- Volume 83, Issue 5, 2002

In order to determine the influence of poliovirus receptor (PVR) expression on poliovirus cell tropism and neuropathogenesis, two transgenic (tg) mouse models were produced in which PVR was expressed under the transcriptional control of the human PVR gene promoter (hg–PVR mice) and the CAG promoter (CAG–PVR mice). Then the pathogenicity of poliovirus after intracerebral inoculation of the type 1 Mahoney strain was compared. These showed completely different clinical and pathological changes. In the former, the expression of PVR in neurons in the central nervous system (CNS) confered susceptibility to poliovirus, and a paralytic disease that resembled the human poliomyelitis occurred. In the latter, PVR expression was detected in glial and ependymal cells in addition to the neurons. Paralysis of the limbs and death were rarely observed and mice survived without showing substantial clinical abnormality. Histopathological examination revealed that glial and ependymal cells also became susceptible to poliovirus infection. Poliovirus antigens were mainly detected in ependymal and glial cells and hippocampal neurons near the lateral ventricles in the brain, but were not frequently detected in neurons in the brainstem unlike in the hg–PVR mice. The levels of viral antigens and virus recovered from the CNS of CAG–PVR mice began to decrease as early as 2 days after inoculation, which suggested induction of a fast immune response. These results suggest that the neuropathogenicity of poliovirus changes markedly depending on the specific expression of the PVR molecule in the CNS.

A survey of poliovirus in river and sewage water was conducted from October 1993 to September 1995 in Toyama Prefecture, Japan. In this study, 25 isolates differentiated as type 2 vaccine-derived polioviruses (VDPVs) were characterized using mutant analysis by PCR and restriction-enzyme cleavage (MAPREC) to estimate the ratio of 481-G revertants correlated to neurovirulence in a virus population. Of these isolates, 23 (92%) comprised between 44 and 96% 481-G revertants by MAPREC. The other two isolates had revertant percentages close to the 0·6% of the attenuated reference strain. It was presumed that these 23 isolates would be variant with potential neurovirulence by MAPREC analysis. Of the 23 isolates, three were isolated from river water. Moreover, our results by MAPREC showed that type 2 poliovirus was phenotypically more variable than type 1 (69%) or type 3 (55%), as determined in previous studies. The prevalence of virulent-type VDPVs in river and sewage water suggested that the oral poliovaccine itself had led to wide environmental pollution in nature. To terminate the cycle of virus transmission in nature, the ecology of VDPVs should be studied further. A hygiene programme, inactivated poliovirus vaccine immunization and well-maintained herd immunity may play key roles in reducing the potential risk of infection by virulent VDPVs.

Efficient internal initiation of translation from the hepatitis C virus (HCV) internal ribosome entry site (IRES) requires sequences of domain II, but the precise role of these sequences is still unknown. In this study, the formation of RNA–RNA complexes in the HCV IRES was evaluated. Using transcripts that contain the sequences of the structural HCV IRES domains II, IIIabcd, IIIabc, IV and IIIef-IV, specific long-range interactions between domains II and IV, as well as domains II and IIIabcd, have been found. These interactions were readily detected in a gel mobility-shift assay and required the presence of magnesium ions. A high concentration of nonspecific competitors, an 80 nt fragment of 18S rRNA or poly(I:C), did not interfere with the formation of RNA complexes. Interestingly, an RNA oligonucleotide bearing the sequence of stem–loop IIId interacted with domain II but not with domain IV or IIIef-IV, strongly suggesting that the interaction between domains II and IIIabcd was mediated by the IIId hairpin. Interaction between domains IIIabcd and IV was barely detected, consistent with the result that the apical part of domain III folds independently of the rest of the IRES. Moreover, the addition of stem–loop IIIef sequences to domain IV significantly reduced its ability to interact, which is in agreement with the formation of a compact RNA structure of domain IV with IIIef. The interactions observed in the absence of proteins between domains II and IV as well as stem–loop IIId and domain II may be transient, having a regulatory role in the translation efficiency of the HCV IRES.

