Abstract:
Factors that determine the molecular basis of African horsesickness virus (AHSV) virulence are unclear. Several proteins and different events in the viral replication cycle together with different environmental factors are likely to contribute to the virulence phenotype. The aim of this investigation was to study the possible contribution of AHSV nonstructural protein NS3 to virus virulence. NS3 is a cytotoxic protein that localizes in areas of plasma membrane disruption and is assumed to be associated with the release of virus particles from infected cells. A conserved feature in all AHSV NS3 proteins is the presence of two hydrophobic domains. To investigate whether these hydrophobic domains interact with membranes, cell surface localization and membrane association studies were conducted on NS3 and the two hydrophobic domain mutant forms of NS3. Results indicated that mutations in either hydrophobic domain did not prevent membrane targeting but abolished membrane anchoring. This prevented cell surface localization and obviated the cytotoxic effect of NS3. AHSV NS3 is a highly variable protein. This level of variation may influence virulence properties. To compare the NS3 sequence variation level of South African AHSVs and bluetongue viruses (BTV/s) with those previously described, the NS3 protein sequences of field isolates, reference and vaccine strains were determined and analysed. The variation level of AHSV NS3 was found to be higher than previously reported. Comparison of the AHSV vaccine and field strain NS3 sequences revealed no obvious NS3 virulence marker. The level of AHSV NS3 sequence variation could distinguish between field isolates and vaccine strains and differentiate between sub-populations within a serotype. The inferred phylogeny of AHSV NS3 corresponded well with the described NS3 phylogenetic clusters with exception of AHSV-8 and one AHSV-6 encoded NS3. BTV NS3 inferred phylogeny indicated that the three BTV NS3 clusters that occur as geographically defined entities are all simultaneously present in S.A. This would suggest that BTV originated in southern Africa. To investigate the impact of minor NS3 variation on AHSV virulence, the membrane permeabilization properties of AHSV-2 vaccine and virulent NS3 proteins were compared. The AHSV-2 S10 genes were cloned and NS3 was expressed using the BAC-TO-BAC system. The membrane permeabilization effect of NS3 expression was monitored by calcium influx into insect cells. Expression of either the vaccine or virulent associated AHSV-2 NS3 protein equally increased membrane permeability. The NS3 sequence variation therefore had a limited influence on membrane permeabilization properties and the virulence status of the vaccine and virulent AHSV-2 strains in this study. The possible effect of larger NS3 sequence variation levels on AHSV virulence was investigated by NS3 associated properties such as virus release and membrane permeabilization. AHSV infections increased the permeability of cell membranes and the different strains had different virus release properties. Virus release was not exclusively related to increased membrane permeability of infected cells or to the yield of the virus in the infected cells. The level of NS3 variation may influence AHSV release mechanisms and membrane permeabilization properties thereby contributing to AHSV virulence properties.