Perturbation Response Scanning Method for Identifying Allosteric Transitions and Utilizing in Flexible Docking

Abstract

We have recently developed coarse-grained method; perturbation response scanning (PRS) that couples elastic network models with linear response theory (LRT). It computes the response of the protein structure (i.e. displacement vector) upon exerting directed random forces on selected residues. The method has proven successful in reproducing residue displacements for a set of 25 proteins that display a variety of conformational motions upon ligand binding[1]. Using PRS we analyzed two PDZ domain proteins (PSD95 PDZ3 domain and hPTP1E PDZ2 domain) whose allosteric behavior play a key role in signaling. By PRS, we first identified the residues that give the highest response upon perturbing the binding sites. Strikingly, we observe that the residues that give the highest response agree with experimentally determined residues involved in allosteric pathways. Second, we constructed the allosteric pathways by clustering the residues giving same type of response upon perturbation of the binding sites. Interestingly our analysis provided molecular understanding of experimentally observed hidden allostery of PSD95. We have shown that removing the distal alpha helix from the binding site alters the allosteric pathway and decreases the binding affinity. Overall, these results indicate that (i) PRS is successful in capturing the conformational changes upon binding[1], (ii) it can identify key residues that mediate long-range communication in PDZ domain proteins[2], (iii) we can construct the allosteric pathways and show how the allosteric pathway changes upon minor alteration in the fold with PRS[2]. Utilizing these exceptional features of PRS, we have recently developed a flexible docking scheme which predicts the peptide-protein interactions, 1. Atilgan C, Gerek ZN, Ozkan SB, Atilgan, AR, Biophys J 99, 933-943 (2010). 2. Gerek ZN and Ozkan SB, submitted to PLoS Comp Biol, (2010).