Toward understanding whole enzymatic reaction cycles using multi-scale molecular simulations
Shingo Ito, Chigusa Kobayashi, Kiyoshi Yagi, Yuji Sugita
Abstract
Enzymes effectively catalyze chemical reactions at their active sites. The reactions involve three microscopic events at the active sites: substrate binding, multi-step chemical reactions, and product release. These events are often coupled with enzyme conformational changes, making theoretical and computational analyses more challenging. Advanced molecular simulations, involving molecular dynamics (MD) and hybrid quantum mechanics/molecular mechanics (QM/MM), are now utilized to investigate the functions of enzymes such as tryptophan synthase and P-type ATPases. Here, we summarize recent multiscale molecular simulations that incorporate multiple microscopic events in enzyme functions. The coupling of enzyme conformational changes and chemical reactions can predict a proper direction in enzymatic reaction cycles, which requires accurate predictions of the free energy changes between different physiological states. Using machine learning (ML) methods, all the microscopic events in enzyme catalysis could be described with the same accuracy as quantum chemistry. We also discuss recent developments in ML/MM simulations for enzyme catalysis.