E. Coli AlkB catalyzes the direct dealkylation of various alkylated bases in damaged DNA. The diffusion of molecular oxygen to the active site in AlkB is an essential step for the oxidative dealkylation activity. Despite detailed studies on the stepwise oxidation mechanism of AlkB, there is no conclusive picture of how O2 molecules reach the active site of the protein. Yu et al. (Nature, 439, 879) proposed the existence of an intra-molecular tunnel based on their initial crystal structures of AlkB. We have employed computational simulations to investigate possible migration pathways inside AlkB for O2 molecules. Extensive molecular dynamics (MD) simulations, including explicit ligand sampling and potential of mean force (PMF) calculations, have been performed to provide a microscopic description of the O2 delivery pathway in AlkB. Analysis of intra-molecular tunnels using the CAVER software indicates two possible pathways for O2 to diffuse into the AlkB active site. Explicit ligand sampling simulations suggests that only one of these tunnels provides a viable route. The free energy path for an oxygen molecule to travel along each of these tunnels has been determined with AMBER and AMOEBA. Both PMFs indicate passive transport of O2 from the surface of the protein. However, the inclusion of explicit polarization shows a very large barrier for diffusion of the co-substrate out of the active site, compared with the non-polarizable potential. In addition, our results suggest that the mutation of a conserved residue along the tunnel, Y178, has dramatic effects on the dynamics of AlkB and on the transport of O2 along the tunnel.

It is a long-standing mechanistic consensus that the mutation of the proton-shuttle mediator Threonine (T) in Cytochrome P450 enzymes severs the water channel and thereby quenches the formation of the active species: the high-valent iron(iv)-oxo porphyrin π-cation radical species, compound I (Cpd I). Using MD simulations and hybrid QM/MM calculations of P450BM3 we demonstrate that this is not the case. Thus, while the original water channel is disrupted in the T268A mutant of the enzyme, a new channel is formed that generates Cpd I. With this new understanding, we address the puzzling regiochemical and kinetic-isotope effect (KIE) results (Volz et al., J. Am. Chem. Soc., 2002, 124, 9724-9725) on the sulfoxidation and N-dealkylation of dimethyl-(4-methylsulfanyl-phenyl)-amine by wild type (WT) P450BM3 and its T268A vs. F87A mutants. We show that the observed variable ratio of S/Me oxidation for these enzymes, vis-à-vis the constant KIE, originates from Cpd I being the sole oxidant. Thus, while the conserved KIE probes the conserved nature of the transition state, the variable regiochemical S/Me ratio reflects the active-site reorganization in the mutants: the shifted location of the new water channel in T268A tightens the binding of the S-end by Cpd I and increases the S/Me ratio, whereas the absence of π-interaction with the S-end in F87A creates a looser binding that lowers the S/Me ratio. Our results match the experimental findings. As such, this study sheds light on puzzling experimental results, and may shift a central paradigm in P450 research. The broader implication on enzymatic research is that a single-site mutation is not a localised alteration but one that may lead to a profound change in the active site, sufficiently so as to change the chemoselectivity of catalyzed reactions.