Analysis Of Influenza Hemagglutinin Ligand-binding From Mutational Data And Molecular Motion

Abstract

Influenza hemagglutinin binds to sialic-acid-terminated glycans on the surface of target cells. This ligand-binding specificity of hemagglutinin is believed an important determinant of which host species it infects. Site-directed mutagenesis of hemagglutinin, expression, and determination of ligand-binding affinity are feasible but technically involved, so only a few hemagglutinin mutations have been tested in this manner. To understand the mechanism of binding specificity and guide further experiments, we have analyzed H5N1 avian influenza isolates to predict residues important to ligand binding and ultimately to ligand specificity. We employed sequence data from all available isolates in combination with analysis of protein residues that correlate with ligand conformation in molecular dynamics simulations to generate candidate sites for mutation. Using this combined analysis, we have predicted five residues both in the sialic-acid-binding site of hemagglutinin and more distant from it. In an initial evaluation, we have performed extensive molecular dynamics simulation of twelve point mutations at these sites. We simulated each of the 12 mutants in 3x100 ns and 200x10 ns simulations to obtain more robust statistical estimates of ligand dissociation. These simulations indicate a greatly increased dissociation rate from the mutants compared to simulations of wild-type H5N1 hemagglutinin VN1194, indicating that the mutations may disrupt ligand binding as expected. This analytic technique may thus provide an important means of screening potential binding-specificity mutants of influenza hemagglutinin as well as a more general tool to assess residues involved in ligand binding.