Tracking Proton Transfer through Titratable Amino Acid Side Chains in Adaptive QM/MM Simulations

J Chem Theory Comput. 2019 Nov 12;15(11):5794-5809. doi: 10.1021/acs.jctc.9b00649. Epub 2019 Oct 7.

Abstract

Ubiquitous throughout biological processes, proton transport usually occurs through the Grotthuss shuttling mechanism, where the hydrated proton exists as a charge or structural defect and is propagated quickly through a network of hydrogen bonds with minimal perturbation to the positions of the involved heavy atoms. The rapid reorganization of the bonding network and changing identity of the migrating proton can cause difficulties in molecular dynamics (MD) simulations. Previously, we formulated a proton indicator that tracks the proton as a structural defect for proton exchange between water molecules. In this work, we extend the proton indicator to treat proton transfer between water and amino acid side chains with titratable functional groups. Of particular interest are histidine, glutamate, and arginine, all of which have titratable groups featuring multiple protonation sites. Comparison with the modified center of excess charge (mCEC) suggests that the proton indicator and mCEC are both comparable in approximating the location of the proton. The location of the proton indicator was then used as the center of the QM subsystem in adaptive quantum-mechanical/molecular-mechanical (QM/MM) simulations of proton transport through a model channel along a path consisting of water and titratable amino acid side chains. In the adaptive QM/MM simulations, atoms were reclassified on the fly in a continuous and smooth manner as QM or MM depending on their distances from the proton indicator. Employing a small, mobile QM subsystem, the adaptive QM/MM simulations were found to be much more efficient than the conventional QM/MM simulations with a large QM subsystem that covered the entire pathway for proton relay.