Computational Approaches for Understanding and Redesigning Enzyme Catalysis

Abstract

The remarkable specificity and catalytic efficiency of many enzymes make them attractive for applications ranging from therapeutics to chemical manufacturing. However, it remains challenging to identify the specific structural and dynamic mechanisms underlying the catalytic power of enzymes, which has limited our ability to re-engineer catalytic properties. In this thesis, I address these shortcomings by developing and demonstrating computational strategies comprised of techniques spanning statistical mechanics, machine learning, and protein design, and I apply them to the enzyme ketol-acid reductoisomerase (KARI), whose economic viability for the production of isobutanol would be strengthened by enhancing its activity on one of its two native substrates: 2-acetolactate (ACL). While computational enzyme redesign strategies for increased activity have traditionally focused on decreasing the energetic gap between the enzyme-substrate ground state and transition state, this thesis postulates and evaluates whether a more holistic treatment including the dynamics of complete turnover events could further elucidate properties affecting turnover efficiency and guide the identification of mutants with enhanced catalytic function.

In the first study, we describe a novel redesign strategy for enhanced specific activity (turnover number) based on analysis of enzyme-substrate turnover dynamics. The approach combined statistical mechanical path sampling algorithms and machine learning methods to identify the structural characteristics of enzyme-substrate complexes primed for successful conversion of substrate to product, which were then energetically stabilized by mutating KARI. A subset of candidate mutants were tested using path sampling-based reaction rate constant calculations, and eight mutants were identified with computed improvements in turnover number of up to four orders of magnitude for the isomerization of ACL. Further analysis revealed structural mechanisms by which enhanced activity was attained. In the second study, we examine the effects of these same mutations on the isomerization of KARI's other native substrate: 2-aceto-2-hydroxybutyrate (AHB), and we find that the mutants selected for increased activity on ACL had varied levels of activity on AHB. These variations in mutant activity on AHB were explained by analysis of WT-AHB simulations, which showed that only some of the structural mechanisms related to enhanced ACL catalysis transferred to, and thereby facilitated, AHB catalysis. This thesis highlights the influence of conformational states that are visited during the dynamics of substrate turnover and their role on enzyme catalysis, and it furthermore suggests a framework with which researchers may consider and apply these effects when engineering catalytic function.