- Volume 88, Issue 11, 2007

Hepatitis C virus (HCV) is a major cause of liver disease worldwide and there is a pressing need for the development of a preventative vaccine as well as new treatments. It was recently demonstrated that the mouse monoclonal antibody (mAb) AP33 potently neutralizes infectivity of HCV pseudoparticles (HCVpp) carrying E1E2 envelopes representative of all of the major genotypes of HCV. This study determined the prevalence of human serum antibodies reactive to the region of HCV E2 recognized by AP33. Antibodies recognizing this region were present in less than 2.5 % of sera obtained from individuals with chronic HCV infection. A similar prevalence was found in a smaller cohort of individuals who had experienced an acute infection, suggesting that AP33-like antibodies do not play a major role in natural clearance of HCV infection. Sera exhibited different patterns of reactivity to a panel of peptides representing circulating variants, highlighting the presence of distinct epitopes in this region. Only two sera contained antibodies that could recognize a specific AP33-reactive peptide mimotope. AP33-like antibodies made a measurable contribution to the ability of these sera to inhibit E2–CD81 interaction, but not to the overall neutralization of cell entry. Together, these data show that antibodies to the AP33 epitope are not commonly generated during natural infection and that generation of such antibodies via vaccination may require modified immunogens to focus the generation of specific antibodies. Importantly, individuals harbouring AP33-like antibodies are an important potential source of human mAbs for future therapeutic development.

Classical swine fever virus (CSFV) is a member of the genus Pestivirus in the family Flaviviridae. The Npro product of CSFV targets the host's innate immune response and can prevent the production of type I interferon (IFN). The mechanism by which CSFV orchestrates this inhibition was investigated and it is shown that, like the related pestivirus bovine viral diarrhea virus (BVDV), this involves the Npro protein targeting interferon regulatory factor-3 (IRF-3) for degradation by proteasomes and thus preventing IRF-3 from activating transcription from the IFN-β promoter. Like BVDV, the steady-state levels of IRF-3 mRNA are not reduced markedly by CSFV infection or Npro overexpression. Moreover, IFN-α stimulation of CSFV-infected cells induces the antiviral protein MxA, indicating that, as in BVDV-infected cells, the JAK/STAT pathway is not targeted for inhibition.

Classical swine fever virus (CSFV)–macrophage interactions during infection were analysed by examining macrophage transcriptional responses via microarray. Eleven genes had increased mRNA levels (>2.5-fold, P<0.05) in infected cell cultures, including arginase-1, an inhibitor of nitric oxide production, phosphoinositide 3-kinase, chemokine receptor 4 and interleukin-1β. Lower levels of nitric oxide and increased arginase activity were found in CSFV-infected macrophages. These changes in gene expression in macrophages suggest viral modulation of host expression to suppress nitric oxide production.

The human MxA protein is a type I and III interferon (IFN)-induced protein with proven antiviral activity against RNA viruses. In this study, we investigated the effect of MxA expression on the replication of West Nile Virus strain Kunjin (WNVKUN). Pretreatment of A549 cells with IFN-α lead to increased expression of MxA, which contributed to inhibition of WNVKUN replication and secretion. However, in Vero cells stably expressing the MxA protein, WNVKUN replication, maturation and secretion was not inhibited. Biochemical and subcellular localization studies of WNVKUN proteins and MxA suggest that the MxA activity was not compromised by a flavivirus-encoded antagonist. Instead, we show that characteristic membranous structures induced during WNVKUN replication provide partial protection from MxA, possibly by ‘hiding’ WNVKUN replication components. This distinct compartmentalization of viral replication and components of the cellular antiviral response may be an evolutionary mechanism by which flaviviruses can hide from host surveillance.

Alpha- and flaviviruses contain class II fusion proteins, which form ion-permeable pores in the target membrane during virus entry. The pores generated during entry of the alphavirus Semliki Forest virus have been shown previously to be blocked by lanthanide ions. Here, analyses of the influence of rare earth ions on the entry of the flaviviruses West Nile virus and Uganda S virus revealed an unexpected effect of lanthanide ions. The results showed that a 30 s treatment of cells with an appropriate lanthanide ion changed the cellular chemistry into a state in which the cells no longer supported the multiplication of flaviviruses. This change occurred in cells treated before, during or after infection, did not inhibit multiplication of Semliki Forest virus and did not interfere with host-cell multiplication. The change was generated in vertebrate and insect cells, and was elicited in the presence of actinomycin D. In vertebrate cells, the change was elicited specifically by La3+, Ce3+, Pr3+ and Nd3+. In insect cells, additional lanthanide ions had this activity. Further analyses showed that lanthanide ion treatment blocked the ability of the host cell to support the replication of flavivirus RNA. These results open two areas of research: the study of molecular alterations induced by lanthanide ion treatment in uninfected cells and the analysis of the resulting modifications of the flavivirus RNA replicase complex. The findings possibly open the way for the development of a general chemotherapy against flavivirus diseases such as Dengue fever, Japanese encephalitis, West Nile fever and yellow fever.

