Compound I reactivity defines alkene oxidation selectivity in cytochrome P450cam.

Prediction of the chemoselectivity of drug oxidation by the human cytochrome P450 enzymes will aid in the avoidance of adverse drug reactions. The chemoselectivity of alkene oxidation is an important problem to address, as it can result in the formation of epoxides, which can have toxic effects. In this paper the epoxidation and hydroxylation of cyclohexene and propene by the bacterial P450(cam) isoform are modeled with hybrid quantum mechanical/molecular mechanical (QM/MM) methods. Snapshots for QM/MM modeling are chosen from molecular dynamics trajectories, to sample the different conformations of the enzyme-substrate complex. The energy barriers obtained for these processes are in qualitative agreement with experimental work, supporting the use of QM/MM methods in the study of selectivity for this class of enzyme. This work highlights the complexity involved in modeling these systems with QM/MM and the importance in the selection of starting geometries.