Since 2013, equine-like G3 rotavirus (eG3) strains have been detected throughout the world, including in Japan, and the strains were found to be dominant in some countries. In 2016, the first eG3 outbreak in Japan occurred in Tomakomai, Hokkaido prefecture, and the strains became dominant in other Hokkaido areas the following year. There were no significant differences in the clinical characteristics of eG3 and non-eG3 rotavirus infections. The eG3 strains detected in Hokkaido across 2 years from 2016 to 2017 had DS-1-like constellations (i.e. G3-P[8]-I2-R2-C2-M2-A2-N2-T2-E2-H2), and the genes were highly conserved (97.5-100 %). One strain, designated as To16-12 was selected as the representative strain for these strains, and all 11 genes of this strain (To16-12) exhibited the closest identity to one foreign eG3 strain (STM050) seen in Indonesia in 2015 and two eG3 strains (IS1090 and MI1125) in another Japanese prefecture in 2016, suggesting that this strain might be introduced into Japan from Indonesia. Sequence analyses of VP7 genes from animal and human G3 strains found worldwide did not identify any with close identity (>92 %) to eG3 strains, including equine RV Erv105. Analysis of another ten genes indicated that the eG3 strain had low similarity to G2P[4] strains, which are considered traditional DS-1-like strains, but high similarity to DS-1-like G1P[8] strains, which first appeared in Asia in 2012. These data suggest that eG3 strains were recently generated in Asia as mono-reassortant strain between DS-1-like G1P[8] strains and unspecified animal G3 strains. Our results indicate that rotavirus surveillance in the postvaccine era requires whole-genome analyses.
Keywords: DS-1-like constellation; Japan; equine-like G3; outbreak; rotavirus; surveillance.