Modeling Nonnatural Flavoenzyme Catalysis with Molecular Dynamics and Quantum Chemical Methods

Abstract

Enzyme catalysts have become an integral part of organic synthesis over the past few decades. One of several potential advantages over metal catalysts is that enzymes are often able to access challenging asymmetric transformations with higher enantiopurity and yield than achiral metal catalysts. In this work, two particular radical enzyme catalysts, OPR1 and GluER-T36A, were modeled with molecular dynamics and quantum chemical methods to better understand mechanistic details that underly both the stereoselectivity and yield of these reactions. Specifically, ab initio calculations of substrate binding free energies were calculated both by umbrella sampling and by trajectory-average MM-PBSA and MM-GBSA. Correlations were found between calculated binding affinities and experimental reaction yields which suggest that enzyme catalytic efficiency is dictated by substrate-enzyme binding. These correlations demonstrate that optimization of these catalytic systems could be computationally guided by binding free energy calculations for a large number of substrate and enzyme pairings. Similar methods were also used to elucidate the chirality of GluER-T36A and OPR1 catalysts. Contrary to the stepwise mechanism proposed in this work, simulation data suggest a more concerted reaction that does not involve a true radical intermediate. Quantum chemical calculations of the GluER-T36A substrate complex are also presented, corroborating empirical spectral data and offering a qualitative picture for the mechanism of photoinitiated electron transfer. Elaborations on this work are proposed to make such quantum chemical models more rigorous and quantitative.