Cooperative changes in solvent exposure identify cryptic pockets, conformational switches, and allosteric coupling

Abstract

Conformational changes can dramatically alter a protein's function by changing the surfaces that are accessible to interact with binding partners. However, it is often difficult to hone in on the most relevant conformational changes from the cartesian coordinates of atoms on the protein's surface. Instead, we describe a protein's surface in terms of groups of residues that undergo cooperative changes in their solvent exposure. We term these groups exposons. We demonstrate that Markov state models (MSMs) elegantly identify the conformational transitions that give rise to an exposon, enabling users to rapidly find the most interesting conformational changes in their system. For example, this approach readily identifies previously-known cryptic allosteric sites and other functionally-relevant conformational transitions. Moreover, it predicts a cryptic allosteric site in an important target for combating antibiotic resistance that lacks known cryptic pockets. Experimental tests confirm that targeting this site reduces catalytic efficiency 15-fold.

Enspara: Modeling molecular ensembles with scalable data structures and parallel computing

“…For example, the opening of druggable cryptic allosteric sites can occur many orders of magnitude faster than the global unfolding process. 25,26 Thus, for biological questions where the underlying physical chemistry is irreducibly high-dimensional or the features in which it is low-dimensional are not known, building models with a large number of states is an effective strategy for ensuring that important states are not overlooked.…”