P450 oxidoreductase (POR) transports electrons from NADPH to all microsomal cytochrome P450 enzymes, including steroidogenic P450c17, P450c21 and P450aro. Severe POR mutations A287P (in Europeans) and R457H (in Japanese) cause the Antley-Bixler skeletal malformation syndrome (ABS) plus impaired steroidogenesis (causing genital anomalies), but the basis of ABS is unclear. We have characterized the activities of ~40 POR variants, showing that assays based on P450c17 activities, but not cytochrome c assays, correlate with the clinical phenotype. The human POR gene is highly polymorphic: the A503V sequence variant, which decreases P450c17 activities to ~60%, is found on ~28% of human alleles. A promoter polymorphism (~8% of Asians and ~13% of Caucasians) at −152 reduces transcriptional activity by half. Screening of 35 POR variants showed that most mutants lacking activity with P450c17 or cytochrome c also lacked activity to support CYP1A2 and CYP2C19 metabolism of EOMCC (a fluorogenic non-drug substrate), although there were some remarkable differences: Q153R causes ABS and has ~30% of wild-type activity with P450c17 but had 144% of WT activity with CYP1A2 and 284% with CYP2C19. The effects of POR variants on CYP3A4, which metabolizes nearly 50% of clinically used drugs, was examined with multiple, clinically-relevant drug substrates, showing that A287P and R457H dramatically reduce drug metabolism, and that A503V variably impairs drug metabolism. The degree of activity can vary with the drug substrate assayed, as the drugs can influence the conformation of the P450. POR is probably an important contributor to genetic variation in both steroidogenesis and drug metabolism.

Objectives P450 oxidoreductase (POR) donates electrons to all microsomal cytochrome P450s, including drug-metabolizing and steroidogenic enzymes. Severe POR mutations cause skeletal malformations and disordered steroidogenesis. The POR polymorphism A503V is found on ~28% of human alleles and decreases activities of CYP3A4 and steroidogenic CYP17, but not the activities of steroidogenic CYP21 or drug-metabolizing CYP1A2 and CYP2C19. CYP2D6 metabolizes about 25% of clinically used drugs; we assessed the capacity of POR variants to support the activities of human CYP2D6. Methods N-27 forms of wild type (WT), Q153R, A287P, R457H and A503V POR, and WT CYP2D6 were expressed in E.coli. POR proteins in bacterial membranes were reconstituted with purified CYP2D6. Support of CYP2D6 was measured by metabolism of EOMCC (2H-1-benzopyran-3-carbonitrile,7-(ethoxy-methoxy)-2-oxo-(9Cl)), dextromethorphan and bufuralol. Km and Vmax were determined in three triplicate experiments for each reaction; catalytic efficiency is expressed as Vmax/Km. Results Compared to WT POR, disease-causing POR mutants A287P and R457H supported no detectable CYP2D6 activity with EOMCC, but A287P supported ~ 25% activity with dextromethorphan and bufuralol. Q153R had increased function with CYP2D6 (128% with EOMCC; 198% with dextromethorphan; 153% with bufuralol). A503V supported decreased CYP2D6 activity: 85% with EOMCC, 62% with dextromethorphan and 53% with bufuralol. Conclusions POR variants have different effects depending on the substrate metabolized. Disease-causing POR mutations R457H and A287P had poor activities, suggesting that diminished drug metabolism should be considered in affected patients. The common A503V polymorphism impaired CYP2D6 activities with two commonly-used drugs by 40–50%; potentially explaining some genetic variation in drug metabolism.

POR variants affect CYP2C9 activities. The impact of a POR variant on catalysis varies with the isoform of CYP2C9 and the assay substrate.

P450 oxidoreductase (POR) is a two-flavin protein that reduces microsomal P450 enzymes and some other proteins. Preparation of active bacterially expressed human POR for biochemical studies has been difficult because membrane-bound proteins tend to interact with column matrices. To reduce column-protein interactions and permit more vigorous washing, human POR lacking 27 N-terminal residues (N-27 POR) was modified to carry a C-terminal Gly3His6-tag (N-27 POR-G3H6). When expressed in Escherichia coli, N-27 POR-G3H6 could be purified to apparent homogeneity by a modified, single-step nickel-nitrilotriacetic acid affinity chromatography, yielding 31 mg POR per liter of culture, whereas standard purification of native N-27 POR required multiple steps, yielding 5 mg POR per liter. Both POR proteins had absorption maxima at 375 and 453 nm and both reduced cytochrome c with indistinguishable specific activities. Using progesterone as substrate for bacterially expressed purified human P450c17, the Michaelis constant for 17α-hydroxylase activity supported by N-27 POR or N-27 POR-G3H6 were 1.73 or 1.49 μm, and the maximal velocity was 0.029 or 0.026 pmol steroids per picomole P450 per minute, respectively. Using 17-hydroxypregnenolone as the P450c17 substrate, the Michaelis constant for 17,20 lyase activity using N-27 POR or N-27 POR-G3H6 was 1.92 or 1.89 μm and the maximal velocity was 0.041 or 0.042 pmol steroid per picomole P450 per minute, respectively. Thus, N-27 POR-G3H6 is equally active as native N-27 POR. This expression and purification system permits the rapid preparation of large amounts of highly pure, biologically active POR and may be generally applicable for the preparation of membrane-bound proteins.