Atypical kinetics of cytochrome P450 2J2: Epoxidation of arachidonic acid and reversible inhibition by xenobiotic inhibitors

Abstract

Extrahepatic CYP2J2 metabolism of arachidonic acid (AA) to bioactive regioisomeric epoxyeicosatrienoic acids (EETs) is implicated in both physiological and pathological conditions. Here, we aimed to characterize atypical substrate inhibition kinetics of this endogenous metabolic pathway and its reversible inhibition by xenobiotic inhibitors when AA is used as the physiologically-relevant substrate vis-à-vis conventional probe substrate astemizole (AST). As compared to typical Michaelis-Menten kinetics observed for AST, complete substrate inhibition was observed for CYP2J2 metabolism of AA to 14,15-EET whereby velocity of the reaction declined significantly at concentrations of AA above 20-30 µM with an estimated substrate inhibition constant (Ks) of 31 µM. In silico sequential docking of two AA substrates to orthosteric (OBS) and adjacent secondary binding sites (SBS) within a 3-dimensional homology model of CYP2J2 revealed favorable and comparable binding poses of glide-scores -3.1 and -3.8 respectively. Molecular dynamics (MD) simulations ascertained CYP2J2 conformational stability with dual AA substrate binding as time-dependent root mean squared deviation (RMSD) of protein Cα atoms and ligand heavy atoms stabilized to a plateau in all but one trajectory (n=6). The distance between heme-iron and ω6 (C14, C15) double bond of AA in OBS also increased from 7.5 ± 1.4 Å to 8.5 ± 1.8 Å when CYP2J2 was simulated with only AA in OBS versus the presence of AA in both OBS and SBS (p<0.001), supporting the observed in vitro substrate inhibition phenomenon. Poor correlation was observed between inhibitory constants (Ki) determined for a panel of nine competitive and mixed mode xenobiotic inhibitors against CYP2J2 metabolism of AA as compared to AST, whereby 4 out of 9 drugs had a greater than 5-fold difference between Ki values. Nonlinear Eadie-Hofstee plots illustrated that complete substrate inhibition of CYP2J2 by AA was not attenuated even at high concentrations of xenobiotic inhibitors which further corroborates that CYP2J2 may accommodate three or more ligands simultaneously. In light of the atypical kinetics, our results highlight the importance of using physiologically-relevant substrates in in vitro enzymatic inhibition assays for the characterization of xenobiotic-endobiotic interactions which is applicable to other complex endogenous metabolic pathways beyond CYP2J2 metabolism of AA to EETs. The accurate determination of Ki would further facilitate the association of xenobiotic-endobiotic interactions to observed therapeutic or toxic outcomes.