Diversity and function of mutations in p450 oxidoreductase in patients with Antley-Bixler syndrome and disordered steroidogenesis

Abstract

P450 oxidoreductase (POR) is the obligatory flavoprotein intermediate that transfers electrons from reduced nicotinamide adenine dinucleotide phosphate (NADPH) to all microsomal cytochrome P450 enzymes. Although mouse Por gene ablation causes embryonic lethality, POR missense mutations cause disordered steroidogenesis, ambiguous genitalia, and Antley-Bixler syndrome (ABS), which has also been attributed to fibroblast growth factor receptor 2 (FGFR2) mutations. We sequenced the POR gene and FGFR2 exons 8 and 10 in 32 individuals with ABS and/or hormonal findings that suggested POR deficiency. POR and FGFR2 mutations segregated completely. Fifteen patients carried POR mutations on both alleles, 4 carried mutations on only one allele, 10 carried FGFR2 or FGFR3 mutations, and 3 patients carried no mutations. The 34 affected POR alleles included 10 with A287P (all from whites) and 7 with R457H (four Japanese, one African, two whites); 17 of the 34 alleles carried 16 "private" mutations, including 9 missense and 7 frameshift mutations. These 11 missense mutations, plus 10 others found in databases or reported elsewhere, were recreated by site-directed mutagenesis and were assessed by four assays: reduction of cytochrome c, oxidation of NADPH, support of 17alpha-hydroxylase activity, and support of 17,20 lyase using human P450c17. Assays that were based on cytochrome c, which is not a physiologic substrate for POR, correlated poorly with clinical phenotype, but assays that were based on POR's support of catalysis by P450c17--the enzyme most closely associated with the hormonal phenotype--provided an excellent genotype/phenotype correlation. Our large survey of patients with ABS shows that individuals with an ABS-like phenotype and normal steroidogenesis have FGFR mutations, whereas those with ambiguous genitalia and disordered steroidogenesis should be recognized as having a distinct new disease: POR deficiency.

Figure 1

Purification of POR and cytochrome b₅. Left panel, 10% SDS-polyacrylamide gel showing purified preparations of wild-type POR used for quantitation and normalization of POR mutants; right panel, 12% SDS-polyacrylamide gel showing purified cytochrome b₅ used in assays of the 17,20 lyase reaction. Both gels were purposely overloaded to permit identification of trace contamination. The gels were stained with Coomassie Brilliant Blue.

Figure 2

Sequences of POR mutations. The mutants shown are from the patients in the present study who had disordered steroidogenesis. Each mutant is shown as a heterozygote in association with a wild-type allele, except for Q153R, which was only seen in a homozygously affected patient. The mutated nucleotides, indicated by arrows, are 424A→G, causing T142A; 458A→G, causing Q153R; 787A→G, causing M263V; 859G→C, causing A287P; 1370G→A, causing R457H; 1375T→C, causing Y459H; 1615G→A, causing G539R; 1694T→C, causing L565P; 1845C→T, causing R616X; and 1937-1939delTCT, causing F646del. The location of each mutation is shown on a linear diagram of the POR protein, as described by Wang et al. (1997), which illustrates the crystallographically inferred boundaries of the functional domains of POR.

Figure 3

Model of human POR, showing the location of the mutants. The model is based on the X-ray crystal structure of rat POR and was created using the programs DeepView and POVRAY (see the Persistence of Vision Raytracer Web site). The α-carbon backbone is shown as a narrow ribbon, permitting visualization of buried residues. The missense residues identified by sequencing are depicted by charged packed sphere images. Mutations retaining >50% of activity are shown in green, those retaining 25%–50% are shown in blue, and those retaining <25% are shown in red. Ball-and-stick models are used to represent the FMN (yellow), FAD (yellow) and NADPH (cyan) moieties.

Figure 4

Amino acid sequence alignment of POR from five species. The protein sequences used are those of human (NCBI accession number NP_000932.1), rat (NCBI accession number NP_113764.1), Xenopus (NCBI accession number AAH59318.1), yeast (NCBI accession number NP_596046.1) and Drosophila (NCBI accession number NP_477158.1). Alignments were made with ClustalW and are displayed with GeneDoc. Residues that are identical in all species (when the residue exists) are highlighted in black, those identical in four of five species are shown in dark gray, and those identical in three of five species are shown in light gray. Amino acid numbers are shown at the right. Note that the human POR sequence has a three-residue amino-terminal extension. The functional domains and individual residues discussed in the text are shown above the human sequence. The location of the truncation used to create the N-27 POR used in the functional assays is shown.