Theoretical methods will be applied to understand the structural consequences of mutations acquired during a directed evolution experiment. The iterative randomisation of active site residues was used to evolve the active site of the Aspergillus niger epoxide hydrolase towards higher enantioselectivity and provided a ninefold mutant. The goal of our computational investigation is the molecular understanding and the computational analysis of the contribution of the mutated residues on substrate recognition and biocatalysis. Classical molecular dynamics simulations will be used to investigate the Michaelis-Menten complex and the ester intermediates involved in the catalytic cycle. The primary goal of this project is a detailed molecular understanding of directed evolution in epoxide hydrolases through realistic MM and later QM/MM calculations, but we shall also investigate other substrates. It is hoped that this theoretical work will help to rationalize remote and cooperative effects of mutations, to identify structurally important regions that govern enantioselectivity, and to guide the planning of directed evolution experiments. The current proposal extends an ongoing collaboration with the experimental groups of M. T. Reetz and M. Arand.The second part of project will analyze the divergent evolution of monofunctional ((()8-barrel isomerases from the histidine (HisA) and tryptophan (TrpF) biosynthesis pathway applying our theorethical methods. The analysis of a bifunctional isomerase (PriA) which catalyses both the HisA and TrpF reaction should shed further light on the divergent evolution of different activites on a stable fold. The X-ray structures of PriA in complex with the reaction products solved in the group of M. Wilmanns in the framework of this priority program will be the basis for the MM and later also QM/MM modeling of the substrate binding, which is not easily accessible by experimental methods. Furthermore we will study the conformational flexibility of the loops in the active site using molecular dynamics methods. We hope to develop a predictive method for the simulation of ((()8-barrel enzyme mutants to perform a computational screening for new catalytic activities on the stable ((()8-barrel scaffold.
