Toward a systems approach to the human cytochrome P450 ensemble: interactions between CYP2D6 and CYP2E1 and their functional consequences

Abstract

Functional cross-talk among human drug-metabolizing cytochrome P450 through their association is a topic of emerging importance. Here, we studied the interactions of human CYP2D6, a major metabolizer of psychoactive drugs, with one of the most prevalent human P450 enzymes, ethanol-inducible CYP2E1. Detection of P450-P450 interactions was accomplished through luminescence resonance energy transfer between labeled proteins incorporated into human liver microsomes and the microsomes of insect cells containing NADPH-cytochrome P450 reductase. The potential of CYP2D6 to form oligomers in the microsomal membrane is among the highest observed with human cytochrome P450 studied up to date. We also observed the formation of heteromeric complexes of CYP2D6 with CYP2E1 and CYP3A4, and found a significant modulation of these interactions by 3,4-methylenedioxymethylamphetamine, a widespread drug of abuse metabolized by CYP2D6. Our results demonstrate an ample alteration of the catalytic properties of CYP2D6 and CYP2E1 caused by their association. In particular, we demonstrated that preincubation of microsomes containing co-incorporated CYP2D6 and CYP2E1 with CYP2D6-specific substrates resulted in considerable time-dependent activation of CYP2D6, which presumably occurs via a slow substrate-induced reorganization of CYP2E1-CYP2D6 hetero-oligomers. Furthermore, we demonstrated that the formation of heteromeric complexes between CYP2E1 and CYP2D6 affects the stoichiometry of futile cycling and substrate oxidation by CYP2D6 by means of decreasing the electron leakage through the peroxide-generating pathways. Our results further emphasize the role of P450-P450 interactions in regulatory cross-talk in human drug-metabolizing ensemble and suggest a role of interactions of CYP2E1 with CYP2D6 in pharmacologically important instances of alcohol-drug interactions.

Keywords: allosteric regulation; cytochrome P450; drug metabolism; membrane proteins; protein–protein interactions.

Fig 1

Interactions of CYP2D6-ERIA with CYP2D6-DYM mutants in ICM(CPR) microsomes studied with LRET. Main panel shows a series of spectra of delayed emission recorded during incubation of a 0.4 μM suspension of CYP2D6-ERIA incorporated into ICM(CPR) at R_L/P=300 with 0.4 μM CYP3A4(C166)-DYM. (The R_L/P ratio in the membrane after incorporation of DYM-labeled enzyme is equal to 150). The inset shows the time dependencies of normalized intensity of donor fluorescence obtained in the experiments at the R_L/P of 150, 225, 300 and 1200. Solid lines represent the approximation of the kinetic curves with a bi-exponential equation.

Fig. 2

The dependencies of the LRET amplitude on P450 concentration in the membrane of ICM(CPR) microsomes for the interactios in CYP2D6-ERIA/CYP2D6-DYM (a), CYP2D6-ERIA/CYP2E1-DYM (b), and CYP2D6-ERIA/CYP3A4(C58, C64)-DYM (c) pairs. The data sets shown in circles were obtained at no substrate added, while the data sets shown in trriangles were recorded in the presence of 50 μM MDMA. Solid and dashed lines show the approximations of the data sets with the equations (2) and (1), respectively.

Fig. 3

Oligomerization of CYP2D6 and its interactions with CYP2E1 in human liver microsomes studied with LRET. Main panel shows a series of spectra of delayed emission recorded during incubation of a 0.1 μM suspension of CYP2E1-ERIA incorporated into HLM with 0.1 μM CYP2E1-DYM. The inset shows the time dependencies of normalized intensity of donor fluorescence obtained in the course of interactions of CYP2D6-ERIA with DYM-labeled CYP2D6 (circles) and CYP2E1 (triangles) in the absence (open symbols) and in the presence of MDMA (closed symbols).

Fig. 4

The effect of CYP2D6-CYP2E1 interactions in ICM(CPR) microsomes on CYP2E1-dependent O-demethylation of 7-MFC (a) and CYP2D6-dependent O-demethylation of AMMC (b). The data sets obtained with the microsomes containing individual CYP2E1 and CYP2D6 enzymes are shown in circles and solid lines, triangles and dashed lines show the data obtained with a mixture of the two enzymes. Filled circles in panel b designate the data obtained after 15 min of preincubation with substrate, while the open circles show the data obtained without preincubation. The lines represent the approximations of the data sets with the Hill equation.

Fig. 5

Kinetics of activation of CYP2D6 during preincubation with substrate in mixed CYP2D6/CYP2E1 systems. Panel a shows the activation of AMMC O-demethylation during preincubation with AMMC in ICM(CPR) microsomes (circles, solid line) and CYP2D6-containing ICM(2D6) (triangles, dashed line) with co-incorporated CYP2E1. In these data sets the initial (100%) levels of activity correspond to turnover rates of 0.24 and 1.0 min⁻¹ for the data shown in circles and triangles, respectively. Panel b shows the result obtained with AMMC O-demethylation by CPR-containing proteoliposomes with co-incorporated CYP2D6 and CYP2E1 (R_L/P=150) upon their preincubation with AMMC (triangles) and MDMA (circles). In both datasets shown in this figure the initial (100%) rates of turnover were equal to 0.042 min⁻¹. The lines show the approximation of the data by an exponential equation.

Fig. 6

Schematic illustrating the hypothesis of positional heterogeneity in P450 oligomers and the proposed mechanism of time-dependent activation of CYPD6 (green) by its substrates in the presence of CYP2E1 (magenta). The proteins are present in equilibrium between oligomers (trimers in this figure) and monomers. The architecture of the oligomer gives rise to the positional dissimilarity of the subunits, which are stabilized in “active” and “hindered” conformations. The “hindered” subunits are hypothesized to be unable to bind substrates and interact with CPR. In the absence of substrate the distribution of CYP2D6 and CYP2E1 between the two subunit locations is random. However, addition of a substrate specific to CYP2D6 (bufuralol or MDMA in this study) results in its binding to the molecules of CYP2D6 occupying active positions, as well as to the CYP2D6 monomers. These interactions displace the conformational equilibrium in monomers towards the “active” conformation, so that after a slow dissociation and reassembly of the oligomers, the CYP2D6 molecules will eventually occupy all “active” positions and CYP2E1 will be completely displaced to the “hindered” locations. The bottom part of the scheme illustrates the situation with complete saturation of CYP2D6 with its substrate, so that all CYP2D6 molecules adopting the “active” conformation are presumed to be associated with substrate.