Search

Murray Valley encephalitis virus (MVEV) is a mosquito-borne flavivirus endemic to Australia and Papua New Guinea. Most strains of MVEV cause potentially fatal cases of encephalitis in humans and horses, and have been shown to be highly neuroinvasive in weanling mice. In contrast, the naturally occurring subtype Alfuy virus (ALFV) has never been associated with human disease, nor is it neuroinvasive in weanling mice, even at high doses. To identify viral factors associated with ALFV attenuation, a chimeric infectious clone was constructed containing the structural genes premembrane (prM) and envelope (E) of ALFV swapped into the MVEV genome. The resulting virus (vMVEV/ALFVstr) was no longer neuroinvasive in mice, suggesting that motifs within prM–E of ALFV confer attenuation. To define these motifs further, mutants were constructed by targeting divergent sequences between the MVEV and ALFV E proteins that are known markers of virulence in other encephalitic flaviviruses. MVEV mutants containing a unique ALFV sequence in the flexible hinge region (residues 273–277) or lacking the conserved glycosylation site at position 154 were significantly less neuroinvasive in mice than wild-type MVEV, as determined by delayed time to death or increased LD50. Conversely, when the corresponding MVEV sequences were inserted into the vMVEV/ALFVstr chimera, the mutant containing the MVEV hinge sequence was more neuroinvasive than the parental chimera, though not to the same level as wild-type MVEV. These results identify the hinge region and E protein glycosylation as motifs that contribute to the attenuation of ALFV.

The West Nile virus (WNV) NS5 protein contains a methyltransferase (MTase) domain involved in RNA capping and an RNA-dependent RNA polymerase (RdRp) domain essential for virus replication. Crystal structures of individual WNV MTase and RdRp domains have been solved; however, the structure of full-length NS5 has not been determined. To gain more insight into the structure of NS5 and interactions between the MTase and RdRp domains, we generated a panel of seven monoclonal antibodies (mAbs) to the NS5 protein of WNV (Kunjin strain) and mapped their binding sites using a series of truncated NS5 proteins and synthetic peptides. Binding sites of four mAbs (5D4, 4B6, 5C11 and 6A10) were mapped to residues 354–389 in the fingers subdomain of the RdRp. This is consistent with the ability of these mAbs to inhibit RdRp activity in vitro and suggests that this region represents a potential target for RdRp inhibitors. Using a series of synthetic peptides, we also identified a linear epitope (bound by mAb 5H1) that mapped to a 13 aa stretch surrounding residues 47 and 49 in the MTase domain, a region predicted to interact with the palm subdomain of the RdRp. The failure of one mAb (7G6) to bind both N- and C-terminally truncated NS5 recombinants indicates that the antibody recognizes a conformational epitope that requires the presence of residues in both the MTase and RdRp domains. These data support a structural model of the full-length NS5 molecule that predicts a physical interaction between the MTase and the RdRp domains.