Prolyl-4-hydroxylase (P4H) is a non-heme iron hydroxylase that regio- and stereospecifically hydroxylates proline residues in a peptide chain into R-4-hydroxyproline, which is essential for collagen cross-linking purposes in the human body. Surprisingly, in P4H, a strong aliphatic C-H bond is activated, while thermodynamically much weaker aliphatic C-H groups, that is, at the C and C positions, are untouched. Little is known on the origins of the high regio- and stereoselectivity of P4H and many non-heme and heme enzymes in general, and insight into this matter may be relevant to Biotechnology as well as Drug Development. The active site of the protein contains two aromatic residues (Tyr and Trp) that we expected to be crucial for guiding the regioselectivity of the reaction. We performed a detailed quantum mechanics/molecular mechanics (QM/MM) and molecular dynamics (MD) study on wild-type and mutant structures. The work shows that Trp is involved in key protein loop-loop interactions that affect the shape and size of the substrate binding pocket and its mutation has major long-range effects. By contrast, the Tyr residue is shown to guide the regio- and stereoselectivity by holding the substrate and ferryl oxidant in a specific orientation through hydrogen bonding and π-stacking interactions. Compelling evidence is found that the Tyr residue is involved in expelling the product from the binding pocket after the reaction is complete. It is shown that mutations where the hydrogen bonding network that involves the Tyr and Trp residues is disrupted lead to major changes in folding of the protein and the size and shape of the substrate binding pocket. Specifically, the Trp residue positions the amino acid side chains of Arg and Glu in specific orientations with substrate. As such, the P4H enzyme is a carefully designed protein with a subtle and rigid secondary structure that enables the binding of substrate, guides the regioselectivity, and expels product efficiently.

The cytochromes P450 are drug metabolizing enzymes in the body that typically react with substrates through a monoxygenation reaction. During the catalytic cycle two reduction and protonation steps generate a high-valent iron (IV)-oxo heme cation radical species called Compound I. However, with sufficient reduction equivalents present, the catalytic cycle should be able to continue to the reduced species of Compound I, called Compound II, rather than a reaction of Compound I with substrate. In particular, since electron transfer is usually on faster timescales than atom transfer, we considered this process feasible and decided to investigate the reaction computationally. In this work we present a computational study using density functional theory methods on active site model complexes alongside quantum mechanics/molecular mechanics calculations on full enzyme structures of cytochrome P450 enzymes. Specifically, we focus on the relative reactivity of Compound I and II with a model substrate for O–H bond activation. We show that generally the barrier heights for hydrogen atom abstraction are higher in energy for Compound II than Compound I for O–H bond activation. Nevertheless, for the activation of such bonds, Compound II should still be an active oxidant under enzymatic conditions. As such, our computational modelling predicts that under high-reduction environments the cytochromes P450 can react with substrates via Compound II but the rates will be much slower.