During enterovirus infection, host cell membranes are rigorously rearranged and modified. One ubiquitously expressed lipid-modifying enzyme that might contribute to these alterations is phospholipase D (PLD). Here, we investigated PLD activity in coxsackievirus-infected cells. We show that PLD activity is not required for efficient coxsackievirus RNA replication. Instead, PLD activity rapidly decreased upon infection and upon ectopic expression of the viral 3A protein, which inhibits the PLD activator ADP-ribosylation factor 1. However, similar decreases were observed during infection with coxsackieviruses carrying defective mutant 3A proteins. Possible causes for the reduction of PLD activity and the biological consequences are discussed.

Cells in which the appropriate viral receptor cannot be detected may paradoxically act as a host to the virus. For example, RD cells are often considered to be non-permissive for infection with coxsackievirus and adenovirus receptor (CAR)-dependent group B coxsackieviruses (CVB), insofar as inoculated cell monolayers show little or no cytopathic effect (CPE) and immunohistological assays for CAR have been consistently negative. Supernatants recovered from RD cells exposed to CVB, however, contained more virus than was added in the initial inoculum, indicating that productive virus replication occurred in the monolayer. When infected with a recombinant CVB type 3 (CVB3) chimeric strain expressing S-Tag within the viral polyprotein, 4–11 % of RD cells expressed S-Tag over 48 h. CAR mRNA was detected in RD cells by RT-PCR, and CAR protein was detected on Western blots of RD lysates; both were detected at much lower levels than in HeLa cells. Receptor blockade by an anti-CAR antibody confirmed that CVB3 infection of RD cells was mediated by CAR. These results show that some RD cells in the culture population express CAR and can thereby be infected by CVB, which explains the replication of CAR-dependent CVB in cell types that show little or no CPE and in which CAR has not previously been detected. Cells within cultures of cell types that have been considered non-permissive may express receptor transiently, leading to persistent replication of virus within the cultured population.

Hepatitis A virus (HAV) is a hepatotropic member of the family Picornaviridae. Previous studies suggested that HAV may evolve more slowly than other members of the family. To estimate HAV substitution rates precisely, we used a Bayesian Markov chain Monte Carlo (MCMC) approach on temporally sampled HAV VP1 full-length sequences from strains isolated in France. A mean rate of evolutionary change of 9.76×10−4 nucleotide substitution per site per year was found. The results also revealed that the synonymous rate found for HAV is lower than that of other members of the family. Bayesian skyline plots revealed a sharp decline in the effective number of infections in 1996, coinciding with the introduction of HAV vaccine.

Internal initiation of translation from the human rhinovirus-2 (HRV-2) internal ribosome entry site (IRES) is dependent upon host cell trans-acting factors. The multiple cold shock domain protein Unr and the polypyrimidine tract-binding protein have been identified as synergistic activators of HRV-2 IRES-driven translation. In order to investigate the mechanism by which Unr acts in this process, we have mapped the binding sites of Unr to two distinct secondary structure domains of the HRV-2 IRES, and have identified specific nucleotides that are involved in the binding of Unr to the IRES. The data suggest that Unr acts as an RNA chaperone to maintain a complex tertiary IRES structure required for translational competency.

Internal ribosome entry site (IRES) elements were described in picornaviruses as an essential region of the viral RNA. Understanding of IRES function requires a detailed knowledge of each step involved in the internal initiation process, from RNA folding and IRES–protein interaction to ribosome recruitment. Thus, deciphering IRES accessibility to external agents due to RNA structural features, as well as RNA–protein protection within living cells, is of primary importance. In this study, two chemical reagents, dimethylsulfate (DMS) and aminomethylpsoralen, have been used to footprint the entire IRES of foot-and-mouth disease virus (FMDV) in living cells; these reagents enter the cell membrane and interact with nucleic acids in a structure-dependent manner. For FMDV, as in other picornaviruses, viral infection is dependent on the correct function of the IRES; therefore, the IRES region itself constitutes a useful target of antiviral drugs. Here, the in vivo footprint of a picornavirus IRES element in the context of a biologically active mRNA is shown for the first time. The accessibility of unpaired adenosine and cytosine nucleotides in the entire FMDV IRES was first obtained in vitro by DMS probing; subsequently, this information was used to interpret the footprint data obtained in vivo for the mRNA encompassing the IRES element in the intercistronic space. The results of DMS accessibility and UV–psoralen cross-linking studies in the competitive cellular environment provided evidence for differences in RNA structure from data obtained in vitro, and provided essential information to identify appropriate targets within the FMDV IRES aimed at combating this important pathogen.

