Evidence for Genetic Reassortment between Human Rotaviruses by Full Genome Sequencing of G3P[4] and G2P[4] Strains Co-circulating in India

Abstract

Rotavirus A causes severe diarrhoea in infants and young children worldwide. Many unusual combinations of G and P genotypes have been observed in rotaviruses circulating in developing countries. Mixed infection of a single individual with more than one strain is a mechanism by which genetic reassortants are formed with unusual G and P combinations. However, few studies have provided direct evidence for the formation of such unusual strains as a result of co-infection of co-circulating strains. Here, we used full-genome sequencing to re-analyze a G3P[4] strain (107E1B) and a G2P[4] strain (116E3D) detected in India in 1993 and showed that 107E1B had virtually an identical nucleotide sequence with 116E3D, except the VP7 gene. Phylogenetic analysis revealed that the 107E1B VP7 gene was of typical human rotavirus origin, with a 99.3% nucleotide sequence identity with another Indian G3 VP7 gene. Thus, this study provided robust evidence for the formation of the G3P[4] strain through genetic reassortment in which a G2P[4] strain with a typical DS-1 genogroup background acquired the VP7 gene from a co-circulating G3 human rotavirus strain. This study established a basis on which to facilitate full genome sequence analysis of an increasing number of G3P[4] strains in China and elsewhere in the world.

Keywords: G3P[4]; full genome sequencing; human rotavirus; mixed infection; reassortment.

Fig. 1.

An abbreviated phylogenetic tree for the G3 VP7 genes from human and animal rotavirus strains available in the DNA databases. This phylogenetic tree was constructed by the neighbor-joining method, the genetic distances were computed according to the Kimura 2-parameter model, bootstrap values were obtained after 1000 replicate trials, and the VP7 gene of the strain 116E3D (G2P[4]) was used as the outgroup. The bootstrap values lower than 70% were omitted. The phylogenetic tree was constructed from the 486 sequences of entire coding region (981 nucleotides in length including the stop codon) of VP7 gene of G3 strains with the exception of the sequences of three porcine strains (*) with 91 nucleotides shorter at the 3’ end. In total, there were 405 and 81 sequences from G3 human and animal rotavirus strains, respectively.

Fig. 2.

Distribution of pairwise nucleotide sequence identity (%) for the coding region of the VP7 gene between 107E1B and other G3 human and animal rotavirus strains. Pairwise nucleotide sequence identity was calculated as (1 – p-distacnces) × 100, and p-distance was computed with the aid of MEGA v.5.0 software [27].