Four human monocyte-derived macrophage membrane proteins, with apparent molecular masses of 27, 45, 67 and 87 kDa, were identified as possible receptors for dengue virus serotype 2 (DEN-2) (Mexican isolate 200787/1983), based on affinity chromatography, immunofluorescence, virus overlay protein-binding assays and Western blotting. Additionally, mouse polyclonal antibodies raised against each of the four proteins were capable of partially inhibiting in vitro DEN-2 infection of monocyte-macrophages, thus supporting the notion of a role for such proteins as DEN-2 receptors. Parallel studies were carried out using the human promonocytic U-937 cell line, both as undifferentiated cells and as monocyte-like phorbol myristate acetate (PMA)-differentiated cells, as target cells. Whereas interaction between DEN-2 and undifferentiated U-937 cells was almost negligible, PMA-differentiated U-937 cells were shown to harbour putative receptors (with molecular masses of 45 and 67 kDa) for DEN-2, similar to those found in human monocyte-derived macrophages. To our knowledge, this is the first report that describes putative receptors for DEN-2 in primary cultures of human macrophages.

Nine different CTL epitopes, conserved in both Hantaan virus (HTNV) and Sin Nombre virus (SNV), were selected for study. The binding affinity of each peptide with HLA-A2.1 molecules in vitro was determined and antigen-specific responses from seven donors who had a previous field infection with HTNV were examined. Although the strength or frequency of CTL activity showed different patterns in the seven patients, five of seven patients showed significant activity against at least one or more epitope peptides. In particular, the peptide ILQDMRNTI (HTNV, aa 334–342; SNV, aa 333–341), which elicited CTL activity in five patients, was shown to be specifically HLA-A2.1-restricted in partially cloned CD8+ T cells and also induced activated and effector CD8+ T cell-producing T cytotoxic (Tc) type 1 cytokines, such as IL-2 and IFN-γ. The results suggest that this epitope would serve as a useful component for the intervention of both HTNV and SNV infection.

The haemagglutinin (HA) of influenza A/H2N2 virus possesses six antigenic sites (I-A to I-D, II-A and II-B), and sites I-A, I-B and I-C are located in the regions corresponding to sites A, B and D on the H3 HA. We demonstrated previously that most escape mutants selected by mAbs to site I-A, I-B or I-C had acquired a new oligosaccharide at position 160, 187 or 131, respectively, but this has never occurred during circulation of A/H2N2 virus in humans. Here, to examine whether the H2 HA has the potential to gain two new oligosaccharides on its tip, 31 double escape mutants were isolated by using a single escape mutant with an oligosaccharide at position 160, 187 or 131 as a parental virus and a mAb to an antigenic site different from that to which the mAb used for selection of the parental virus was directed as a selecting antibody, but there were no mutants with two new oligosaccharides. Glycosylation-site HA mutants containing one to three oligosaccharides at positions 160, 187 and 131 were also constructed and their intracellular transport and biological activities were analysed. The results showed that all of the mutant HAs were transported to the cell surface but exhibited a decrease in both receptor-binding and cell-fusing activities. Thus, influenza A/H2N2 virus may have failed to increase the number of oligosaccharides on the HA because, if this happens, the biological activities of the HA are reduced, decreasing the ability of the virus to replicate in humans.

According to their cellular receptor use, measles virus (MV) strains can be separated into two phenotypes, CD46-using and CD46-non-using. A long chimeric receptor, CD46CD[55–46], was generated from the CD46 backbone, encompassing the four short consensus repeat (SCR) domains of CD46 linked via a flexible glycine hinge to SCR1 and SCR2 of CD55, SCR3 and SCR4 of CD46 and the STP, transmembrane and cytoplasmic tail of CD46. This chimeric receptor was proficient for MV binding but deficient in mediating MV-induced cell-to-cell fusion and virus replication, possibly due to the extended distance between the MV haemagglutinin (H) binding site (CD46 SCR1–SCR2) and the cell membrane. When coexpressed with either wild-type CD46 or CD150, this fusion-incompetent receptor exerted a dominant negative effect and inhibited both cell-to-cell fusion and entry of MV with CD46-using, but not CD46-non-using, phenotype. A soluble octameric CD46–C4bpα exhibited similar CD46- and CD150-mediated fusion inhibition properties only against CD46-using MV. This suggests that the long CD46CD[55–46] receptor acts by sequestering incoming MV prior to its binding to the shorter functional CD46 or CD150 receptor.

