Theoretical modeling of ligand-gated ion channels

The goal of the studies described in this thesis was to develop computational models of ligand gated ion channels to better understand their functionality.; Ligand-gated ion channels (LGIC) play many vital roles in living cells. They are complex biological machines that combine several functionally distinctive domains. Typically, the two most important domains of LGIC are the ligand binding (LB) domain and the transmembrane (TM) domain which forms a channel. LGICs function by regulating conformational state of the TM domain in response to a ligand binding event.; This thesis concerns the methods to investigate the ligand-protein interactions on atomic and molecular levels as well as methods to predict the three-dimensional structure of ligand-gated ion channel domains. In the first part of the manuscript we describe the development of a method to calculate vibrational frequencies of a ligand residing in the inhomogeneous dielectric environment of the ligand binding site. The LB domain of the Glutamate receptor of AMPA type and its interactions with the agonist ligand glutamate has been then studied using our hybrid molecular dynamics (MD)/quantum mechanics (QM) method, as well as using the hybrid continuum electrostatics/molecular dynamics method to calculate ligand binding energy, electrostatic potential, and the role of the structured water molecules in the LB pocket. Conformations of the peptides connecting the LB and TM domains of the Glutamate receptor were modeled using Replica Exchange Molecular Dynamics (REMD).; The remainder of the document describes studies which involve knowledge-based approaches. A homology model of the LB domain of the Glycine receptor (another proteins of the LGIC class) had been proposed and its stability verified in MD simulations. Finally, a novel knowledge-based potential that can be used for prediction of the TM domain structure in membrane proteins has been developed and tested in this work.; The results of this thesis can be applied for predictions of ligand-protein interactions, 3D prediction of structural elements, and understanding the mechanism of ligand gating.