Ensemble Networks for Allosteric Drug Design

Wesleyan University, Middletown CT

Investigators

Linked publications, trials & patents

Abstract

The development of allosteric drugs, an entirely new class of drugs operating by a mechanism widely used in biology to modulate protein activity, is stalled because a means of identifying allosteric binding sites has not yet been developed. The overarching goal of this proposal is to enable the development of such allosteric drugs through developing the needed allosteric site algorithm. The objective of this proposal is to develop and apply an algorithm that robustly identifies sites of allosteric control than can be used to rationally design drugs. The central hypothesis is that networks of covarying residues determined by Molecular Dynamics simulations identify residues in the protein that are allosterically linked to the active site and are thus suitable targets for allosteric drugs that will control the active site of the protein. The three aims of the project will realize rational design of a new class of therapeutics. First, the new Allosteric Residues of Control method (ARC) will be developed and implemented. A network of spatially and/or energetically covarying residues will be identified from Molecular Dynamics simulations. Residues in the network will be screened to identify the ones on the surface as likely candidates for interaction with drugs. They will then be tested by mutational MD analysis and/or rotameric perturbation. Preliminary results indicate the ability of both of these methods to test whether residues can allosterically affect the distant active site. Second, proof of concept will be demonstrated on three systems having known allosteric regulators with known points of allosteric control. The ARC procedure will be carried out blindly, and identified points of allosteric control will be compared to the known allosteric binding sites. Third, the method will be applied to two proteins involved in cancer: MutS, a central DNA repair protein whose mutations have no treatment, and p53, the guardian of the genome responsible for initiating apoptosis or repair of damaged cells. This work will significantly advance the development of the new class of allosteric drugs by creating an algorithm to predict residues of proteins to enable rational design of new therapeutics to these allosteric binding sites. This will enable the development of allosteric therapeutics for all allosteric proteins, thereby allowing drugs to be developed for targets that have not yet been able to be drugged by conventional means. This will lead to cures to diseases such as cancer and beyond. The project incorporates two major innovative angles of attack. The ARC method moves identification and testing of allosteric points of control from expensive wet lab screens into a computer based strategy, thus providing a major savings in monetary cost. The method also implements the novel MD based spectral analysis of protein dynamics to identify the allosteric network as an MD sector. The key innovation lies in rooting the prediction of points of allosteric control in the physical ensemble properties of the proteins via statistical mechanics. The innovative approaches will allow broad application of the method to quickly develop allosteric drugs to cure a wide range of diseases.