Implementation of the solvent macromolecule boundary potential and application to model and realistic enzyme systems

The implementation of the solvent macromolecule boundary potential (SMBP) by Benighaus and Thiel (J. Chem. Theory Comput. 2009, 5, 3114) into the program package CHARMM is presented. The SMBP allows for the efficient calculation of solvent effects for large macromolecules using irregularly shaped dielectric boundaries. In contrast to the generalized solvent boundary potential (GSBP) by Roux et al. (J. Chem. Phys. 2001, 114, 2924) from which it is derived, the SMBP is targeted for quantum mechanical/molecular mechanical (QM/MM) setups using ab initio methods for the QM part. After presenting benchmark results for simple model systems, applications of the SMBP for the calculation of geometries, reaction energy barriers, and vibrational frequencies for an alkaline phosphatase (AP) enzyme are discussed. Although the effect of the boundary potential on optimized structures (including the transition state) and vibrational frequencies is relatively small, the energetics of the phosphoryl transfer catalyzed by AP depend significantly on the boundary potential. Finally, to emphasize a unique feature of our implementation, we apply both SMBP and GSBP to the calculation of the energy barrier for a proton transfer reaction in a simple model channel, where the effect of an external transmembrane potential is studied. Due to the dipolar response of the polar environment, the effective charge displacement estimated based on the effect of the membrane potential on the proton transfer energetics deviates from the net charge that passes the membrane.