- Volume 85, Issue 8, 2004

The capsid of birnaviruses contains two proteins, VP2 and VP3, which derive from the processing of a large polyprotein, NH2–pVP2–VP4–VP3–COOH. The proteolytic cascade involved in processing the polyprotein, and in the final maturation of pVP2 (the precursor of VP2), has recently been shown to generate VP2 and four structural peptides in infectious bursal disease virus and blotched snakehead virus. The presence of peptides in infectious pancreatic necrosis virus particles was investigated using mass spectrometry and N-terminal sequencing of virus particles. Three peptides deriving from the C terminus of pVP2 (residues 443–486, 487–495 and 496–508 of the polyprotein) and 14 additional peptides produced by further processing of peptides [443–486] and [496–508] were identified. These results indicate that the presence of several virus-encoded peptides in the virions is a hallmark of birnaviruses.

Genome segment 9 (Seg-9) of Colorado tick fever virus (CTFV) is 1884 bp long and contains a large open reading frame (ORF; 1845 nt in length overall), although a single in-frame stop codon (at nt 1052–1054) reduces the ORF coding capacity by approximately 40 %. However, analyses of highly conserved RNA sequences in the vicinity of the stop codon indicate that it belongs to a class of ‘leaky terminators’. The third nucleotide positions in codons situated both before and after the stop codon, shows the highest variability, suggesting that both regions are translated during virus replication. This also suggests that the stop signal is functionally leaky, allowing read-through translation to occur. Indeed, both the truncated ‘termination’ protein and the full-length ‘read-through’ protein (VP9 and VP9′, respectively) were detected in CTFV-infected cells, in cells transfected with a plasmid expressing only Seg-9 protein products, and in the in vitro translation products from undenatured Seg-9 ssRNA. The ratios of full-length and truncated proteins generated suggest that read-through may be down-regulated by other viral proteins. Western blot analysis of infected cells and purified CTFV showed that VP9 is a structural component of the virion, while VP9′ is a non-structural protein.

The effects of oral inoculation into infant CB17scid mice of two reassortant rotavirus vaccines were compared. The vaccines were Rotashield and WC3-PV, a mixture of five reassortants (G1, G2, G3, G4 and P1; pentavalent reassortant vaccine). Control mice were inoculated with a placebo. At 6 days post-inoculation (p.i.), 8 of 13 (62 %; P<0·005) Rotashield-inoculated mice developed hepatitis and/or bile-duct obstruction compared with none of 11 mice given WC3-PV and none of 14 given placebo. In the Rotashield-inoculated mice, only serotype G3 rhesus rotavirus (RRV) was isolated from multiple sites, including intestine, liver, pancreas, spleen, blood and mesenteric lymph nodes. Recovery of RRV from Rotashield-inoculated mice followed a biphasic pattern. The two peaks of RRV recovery appeared to coincide firstly with replication in the intestine during days 1–3 p.i., and secondly with virus infection of the liver from days 10 to 15 p.i. WC3 reassortants of four different serotypes were detected only at day 1 p.i. in the intestine, liver, pancreas and blood cells from three WC3-PV-inoculated mouse pups. However, WC3-PV did not produce any hepatopathology. Rotashield and WC3-PV appeared to exhibit different biological activity in infant CB17scid mouse pups.

Lepidopteran larvae resist baculovirus infection by selective apoptosis of infected midgut epithelial cells and by sloughing off infected cells from the midgut. Once the infection breaches the midgut epithelial barrier and propagates from infective foci to the haemocoel, however, there are few mechanisms known to account for the resistance and clearance of infection observed in some virus–host combinations. The hypothesis that factors present in the plasma of infected pest larvae act to limit the spread of virus from initial infective foci within the haemocoel was tested. An in vitro bioassay was developed in which Helicoverpa zea single capsid nucleopolyhedrovirus (HzSNPV) was incubated with plasma collected from uninfected Heliothis virescens larvae. Infectious HzSNPV particles were then titrated on HzAM1 cells. Diluted plasma from larval Heliothis virescens exhibited a virucidal effect against HzSNPV in vitro, reducing the TCID50 ml−1 by more than 64-fold (from 4·3±3·6×105 to 6·7±0·6×103). The antiviral activity was heat-labile but was unaffected by freezing. In addition, protease inhibitors and specific chemical inhibitors of phenol oxidase or prophenol oxidase activation added to diluted plasma eliminated the virucidal activity. Thus, in the plasma of larval lepidopterans, the enzyme phenol oxidase may act as a constitutive, humoral innate antiviral immune response.