The influence of measles virus (MV) infection on gene expression by human peripheral blood mononuclear cells (PBMCs) was examined with cDNA microarrays. The mRNA levels of more than 3000 cellular genes were compared between uninfected PBMCs and cells infected with either the Edmonston MV strain or a wild-type MV isolate. The MV-induced upregulation of individual genes identified by microarray analyses was confirmed by RT–PCR. In the present study, a total of 17 genes was found to be upregulated by MV infection. The Edmonston strain grew better in the PBMC cultures than the wild-type MV, and the Edmonston strain was a stronger inducer of the upregulated host cell genes than the wild-type virus. The anti-apoptotic B cell lymphoma 3 (Bcl-3) protein and the transcription factor NF-κB p52 subunit were upregulated in infected PBMCs both at the mRNA and at the protein level. Several genes of the interferon system including that for interferon regulatory factor 7 were upregulated by MV. The genes for a number of chaperones, transcription factors and other proteins of the endoplasmic reticulum stress response were also upregulated. These included the gene for the pro-apoptotic and growth arrest-inducing CHOP/GADD153 protein. Thus, the present study demonstrated the activation by MV of cellular mechanisms and pathways that may play a role in the pathogenesis of measles.

Our previous work has shown that high avidity cytotoxic T lymphocytes (CTL) are optimal for virus clearance in vivo and thus it is necessary that an effective vaccine is capable of eliciting high avidity CTL. To determine if vaccination with the paramyxovirus simian virus 5 (SV5) elicits a high avidity response to a model foreign antigen, a recombinant virus was engineered to express chicken ovalbumin (rSV5–Ova). To compare the CTL response elicited with rSV5–Ova and a recombinant vaccinia virus expressing ovalbumin (rVV–Ova), mice were vaccinated intranasally with various doses of each vector and the Ova-specific CTL response was determined by ELISPOT analysis. Here, it has been shown that rSV5 can be equally as effective as rVV in eliciting antigen-specific CTL, in terms of both the total number of CTL and the number of high avidity cells. This has implications for both the design of vaccine vectors and the route utilized for vaccine administration for the elicitation of high avidity CTL responses. The advantages and future potential use of rSV5 vaccine vectors are discussed.

The human immunodeficiency virus type 1 (HIV-1) Tat protein transduction domain (PTD), which contains a high proportion of arginine and lysine residues, is responsible for highly efficient protein transduction through the plasma membrane. To identify the role of the PTD sequence motif in transduction, various deletions and substitutions were introduced into the PTD. Tat–green fluorescent protein (GFP) fusion proteins, containing various lengths of the Tat PTD, were expressed and the extent of their transduction into mammalian cells was analysed by Western blot analysis and fluorescence microscopy. Deletion analysis of PTD mapped to a nine amino acid motif (residues 49–57: RKKRRQRRR) sufficient for transduction. Further deletion of this Tat basic domain either at the N terminus or at the C terminus significantly decreased transduction efficiency. The transduction efficiencies of GFPs fused to nine consecutive lysine (9Lys–GFP) or arginine (9Arg–GFP) residues were similar to that of Tat(49–57)–GFP. The transduced proteins localized to both the nucleus and the cytosol, as assessed by confocal microscopy and Western blot analysis of subcellular fractions from transduced cells. Thus, the availability of recombinant GFP fusion proteins facilitates the simple and specific identification of protein transduction mediated by these peptide sequences. The modified PTD sequences designed in this study may provide useful tools necessary for delivering therapeutic proteins/peptides into cells.

The positive effect of the co-expression of T helper (Th) cell type 2 cytokine interleukin-5 (IL-5) on nef-deleted simian/human immunodeficiency virus (SHIV) replication in vitro has been observed previously. To analyse whether the growth advantage of IL-5-containing SHIV (NI-IL5) in vitro would be relevant in vivo, the virus was inoculated into monkeys. Three rhesus macaques were inoculated intravenously with 104 TCID50 of NI-IL5. Results were compared with those obtained previously from SHIV NM-3rN (intact) and SHIV-dn (nef-deleted)-infected monkeys. Cytokine production, analysed by IL-5 ELISA, showed a twofold increase in IL-5 concentration in the plasma soon after the peak of virus replication. Virus replication and antibody production were greater in monkeys inoculated with IL-5-expressing SHIV than in monkeys inoculated with nef-deleted SHIV without IL-5. These findings show a stimulation of SHIV replication by co-expression of IL-5 and suggest the important role of Th2-type cytokines in human immunodeficiency virus type 1 infection.

