Geometric constraints dominate the antigenic evolution of influenza H3N2 hemagglutinin
We have carried out a comprehensive analysis of the determinants of human influenza A H3 hemagglutinin evolution, considering three distinct predictors of evolutionary variation at individual sites: solvent accessibility (as a proxy for protein fold stability and/or conservation), experimental epitope sites (as a proxy for host immune bias), and proximity to the receptor-binding region (as a proxy for protein function). We found that these three predictors individually explain approximately 15% of the variation in site-wise dN/dS. The solvent accessibility and proximity predictors were largely independent of each other, while the epitope sites were not. In combination, solvent accessibility and proximity explained 32% of the variation in dN/dS. Incorporating experimental epitope sites into the model added only an additional 2 percentage points. We also found that the historical H3 epitope sites, which date back to the 1980s and 1990s, showed only weak overlap with the latest experimental epitope data. Finally, sites with dN/dS>1, i.e., the sites most likely driving seasonal immune escape, are not correctly predicted by either historical or experimental epitope sites, but only by proximity to the receptor-binding region. In summary, proximity to the receptor-binding region, and not host immune bias, seems to be the primary determinant of H3 evolution.