Testing the limits of model membrane simulations-bilayer composition and pressure scaling.

Studying transfer of bioactive compounds across cell membranes by simulations attracts growing attention. To perform such calculations accurately, it is necessary to verify the validity of computational protocols established for description of unperturbed lipid bilayers also with translocating substances present. The current work reports the results from 1 μs long atomistic molecular dynamics simulations of two types of model plasma membranes-one built of a single phospholipid (DPPC) and one constructed of four types of phospholipids-in the presence of a drug-peptide complex experimentally known to cross cell membranes. The influence of membrane composition and of applied pressure scaling algorithm on the simulations outcome is analyzed with particular focus on membrane structure and on complex-lipid interactions during the initial penetration stage. It is found that the mixed composition of the membrane is important for correct assessment of the interactions with the complex both from purely structural perspective and because of the uneven charge distribution. The structure of the mixed lipid bilayer is affected more markedly by the pressure scaling algorithm. When the pressure is isotropically scaled, lipids are distributed almost homogeneously along the membrane in liquid ordered state. On semi-isotropic scaling, the lipid tails undergo significant rearrangement and a long-range ordered state is established. This results in "freezing" of the membrane and expulsion of the complex. The statistical analysis of the MD data points to the conclusion that a mixed-lipid membrane model with isotropic pressure scaling would be more suitable for describing the process of complex translocation across neoplastic membranes. © 2017 Wiley Periodicals, Inc.