Comparison of functional domains of the cytotoxic protein NS3 of different orbiviruses

Abstract

African horsesickness (AHS), caused by the African horsesickness virus (AHSV) (Coetzer&Erasmus, 1994), has 10 segments of double stranded RNA (Verwoerd et al., 1970). These encode seven structural proteins, namely VP1, VP2, VP3, VP4, VP5, VP6 and VP7 (Verwoerd et al., 1972; Huismans&Van Dijk, 1990) as well as four nonstructural proteins (NS1, NS2, NS3 and NS3A) (Van Dijk&Huismans, 1988). NS3 is cytotoxic when expressed in insect cells causing membrane permeabilisation and cell death (Van Staden et al., 1995). This characteristic was postulated to be associated with the two hydrophobic domains (Van Staden et al., 1998; Van Niekerk et al., 2001(a)). Although NS3 has the same basic structure and function in all orbiviruses it differs greatly on nucleic acid sequence level as well as in the degree of variation that is found within the protein (Sailleau et al., 1997; Huismans et al., 2004). Therefore the main aim of this study was to identify which regions or amino acids remain conserved and might therefore play a role in the structure or function of NS3. The focus was mainly on the two hydrophobic domains as well as the spacer region between the hydrophobic domains. The sequence variation analyses between AHSV, BTV and EHDV NS3 revealed that there is a conserved threonine-serine motif in HD1, a conserved aspartic acid at the beginning of the spacer region as well as a conserved asparagine in HD2. The conservation of these residues between the different serogroups indicates that they might have some functional or structural role in the NS3 protein. NS3 protein sequences in the AHSV ã phylogenetic cluster has three extra amino acids in the spacer region at aa 149-151. There is also a significant difference in the spacer region length between EEV NS3 and AHSV NS3, which might have an effect on the topology of the respective proteins. In the spacer region of AHSV NS3, three regions were targeted for mutation analyses: substitution of the conserved isoleucine (aa 141) with a neutral amino acid, deletion of three amino acids (aa 149-151), and insertion of 15 additional amino acids. Three mutants were generated and assayed: the IÄS mutant, the KGDdel mutant and the 15 aa insert mutant. The cytotoxicity of the IÄS and KGD deletion mutants was exactly the same as the native NS3 protein, while the 15 aa insertion resulted in a slightly decreased cytotoxicity. This was the same as the EEV NS3 cytotoxicity level in insect cells. The fluorescent localisation and subcellular fractionation studies showed that both the IÄS mutation and 15 aa insertion affected the localisation of NS3 within the cell and also increased the solubility of the protein. This indicated that even though the membrane localisation of NS3 was disrupted it still had a cytotoxic role in the cell. It could be that wild type NS3 has both a membrane associated and a soluble fraction within cells, and that the soluble fraction is indeed responsible for the observed cytotoxicity and not the membrane fraction as assumed previously. It is therefore postulated that NS3 might have different domains that have various roles within the cell.