Epidemiology of HFMD Beijing, China, from 2017 to 2019
From 2017 to 2019, a total of 1,721 cases were enrolled in our study with HFMD that were clinically diagnosed (207 in 2017, 732 in 2018, and 782 in 2019). The ratio of male to female was 1.62:1 with 1,064 male and 657 female cases. The age ranged from 53 days to 16 years. The majority of HFMD patients were aged 0-5 years, which accounted for 76.70% (1,320 of 1,721 patients). Among the cases, there were 832 (48.34%) scattered children, 621 (36.08%) kindergarten children, and 268 (15.57%) students. The cases were mainly scattered children and kindergarten children, which accounted for 84.43% (1,453/1,721). HFMD infections mainly occurred between May and October, with a peak in July or August during this 3-year period.
Serotyping of enteroviruses associated with HFMD cases
The EV detection rate of HFMD was 88.32% (1,520/1,721) in this study, and a total of 8 EV types were identified, while 175 EV positive species were not typed (11.51%, 175/1520). The most common EV-A was CV-A6, which accounted for 846 (55.86%) cases, follow by CV-A16 (400, 26.32%), EV-A71 (34, 2.24%), CV-A10 (31, 2.04%), CV-A4 (16, 1.05%), CV-A5 (9, 0.59%), CV-A2 (5, 0.33%), and CV-A8 (1, 0.07%). In 2017 and 2018, the most predominate pathogen of HFMD was CV-A6, and the constitution ratio was 61.88% and 66.82%, respectively. However, CV-A6 and CV-A16 have co-circulated in 2019, with the composition ratio 43.70% and 42.82%, respectively. The detection rate of EV-A71 has dropped significantly, from 16.02% in 2017 to 0.29% in 2019, while the detection rate of CV-A16 has increased year by year, from 7.73% in 2017 to 42.82% in 2019 (Table 1).
The homologous analysis of CV-A6 based on complete VP1 sequences
120 CV-A6 VP1 gene sequences were amplified and sequenced, including 29 in 2017, 65 in 2018, and 26 in 2019. The VP1 gene was 915bp, which encodes 305 amino acids. Nucleotide and amino acid sequence identity analyses were performed on 120 CV-A6 VP1 gene sequences of this study and CV-A6 prototype strains (Gdula, USA/1949). The results indicated that the nucleotide and amino acid sequence identities of 120 CV-A6 VP1 gene sequences were 91.2%–100.0% and 97.7%–100.0%, respectively. Compared with the prototype strain (Gdula), the nucleotide and amino acid sequence identities were 81.7%–84% and 94.7%–96.3%, respectively.
Phylogenetic analysis of CV-A6 based on complete VP1 sequences
The 120 CV-A6 strains from our study and other 50 CV-A6 reference strains worldwide available from GenBank were subjected to phylogenetic analysis based on the complete VP1 gene sequences. The cladogram indicated that all the CV-A6 strains were segregated into four genotypes which were designated as A, B, C, and D. D genotype can be further divided into D1–3 sub-genotype. While sub-genotype D3 can be further subdivided into two evolutionary branches, which were designated as D3a and D3b. The Gdula prototype strain isolated in the United States in 1949 formed a single genotype A. Genotype B contains four strains isolated in mainland of China between 1992 and 2007 and one strain isolated from Japan. Genotype C, represented by two strains from Shandong province of China in 1996 and India in 2008. Sub-genotype D1 comprised strains from Japan, Australia, France, Spain, and Taiwan, China during 1999 to 2010. Sub-genotype D2 comprised strains from mainland of China during 2006 to 2013. Sub-genotype D3 comprised strains from mainland of China and Australia during 2016 to 2019, of which 119 Beijing CV-A6s in this study belonged to D3a evolutionary branch during 2017 to 2019, and the other one belonged to D3b evolutionary branch in 2017. The 119 Beijing CV-A6s obtained in this study were closely related to the CV-A6 stains isolated from some other provinces and cities of China, such as Guangxi province, Guangdong province, Shandong province, Liaoning province, Yunnan province, Beijing, Shenzhen, and Shanghai, which clustered in the evolutionary branch D3a, but gathered in different clusters. These findings indicated that D3a were the predominant evolutionary branch in Beijing from 2017 to 2019, and has co-circulated with other CV-A6s of China, which belonged to different transmission chains (Figure 1).
VP1 Deduced Amino Acid Sequence Analysis of CV-A6
The 120 CV-A6 strains in this study were compared to the Finland/2008 strain (KM114057, the earliest reference strain of D3 sub-genotype) to identify variations at the amino acid level. Many amino acid variations sites were found, including A5T (n=117/120, 97.5%), S27N (n = 120/120, 100%), A29V (n = 38/120, 31.67%), V30A (n = 118/120, 98.33%), V71I (n = 5/120, 4.17%), S137N (n = 110/120, 91.67%), V174I (n = 21/120, 17.5%), N241D (n = 5/120, 4%), V242I (n = 4/120, 3%) and T283A (n = 95/120, 79.17%) (Table 2).
Phylogenetic and Recombination Analysis of the CV-A6 genome sequences
Phylogenetic trees were constructed based on the nucleotide sequences of VP1, P1, P2, and P3 coding region of the fourteen CV-A6 strains among these prototypes of EV-A species from GenBank database. Consistent with the phylogenetic tree of VP1 coding region, the P1 phylogenetic tree indicated that the fourteen CV-A6 strains clustered together with the CV-A6 prototype (Gdula). The phylogenetic trees based on P2 and P3 coding regions showed that the fourteen CV-A6 strains were clustered together with the prototype strains of EV-A114, CV-A5, CV-A16, CV-A4, CV-A14, and EV-A114, CV-A16, CV-A4, CV-A14, respectively (Figure 2a–d).
The nucleotide homology of the complete genome sequences among the fourteen CV-A6 strains was 95.6%–100%. Compared with CV-A6 prototype strain (Gdual), the nucleotide homology was 80.4%–99.3%. Potential evidence of recombination in the genome of the fourteen CV-A6 strains was investigated via similarity plot and boot scanning analysis. The results showed that CV-A6 strains prevalent in Beijing from 2017 to 2019 had higher similarity with the CV-A6 prototype strain (Gdual) in P1 coding region, which is consistent with the results of the above-mentioned phylogenetic analysis. However, in P2 and P3 non-structural protein coding regions, CV-A6 strains of this study had higher similarity with the CV-A16, CV-A14, EV-A114, and CV-A4. The boot scanning analysis results indicated that recombination may have occurred between CV-A6 representative strains and EV-A114 strain in 2B and 3D coding regions. (Figure 3). Previous study indicated that HFMD outbreaks associated with CV-A6 resulted from the evolutionary dynamics of CV-A6 and the appearance of novel recombinant forms (RFs)[14]. And the CVA6 strains were assigned into recombinant form (RF)-A to -H and (RF)-J to -K based on 3D polymerase (3Dpol) phylogeny[15]. Referring to the rule from previous research, all of the fourteen CVA6s prevalent in Beijing were assigned into RF-A from 2017 to 2019 in this study (Figure 4).