Computational design and screening of enzyme enantioselectivity

The aim of this thesis is to explore computational methodologies for the design and screening of enzyme variants for use in green chemistry. Methods such as docking calculations, Rosetta enzyme redesign, molecular dynamics simulations and quantum mechanical calculations are used to predict the catalytic activity and enantioselectivity of enzyme variants. Pipelines incorporating such methods will reduce the amount of costly and time-consuming laboratory work that is needed for biocatalyst development by protein engineering. Target enzymes are cytochrome P450s and epoxide hydrolases, both of which are important for industrial synthesis of fine chemicals and pharmaceuticals.
In general the activity and selectivity of enzymes depends on multiple steps of the catalytic cycle, e.g. formation and dynamics of enzyme-substrate complexes, energy of transition states, and release of products. This complexity may hinder the use of computational methods for design and prediction of better enzymes, but simplifications are possible. For example, enantioselectivity is mainly determined by the conformation of substrate bound in the active site, which in this thesis was controlled using Rosetta enzyme design software. Because Rosetta often gives a number of solutions that is too large for experimental evaluation, ranking of the predicted designs is important. The thesis shows that molecular dynamics simulations can be used for that purpose. By scoring the frequency of reactive poses, molecular dynamics reveals enzyme variants in which substrate binds in a position that supports the desired activity or selectivity. In cases where major changes in the geometry of the bound substrate occur during the reaction, quantum mechanical methods (hybrid DFT) were used to further examine the selectivity.
Based on the results, the thesis recommends a pragmatic balance between accuracy and cost of computational methods when developing in silico pipelines for the design of tailored enantioselective biocatalysts.