A study of the African horse sickness virus using High Resolution Melt, multivariate and phylogenetic analyses for a potential serotyping assay.

Abstract

African horse sickness (AHS) is a viral disease that afflicts all equine species and has a 90% mortality rate in unvaccinated horses. The disease has a devastating effect on the national herd of South Africa each year and affects both the sport and racehorse industries, including the export of prized bloodstock, as well as the rural and subsistence economies that depend on animal traction. Transmitted by the Culicoides spp. of biting midge, the virus belongs to the Orbivirus genus of the Reoviridae family with nine known serotypes and ten genome segments. Segment 2 (which encodes VP2) is responsible for serotype determination while segment 10 (which encodes NS3) is merely serotype-divergent. Knowledge of the seroprevalence of the virus is poor. The increasing reluctance of horse owners to use the registered vaccine due to perceived inefficacy is of concern. As a means to increase knowledge output in this regard, and potentially provide a service to horse owners, a rapid serotyping assay is sought based on High Resolution Melt (HRM) analysis. HRM analysis is a powerful tool that is based on the release of a DNA intercalating dye from polymerase chain reaction products through gradual and controlled heating. The dye is released at a specific point that is dependent on the unique sequence of the amplicon. It can thus be used to distinguish, very sensitively, differences in divergent amplicons. Using a range of freely available bioinformatics software, such as Clustal X2, Primaclade, Treeview and BLAST analysis, primers were designed based on segment 2 that sought to differentiate the individual serotype from previously defined clades based on a pair of segment 10 primers. Reference and field isolates of the AHS virus were obtained from the National Institute of Communicable Disease and the Onderstepoort Veterinary Institute, South Africa, and were propagated on Vero cell monolayers. Total RNA was extracted using guanidine-thiocyanate and verified as containing AHSV genomic material using primers recognised by the World Organisation for Animal Health that target the genome segment encoding VP7. Variable amounts of total RNA did not influence the downstream analysis as individual serotypes were easily distinguished using HRM despite wide ranging template concentrations. Through testing the primers designed in the present study, various serotype anomalies were discovered with regard to the isolates obtained from the Onderstepoort Veterinary Institute. Serotype-specific primers and the segment 10 primers were used to interpret the serotype anomalies through High Resolution Melt analysis. Sequencing confirmed the anomalies: serotype 2 isolates were serotype 6 isolates and a serotype 5 isolate was serotype 8. A proposed protocol for a rapid serotyping assay was investigated. This involved an initial PCR to determine into which clade of segment 10 the sample fits. Following this, the serotype was elucidated for each clade using segment 2 clade-specific primers. These reactions were performed in the Corbett Rotor-Gene™ 6000 and its in-built software was used. However, limitations of the software soon became apparent, as it was not able to completely distinguish the serotypes. Alternate methods were sought and included ScreenClust HRM® Software, principal component analysis (PCA) and discriminant analysis (DA). The use of normalised HRM fluorescence curves for PCA and DA was effective in standardising the template concentrations. These methods were successful in determining the serotype and rendered results with greater statistical confidence. Finally, a phylogenetic analysis was performed of all the available AHS virus sequences to determine the degree of possible genetic drift in the virus that may give rise to a new serotype. Minimal to no genetic drift could be found comparing sequences from the 1960s to sequences from the 2000s. This study presents the ability of HRM analysis to recognise and define AHS virus serotype anomalies, provide a new protocol for the serotyping of the virus with an extensive statistical analysis, the first for an orbivirus. Furthermore, the protocols described can be extrapolated to other orbiviruses.