Engineering of Cyclohexanone Monooxygenase for the Enantioselective Synthesis of (S)‑Omeprazole

• Published in: ACS sustainable chemistry & engineering 2019-04, Vol.7 (7), p.7218-7226
• Main Authors: Zhang, Yan, Wu, Yin-Qi, Xu, Na, Zhao, Qian, Yu, Hui-Lei, Xu, Jian-He
• Format: Article
• Language: English
• ISSN: 2168-0485; 2168-0485
• DOI: 10.1021/acssuschemeng.9b00224

Enzymatic asymmetric sulfoxidation using molecular oxygen as the oxidant is a promising green chemistry approach to chiral sulfoxide production. Despite the broad substrate spectrum of cyclohexanone monooxygenases (CHMOs), some unnatural substrates with bulky functional groups, such as the pharmaceutically relevant omeprazole sulfide, cannot be effectively accepted by CHMOs. Herein, we describe a set of variants derived from an Acinetobacter calcoaceticus CHMO (AcCHMO), whose active sites adjacent to the substrate tunnel were altered to shift the substrate specificity from cyclohexanone monooxygenation toward omeprazole sulfide sulfoxidation. We performed homologous modeling and molecular docking to identify key residues that might affect the substrate specificity. Two libraries of residues lining the active center of AcCHMO were then constructed and screened by an effective halo-based selection method using the solubility difference between the substrate (omeprazole sulfide) and product (esomeprazole). Functional evaluation of the resultant variants showed that the substrate specificity of AcCHMO was markedly altered from the small natural substrate (cyclohexanone) toward the desired bulky substrate (omeprazole sulfide) despite the extremely poor activity detected even for the best variant, M2 (0.61 U/gprot). The crystal structure of M2 complexed with a flavin adenine dinucleotide (FAD) prosthetic group was determined, which provided insight into the altered substrate specificity. To improve the activity of enzyme M2 toward pharmaceutical precursor omeprazole sulfide, we performed both local and global protein engineering among the two CASTing libraries surrounding FAD+ and NADP+ prosthetic groups and an error-prone PCR library of the full-length AcCHMO. As a result, variant M6 was obtained, giving a 50-fold higher activity compared to M2. This structure-guided protein engineering of AcCHMO provided a promising candidate for converting omeprazole sulfide into (S)-omeprazole using a green biocatalytic method.