Three Key Regions in Hemagglutinin Determine Binding Selectivity and Stability: Insights into Antiviral Drug Design

The receptor binding domain of hemagglutinin (HA) of influenza viruses contains three key regions for binding its endogenous carbohydrate receptors: loop-130, helix-190, and loop-220. To effectively predict the binding of HA with endogenous glycan ligands or designed inhibitors, the present study proposed a hypothesis that these ligands need to form stable interactions with at least two of the three critical regions simultaneously in the binding site. The testing of the hypothesis employed multiple HA variants, including H1, H3, H7, H17, and H18, with both α-2,6 and α-2,3-linked sialosides. Observations from molecular dynamics simulations are consistent with the experimentally discovered binding preferences for HA. To extend the proposed hypothesis to the antiviral drug design, it was further tested by using a noncarbohydrate receptor that formed a cocrystal complex with H5, <i>N</i>-cyclohexyltaurine (NCT), and an experimentally measured inhibitor, curcumin. Observations from the molecular models for these structurally distinctive molecules provided further test for the hypothesis and extended the applicability to noncarbohydrate ligands. The proposed hypothesis provided an alternative explanation for the binding preference of HA proteins, a fast approach to determine the binding stability of a ligand, and insights into the design of antiviral drug molecules targeting HA.