Modeling Hydrophobic and Hydrophilic Interactions

Columbia Univ New York Morningside, New York NY

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Biological processes often depend on ligand binding to receptors yet accurate calculation of the associated binding free energy remains a significant challenge of central importance to structure-based drug design. Water molecules participate in biological processes, including ligand binding to proteins, and the mechanical stability and aggregation of proteins. We propose to use molecular simulations and theory to provide deeper insights into water's role in the above processes. We propose to develop methods for the accurate calculation of protein-ligand binding affinities based on our recent progress towards developing a displaced solvent functional methodology (DSM). We propose to develop our WaterMap method into a practical tool for medicinal chemistry, by extending it to heterogeneous protein receptor interfaces with mixtures of hydrophilic, hydrophobic and neutral residues and to apply it to libraries of receptors and ligands. This work is part of an ongoing fruitful collaboration with the Friesner group at Columbia. To expedite this research we propose to combine our new colored noise multiple time step MD algorithm (CN-RESPA) with our replica exchange with solute tempering (REST) algorithm and the existing ;-hopping free energy perturbation algorithm (;-hopping FEP) to provide an extremely powerful methodology for calculating binding efficiencies and sampling conformational states in the above research. Huntington's disease is caused by a genetic expansion of a region containing the CAG codon that leads to an increase in the length of the protein huntingtin's polyglutamine tract and a subsequent triggering of the formation of cytotoxic polyglutamine amyloid fibers and plaques. It is thought that a single rare misfolded polyglutamine polypeptide serves as a nucleus for the rapid formation of aggregated oligomers and finally the characteristic amyloid, but structural biologists have been unable to determine structures of these misfolded conformations. In preliminary simulations we found examples of highly collapsed long lived polyQ double-back structures that are mechanically resilient against large pulling forces (as is seen in AFM experiments). We propose to perform long enough molecular simulations using powerful computers like ANTON to pin down the structures of the mechanically resilient misfolded conformations of polyQ, to explore the balance between residue-water and residue-residue interactions to better understand why polyQ with its highly polar side-chain groups forms such compact structure, and, then, to get insights into conformational transitions of polyQ during aggregation.