Simulation Of The Influenza Fusion Peptide Pre-Pore Structure

Influenza virus dependent membrane fusion requires facilitation by a fusion peptide. Membrane deformation theory predicts that amphipathic peptide-induced lipidic pores are less stable in membranes containing lipids that generate negative spontaneous curvature, such as cholesterol. To explicitly test predictions of this theory, the membrane permeability of giant unilamellar vesicles (GUV), with and without cholesterol, were investigated experimentally using an influenza fusion peptide analog. The peptide formed stable membrane pores in POPC lipid GUVs, while pore formation was inhibited in 50:50 mol% cholesterol:POPC membranes. Complimentary atomistic molecular dynamics simulations presented here were performed containing either a single copy of the fusion peptide or an aggregate of 21 fusion peptides interacting with model POPC or 1:1 chol:POPC membranes. Spontaneous curvature induction by fusion peptide aggregates differed both quantitatively and qualitatively from that of a single fusion peptide. Furthermore, fusion peptide aggregates generated explicit negative curvature in the opposing leaflet, which we hypothesize represents a relevant pre-pore structure. These results indicate a fundamental difference between the action of a single fusion peptide and an aggregate, and describe how the intrinsic spontaneous curvature of the membrane, modulated by negative curvature-generating lipids such as cholesterol, tempers this response.