Peter M. Kasson, Departments of Chemistry and Structural Biology, Stanford University, 318 Campus Drive MC 5447, Stanford, CA 94305-5447 and Vijay S. Pande, Department of Chemistry, Stanford University, 318 Campus Drive MC 5447, Stanford, CA 94305-5447.
Lipid vesicle fusion is an important therapeutic target for control of neurotransmission, viral infection, and endocrine secretion. Synthetic fusion systems also provide an attractive means of drug delivery. One challenge in designing systems for the control of membrane fusion is understanding how changes to the physical properties of lipid membranes affect fusion behavior. Biological mechanisms for controlling fusion behavior are thought to include changes to bilayer curvature, control of lipid composition, and induction of local lipid disorder by fusion proteins. We have used ensemble molecular dynamics simulation to develop physical models for each of these mechanisms and its effects on the fusion process. By analyzing thousands of such simulations, we can derive a more general model for how lipid composition and curvature control fusion kinetics. Here we report simulations where both the composition and the curvature of lipid membranes were varied, with curvature regimes ranging from small 15-nm vesicles to larger vesicles to planar bilayers. We also consider the effects of curvature fluctuations in bilayers and changes to bilayer structure induced by fusion peptides. We find a pronounced effect of curvature on fusion rates in our simulations; in addition, the effects of lipid composition remain relatively independent of curvature over the systems simulated. We then apply these models to interpret the effects of fusion proteins on membrane curvature.