To determine the predominance of the unusual G9P[4] circulating in Indonesia, we performed aa analyses of RVA interaction sites on host cells, focusing on the outer and inner capsid proteins G9-VP7 and I2-VP6. Aa substitutions were deduced from 9 VP7-G9 and VP6-I2 reference strains from the GenBank database (2009-2015). The aa substitutions of T78I and T108I were determined in the VP7 gene (Table 3). Additionally, a single aa substitution of L291S was detected in the VP6 gene (Table 4).
DISCUSSION
Previously, we reported dynamic changes in RVA genotypes from equine-like G3 to human G1/G312,25. Complex RVA genotype diversity is more common in developing countries than in developed countries21,26,27. Inter-genogroup reassortant strains highlight the ongoing spread of the unusual RVA strains throughout Asia and other countries28. It has been suggested that evolution occurs primarily through the selection of point mutants in the antigenic site or by the reassortant of each segment of the RVA genome3. The mutation rate of RVA was estimated to be around the value of 5×10-5 per nucleotide during genome synthesis, suggesting that approximately one mutation emerges in each new copy10.
In this study, we examined the whole genome sequences of the unusual G9P[4] RVA detected in 2021 in Indonesia. This is the first report of the whole genome sequence of the unusual G9P[4] RVA in Indonesia. The newly identified unusual Indonesian G9P[4] strains exhibit genotypes within the G9P[4] constellation distinct from the sporadic G9P[4] strains detected in India, Bangladesh, Japan, and South Korea from 2007 to 2012. The more prevalent G9 genotype in the Americas (18%) and Europe (13%) underwent recombination with the P[8] genotype in 2003 –200410. Human RVA G9P[4] was first identified in Brazil in 1999 as an uncommon genotype14. Reassortant in the G9P[4] genotype variation was found in Paraguay with mixed infection of G2G9P[4] from Bangladesh in 2005 and from India in 20081,18. The unusual G9P[4] with a genomic constellation backbone of DS-1-like was reported in Indonesia and was detected in several countries such as the USA, Japan, India, Italy, Mexico, and Bangladesh21,27-31. Human RVA reassortant with genomic constellation backbone of Wa-like or DS-1-like were found more frequently16. The double-reassortant strain (G9-P[4]-I2-R2-C2-M2-A2-N2/N1-T2-E6-H2) detected in India and in the Czech Republic have the same genotypic constellation as the Indonesian strain15,32. In 2013, triple-reassortant strains (G9-P[4]-I2-R2-C2-M2-A2-N1-T2-E6-H2) were found, following the detection of unusual G9P[4] strains in India during 2011-201315.
A time-scaled Bayesian phylogenetic tree was constructed to examine the evolution of the RVA genotype14,32. The evolutionary rate of VP7-G9 in this study was estimated to be 6.43×10-3 nucleotide substitutions/site/year, which was similar to those of the VP7-G9 genotype with 1.38×10−3, 1.87×10−3, and 1.609×10−3 nucleotide substitutions/site/year in China and Ghana3,16,33. The evolutionary rate of the VP4-P[4] in this study was 2.53×10-3 nucleotide substitutions/site/year, comparable to the Chinese strain’s evolutionary rate of 1.172×10-3 nucleotide substitutions/site/year33. Conversely, the VP4 gene in Ghana showed a higher evolutionary rate at 8 ×10-4nucleotide substitutions/sites/year21. These results align with the evolutionary rates typical of many RNA viruses, which evolve at a rate of approximately 1×10-3 nucleotide substitutions/site/year4. The parent viruses spread regionally within 1-3 years, whereas the ancestral virus takes 2-6 years to propagate13. In the present study, the evolutionary rates of both the VP7/VP4 G9P[4] genes of RVA show similar evolutionary speed to those of G9 from China and Ghana. However, the evolution of the P[4] genotype happened more rapidly than previously observed in Ghana. It is essential to monitor the evolutionary rates of RVA genotypes to identify trends in RVA genotype evolution.