The complete nucleotide sequence of a novel virus is presented here together with serological evidence that it belongs to Kashmir bee virus (KBV). Analysis reveals that KBV is a cricket paralysis-like virus (family Dicistroviridae: genus Cripavirus), with a non-structural polyprotein open reading frame in the 5′ portion of the genome separated by an intergenic region from a structural polyprotein open reading frame in the 3′ part of the genome. The genome also has a polyadenylated tail at the 3′ terminus. KBV is one of several related viruses that also includes acute bee paralysis virus (ABPV). Although KBV and ABPV are about 70 % identical over the entire genome, there are considerable differences between them in significant areas of the genome, such as the 5′ non-translated region (42 % nucleotide identity), between the helicase and 3C-protease domains of the non-structural polyprotein (57 % amino acid identity) and in a 90 aa stretch of the structural polyprotein (33 % amino acid identity). Phylogenetic analyses show that KBV and ABPV isolates fall into clearly separated clades with moderate evolutionary distance between them. Whether these genomic and evolutionary differences are sufficient to classify KBV and ABPV as separate species remains to be determined.

Human rhinoviruses (HRVs; family Picornaviridae), the most frequent causative agents of respiratory infections, comprise more than 100 distinct serotypes. According to previous phylogenetic analysis of the VP4/VP2-coding sequences, all but one of the HRV prototype strains distribute between the two established species, Human rhinovirus A (HRV-A) and Human rhinovirus B (HRV-B). Here, partial sequences of the RNA-dependent RNA polymerase (3D polymerase)-coding gene of 48 HRV prototype strains and 12 field isolates were analysed. The designated division of the HRV strains into the species HRV-A and HRV-B was also seen in the 3D-coding region. Phylogenetically, HRV-B clustered closer to human enterovirus (HEV) species HEV-B, HEV-C and poliovirus than to HRV-A. Intraspecies variation within both HRV-A and HRV-B was greater in the 3D-coding region than in the VP4/VP2-coding region, with the difference maxima reaching 48 % at the nucleotide level and 36 % at the amino acid level in HRV-A and 53 and 35 %, respectively, in HRV-B. Within both species, a few strains formed a separate cluster differing from the majority of strains as much as HEV-B from HEV-C. Furthermore, the tree topology within HRV-A differed from that for VP4/VP2, suggesting possible recombination events in the evolutionary history of the strains. However, all 12 field isolates clustered similarly, as in the capsid region. These results showed that the within-species variation in the 3D region is greater in HRV than in HEV. Furthermore, HRV variation in the 3D region exceeds that in the capsid-coding region.

Efficient translation of human rhinovirus-2 (HRV-2) RNA from its internal ribosome entry site (IRES) depends on the presence of cellular trans-acting factors upstream of N- ras (unr) and polypyrimidine-tract-binding protein. unr contains five cold-shock domains (CSDs) and is predicted to act as an RNA chaperone, allowing the HRV-2 IRES to attain the correct conformation for ribosome binding. To investigate the role of each of the CSDs in IRES-dependent translation, five unr mutants, each harbouring a point mutation in a different CSD, were generated. All five mutants were severely impaired in their ability to bind to the IRES and to stimulate translation from it. This showed that the ability of unr to function as an activator of HRV-2 RNA translation requires the RNA-binding activity of all five CSDs.

Foot-and-mouth disease virus (FMDV) variants adapted to BHK-21 cells showed an expanded host-cell tropism that extended to primate and human cell lines. Virus replication in human HeLa and Jurkat cells has been documented by titration of virus infectivity, quantification of virus RNA, expression of a virus-specific non-structural antigen, and serial passage of virus in the cells. Parallel serial infections of human Jurkat cells with the same variant FMDVs indicates a strong stochastic component in the progression of infection. Chimeric viruses identified the capsid as a genomic region involved in tropism expansion. These results indicate that, contrary to theoretical predictions, replication of an RNA virus in a constant cellular environment may lead to expansion of cellular tropism, rather than to a more specialized infection of the cellular type to which the virus has been adapted.

