Molecular dynamics study of the human voltage-gated proton channel, hHv1

A membrane protein, the human voltage-gated proton channel ( hHV1) is crucial for many physiological processes of a wide range of cell types. Such processes include bacterial killing by white blood cells, allergic responses by basophils, airway pH regulation in respiratory tracts, and sperm maturation. hHV1 is homologous to the voltage sensor domains (VSD) of voltage-gated cation channels. As such, hHV1 is predicted to adopt a four-helix (S1-S4) bundle structure with three highly-conserved Arg residues (R1-R3) on S4 as well as a number of other charged residues on S1-S3. As a channel, hHV1displays a remarkable diversity in ion selectivity upon mutating residue Asp112 on S1, the putative selectivity filter of the pore. hHV1carrying single or double point mutations of Asp112 along with Val116 makes the channel either proton-selective, anion-selective, or nonconducting. The molecular basis for ion and charge selectivity in hHV1remains unexplained. The lack of an experimentally-resolved native structure for hHV1 has hindered not only the uncovering of the molecular basis for ion permeation and charge selectivity in hHV1, but also the advancement of its potential applications as a drug target for the treatment of cerebral damage from ischemic stroke as well as of colorectal and breast cancers. Here, I investigate the molecular basis for ion and charge selectivity of hHV1, using molecular dynamics simulations. I first present ROMP, a general, computationally-efficient simulation approach to predict the solvation and the orientation of peptides and proteins in lipid bilayers. The method, which employs a biphasic membrane mimetic to accelerate the convergence of solvent properties, is shown to reproduce the solvation and structural fluctuations of various types of membrane proteins and peptides in lipid bilayers. I then describe the construction and the validation of two homology models (HM) of hHV1 based on three VSD templates and differing in the registry of S4 with respect to S1-S3 helices. By incorporating ROMP, the relative quality of these two models is evaluated using massively-repeated simulations in the biphasic membrane mimetic. The structural properties of the better model are consistent with experimental accessibility data and a priori features of a proton-conducting channel, with a putative proton permeation pathway comprised of an intermittent hydrogen-bonded water chain in a narrow non-polar bottleneck connecting two water-filled crevices lined with charged residues. Next, I carry out a comparative assessment of the structure and dynamics of wild-type and mutant channels in a lipid bilayer. This analysis reveals a reorganization of the salt-bridge network involving primarily R2 and mutations at positions 112 and/or 116. The presence or absence of charge compensation of R2 by mutations modulates the electrostatic properties of the pore consistently with the proton or anion selectivity of the mutants, respectively. Finally, I use free energy simulations to investigate the thermodynamic basis for the movement of Na+ and Cl- in the wild-type and two anion-selective mutant channels. Results indicate that the movement of non-proton ions in the wildtype is opposed at the constriction site of the pore, where both types of ions face a substantial desolvation penalty. Although the barrier for Cl? is significant even in the anion- selective mutants, limited structural rearrangements in the constriction site and, as previously shown by our electrostatic calculations, the absence of a negatively charged residue at position 112, lower the barrier for Cl- in mutants compared to the wildtype channel. Taken together, these findings provide clues to refine our structural model as well as insight that can be used to further the elucidation of ion and charge selectivity and the molecular mechanism of proton translocation in hHV1.