The conformational space of the SARS-CoV-2 main protease active site loops is determined by ligand binding and interprotomer allostery.

Abstract

In this study, we characterize two SARS-CoV-2 main protease (Mpro) active site loops that play a role in catalysis using large scale all-atom molecular dynamics simulations. Machine-learning methods are used to identify open and closed states in both the upper and lower loops, as well as an additional intermediate state in the lower loop. In the closed and intermediate states, the positions of the catalytic Cys-His dyad are stabilized, bringing Asp187 into position to mediate charge from a proton transfer during catalysis. Simulations of a protease substrate mimetic complex exhibit an increased frequency of closed and intermediate states and reveals that a substrate binding to one protomer of the homodimer causes its apo partner to more closely resemble the structural ensemble of the bound state. We use dynamic network analysis to reveal the optimal allosteric path between active sites, which travels through a pathway bridged by the N-terminus, providing useful information for the development of potential allosteric inhibitors. This study offers insight into relationships between the flexible loops and the substrate binding in a prime drug target.