Hepatitis C virus (HCV) often causes persistent infection in humans. This could be due in part to the effect of viral proteins on cellular gene expression. Earlier observations suggest that the HCV core protein expressed from genotype 1a modulates important cellular genes at the transcriptional level, affects programmed cell death (apoptosis) and promotes cell growth. Recently, different groups of investigators have reported the translation of an ∼16 kDa protein (named F/ARFP/core+1 ORF) from an alternate open reading frame of the HCV core-encoding genomic region. The functional significance of this F protein is presently unknown. Thus, whether the F and core proteins have both shared and distinct functions was investigated here. The experimental observations suggested that the F protein does not significantly modulate c-myc, hTERT and p53 promoter activities, unlike the HCV core protein. Interestingly, the F protein repressed p21 expression. Further studies indicated that the F protein does not inhibit tumour necrosis factor alpha-mediated apoptosis of HepG2 cells or promote rat embryo fibroblast growth. Taken together, these results suggest that the F protein does not share major properties identified previously for the HCV core protein, other than regulating p21 expression.

Dengue virus causes dengue haemorrhagic fever or dengue shock syndrome with a high mortality rate. The genome of dengue virus is a positive-sense, single-stranded RNA encoding three structural and seven non-structural proteins. The core protein is one of the three structural proteins and is the building block of the nucleocapsid of dengue virus. The core protein of dengue virus type 2 (DEN2) is composed of 100 aa with four α-helix domains. An internal hydrophobic domain located at aa 44–60 was identified. The DEN2 core protein was shown to form homodimers. Deletion of aa 1–36 or 73–100 decreased but did not completely abolish the core-to-core homotypic interaction, whereas deletion of a portion (aa 44–60) within aa 37–72 completely abolished the ability of the DEN2 core proteins to interact with each other. A recombinant DEN2 core protein corresponding to aa 37–72 was able to undergo homotypic interaction and bound to a native DEN2 core protein. The results of this study indicated that the homotypic interaction domain of the DEN2 core protein is located at aa 37–72 and that the internal hydrophobic domain located at aa 44–60 plays a pivotal role in core-to-core homotypic interaction.

The interferon-induced murine Mx1 GTPase is a nuclear protein. It specifically inhibits influenza A viruses at the step of primary transcription, a process known to occur in the nucleus of infected cells. However, the exact mechanism of inhibition is still poorly understood. The Mx1 GTPase has previously been shown to accumulate in distinct nuclear dots that are spatially associated with promyelocytic leukaemia protein (PML) nuclear bodies (NBs), but the significance of this association is not known. Here it is reported that, in cells lacking PML and, as a consequence, PML NBs, Mx1 still formed nuclear dots. These dots were indistinguishable from the dots observed in wild-type cells, indicating that intact PML NBs are not required for Mx1 dot formation. Furthermore, Mx1 retained its antiviral activity against influenza A virus in these PML-deficient cells, which were fully permissive for influenza A virus. Nuclear Mx proteins from other species showed a similar subnuclear distribution. This was also the case for the human MxA GTPase when this otherwise cytoplasmic protein was translocated into the nucleus by virtue of a foreign nuclear localization signal. Human MxA and mouse Mx1 do not interact or form heterooligomers. Yet, they co-localized to a large degree when co-expressed in the nucleus. Taken together, these findings suggest that Mx1 dots represent distinct nuclear domains (‘Mx nuclear domains’) that are frequently associated with, but functionally independent of, PML NBs.

