Circulation of classic and recombinant human astroviruses detected in South Africa: 2009 to 2014

• Published in: Journal of clinical virology 2021-02, Vol.135, p.104719, Article 104719
• Main Authors: Nadan, Sandrama, Taylor, Maureen B., Page, Nicola A.
• Format: Article
• Language: English
• Subjects: Classification; Diversity; Human astrovirus; Recombinants; South Africa
• ISSN: 1386-6532; 1873-5967
• DOI: 10.1016/j.jcv.2020.104719

•More recombinant than classic strains were detected.•Astrovirus surveillance in non-human sources will identify reservoirs for recombination.•Challenges in classifying recombinant strains exist in the current characterisation system.•The Vesikari severity scale must be used with caution when rating the severity of astrovirus-associated diarrhoea.•Genotype HAstV-5 requires further in-depth investigation. Astroviruses (AstVs) are associated with diarrhoeal and extra-intestinal infections in human, animal and avian species. A prevalence of 7% was reported in selected regions in SA while AstVs detected from clinical stool specimens were almost identical phylogenetically to strains identified in environmental and water samples. This study investigated the molecular diversity of astroviruses circulating between 2009 and 2014 in South Africa (SA). Astroviruses detected in stool specimens collected from hospitalised children were investigated retrospectively. Astroviruses were characterised using type-specific RT-PCR, partial nucleotide sequence analyses in ORF1 and ORF2 and whole genome sequencing. Different genotypes were compared with clinical features to investigate genotype-related associations. The Vesikari severity scale (VSS) was evaluated for scoring astrovirus diarrhoeal infections. Of 405 astroviruses detected, 49.9 % (202/405) were characterised into 32 genotypes comprising 66.3 % (134/202) putative-recombinants and 33.7 % (68/202) classic strains. No trends by year of collection, age or site were observed. Whole genome analysis in eight strains revealed that genotypes assigned by partial nucleotide sequence analyses to five astroviruses were incorrect. Bivariate analyses showed there were no significant associations between genotypes and clinical symptoms or severity of infection. A comparison of Vesikari parameters with astrovirus-positive proxy values demonstrated that Vesikari scores for duration of diarrhoea and admission temperatures would result in a milder infection rating in astrovirus-positive cases. Diverse genotypes co-circulated with putative-recombinants predominating. Astrovirus classification was complicated by the lack of a consistent characterisation system and reliable reference database. The VSS should be used cautiously to rate astrovirus diarrhoea. While surveillance in communities and out-patient clinics must be continued, screening for human astroviruses in alternate hosts is needed to determine the reservoir species.