Computation-Aided Engineering of Cytochrome P450 for the Production of Pravastatin

Mark A. Ashworth, Elvira Bombino, René M. De Jong, Hein J. Wijma, Dick B. Janssen, Kirsty J. McLean, Andrew W. Munro

Research output: Contribution to journalArticlepeer-review

Abstract

CYP105AS1 is a cytochrome P450 from Amycolatopsis orientalis that catalyzes monooxygenation of compactin to 6-epi-pravastatin. For fermentative production of the cholesterol-lowering drug pravastatin, the stereoselectivity of the enzyme needs to be inverted, which has been partially achieved by error-prone PCR mutagenesis and screening. In the current study, we report further optimization of the stereoselectivity by a computationally aided approach. Using the CoupledMoves protocol of Rosetta, a virtual library of mutants was designed to bind compactin in a pro-pravastatin orientation. By examining the frequency of occurrence of beneficial substitutions and rational inspection of their interactions, a small set of eight mutants was predicted to show the desired selectivity and these variants were tested experimentally. The best CYP105AS1 variant gave >99% stereoselective hydroxylation of compactin to pravastatin, with complete elimination of the unwanted 6-epi-pravastatin diastereomer. The enzyme-substrate complexes were also examined by ultrashort molecular dynamics simulations of 50 × 100 ps and 5 × 22 ns, which revealed that the frequency of occurrence of near-attack conformations agreed with the experimentally observed stereoselectivity. These results show that a combination of computational methods and rational inspection could improve CYP105AS1 stereoselectivity beyond what was obtained by directed evolution. Moreover, the work lays out a general in silico framework for specificity engineering of enzymes of known structure.

Original languageEnglish
Pages (from-to)15028-15044
Number of pages17
JournalACS Catalysis
Volume12
Issue number24
Early online date28 Nov 2022
DOIs
Publication statusPublished - 16 Dec 2022

Cite this