The current dogma of influenza accepts that feral aquatic birds are the reservoir for influenza A viruses. Although the genomic information of human influenza A viruses is increasing, little of this type of data is available for viruses circulating in feral waterfowl. This study presents the genetic characterization of 35 viruses isolated from wild Canadian ducks from 1983 to 2000, as the first attempt at a comprehensive genotypic analysis of influenza viruses isolated from feral ducks. This study demonstrates that influenza virus genes circulating in Canadian ducks have achieved evolutionary stasis. The majority of these duck virus genes are clustered in distinct North American clades; however, some H6 and H9 genes are clustered with those from Eurasian viruses. Genes appeared to reassort in a random fashion. None of the genotypes identified remained present throughout all of the years examined and most PA and PB2 genes that crossed over into swine were clustered in one phylogenetic grouping. Additionally, matrix genes were identified that branch very early in the evolutionary tree. These findings demonstrate the diversity of the influenza virus gene pool in Canadian ducks, and suggest that genes which cluster in specific phylogenetic groupings in the PB2 and PA genes can be used for markers of viruses with the potential for crossing the species barrier. A more comprehensive study of this important reservoir is needed to provide further insight into the genomic composition of viruses that crossover the species barrier, which would be a useful component to pandemic planning.

The kinetics of infectious virus (p.f.u.), total virus and virus–Ig complex formation following influenza A/PR8 (H1N1) viral infection was examined in the nasal secretions of naive mice and mice immunized with A/PR8, A/Yamagata (H1N1), A/Guizhou (H3N2) and B/Ibaraki influenza viruses. The total number of virus particles and the number within virus–Ig complexes, captured in advance using an anti-mouse Ig-coated plate, were determined on the basis of viral genome copy number using quantitative RT-PCR. The kinetics of infectious and total virus particle formation, the latter of which increased by 103–104-fold above infectious virus numbers, showed that virus elimination from the nasal area was earlier in A/PR8, A/Yamagata and A/Guizhou-X virus-immunized mice, in decreasing order, compared with naive mice. Early virus elimination correlated with the level of A/PR8 virus-reactive antibodies in immunized mice. Virus elimination coincided with the appearance of virus–Ig complexes shortly after infection. This result suggested that antibodies led to the formation of immune complexes in a dose-dependent manner together with a reduction in number of infectious virus particles. The fact that a large number of virus particles was observed in immune complexes for a wide range antibody levels made it difficult to detect slight differences in virus number within the immune complexes, depending on antibody level. These results suggested that the formation of virus–Ig complexes in virus-immunized mice shortly after infection is involved in early virus elimination, which is determined by the strength of protective immunity against challenge viruses.

Influenza virus still poses a major threat to human health. Despite widespread vaccination programmes and the development of drugs targeting essential viral proteins, the extremely high mutation rate of influenza virus still leads to the emergence of new pathogenic virus strains. Therefore, it has been suggested that cellular cofactors that are essential for influenza virus infection might be better targets for antiviral therapy. It has previously been reported that influenza virus efficiently infects Epstein–Barr virus-immortalized B cells, whereas Burkitt's lymphoma cells are virtually resistant to infection. Using this cellular system, it has been shown here that an active NF-κB signalling pathway is a general prerequisite for influenza virus infection of human cells. Cells with low NF-κB activity were resistant to influenza virus infection, but became susceptible upon activation of NF-κB. In addition, blocking of NF-κB activation severely impaired influenza virus infection of otherwise highly susceptible cells, including the human lung carcinoma cell lines A549 and U1752 and primary human cells. On the other hand, infection with vaccinia virus was not dependent on an active NF-κB signalling pathway, demonstrating the specificity of this pathway for influenza virus infection. These results might be of major importance for both the development of new antiviral therapies and the understanding of influenza virus biology.

NK cells participate in innate immune responses by secreting gamma interferon (IFN-γ) and by destroying virus-infected cells. Here the interaction between influenza A or Sendai virus-infected macrophages and NK cells has been studied. A rapid, cell–cell contact-dependent production of IFN-γ from NK cells cultured with virus-infected macrophages was observed. Expression of the MHC class I-related chain B (MICB) gene, a ligand for NK cell-activating receptor NKG2D, was upregulated in virus-infected macrophages suggesting a role for MICB in the activation of the IFN-γ gene in NK cells. IL12Rβ2, IL18R and T-bet mRNA synthesis was enhanced in NK cells cultured with virus-infected macrophages. Upregulation of these genes was dependent on macrophage-derived IFN-α. In contrast to IL12Rβ2, expression of WSX-1/TCCR, a receptor for IL27, was reduced in NK cells in response to virus-induced IFN-α. In conclusion, these results show that virus-infected macrophages activate NK cells via cytokines and direct cellular interactions and further emphasize the role of IFN-α in the activation of innate immunity.

