Reengineered carbonyl reductase for reducing methyl-substituted cyclohexanones.

The carbonyl reductase from Candida parapsilosis (CPCR2) is a versatile biocatalyst for the production of optically pure alcohols from ketones. Prochiral ketones like 2-methyl cyclohexanone are, however, only poorly accepted, despite CPCR2's large substrate spectrum. The substrate spectrum of CPCR2 was investigated by selecting five amino positions (55, 92, 118, 119 and 262) and exploring them by single site-saturation mutagenesis. Screening of CPCR2 libraries with poor (14 compounds) and well-accepted (2 compounds) substrates showed that only position 55 and position 119 showed an influence on activity. Saturation of positions 92, 118 and 262 yielded only wild-type sequences for the two well-accepted substrates and no variant converted one of the 14 other compounds. Only the variant (L119M) showed a significantly improved activity (7-fold on 2-methyl cyclohexanone; vmax = 33.6 U/mg, Km = 9.7 mmol/l). The L119M substitution exhibited also significantly increased activity toward reduction of 3-methyl (>2-fold), 4-methyl (>5-fold) and non-substituted cyclohexanone (>4-fold). After docking 2-methyl cyclohexanone into the substrate-binding pocket of a CPCR2 homology model, we hypothesized that the flexible side chain of M119 provides more space for 2-methyl cyclohexanone than branched L119. This report represents the first study on CPCR2 engineering and provides first insights how to redesign CPCR2 toward a broadened substrate spectrum.