Enantioselective oxidation of unactivated C–H bonds in cyclic amines by iterative docking-guided mutagenesis of P450BM3 (CYP102A1)

Abstract

Selective oxidation of ring C–H bonds is an attractive route to functionalized cyclic amines, which are versatile intermediates in drug synthesis and important fragment molecules in drug discovery. Here we report a combined substrate and enzyme engineering approach to achieve enantioselective functionalization of all unactivated C–H bonds of azepane, azocane, 7-azabicyclo[2.2.1]heptane and 8-azaspiro[4.5]decane by cytochrome P450BM3 (CYP102A1). Different N-modifying groups provide product diversity at high enantioselectivity (up to 99% e.e.) from a panel of just 48 variants of P450BM3. Substrate docking into molecular-dynamics-simulated structures of enzyme variants is shown to be useful for designing mutations to increase enantioselectivity by disfavouring binding poses leading to the unwanted enantiomer, and to increase enzymatic activity by disfavouring non-productive poses from ten or so variants per generation. The synthetic application of remote C–H activation within cyclic amines is exemplified by the synthesis of anisodamine via enantioselective hydroxylation of N-Boc-nortropinone.