Directed Evolution of a Cytochrome P450 Carbene Transferase for Selective Functionalization of Cyclic Compounds

Abstract

Transfers of carbene moieties to heterocycles or cyclic alkenes to obtain C(sp^2)–H alkylation or cyclopropane products are valuable transformations for synthesis of pharmacophores and chemical building blocks. Through their readily tunable active-site geometries, hemoprotein “carbene transferases” could provide an alternative to traditional transition metal catalysts by enabling heterocycle functionalizations with high chemo-, regio-, and stereocontrol. However, carbene transferases accepting heterocyclic substrates are scarce; the few enzymes capable of heterocycle or cyclic internal alkene functionalization described to date are characterized by low turnovers or depend on artificially introduced, costly iridium–porphyrin cofactors. We addressed this challenge by evolving a cytochrome P450 for highly efficient carbene transfer to indoles, pyrroles, and cyclic alkenes. We first developed a spectrophotometric high-throughput screening assay based on 1-methylindole C3-alkylation that enabled rapid analysis of thousands of P450 variants and comprehensive directed evolution via random and targeted mutagenesis. This effort yielded a P450 variant with 11 amino acid substitutions and a large deletion of the non-catalytic P450 reductase domain, which chemoselectively C_3-alkylates indoles with up to 470 turnovers per minute and 18 000 total turnovers. We subsequently used this optimized alkylation variant for parallel evolution toward more challenging heterocycle carbene functionalizations, including C_2/C_3 regioselective pyrrole alkylation, enantioselective indole alkylation with ethyl 2-diazopropanoate, and cyclic internal alkene cyclopropanation. The resulting set of efficient biocatalysts showcases the tunability of hemoproteins for highly selective functionalization of cyclic targets and the power of directed evolution to enhance the scope of new-to-nature enzyme catalysts.