBV3 and EV1, respectively, as well as with the expression of the PV receptor on the cell surface. To analyse virus replication in the cells, viral protein and nucleic acid syntheses were studied by immunoprecipitation and RT-PCR. CBV3 was able to replicate in Raji and Molt-4 cells, even though no expression of DAF was detected, whereas in the monocytic cell line, viral protein synthesis was detected only after transfection of virus RNA. Neither CBV3 nor EV1 replicated in PBMC and all haematopoietic cells studied seemed to be poorly permissive for PV1 replication.

In recent years a wealth of data has become available about the caliciviruses that infect humans, as well as those which infect a range of animal species, notably cats, rabbits, pigs and marine animals. However, in the two decades since the earliest reports of calicivirus infection in dogs, very little has become known about the epidemiology, pathogenicity and molecular biology of the caliciviruses that may infect canines. In 1990, a canine calicivirus (CaCV) was isolated from a 2-month-old diarrhoeic domestic dog in Japan. This virus, which can be grown in cultured cells of canine origin, has the classic ‘Star of David’ morphology of caliciviruses, and the one major structural protein was shown to be immunogenic in dogs. In this study, a 3.8 kb region of the genome of this CaCV isolate from the RNA polymerase gene to the 3′ poly(A) tail was cloned and sequenced, and phylogenetic analysis was undertaken in order to establish the relationship of CaCV to other animal and human caliciviruses. This CaCV isolate had a nucleotide sequence, genomic organization and phylogenetic position closest to, but clearly distinct from, both feline calicivirus and San Miguel sea lion virus isolates. These findings suggest that CaCV represents a new clade of animal caliciviruses, presumably as a member of the recently proposed new genus Vesivirus.

The nucleotide sequence of the entire genome of Chuzan virus, which belongs to the Palyam serogroup orbiviruses and causes congenital abnormalities of cattle, has been completed by analysis of the genes encoding minor core proteins (VP1, VP4 and VP6) and non-structural proteins (NS1, NS2 and NS3). The genome of Chuzan virus is 18,915 bp in length and the coding capacity of its open reading frames is 6071 aa. Comparative sequence analysis with other serogroups of the genus Orbivirus indicated that the outer capsid protein VP2, which is the neutralizing antigen, appears to be the most variable and the major core protein VP3 is the most conserved. Overall, the structural proteins, with the exception of VP2, are more conserved than the non-structural proteins among orbiviruses. Chuzan virus is phylogenetically most related to African horsesickness virus.

The outer capsid spike protein VP4 is the main rotavirus cell attachment protein, but the cellular receptor used by rotavirus to establish a productive infection remains unknown. Sialic acid (SA) residues on the cell surface have been shown to be required for efficient binding and infectivity of animal rotaviruses (ARVs), but not of human rotaviruses (HRVs). Since the SA dependence of only a limited number of strains has been tested to date, in this study a larger number of strains were tested to further investigate the involvement of SA in rotavirus infectivity. Following treatment of African green monkey kidney cell (MA104) monolayers with neuraminidase, productive infection of rotavirus was measured by immunofluorescence. The infectivity of all 14 HRVs tested was SA-independent. Ten of 15 ARVs tested were SA-independent, while only five were SA-dependent. These results indicate that most ARVs, like HRVs, infect permissive cells in an SA-independent manner, probably by a common cellular receptor.

Poxviruses encode a broad range of proteins that interfere with host immune functions such as soluble versions of cytokine receptors. Soluble virus tumour necrosis factor receptors (vTNFRs) were described originally in myxoma and Shope fibroma viruses. Cowpox virus (CPV) encodes three vTNFRs (CrmB, CrmC and CrmD). The genes equivalent to CrmB and CrmC in vaccinia virus (VV) Copenhagen are mutated and are named B28R/C22L and A53R, respectively. CrmD was identified recently in CPV and ectromelia virus but the gene is absent in VV Copenhagen. We have tested for expression of soluble binding activity for human TNF in cultures infected with 18 orthopoxviruses and have found that TNFRs are mostly absent but are produced by VV strains Lister, USSR and Evans, by the CPV elephantpox and by camelpox virus. Interestingly, we also found TNFR activity on the surface of cells infected with VV Lister, USSR and Evans. Sequence analysis of the relevant regions in VV Lister identified an intact A53R gene and an inactive B28R gene. Expression of VV Lister A53R in baculovirus and VV Western Reserve demonstrated that gene A53R encodes an active soluble vTNFR of 22 kDa. Expression and characterization of recombinant vTNFRs from VV Lister (A53R) and CPV (CrmB and CrmC) showed a similar binding specificity, with each receptor binding TNF from man, mouse and rat, but not human lymphotoxin-alpha. Lastly, the VV Lister and CPV vTNFRs bind human TNF with high affinity and prevent the binding of TNF to cellular receptors.