CD8+ T lymphocytes play a major role in the clearance of respiratory syncytial virus (RSV) infections. To be able to study the primary CTL response in RSV-infected children, epitopes presented by a set of commonly used HLA alleles (HLA-A1, -A3, -B44 and -B51) were searched for. Five epitopes were characterized derived from the matrix (M), non-structural (NS2) and second matrix (M2) proteins of RSV. All epitopes were shown to be processed and presented by RSV-infected antigen-presenting cells. HLA-A1 tetramers for one of these epitopes derived from the M protein were constructed and used to quantify and phenotype the memory CD8+ T cell pool in a panel of healthy adult donors. In about 60 % of the donors, CD8+ T cells specific for the M protein could be identified. These cells belonged to the memory T cell subset characterized by expression of CD27 and CD28, and down-regulation of CCR7 and CD45RA. The frequency of tetramer-positive cells varied between 0·4 and 3 per 104 CD8+ T cells in PBMC of healthy asymptomatic adult donors.

Many paramyxoviruses encode non-essential accessory proteins that are involved in the regulation of virus replication and inhibition of cellular antiviral responses. It has been suggested that the P gene mRNA of Newcastle disease virus (NDV) encodes an accessory protein – the so-called X protein – by translation initiation at a conserved in-frame AUG codon at position 120. Using a monoclonal antibody that specifically detected the P and X proteins, it was shown that an accessory X protein was not expressed in NDV-infected cells. Recombinant NDV strains in which the AUG was changed into a GCC (Ala) or GUC (Val) codon were viable but showed a reduction in virulence, probably because the amino acid change affected the function of the P and/or V protein.

Dendritic cells (DCs) have been used successfully to induce CD8 T cells that control virus infections and growth of tumours. The efficacy of DC-mediated immunization for the control of neurotropic Borna disease virus (BDV) in mice was evaluated. Certain strains of mice only rarely develop spontaneous neurological disease, despite massive BDV replication in the brain. Resistance to disease is due to immunological ignorance toward BDV antigen in the central nervous system. Ignorance in mice can be broken by immunization with DCs coated with TELEISSI, a peptide derived from the N protein of BDV, which represents the immunodominant cytotoxic T lymphocyte epitope in H-2k mice. Immunization with TELEISSI-coated DCs further induced solid protective immunity against intravenous challenge with a recombinant vaccinia virus expressing BDV-N. Interestingly, however, this immunization scheme induced only moderate protection against intracerebral challenge with BDV, suggesting that immune memory raised against a shared antigen may be sufficient to control a peripherally replicating virus, but not a highly neurotropic virus that is able to avoid activation of T cells. This difference might be due to the lack of BDV-specific CD4 T cells and/or inefficient reactivation of DC-primed, BDV-specific CD8 T cells by the locally restricted BDV infection. Thus, a successful vaccine against persistent viruses with strong neurotropism should probably induce antiviral CD8 (as well as CD4) T-cell responses and should favour the accumulation of virus-specific memory T cells in cervical lymph nodes.

The rate of mutation during retrovirus replication is high. Mutations can occur during transcription of the viral genomic RNA from the integrated provirus or during reverse transcription from viral RNA to form viral DNA or during replication of the proviral DNA as the host cell is dividing. Therefore, three polymerases may all contribute to retroviral evolution: host RNA polymerase II, viral reverse transcriptases and host DNA polymerases, respectively. Since the rate of mutation for host DNA polymerase is very low, mutations are more likely to be caused by the host RNA polymerase II and/or the viral reverse transcriptase. A system was established to detect the frequency of frame-shift mutations caused by cellular RNA polymerase II, as well as the rate of retroviral mutation during a single cycle of replication in vivo. In this study, it was determined that RNA polymerase II contributes less than 3 % to frame-shift mutations that occur during retrovirus replication. Therefore, the majority of frame-shift mutations detected within the viral genome are the result of errors during reverse transcription.