Although muscovy duck reovirus (DRV) shares properties with the reovirus isolated from chicken, commonly named avian reovirus (ARV), the two virus species are antigenically different. Similar to the DRV σB-encoded gene (1201 bp long) previously identified, the three other double-stranded RNA small genome segments of DRV have been cloned and sequenced. They were 1325, 1191 and 1124 bp long, respectively, and contained conserved terminal sequences common to ARVs. They coded for single expression products, except the smallest (S4), which contained two overlapping open reading frames (ORF1 and ORF2). BLAST analyses revealed that the proteins encoded by the 1325 and 1191 bp genes shared high identity levels with ARV σA and σNS, respectively, and to a lesser extent with other orthoreovirus counterparts. No homology was found for the S4 ORF1-encoded p10 protein. The 29·4 kDa product encoded by S4 ORF2 appeared to be 25% identical to ARV S1 ORF3-encoded σC, a cell-attachment oligomer inducing type-specific neutralizing antibodies. Introduction of large gaps in the N-terminal part of the DRV protein was necessary to improve DRV and ARV σC amino acid sequence alignments. However, a leucine zipper motif was conserved and secondary structure analyses predicted a three-stranded α-helical coiled-coil feature at this amino portion. Thus, despite extensive sequence divergence, DRV σC was suggested to be structurally and probably functionally related to ARV σC. This work provides evidence for the diversity of the polycistronic S class genes of reoviruses isolated from birds and raises the question of the relative classification of DRV in the Orthoreovirus genus.

The near-full-length genome of a TT virus (TTV) (HEL32), closely related to the previously uncharacterized genotype 6, was cloned and sequenced. The genomic organization of HEL32 was compared to 41 published near-full-length TTV sequences representing 17 genotypes. In the majority of genomes, the open reading frame (ORF) 2 region was divided into two separate ORFs, 2a and 2b. The ORF2a sequence was conserved among all genotypes, while the ORF2b region showed more variability. The two corresponding putative proteins of HEL32 were expressed in prokaryotes and their antigenic potential was studied. IgM and IgG antibodies to the respective ORF2-encoded proteins, fp2a and fp2b, and the presence of TTV DNA were studied in the sera of 89 constitutionally healthy adults. By immunoblot using the small TTV proteins as antigens, strong IgM and IgG reactivities were found in 9 and 10% of subjects, respectively. Follow-up studies for 12–15 years of three subjects showed either a persistent coexistence of IgM and TTV DNA or the appearance of viral DNA regardless of pre-existing antibodies. The low prevalence of IgG could be due to the weak immunogenicity of these probably non-structural proteins or to a genotype-specific antibody response. By nested PCR of the conserved ORF2a region, the prevalence of TTV DNA was 85%. TTV genotype 6 sequences were found by specific PCR in 3 of 35 (8·6%) subjects. The low prevalence of TTV IgG compared to the high TTV DNA prevalence, the coexistence of antibodies and viral DNA and the appearance of TTV DNA regardless of pre-existing antibodies suggest that the B-cell immunity against these minor TTV proteins would not be cross protective.

The function of the X protein (pX) in the replication cycle of mammalian hepadnaviruses is enigmatic. Using tissue culture experiments it has been shown that the X gene product is not central to hepatitis B virus (HBV) replication and virion export. However, at present it is still unclear whether this also applies to the in vivo situation. Using a terminally redundant X-deficient HBV DNA construct, transgenic mice were established that exhibited high-level expression of the viral core protein in liver and kidneys. Importantly, replicative DNA intermediates and mature viral genomes could be detected in the liver and serum of these mice, respectively. These findings indicate that, in the in vivo model of transgenic mice, the HBV X (HBx) gene product is not required for HBV replication and virion secretion.

The role of the products of the UL10 and the UL49.5 homologous genes of Marek’s disease virus serotype 1 (MDV-1) in virus replication was investigated. Deletion of either open reading frame in an infectious bacterial artificial chromosome clone (BAC20) of MDV-1 resulted in progeny viruses that were unable to spread from cell to cell. After transfection of UL10- or UL49.5-negative BAC20 DNA into chicken or quail cells, only single infected cells were observed by indirect immunofluorescence analysis. In contrast, plaque formation was restored when mutant BAC DNAs were co-transfected with the corresponding expression plasmid encoding either the UL10-encoded gM or the UL49.5 gene product. These data demonstrate that gM and its putative complex partner, the UL49.5 homologous protein, are essential for MDV-1 growth in cultured cells. Thus, MDV-1 represents the first example of a member of the family Herpesviridae for which the highly conserved membrane proteins are indispensable for cell-to-cell spread.

