Revisiting strategies and their combinatorial effect for introducing peroxygenase activity in CYP102A1 (P450BM3)

Abstract

Different strategies have previously been reported to convert cytochrome P450 monooxygenases to peroxygenases, allowing H2O2-driven oxyfunctionalization reactions. Comparison of the BM3 (CYP102A1) peroxygenase variant 21B3, obtained through enlargement of the active site by the F87A mutation followed by mutational stabilization towards H2O2, with the BM3_T268E variant, with an acid-base catalyst introduced, showed 21B3 to be the superior peroxygenase. A combination of these two strategies (21B3_T268E combinatorial mutant), however, resulted in reduced peroxygenase activity. The further introduction of the F(A)87V and A328F mutations (87–328 variants), previously reported to improve the regioselectivity of BM3 on n-alkanes, resulted in a loss of activity towards dodecanoic acid, a substrate commonly used to evaluate the activity of BM3. Although a reduction in activity was observed for styrene, the combinatorial mutants yielded higher enantioselectivities for R-styrene oxide. The activity towards α-olefins were, however, comparable between the different peroxygenase variants of BM3, but with absolute selectivity for epoxidation observed with the combinatorial mutants 21B3_87-328 and 21B3_T268E_87-382. Structural investigation of the active site architecture showed significant differences in the I-helix and heme accessibility. According to these results, future directed evolution studies will require selective pressure not only for H2O2, but also the specific substrate to be activated rather than surrogate substrates.