(700c) Multiscale Computational Investigations of Antimicrobial Action

Protegrins are a class of potent antimicrobial peptides (AMPs) that exhibit a broad spectrum of activity against harmful bacteria. The balance of evidence suggests that the primary target of protegrins is the bacterial cell membrane. In order to elucidate this mechanism of action, we have carried out theoretical and computational studies of several aspects of the action of protegrin, which are synthesized in the present work to provide a comprehensive, quantitative description of the mechanism of action of this peptide. We have used fully atomistic molecular dynamics simulations to obtain the potentials of mean force governing protegrin membrane association, dimerization and insertion. Based on this information, we have constructed a larger scale model for the equilibrium behavior based on statistical thermodynamics that includes the effects of peptide crowding on the membrane, membrane area expansion by inserted peptides and diffuse double-layer effects on adsorption. The resulting membrane expansion isotherms are in good agreement with experimental measurements of monolayer area expansion. Following membrane adsorption and insertion, the second key step in the action of protegrin is permeabilization of the bacterial membrane. NMR investigations have suggested that inserted protegrin peptides aggregate and form pores that allow the free passage of ions through the membrane. We have used all-atom molecular dynamics simulations of such pore structures embedded in model lipid bilayers in order to refine the structures suggested by NMR constraints. Based on structures extracted from these simulations, we have used an approximate continuum theory (Poisson-Nernst-Planck) to model the steady-state flow of ions through these membranes, and have been able to obtain quantitatively good agreement with experimentally determined single-pore conductance values. We then used the single-pore permeability values obtained this way as input to a larger scale model of transient transport from an entire bacterial cell. Using only the number of pores as a fitting parameter, we have been able to fit experimentally observed potassium leakage data obtained from protegrin-treated E. coli cells, and thereby determined that approximately one hundred protegrin pores are sufficient to cause cell death. Overall, this work provides a quantitative link between some of the crucial steps in the action of protegrin peptides. The tools that we present are applicable to a wide range of membrane-active peptides and interfacial phenomena in biological systems.