DAPC and Mutation of Surface Genes
The HA and NA proteins of AIVs play a critical role in virulence, transmissibility, and antigenicity. The variation and combinations between HA and NA also primarily determined the pandemic strain of the virus. We first divided the 2016-2019 strain into epidemic seasons (that is, 2015.7−2016.6 as a popular season, and 2016.7-2017.6 as the next popular season). We calculated the changes of HA and NA in four popular seasons from 2016 to 2019 through DAPC analysis. The horizontal axis represents the first principal component, and the vertical axis represents the second principal component. It can be seen that the strains are clustered in each epidemic season [Figure 3].
The HA principal component analysis showed a significant difference in the first principal component in 2016, 2017, and 2018. The 2019 strain showed a greater distance from the 2018 strain in the second principal component, while the difference in the first principal component was smaller [Figure 3(A)]. This means that the variation of the strain varies significantly from 2016 to 2018 yearly, in contrast to the smaller variation in 2019. This variation also largely represents the evolution and antigenicity of the strain. The principal component analysis of NA was similar to that of HA. The difference in the first principal component between 2016 and 2017 is increasing, There is no significant difference in the first principal component between the 2018 strain and the 2019 strain, but there are some significant differences in the second principal component [Figure 3(B)].
A total of 22 amino acid sites in the HA gene and 13 amino acid sites in the NA gene with obvious variation were identified from 2016 to 2019 and were compared and analyzed [Figure 4(A)(B)]. The results showed that the amino acid sites changed regularly. The HA amino acid changes were mainly focused on the HA1 area, positions 83, 115, 120, 123, 126, 127, 138, 140, 141, and 269 have been proven to cause antigenic drift in other strains of research (Beato et al., 2013; Velkov et al., 2013), which may cause significant changes in the antigenicity of the virus.
In summary, we believe that the virus has undergone a rapid evolution from 2016 to 2018, that the amino acid substitutions were basically completed in 2017−2018 [Figures 3, 4], and that the strain has basically evolved to adapt itself to the environment. Consequently, the strain changed considerably less in 2019. We suppose that the reason for the big difference in strains from 2016 to 2018 may be related to the current vaccination environment. Vaccines are constantly being updated. From 2017, the H5-H7 dual vaccine was mainly used for H5 and H7N9. Research by Shi et al. showed that the vaccine controlled the outbreak of H7N9 to a large extent and greatly reduced the cases of human infection with the virus(Shi et al., 2018). We speculate that H5N6 has evolved to adapt to the current vaccine immune environment.