Enzymatic enantioselective reduction is an excellent technique in pharmaceutical manufacturing, enabling elaborate chirality tuning. KREDs have been implemented into practice for producing many chiral drug intermediates, while multihalogen-substituted acetophenones persist as a monolith in the path toward on-demand inversion of the chirality. (S)-2-Chloro-1-(2,4-difluorophenyl) ethanols constitute the key chiral fragment of voriconazole, of which the S-enantiomer is more rarely achieved than its counterpart through industrial reductases, i.e., LfSDR1s. In this regard, a selection strategy based on residues with opposing physicochemical properties was developed to construct a minimal but highly effective library of LfSDR1, where five amino acids (namely, A, W, T, S, and L) were classified and selected by hydrophobicity and steric hindrance. LfSDR1-E141T-G92S-V186W (denoted as TSW, S-preferred) and LfSDR1-E141L-G92W-V186A (denoted as LWA, R-preferred) were screened out from a 17-membered library, achieving S-specific reduction with 168 mM 1a in 20 min. Furthermore, various halogenated 2-phenylethanol derivatives that could constitute diverse drug intermediates were stereoselectively prepared with optimized LfSDR1 variants. This success highlights the advantages of devising a labor-saving yet effective approach toward asymmetric reduction of target substrates, balancing activities and enantioselectivities. Our findings offer a powerful choice for shortening the timelines of enzyme engineering works.
Wang et al. (Fri,) studied this question.