Temporal variations of human and animal Rotavirus A genotypes in surface water used for drinking water production

Abstract

Rotavirus is a major cause of gastroenteritis among infants and children. In this study, nested PCR assays were developed to amplify partial regions of the VP7, VP4, and VP6 genes of Rotavirus A (RVA) for amplicon-based Illumina MiSeq sequencing to investigate RVA genotypes in environmental water samples. Eight sets of inner primers were first designed and screened for use in the nested PCR assays, and four sets of them could produce amplicons. Six sets of outer primers were then designed and combined with the four sets of inner primers that worked. The assays were evaluated for sensitivity using raw water samples collected from one drinking water treatment plant between April 2019 and March 2020 (Sample Set 1; N = 12) and seven DWTPs between 2018 and 2020 (Sample Set 2; N = 18). In total, 43 amplicons from Set 1 were sequenced and diverse sequences from human, porcine, bovine, equine, and feline RVA were observed. Human G8, G3, and G2 genotypes were obtained, with G8 predominant (relative abundance, 36-87%) in samples taken during the rotavirus epidemic season between April and June. Porcine G5, G11, and G4, and bovine G10 and G6 genotypes were also detected. VP4 sequence analysis revealed that the human P[8] genotype was present throughout the year, whereas P[4] and P[9] were present only in the epidemic season. The vaccine strains P[5] and P[8] (RotaTeq^®) were also detected. Our approach enables the identification of prevalent human and animal RVA genotypes and their host species that potentially caused fecal contamination in water sources.

Keywords: drinking water source; massive parallel sequencing; molecular epidemiology; rotavirus; surface water.

Figure 1

Concentration and genetic diversity of Rotavirus A (RVA) in water samples from DWTP E. The RVA NSP3 gene was quantitatively detected by real-time RT-PCR and plotted together with cases of rotavirus gastroenteritis reported for the week of sample collection in the prefecture upstream (A). Relative abundance (%) was calculated for RVA G genotypes determined by the VP7 gene (B), P genotypes determined by the VP4 gene (C), and I genotypes determined by the VP6 gene amplified in human (D) and animal (E) nested PCR assays. The value above each column of relative abundance represents the number of OTUs with > 500 reads.

Figure 2

Phylogenetic trees of Rotavirus A strains detected in surface water samples collected at DWTP E. Trees were built for partial regions of the VP7 (A), VP4 (B), and VP6 genes (C,D) obtained using human and animal nested PCR assays with maximum likelihood and bootstrapped with 1,000 repetitions. The sequence lengths of VP7, VP4, and human and animal VP6 were 242, 230, 239, and 274  bp, respectively. Circles with branch colors represent reference sequences that are closely related to OTU sequences. White and gray circles represent reference sequences for RotaTeq^® and closely-related OTU sequences. The information of reference sequences is provided in Supplementary Table S3 in Supplementary material. Triangles at the nodes represent bootstrap rates of >0.8. The name of each sub-cluster consists of the genotype, followed by host or RotaTeq^®, 1 (cluster with reference sequences) or 2 (cluster with only OTU sequences), and a or b for more than two sub-clusters. The values represent the minimum/maximum sequence similarity within each sub-cluster.