During human cytomegalovirus (HCMV) infection, a rapid increase in AP-1 activity is detected. In this study, activation of transcription from promoters containing AP-1-binding sites by the IE1 protein of HCMV was examined. In transient transfection assays with reporter plasmids, it was found that IE1 strongly induced AP-1-driven transcription. Cells stably expressing IE1 also showed higher levels of AP-1 activity than did control cells. IE1 expression did not raise levels of c-jun and c-fos RNA, as determined by quantitative RT-PCR. AP-1 induction by IE1 was blocked efficiently by protein kinase inhibitors in a cell type-dependent manner; for example, by staurosporine in the human microglial cell line U373MG and by H7 in the human promonocytic cell line U937. IE1-driven activation of AP-1 was increased dramatically by mitogen-activated protein kinase/extracellular signal-regulated kinase kinase kinase 1 (MEKK1). The results of this study indicate that IE1 activates AP-1 at the post-transcriptional level and that MEKK1 may play an important role in this process.

Evidence for the existence of porcine gammaherpesviruses was obtained by PCR and sequence analysis. Initially, samples of peripheral blood mononuclear cells (PBMC), spleens, lungs, kidneys and livers of pigs from Germany and Spain were tested with a PCR assay which targets conserved regions of the herpesvirus DNA polymerase gene with degenerate and deoxyinosine-substituted primers. Amplicons of identical sequence were obtained from one spleen and two PBMC samples. This sequence showed a high percentage of identity with the DNA polymerase genes of herpesviruses of the oncogenic subfamily Gammaherpesvirinae. Alignment of amino acid sequences showed the highest identity values with bovine gammaherpesviruses, namely alcelaphine herpesvirus type 1 (68%), ovine herpesvirus type 2 (68%) and bovine lymphotropic herpesvirus (67%). Comparison with pseudorabies virus and porcine cytomegalovirus, which are the only porcine herpesvirus species presently known, showed values of only 41%. PCR analysis of PBMC (n = 39) and spleen (n = 19) samples from German pigs, using primers specific for the novel sequence, revealed a prevalence of 87 and 95%, respectively. In this analysis, three out of eight spleen samples from Spanish pigs were also positive. Subsequent sequencing of the amplicons revealed the presence of two closely related gammaherpesvirus sequences, differing from each other by 8% at the amino acid level. The putative novel porcine herpesviruses, from which these sequences originated, were tentatively designated porcine lymphotropic herpesvirus type 1 and type 2 (PLHV-1 and PLHV-2). When using pig organs for xenotransplantation, the presence of these viruses has to be considered.

The linear virion DNA of bovine herpesvirus type 4 (BHV-4) is flanked by tandem repeats designated polyrepetitive DNA (prDNA). To investigate the structure and functional role of the prDNA for cleavage/packaging of progeny viral DNA, the complete nucleotide sequence (2267 bp) of a cloned prDNA unit of BHV-4 was determined. Moreover, the terminal fragments of the genome and the junctions between prDNA and the central unique DNA were analysed. In order to characterize the function of the prDNA of BHV-4, a transient packaging assay was developed. The prDNA has a G+C content of 71.1%. Its structure is composed of numerous internal repeats and every unit contains the conserved sequence of the cleavage/packaging signal. A fragment of 443 bp comprising the cleavage/packaging signal was found to be sufficient for cleavage and encapsidation of replicated concatemeric viral DNA. These results suggest that prDNA is a functionally important region of the genome of BHV-4.