Exploring pKa Values for Buried Residues in Membrane Bilayers using Constant pH Molecular Dynamics Simulations

Abstract

Environmental conditions such as pH are considered to play profound roles in determining proteins' structures and functions. In recent years, several experimental and computational techniques have been developed to improve our understanding of the effects of pH on proteins. However, the implicit representation of solvents in the current constant pH molecular dynamics (CpHMD) methods can lead to undesirable side effects in calculating pKa values of titratable side chains. These effects can be even more pronounced for membrane proteins where the rigid nature of the current implicit membrane models causes unrealistic pKa shifts in favor of neutral states. Inspired by recently published pKa values for Lys residues in model transmembrane peptides (Gleason NJ et al. (2013) PNAS 110:1692-1695), here we explore a newly developed CpHMD technique based on an all-atom representation of solvent/lipid molecules and proteins for simple transmembrane peptide models with a titratable side chain. The pKa values of the side chains along with the pH dependent configuration of these peptides in the membrane bilayer are investigated and compared with the available experimental values. Motivated by the excellent agreement between the calculated pKa values and the experiment for model peptides, the purely explicit CpHMD is also applied to study the pKa values of titratable side chains of influenza A proton channel (M2) and its pH dependent structural change in the membrane bilayer.