Computational Methods for Analysis and Redesign of Enzyme Active Sites
The rational design of enzymes and understanding of enzyme mechanisms both present a tremendous challenge in terms of both computational and experimental methodologies. In this work, we review the current state-of-the-art techniques in enzyme design and present a new computational method for active site redesign, called SABER (Selection of Active/Binding sites for Enzyme Redesign). This program was used to analyze both an existing enzyme redesign from the literature, o-succinyl benzoate synthase, and to predict new scaffolds that might be redesigned function similarly to a designed Kemp elimination enzyme. The logic behind the program is discussed in detail with examples taken from the code. Next, SABER was applied to the problem of predicting catalytic functionality in enzyme active sites. This methodology was used to analyze the active sites from two well-studied enzyme families, the serine proteases and the tyrosine phosphatases, and was found to predict catalytic function with high accuracy. It was then used to analyze the active site of orotidine 5'-monophosphate decarboxylase, an enzyme with an unknown mechanism, and used to support of an imminium-based mechanism of catalysis. Finally, a very challenging rational enzyme design process to catalyze an aromatic Claisen rearrangement is discussed. This led to the design and synthesis of six enzymes, none of which showed any rate acceleration versus the background reaction. During this work, an aromatic Claisen substrate that rearranges rapidly in water was discovered and characterized. This was used to highlight the difficulties in both the enzyme design process and in prediction of hydrogen-bond catalyzed reaction rates in water.