Human cytomegalovirus (HCMV) UL53 belongs to a family of conserved herpesvirus genes. In this work, the expression and localization of the UL53 gene product was analysed. Results obtained showed that pUL53 is a new structural protein. In infected human fibroblasts, pUL53 localizes in cytoplasmic perinuclear granular formations together with other structural viral proteins. In the nucleus, pUL53 forms patches at the nuclear periphery and co-localizes with lamin B at the internal nuclear membrane level. Immunoelectron microscopy studies have disclosed that nuclear pseudo-inclusions are labelled, whereas nucleocapsid formations within the intranuclear skein are negative. Furthermore, the mature virus particle maintains pUL53 at its tegumental level. These data suggest that pUL53 could be involved either in nucleocapsid maturation or in the egress of nucleocapsids from the nucleus to the cytoplasm through the nuclear membrane, a role compatible with the function hypothesized for UL31, its positional homologue in herpes simplex virus type 1.

The protein kinase pUL97, encoded by human cytomegalovirus (HCMV), is an important determinant of virus replication. Recently, indolocarbazoles were identified as a class of substances that inhibit the pUL97 kinase activity in vitro. In parallel, it was shown that indolocarbazoles interfere with HCMV replication; however, the causal relationship between inhibition of pUL97 kinase activity and virus replication has not been clarified. Here evidence is provided that indolocarbazole-mediated inhibition of virus replication is a direct result of diminished pUL97 protein kinase activity. In cell culture infections, a strong and selective antiviral activity was measured with respect to several strains of HCMV in contrast with other related or non-related viruses. For fine quantification, recombinant HCMVs expressing green fluorescent protein were used, demonstrating the high sensitivity towards compounds NGIC-I and Gö6976. Interestingly, a ganciclovir-resistant virus mutant (UL97-M460I) showed increased sensitivity to both compounds. Supporting this concept, transfection experiments with cloned pUL97 revealed that ganciclovir-resistant mutants were characterized by reduced levels of autophosphorylation compared with wild-type and possessed particularly high sensitivity to indolocarbazoles. Moreover, the Epstein–Barr virus-encoded homologous kinase, BGLF4, which showed a similar pattern of autophosphorylation and ganciclovir phosphorylation activities, was not inhibited. Importantly, a cytomegalovirus deletion mutant, lacking a functional UL97 gene and showing a severe impairment of replication, was completely insensitive to indolocarbazoles. Thus, our findings indicate that a specific block in the activity of pUL97 is the critical step in indolocarbazole-mediated inhibition of virus replication and that pUL97 might be targeted very efficiently by a novel antiviral therapy.

Epstein–Barr virus is a human gammaherpesvirus that infects and establishes latency in B lymphocytes in vivo. The latent membrane protein 2 (LMP2) gene is expressed in latently infected B cells and encodes two protein isoforms, LMP2A and LMP2B, that are identical except for an additional N-terminal 119 aa cytoplasmic domain which is present in the LMP2A isoform. A panel of fusion proteins was constructed in which the fluorescent enhanced green fluorescent protein and DsRed protein domains were fused to the N- and C-termini of LMP2A and LMP2B. By fluorescence microscopy, LMP2B localized to perinuclear regions of both live and fixed transiently transfected cells. Co-localization was detected with markers for the endoplasmic reticulum and the trans-Golgi network. No evidence of co-localization of LMP2B with endosomes or surface expression was obtained. Transiently expressed LMP2B co-localized with transiently or constitutively expressed LMP2A. Confocal microscopy confirmed that LMP2A proteins localized to intracellular perinuclear compartments with markers for the trans-Golgi network. Only LMP2A proteins with C-terminal truncations were detected in the plasma membrane with extracellular loop1 epitope tags. These results indicate that the transmembrane domain of LMP2 proteins possess intracellular retention signals and suggest that LMP2A-mediated signalling effects are likely to be ectopic, originating from sites inside the cell close to the nucleus.