NS4B x NS3/NS4A

West Nile virus (WNV) is a single-stranded positive-sense RNA virus (+ssRNA) belonging to the genus Orthoflavivirus. Its enzootic cycle involves mosquito vectors, mainly Culex, and wild birds as reservoir hosts, while mammals, such as humans and equids, are incidental dead-end hosts. It was first discovered in 1934 in Uganda, and since 1999 has been responsible for frequent outbreaks in humans, horses and wild birds, mostly in America and in Europe. Virus spread, as well as outbreak severity, can be influenced by many ecological factors, such as reservoir host availability, biodiversity, movements and competence, mosquito abundance, distribution and vector competence, by environmental factors such as temperature, land use and precipitation, as well as by virus genetic factors influencing virulence or transmission. Former studies have investigated WNV factors of virulence, but few have compared viral genetic determinants of pathogenicity in different host species, and even fewer have considered the genetic drivers of virus invasiveness and excretion in Culex vector. In this study, we characterized WNV genetic factors implicated in the difference in virulence observed in two lineage 1 WNV strains from the Mediterranean Basin, the first isolated during a significant outbreak reported in Israel in 1998, and the second from a milder outbreak in Italy in 2008. We used an innovative and powerful reverse genetic tool, e.g., ISA (infectious subgenomic amplicons) to generate chimeras between Israel 1998 and Italy 2008 strains, focusing on non-structural (NS) proteins and the 3'UTR non-coding region. We analyzed the replication of these chimeras and their progenitors in mammals, in BALB/cByJ mice, and vector competence in Culex (Cx.) pipiens mosquitoes. Results obtained in BALB/cByJ mice suggest a role of the NS2B/NS3/NS4B/NS5 genomic region in viral attenuation in mammals, while NS4B/NS5/3'UTR regions are important in Cx. pipiens infection and possibly in vector competence.

Replication of the genotype 2 hepatitis C virus (HCV) requires hyper-phosphorylation of the nonstructural protein NS5A. It has been known that NS5A hyper-phosphorylation results from the phosphorylation of a cluster of highly conserved serine residues (S2201, S2208, S2211, and S2214) in a sequential manner. It has also been known that NS5A hyper-phosphorylation requires an NS3 protease encoded on one single NS3-5A polyprotein. It was unknown whether NS3 protease participates in the above sequential phosphorylation process. Using an inventory of antibodies specific to S2201, S2208, S2211, and S2214 phosphorylation, we found that protease-dead S1169A mutation abrogated NS5A hyper-phosphorylation and phosphorylation at all serine residues measured, consistent with the role of NS3 in NS5A sequential phosphorylation. These effects were not rescued by a wildtype NS3 protease provided in trans by another molecule. Mutations (T1661R, T1661Y, or T1661D) that prohibited proper cleavage at the NS3-4A junction, too, abolished NS5A hyper-phosphorylation and phosphorylation at all serine residues, whereas mutations at the other cleavage sites, NS4A-4B (C1715S) or NS4B-5A (C1976F), did not. In fact, any combinatory mutations that prohibited NS3-4A cleavage (T1661Y/C1715S or T1661Y/C1976F) abrogated NS5A hyper-phosphorylation and phosphorylation at all serine residues. In the C1715S/C1976F double mutant that resulted in an NS4A-NS4B-NS5A fusion polyprotein, a hyper-phosphorylated band was observed and was phosphorylated at all serine residues. We conclude that NS3-mediated auto-cleavage at the NS3-4A junction is critical to NS5A hyper-phosphorylation at S2201, S2208, S2211, and S2214 and that NS5A hyper-phosphorylation could occur in an NS4A-NS4B-NS5A polyprotein.Importance For twenty some years, the HCV protease NS3 was implicated in NS5A hyper-phosphorylation. We now show that it was the NS3-mediated cis-cleavage at the NS3-4A junction that permits NS5A hyper-phosphorylation at serine 2201, 2208, 2211, and 2214, leading to hyper-phosphorylation, which is a necessary condition for genotype 2 HCV replication. We further show that NS5A may already be phosphorylated at the above serine residues right after NS3-4A cleavage and before NS5A was released from the NS4A-5A polyprotein. Our data suggest that the dual functional NS3, a protease and an ATP-binding RNA helicase, could have a direct or indirect role in NS5A hyper-phosphorylation.

Flaviviruses such as yellow fever, dengue or Zika viruses are responsible for significant human and veterinary diseases worldwide. These viruses contain an RNA genome, prone to mutations, which enhances their potential to emerge as pathogens. Bamaga virus (BgV) is a mosquito-borne flavivirus in the yellow fever virus group that we have previously shown to be host-restricted in vertebrates and horizontally transmissible by Culex mosquitoes. Here, we aimed to characterise BgV host-restriction and to investigate the mechanisms involved. We showed that BgV could not replicate in a wide range of vertebrate cell lines and animal species. We determined that the mechanisms involved in BgV host-restriction were independent of the type-1 interferon response and RNAse L activity. Using a BgV infectious clone and two chimeric viruses generated as hybrids between BgV and West Nile virus, we demonstrated that BgV host-restriction occurred post-cell entry. Notably, BgV host-restriction was shown to be temperature-dependent, as BgV replicated in all vertebrate cell lines at 34°C but only in a subset at 37°C. Serial passaging of BgV in Vero cells resulted in adaptive mutants capable of efficient replication at 37°C. The identified mutations resulted in amino acid substitutions in NS4A-S124F, NS4B-N244K and NS5-G2C, all occurring close to a viral protease cleavage site (NS4A/2K and NS4B/NS5). These mutations were reverse engineered into infectious clones of BgV, which revealed that NS4B-N244K and NS5-G2C were sufficient to restore BgV replication in vertebrate cells at 37°C, while NS4A-S124F further increased replication efficiency. When these mutant viruses were injected into immunocompetent mice, alongside BgV and West Nile virus chimeras, infection and neurovirulence were enhanced as determined by clinical scores, seroconversion, micro-neutralisation, viremia, histopathology and immunohistochemistry, confirming the involvement of these residues in the attenuation of BgV. Our studies identify a new mechanism of host-restriction and attenuation of a mosquito-borne flavivirus.