Application of classical dynamics simulations to investigate cellulose microfibril twisting and influenza H1 binding affinity
Hadden, Jodi Ann
MetadataShow full item record
As the most abundant biopolymer on Earth, cellulose serves as an important raw material for many industries, including paper, textiles, and more recently, biofuels. While high-resolution crystallographic data suggest that cellulose microfibrils occur as linearly oriented assemblies of cellulose chains, computational simulations predict a twisted structure. Through investigation of commonly employed theoretical approximations, this work establishes the physical origin of twisting behavior, indicating that it arises from a balance of competing forces. Overall, twisting appears to be driven by attractive van der Waals interactions, while mitigated by both the cellulose intrachain hydrogen bond network and solvent effects at the microfibril surface. As a result, modeling of simulated microfibrils is sensitive to monomeric charge distribution, solvent model, and the application of dummy atoms to mimic the influence of electron lone pairs. Further, analysis of back-calculated diffraction patterns for twisted and linear structures demonstrates that powder diffraction methodology cannot detect subtle twisting in cellulose samples, raising the possibility that crystals employed to resolve the original crystallographic coordinates could have incorporated twisting. Adhesion of the influenza A virus is mediated by its primary surface antigen, hemagglutinin, which recognizes receptor glycans terminating in sialylated galactose. Host range is determined by specificity for the glycosidic linkage displayed within this disaccharide motif, with avian viruses preferring α2-3 linkages and human viruses preferring α2-6 linkages. While experimental characterization of specificity is relatively straightforward, quantification of associated binding affinity represents a challenge due to the inherent multimeric nature of hemagglutinin structure. Through application of computational simulations, which allow investigation of a monomeric binding domain, this work computes highly accurate binding free energies associated with specificity determinants in the H1 hemagglutinin. Results include quantification of the effects of abrogating and specificity-altering point mutations, as well as avian and human receptor glycan contributions to binding affinity. Altogether, these data likely provide the most reasonable theoretical quantifications of binding currently available for the H1 system.