It has been reported that since 2004, the major EV71 strains in mainland China basically belong to the C4 genotype [18, 33]. In this study, it was found that the epidemic strains of EV71 in mainland China have been dominated by the C4 genotype since 2000, mainly the C4b subgenotype from 2000 to 2004, and the proportion of the C4a subgenotype has increased significantly since 2005.From 2005 to 2018, the C4a subtype was dominant; the C4b subtype was the second most common, and other subtypes appeared sporadically; this pattern was similar to previous reports [20, 21, 30, 33]. From 2004 to 2018, most of the EV71 isolates in mainland China were genotype C, and most of them were subtype C4 (especially subtype C4a); other genotypes occurred sporadically (9 strains of genotype B in 2011, 4 strains of genotype A in 2008, 2 strains of genotype A in 2009, 1 strain of genotype An in 2011 and 1 strain of genotype A in 2012). The immune effectiveness of inactivated vaccines mostly depends on the antigenic correlation between epidemic strains and vaccine strains, which are often best for preventing infection of the same subtype virus but are inferior against different subtypes [30]. Recent studies have also shown that the EV71 vaccine (especially in children who receive 2 doses) can effectively prevent and control childhood EV71-associated HFMD but has no protective effect against coxsackievirus (CV) A6 (CVA-6) or CVA16, and there is no explanation for the effectiveness of other subtypes of EV71 (excluding C4a subtypes) [34]. These studies show that vaccine research and development for EV71 combined with CVA6 and CVA16 and other multivalent vaccines might better prevent EV71 infection.
Interestingly, this study found that the EV71 epidemic strains in Taiwan were mainly of the B4 genotype, which was different from those in mainland China; EV71 epidemic strains are also constantly changing, which is consistent with a previous report [26]. According to a human phase 1 clinical trial on adults in 2010, the FI-EV71 vaccine (EV71vac) based on the B4 genotype from Taiwan is safe and induces a high titre of neutralizing antibodies against EV71; it was also highly effective against B1, B5, and C4a strains. However, the titres of neutralizing antibodies against C4b and CVA16 were low in 20% of volunteers, and virus-neutralizing antibodies against the C2 genotype were not detected in 90% of vaccine recipients [26-27]. These studies indicate that it is necessary to strengthen the monitoring of EV71 genotypes; new multivalent and effective vaccines that can cover local strains should be designed and applied according to the genotypes of the local predominant EV71 epidemic strains to ensure that the vaccine is more accurate in controlling HFMD epidemics.
Some studies have shown that the H22Q mutation in the VP1 protein of EV71 can lead to a decrease in the adsorption capacity of the C4 genotype to host cells [35-37]. The amino acid at position 22 of 78.15% of the 3712 strains isolated in China is H (histidine), which suggests that most of the viruses have strong adsorption capacity to host cells. Furthermore, H22Q was detected in 10.99% of all EV71 strains in Taiwan, significantly more prevalent than that in mainland China and Hong Kong (5.50%), suggesting that the adsorption capacity of some strains in Taiwan to host cells is weak in comparison with that of strains in mainland China.
Studies have shown that the A289T EV71VP1 variant is closely related to the occurrence of severe HFMD and that the neurological symptoms caused by EV71 infection are significantly increased when the amino acid at position 289 of VP1 is A (alanine); in contrast, there is low neurotoxicity when the amino acid is T (threonine) [36, 38]. In this study, 76.45% (2838 strains) of the virus strains were found to contain an A (alanine), suggesting that most of these EV71 viruses have high neurotoxicity. Moreover, 10.24% (380 strains) of the strains in mainland China (including Hong Kong) and 10.75% (399 strains) of those in Taiwan contain a T (threonine), suggesting low neurotoxicity. It remains to be further studied whether new mutations such as A289V (valine)/D (aspartic acid)/I (isoleucine) mutations will cause the emergence of severe HFMD.
EV71 can infect human lymphocytes by binding to its receptor molecule P-selectin glycoprotein ligand-1 (PSGL-1). When E (glutamic acid) at position 145 in VP1 is mutated to G (glycine) or Q (glutamine), the virus binds PSGL-1 more readily, whereas its PSGL-1-binding ability is weakened or lost when E is present [39]. In this study, the amino acid at position 145 in most strains was found to be E, with an E145G/Q mutation rate of 6.63%, suggesting that the emergence of this mutation may result in a virus that is more likely to infect human lymphocytes.
It has been reported that the E98K mutation may increase the hydrophobicity of VP1, making it easy for large compounds to enter and interfere with receptor binding, suggesting that E98K mutant viruses are sensitive to larger compounds [40]. Other studies have shown that E145G and N31D mutations are associated with increased virulence of EV71 and may increase the risk of neurological complications but that I262V mutations reduce the risk of neurological complications [41-43]. In this study, the E98K, E145G, N31D and I262V mutation rates were 6.65%, 3.13%, 7.25% and 2.77%, respectively. These findings indicate that these mutations may play an important role in the pathogenicity of mild and severe EV71-associated HFMD.
Humans are the only natural host and source of EV71. Indeed, EV71cannot infect rodents, which is due mainly to the incompatibility between the virus and rodent cells, and the different expression of its scavenger receptor in humans and rodents [44-46]. However, some studies have found that simultaneous substitution of K98E, E145A and L169F in VP1 of EV71 can result in infection in mice [44]. Our study showed that among 3712 strains, the mutation frequencies of K98E, E145A and L169F were93.24%, 0.30% and 0.03%, respectively; however, no strain with all three mutations was found. These findings indicate that humans are still the only host of EV71 in China; nevertheless, the existence of individual mutations does not rule out the emergence over time of strains that can infect other mammals. Therefore, it is important to closely monitor mutation of the key sites of the EV71VP1 protein.
EV71 is the most important pathogen causing severe HFMD in children, which can lead to irreversible sequelae or death, and it is a serious threat to their health [4, 7]. At present, there is no specific treatment for EV71 infection. The development and marketing of an inactivated EV71 vaccine in China is crucial for the prevention of HFMD caused by EV71 infection [26, 47, 48, 49]. Phase III clinical trials of the EV71 inactivated vaccine approved in China in 2015 have shown protective effectiveness against EV71-associated HFMD of more than 90% [47, 50, 51, 52]. However, according to molecular epidemiological studies of EV71, EV71 gene mutations occur frequently, leading to genetic diversity [28, 29, 30, 31, 53]. These studies suggest that there is still a need for strengthening surveillance of EV71 genotypes and the development of new EV71 vaccines.
This study had a retrospective design, and there are some limitations. First, the EV71VP1 gene sequences from China analysed in this study were downloaded from the GenBank database but were not tested by us. Due to time constraints, only the VP1 region wasanalysed and studied. In future studies, we will conduct research on the complete genome sequence of EV71 in China. Second, the Chinese EV71VP1 strains registered in GenBank do not cover all provinces in the country, and the data for some years are missing; thus, some isolates of other genotypes may have been unavailable. Third, it is not clear whether some variations in the amino acid residues found in the study are related to the severity of disease or the route of transmission.