Physiological models to study the effect of molecular crowding on multi-drug bound proteins: insights from SARS-CoV-2 main protease.

Molecular Dynamics simulations are often used in drug design. However, such simulations do not account for the physiological environment of the receptor; hence overlook its impact on biomolecular interactions. To address this lacuna, we identified three objectives to pursue - develop models of physiological environment, study a drug-receptor complex in such environments, and identify methods to analyze these complicated simulations. Two novel physiological models were developed and studied. The first, called 'm10', comprises of 10 of the most abundant cytoplasmic metabolites at physiological concentrations. The second, called 'phy', supplements m10 with an additional crowder protein to elicit macromolecular crowding effect. The main protease (M) of SARS-CoV-2, being essential for viral replication, is an attractive drug target for COVID-19. Hence, we chose M docked with multiple drugs as our model drug-receptor system. With a plethora of compounds, physiological systems can be exceedingly large and complex. A novel Spark-based software (SparkTraj) was developed to rapidly analyze non-specific contacts and water interactions. Our study shows that crowding enhances the difference in the dynamics of apo- vs drug-bound complexes. Metabolites, at times as a cluster, were seen interacting with the protease, drugs, and binding sites in drug-free receptor. Except one that to an adjacent pocket in phy, the drugs remained in their respective pockets in all simulations. Given these observations, we hope that the models and approach presented here would help the optimization, evaluation, and selection of potential drugs. Generic biomolecular dynamics could also benefit from such models and tools.Communicated by Ramaswamy H. Sarma.