Quantitatively interpreted enhanced inhibition of cytochrome P450s by heteroaromatic rings containing nitrogen

It has been known for a long time that certain substructures bind to the heme iron in cytochromes P450. Detection of spectroscopic changes and crystal structures of protein ligand complexes have provided qualitative evidence, including for aromatic nitrogen-containing ligands. Compounds containing these same substructures are more likely to inhibit cytochrome P450s than expected due to lipophilicity. These two sets of observations are not easily linked by experiment, because binding to the iron atom alone is not readily measured. Quantum mechanical (density functional) calculations of binding energies for a number of different aromatic heterocycles to heme iron in a range of oxidation and spin states can provide a quantitative link between the observed structures and the biochemical inhibition that is measured. The studies reported here for a set of heteroaromatic rings containing nitrogen begin with quantum mechanical calculations which provide geometries and binding energies. Subsequently, AstraZeneca's database of cytochrome P450 inhibition assays has been searched to find data that are relevant to the same set of heteroaromatic compounds. These data have been analyzed in a number of fashions to account for both the narrow dynamic range of the assays and the lipophilicity dependence of this kind of inhibition. Finally, crystal structures have provided experimental geometric information. Taken together these different sources suggest that binding to the metal in our inhibition assays is dominated by Fe(III) in its doublet state, most likely occurring when the iron is pentavalent. Computed binding energies to this state contrast with the hydrogen-bond acceptor ability and basicity of the compounds, neither of which are able to correctly account for the effect of the particular environment in which the iron is found. This highlights the value of modeling biochemical events as closely as can be computationally afforded. The computational protocol devised was used to make predictions about a set of as yet unknown heteroaromatic compounds suggested by Pitt et al.