Glutamate–haem ester bond formation is disfavoured in flavocytochrome P450 BM3: characterization of glutamate substitution mutants at the haem site of P450 BM3

Abstract

Bacillus megaterium flavocytochrome P450 BM3 (CYP102A1) is a biotechnologically important cytochrome P450/P450 reductase fusion enzyme. Mutants I401E, F261E and L86E were engineered near the haem 5-methyl group, to explore the ability of the glutamate carboxylates to form ester linkages with the methyl group, as observed for eukaryotic CYP4 relatives. Although no covalent linkage was detected, mutants displayed marked alterations in substrate/inhibitor affinity, with L86E and I401E mutants having lower Kd values for arachidonic acid and dodecanoic (lauric) acid than WT (wild-type) BM3. All mutations induced positive shifts in haem Fe(III)/Fe(II) potential, with substrate-free I401E (-219 mV) being >170 mV more positive than WT BM3. The elevated potential stimulated FMN-to-haem electron transfer ~2-fold (to 473 s-1) in I401E, and resulted in stabilization of Fe(II)O2 complexes in the I401E and L86E P450s. EPR demonstrated some iron co-ordination by glutamate carboxylate in L86E and F261E mutants, indicating structural plasticity in the haem domains. The Fe(II)O2 complex is EPR-silent, probably resulting from antiferromagnetic coupling between Fe(III) and bound superoxide in a ferric superoxo species. Structural analysis of mutant haem domains revealed modest rearrangements, including altered haem propionate interactions that may underlie the thermodynamic perturbations observed. The mutant flavocytochromes demonstrated WT-like hydroxylation of dodecanoic acid, but regioselectivity was skewed towards omega-3 hydroxydodecanoate formation in F261E and towards omega-1 hydroxydodecanoate production in I401E. Our data point strongly to a likelihood that glutamate-haem linkages are disfavoured in this most catalytically efficient P450, possibly due to the absence of a methylene radical species during catalysis.