Two distinct pathways of gamma interferon (IFN-γ) production have been found in mice infected with lactate dehydrogenase-elevating virus. Both pathways involve natural killer cells. The first is mostly interleukin-12-independent and is not controlled by type I interferons. The second, which is suppressed by type I interferons, leads to increased levels of IFN-γ production and requires the secretion of interleukin-12. This regulation of IFN-γ production by type I interferons may help to control indirect pathogenesis induced by this cytokine.

The orf3a (also called X1 or U274) gene is the largest unique open reading frame in the severe acute respiratory syndrome coronavirus genome and has been proposed to encode a protein with three transmembrane domains and a large cytoplasmic domain. Recent work has suggested that the 3a protein may play a structural role in the viral life cycle, although the mechanisms for this remain uncharacterized. Here, the expression of the 3a protein in various in vitro systems is shown, it has been localized to the Golgi region and its membrane topology in transfected cells has been confirmed. Three potential caveolin-1-binding sites were reported to be present in the 3a protein. By using various biochemical, biophysical and genetic techniques, interaction of the 3a protein with caveolin-1 is demonstrated. Any one of the potential sites in the 3a protein was sufficient for this interaction. These results are discussed with respect to the possible roles of the 3a protein in the viral life cycle.

The complete nucleotide sequence of Middelburg virus (MIDV) was determined for strain MIDV-857 from Zimbabwe. The isolation of this virus in 1993 from a horse that died showing severe clinical signs represents the first indication that MIDV can cause severe disease in equids. Full-length cDNA copies of the viral genome were successfully synthesized by an innovative RT-PCR amplification approach using an ‘anchor primer’ combined with the SMART methodology described previously for the synthesis of full-length cDNA copies from genome segments of dsRNA viruses. The MIDV-857 genome is 11 674 nt, excluding the 5′-terminal cap structure and poly(A) tail (which varies in length from approximately 180 to approximately 220 residues). The organization of the genome is like that of other alphaviruses, including a read-through stop codon between the nsP3 and nsP4 genes. However, phylogenetic analyses of the structural protein amino acid sequences suggested that the MIDV E1 gene was generated by recombination with a Semliki Forest virus-like virus. This hypothesis was supported by bootscanning analysis using a recombination-detection program. The 3′ untranslated region of MIDV-857 also contains a 112 nt duplication. This study reports the first full-length sequence of MIDV, which was obtained from a single RT-PCR product.

Five out of sixteen domestic pigeons, inoculated oculo-nasally with a high dose of highly pathogenic avian influenza virus A/chicken/Indonesia/2003 (H5N1), developed clinical signs and neurological lesions leading to death of three pigeons 5–7 days after inoculation [ Klopfleisch, R., Werner, O., Mundt, E., Harder, T. & Teifke, J. P. (2006). Vet Pathol 43, 463–470 ]. H5N1 virus was recovered from all organs sampled from two apparently healthy pigeons at 3 days post-infection and from the three pigeons which died spontaneously. All surviving birds shed virus via the oropharynx and the cloaca at minimal titres and seroconverted. Sentinel chickens reared in direct contact to the pigeons neither developed clinical signs nor seroconverted to the H5N1 virus.

Genetic analysis of all eight genes of two Nanchang avian influenza viruses, A/Duck/Nanchang/1681/92 (H3N8-1681) and A/Duck/Nanchang/1904/92 (H7N1-1904), isolated from Jiangxi province, China, in 1992, showed that six internal genes of H3N8-1681 virus and five internal (except NS gene) genes of H7N1-1904 virus were closely similar to A/Goose/Guangdong/1/96 (H5N1) virus, the first highly pathogenic avian influenza (HPAI) virus of subtype H5N1 isolated in Asia. The neuraminidase (NA) gene of Gs/Gd/1/96 had the highest genetic similarity with A/Duck/Hokkaido/55/96 (H1N1-55) virus. The haemagglutinin (HA) gene of Gs/Gd/1/96 virus might have originated as a result of mutation of H5 HA gene from A/Swan/Hokkaido/51/96 (H5N3-51)-like viruses. The PA gene of H5N3-51 virus had the highest similarity with Gs/Gd/1/96. This study explains the origin of first Asian HPAI H5N1 virus in Guangdong by the reassortment of Nanchang (close to Guangdong) and Hokkaido (Japan) (H1N1-55 and H5N3-51) viruses. Genetic characteristics of donor and recipient viruses were also studied.

