Computational Enzyme Design

Citation

Bolon, Daniel N. (2002) Computational Enzyme Design. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/W7F3-DS14. https://resolver.caltech.edu/CaltechETD:etd-01252002-100801

Abstract

The long-term objective of computational enzyme design is the ability to generate efficient protein catalysts for any chemical reaction. This thesis develops and experimentally validates a general computational approach for the design of enzymes with novel function.

In order to include catalytic mechanism in protein design, a high-energy state (HES) rotamer (side chain representation) was constructed. In this rotamer, substrate atoms are in a HES. In addition, at least one amino acid side chain is positioned to interact favorably with substrate atoms in their HES and facilitate the reaction. Including an amino acid side chain in the HES rotamer automatically positions substrate relative to a protein scaffold and allows protein design algorithms to search for sequences capable of interacting favorably with the substrate. Because chemical similarity exists between the transition state and the high-energy state, optimizing the protein sequence to interact favorably with the HES rotamer should lead to transition state stabilization. In addition, the HES rotamer model focuses the subsequent computational active site design on a relevant phase space where an amino acid is capable of interacting in a catalytically active geometry with substrate.

Using a HES rotamer model of the histidine mediated nucleophilic hydrolysis of p-nitrophenyl acetate, the catalytically inert 108 residue E. coli thioredoxin as a scaffold, and the ORBIT protein design software to compute sequences, an active site scan identified two promising active site designs. Experimentally, both candidate ?protozymes? demonstrated catalytic activity significantly above background. In addition, the rate enhancement of one of these ?protozymes? was the same order of magnitude as the first catalytic antibodies.

Because polar groups are frequently buried at enzyme-substrate interfaces, improved modeling of buried polar interactions may benefit enzyme design. By studying native protein structures, rules have been developed within the scope of protein design that require core polar residues to largely satisfy their hydrogen bonding potential. Using this polar strategy to design the core of thioredoxin resulted in a protein that was thermodynamically stabilized relative to both the wt protein and a protein designed without core polar residues.

The enzyme design procedures presented here may serve as a platform to develop more detailed methods. It is hoped that the development and experimental testing of more detailed methods will continue to improve our understanding of enzyme mechanism and lead to the long-term goal of designing highly efficient enzymes.

Thesis Files