The tMRCA analysis predicts the rate of transmission of RVA genotypes13,16. The tMRCA of the G9-VP7 of the unusual G9P[4] strains in lineage-III was estimated to be in 2004 (95% HPD interval 2001-2021) (Fig. 3), showing similar results to the unusual G9P[4] RVA strain in India15. Monitoring of genotype diversity revealed that rotavirus G9P[4]-VP7 lineage-III was detected in surveillance networks worldwide in the beginning of the millennium2. The tMRCA of the G9P[4]-VP4 in this study was estimated to be in 1994 (95% HPD interval 1990-2021) (Fig.4). The prevalent P[4]-VP4 genotype, G2P[4], emerged in 1959 (95% HPD interval 1923-1984)16,33. This finding suggests that the unusual G9P[4] strain began spreading with the onset of reassortment in the VP4 gene in the 1990s. The NSP2 gene was generated by reassortant between N1 and N2. The ancestor of NSP2-N1 emerged in 1986 (95% HPD interval 1981-2021), while NSP2-N2 appeared in 1998 (95% HPD interval 1996-2021). NSP2-N1 appeared 10 years earlier than NSP2-N2, indicating that the Wa-like strain appeared earlier than the DS-1-like strain and that both strains are still circulating worldwide.
The outer membrane protein of the VP7 gene contains major antigenic epitopes that induce specific neutralizing antibody responses21. In the present study, the deduced aa substitutions in the VP7 gene were T78I and T108I, conserved in G9P[4] RVA in India15. Interestingly, the Indian G9P[4] strain presents a mutation E154K in VP7, whereas the Indonesian G9P[4] strain lacks this mutation. The aa sequence in the VP7 has been deduced and reported2. One of the strains was linked to the segment within aa 271-342 of VP6, which is necessary for interaction with VP734. In addition, a pair of van der Waals interactions between aa 279-281 on VP7 and aa 313 on VP6, as well as a side chain hydrogen bond between aa 305 on VP7 and aa 310 on VP6 were reported as interaction sites between VP7 and VP613. In viral self-defense, the virus was capable of mutating at their interaction sites of the outer capsid protein VP7 and the inner capsid VP615. The aa substitution S291L on the VP6 protein is linked to the interaction between VP7 and VP6 proteins. In this context, the altered relationship between G9 and I2, with a substitution at the VP6 protein interaction site, may be more sustainable than the unusual relationship without substitutions15. In the present study, aa 291 did not receive any substitutions as previously described. This phenomenon is consistent with the notion that the unusual G9P[4] strains circulating in Indonesia can rapidly change into G9P[6] strains.
CONCLUSIONS
The recent discovery of the unusual G9P[4] RVA strains emphasizes the ongoing spread of RVA in Asia and other regions. In particular, the G9P[4]-DS-1-like strain, which had been reported in several countries outside Southeast Asia since 2011, was first detected in Indonesia in 2018. By 2021, the majority of RVA-positive cases (19 out of 21) were identified as G9P[4], with the remaining 3 as G9P[6]. These Indonesian G9P[4]/P[6]-DS-1-like strains, discovered between 2018 and 2022, show genetic variations from strains reported in other countries between 2011 and 2015, particularly in the NSP4 gene (which has E2 genotype instead of E6 genotype). These strains share a common ancestry in VP7 and VP4 genes with previously reported G9P[4] DS-1-like strains but exhibit slight genetic differences in other genes. These results suggest that there were multiple intra- reassortant events between the original G9P[4] strains and the DS-1-like RVA strains that coexist in Indonesia and elsewhere. Continuous monitoring of RVA genotype dynamics is crucial given that G9P[4] strains emerged as the predominant genotype in Indonesia in 2021. This monitoring is essential to assess the prevalence and genetic diversity of circulating RVA strains and to evaluate the efficacy of RVA vaccines.