Communication between distant sites in a protein or other molecule requires a pathway through which a signal—whether charge carriers, electronic excitation, or mechanical force—can be transmitted between those sites. Mechanical coupling between sites, in which a perturbation to a site is transmitted to another site by an ensemble of vibrational modes, is of interest for understanding the mechanisms behind allostery, in which a ligand binding at one site induces structural changes that alter binding affinity for another ligand at a different site. Graph-based methods provide an ideal tool for quantifying mechanical coupling. We have developed a graph-based method that predicts mechanical coupling based on arbitrary structural and dynamics data and provides three complementary metrics that quantify the role of structural elements in the protein in mechanical communication. The mechanical coupling matrix M weights covariance of atomic motions by the strongest-coupled path available, identifying sites that have correlated and connected motions. The single-site centrality parameter χX identifies the most important structural elements for transferring mechanical energy in or out of a site, and the two-site centrality parameter χX,Yidentifies the most important structural elements along paths between two previously identified sites. We validated our method against a systematic mutation study (alanine scan) of human liver pyruvate kinase and found close agreement between residues identified as crucial along pathways between the allosteric and active sites and point mutations that increased or decreased the allosteric effect. We also validated the method against computational data on two G-protein coupled receptors, and have applied it to complex systems such as nitrogenase and respiratory complexes. Our method provides a powerful tool for identifying key components in a protein for understanding its underlying mechanism and potential allosteric effects.

Rapid graph-based determination of mechanical coupling in proteins

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.