Infectious salmon anemia virus (ISAV) is classified in the genus Isavirus of the family Orthomyxoviridae. Although virulence variation of ISAV can be demonstrated experimentally in fish, virus strain identification is ambiguous because the correlates of pathogenicity and/or antigenicity of ISAV are not well defined. Thirteen ISAV isolates characterized for their ability to kill fish were used to search for markers of virulence on the virus surface glycoprotein genes; haemagglutinin-esterase (HE) and fusion (F) protein genes. A single amino acid change N164D in the putative globular head of the HE protein, and a deletion/insertion of ≤13 aa with the presence of a specific motif 352FNT354 in the highly polymorphic region spanning residues 337V to M372 in the HE protein stalk, in combination with a specific motif 265YP266 very close to the trypsin-cleavage site 267RA/G268 of the precursor F0 protein were correlated with reduced cytopathogenicity and reduced virulence for Atlantic salmon. Phylogenetic analysis suggests that the original ancestral ISAV was virulent. The virulence of the North American genotype has not changed much, whereas the European genotype evolved into two genogroups, the real-European genogroup that is still virulent and the European-in-North America genogroup, which is of lower virulence. A novel phylogenetic software program, backtrack, estimated that the North American and European genotypes diverged between 1879 and 1891, whereas the European-in-North America genogroup diverged from the real-European genogroup between 1976 and 1988. This direction of evolution supports insertion of specific motifs in the HE protein, resulting in ISAV attenuation.

Rodent brain-adapted measles virus (MV) strains, such as CAM/RB and recombinant MVs based on the Edmonston strain containing the haemagglutinin (H) of CAM/RB, cause acute encephalitis after intracerebral infection of newborn rodents. We have demonstrated that rodent neurovirulence is modulated by two mutations at amino acid positions 195 and 200 in the H protein, one of these positions (200) being a potential glycosylation site. In order to analyse the effects of specific amino acids at these positions, we introduced a range of individual and combined mutations into the open reading frame of the H gene to generate a number of eukaryotic expression plasmids. The functionality of the mutant H proteins was assessed in transfected cells and by generating recombinant viruses. Interestingly, viruses caused acute encephalitis only if the amino acid Ser at position 200 was coupled with Gly at position 195, whereas viruses with single or combined mutations at these positions, including glycosylation at position 200, were attenuated. Neurovirulence was associated with virus spread and induction of neuronal apoptosis, whereas attenuated viruses failed to infect brain cells. Similar results were obtained by using primary brain-cell cultures. Our findings indicate that a structural alteration in the stem 2 region of the H protein at position 195 or 200 interferes with infectivity of rodent neurons, and suggest that the interaction of the viral attachment protein with cellular receptors on neurons is affected.

Acute-phase sera from >5 % of cases of haemorrhagic fever with renal syndrome occurring annually in Korea have been found to exhibit a fourfold or higher antibody titre to Puumala virus (PUUV) than to Hantaan virus (HTNV) by double-sandwich IgM ELISA, suggesting the existence of a PUUV-related hantavirus. Based on the phylogenetic relationships among arvicolid rodents, the royal vole (Myodes regulus) was targeted as a likely reservoir host of hantavirus. Using RT-PCR, a genetically distinct hantavirus, designated Muju virus (MUJV), was detected in lung tissue of royal voles, captured in widely separated geographical regions in Korea during 1996–2007. Pairwise analysis of the full-length S (1857 nt) and M (3634 nt) segments of MUJV indicated approximately 77 % sequence similarity with PUUV. At the amino acid level, MUJV differed from PUUV by 5.5–6.9 % (nucleocapsid) and 10.0–11.6 % (Gn and Gc envelope glycoproteins). Interstrain variation of MUJV sequences from royal voles captured in different regions suggested geographic-specific clustering. Neutralizing antibody titres against PUUV were two- to sixfold higher than to HTNV in sera of MUJV-infected Myodes regulus. Although virus isolation attempts were unsuccessful, the collective data indicate that MUJV is a distinct hantavirus species.

We previously showed that mouse adaptation of cDNA-derived Borna disease virus (BDV) strain He/80FR was associated exclusively with mutations in the viral polymerase complex. Interestingly, independent mouse adaptation of non-recombinant He/80 was correlated with different alterations in the polymerase and mutations in the viral glycoprotein. We used reverse genetics to demonstrate that changes in the polymerase which improve enzymatic activity represent the decisive host range mutations. The glycoprotein mutations did not confer replication competence in mice, although they slightly improved viral performance if combined with polymerase mutations. Our findings suggest that the viral polymerase restricts the host range of BDV.