Quantitative study of the dynamics and energetics of peptide-lipid membrane interaction using coarse-grained molecular dynamics simulations.

Abstract

Peptide-lipid membrane interactions are fundamental in molecular biology. A quantitative description, at single amino acid level, of the interaction of peptides with lipid membranes is essential in understanding the subtle microscopic mechanisms of membrane protein partitioning and folding. Here we demonstrate that long time-scale coarse-grained molecular dynamics (CG-MD) simulations can be used effectively to study both the conformational dynamics and the energetics of peptides and peripheral membrane proteins interacting with lipid bilayers. First, the penetration depth distribution of individual amino acids, obtained from tens of µs long CG-MD trajectories, provide a detailed description of the binding of the peptide to the lipid membrane. Second, the free energy profile (potential of mean force or PMF), as a function of the peptide-lipid membrane separation, can be calculated using the CG umbrella sampling method. The features of the PMF appear to be consistent with existing experimental results. Furthermore, we find that repeated CG-MD simulations that take into account the microscopic inhomogeneities of the system lead to a stochastic ensemble of PMFs that can be used to identify the dominant dissociation pathways of the peptide from the lipid bilayer. Third, the reconstructed PMF can be used to (i) interpret the peptide-lipid membrane dissociation force distribution measured in AFM force spectroscopy experiments, and (ii) infer the corresponding dissociation rate. We apply the above approach to investigate two exemplary systems, namely: (1) three Wimley-White pentapeptides (Ac-WLXLL, with guest residues X = R,I and L) interacting with POPC and POPG lipid bilayers, and (2) the extreme N-terminus (SecA2-11) of the SecA peripheral membrane protein from E. coli, interacting with a POPC bilayer.