Evolutionary patterns of pH1N1 and H3N2 in relation to vaccine use

TL;DR

This study explores how pH1N1 and H3N2 influenza viruses evolve in response to vaccines, revealing patterns of mutation and stability that could improve vaccine effectiveness.

Contribution

The study introduces a site-based approach to analyze antigenic evolution, revealing vaccine-associated allele frequency changes and their effects on protein stability.

Findings

Nonsynonymous mutations with vaccine-associated allele frequency changes were enriched in epitope regions of pH1N1 and H3N2.

pH1N1 showed more rapid allele-replacement events within a single season compared to H3N2.

Observed mutations in both subtypes increased protein stability, with pH1N1 epitope mutations being more stabilizing.

Abstract

The rapid evolution of viral antigens poses a major challenge to infectious disease control, particularly for pathogens like influenza that undergo frequent antigenic changes. While deep mutational scanning and platforms such as Nextstrain have advanced our understanding of mutation effects and population-level viral dynamics, they often rely on strain-level analyses that may overlook key within-strain antigenic changes. In this study, we adopted a site-based approach to systematically identify and analyze hemagglutinin (HA) mutations in influenza viruses that differed from vaccine strains, using publicly available genomic data. We found that nonsynonymous mutations exhibiting vaccine-associated allele frequency changes were significantly enriched in epitope regions in both pH1N1 and H3N2, and that pH1N1 showed a higher proportion of rapid allele-replacement events occurring within a single influenza season, whereas H3N2 substitutions more often occurred across multiple seasons. Geographically, several mutations displayed allele frequency changes correlated with local vaccination coverage. Phylogenetic analyses further revealed that five nonsynonymous mutations in H3N2 arose independently across multiple clades. Serological assays confirmed reduced neutralization for multiple pH1N1 mutations, and computational protein stability analyses indicated that observed mutations tended to increase protein stability in both subtypes, and that in pH1N1, potential epitope mutations were more stabilizing than those in non-epitope regions. By integrating bioinformatics with experimental validation, our approach provides a refined understanding of how selective pressures shape antigenic evolution at the site level, which could aid future studies on vaccine effectiveness and epidemic trends.

The online version contains supplementary material available at 10.1186/s12864-026-12608-y.

Figures