Extraadrenal 21-hydroxylation by CYP2C19 and CYP3A4: effect on 21-hydroxylase deficiency.
Journal: 2009/February - Journal of Clinical Endocrinology and Metabolism
ISSN: 0021-972X
Abstract:
BACKGROUND
21-Hydroxylase deficiency (21OHD) is caused by CYP21A2 gene mutations disrupting the adrenal 21-hydroxylase, P450c21. CYP21A2 mutations generally correlate well with the 21OHD phenotype, but some children with severe CYP21A2 mutations have residual 21-hydroxylase activity. Some hepatic P450 enzymes can 21-hydroxylate progesterone, but their physiological relevance in modifying 21OHD is not known.
OBJECTIVE
We determined the ability of CYP2C19 and CYP3A4 to 21-hydroxylate progesterone and 17-hydroxyprogesterone (17OHP), determined the impact of the common P450 oxidoreductase (POR) variant A503V on these activities, and examined correlations between CYP2C19 variants and phenotype in patients with 21OHD.
METHODS
Bacterially expressed, N-terminally modified, C-His-tagged human P450c21, CYP2C19, and CYP3A4 were combined with bacterially expressed wild-type and A503V POR. The 21-hydroxylation of radiolabeled progesterone and 17OHP was assessed, and the Michaelis constant (Km) and maximum velocity (Vmax) of the reactions were measured. CYP2C19 was genotyped in 21OHD patients with genotypes predicting severe congenital adrenal hyperplasia.
RESULTS
Compared to P450c21, the Vmax/Km for 21-hydroxylation of progesterone by CYP2C19 and CYP3A4 were 17 and 10%, respectively. With both forms of POR, the Km for P450c21 was approximately 2.6 microm, the Km for CYP2C19 was approximately 11 microm, and the Km for CYP3A4 was approximately 110 microm. Neither CYP2C19 nor CYP3A4 could 21-hydroxylate 17OHP. The CYP2C19 ultrametabolizer allele CYP2C19 17 was homozygous in one of five patients with a 21OHD phenotype that was milder than predicted by the CYP21A2 genotype.
CONCLUSIONS
CYP2C19 and CYP3A4 can 21-hydroxylate progesterone but not 17OHP, possibly ameliorating mineralocorticoid deficiency, but not glucocorticoid deficiency. Multiple enzymes probably contribute to extraadrenal 21-hydroxylation.
Relations:
Content
Citations
(19)
References
(39)
Diseases
(1)
Drugs
(1)
Chemicals
(4)
Genes
(2)
Organisms
(1)
Processes
(1)
Affiliates
(1)
Similar articles
Articles by the same authors
Discussion board
J Clin Endocrinol Metab 94(1): 89-95

Extraadrenal 21-Hydroxylation by CYP2C19 and CYP3A4: Effect on 21-Hydroxylase Deficiency

Subjects and Methods

Cytochrome P450 enzymes

Purified, bacterially expressed human CYP2C19 and CYP3A4 were obtained from Invitrogen (Madison, WI). P450c21 was expressed in bacteria and purified as described (19). Human P450c21 cDNA, with the N-terminal region replaced by the sequence MALLLAVFL (20) and with a C-terminal 6-His-tag, was cloned in pCWori (construct built by Dr. Christa E. Flück). This construct was expressed in Escherichia coli C41(DE3)pLysS, and bacterial membranes were prepared as described (21). For protein purification, membranes were applied to an Ni-NTA agarose column (Sigma, St. Louis, MO), then to a DEAE-Sepharose column (Sigma), and finally to an Sp-Sepharose column (Sigma) (22). The purification was assessed by Coomasie Blue-staining of an SDS-PAGE gel and confirmed by Western blotting using our rabbit antiserum against bacterially expressed human P450c21.

Expression of POR and cytochrome b5

POR lacking 27 N-terminal residues was cloned in pET22b, and the A503V POR variant was generated by site-directed mutagenesis (23). Expression vectors for wild-type and A503V POR were expressed in E. coli C41(DE3)pLysS, and the bacterial membranes were prepared as described (23). Wild-type and A503V POR were quantified by Western blotting, with comparison to a standard curve of purified wild-type POR (24). The POR proteins were separated by SDS-PAGE, transferred to Immobilon-FL transfer membrane (Millipore, Bedford, MA), and incubated with rabbit polyclonal antibody against rat POR (Stressgene, Ann Arbor, MI). The blot was analyzed on an Odyssey Infrared Imaging System (LI-COR Bioscience, Lincoln, NE) (24). Human cytochrome b5 was expressed in E. coli and purified as described (23).

Enzyme assays

For CYP2C19 assays, 5 pmol of purified CYP2C19 was incubated with bacterial membranes containing 10 pmol of wild-type or A503V POR, 10 pmol of cytochrome b5, 10 μg of 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC), and 10 μg of 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) in a buffer mix containing 50 mm HEPES/KOH (pH 7.4), 3 mm reduced glutathione, and 30 mm MgCl2. The assays were performed with 3.0, 8.0, 15, and 30 μm [C]progesterone and [H]17OHP as substrates in a total volume of 200 μl.

For CYP3A4 assays, 5 pmol of purified CYP3A4 was incubated with bacterial membranes containing 10 pmol POR, 10 pmol cytochrome b5, 10 μg of a lipid mixture containing three synthetic phospholipids [DLPC, 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), and 1,2-dilauroyl-sn-glycero-3-phosphoserine (PS)] at a ratio of 1:1:1, 10 μg CHAPS, and the same HEPES buffer mix used for CYP2C19. The assays were performed with 30, 60, 90, and 150 μm [C]progesterone and [H]17OHP as substrates, in a total volume of 200 μl.

For P450c21 assays, 5 pmol of purified P450c21 was incubated with bacterial membranes containing 10 pmol POR, 10 μg DLPC, 10 μg CHAPS, 50 mm potassium phosphate (pH 7.4), 6 mm potassium acetate, 1 mm reduced glutathione, and 10 mm MgCl2. The assays were performed with 0.3, 1, 3, and 5 μm [C]progesterone and [H]17OHP as substrates, in a total volume of 200 μl.

The reactions were started by adding 2 mm nicotinamide adenine dinucleotide phosphate and were stopped after 1 h of incubation at 37 C, by adding 450 μl ethylacetate/isooctane 1:1. Steroids were analyzed by thin layer chromatography (TLC), and steroids were identified by cochromatography, with the labeled progesterone and 17OHP located by autoradiography, and with unlabeled DOC, 11-deoxycortisol, 6β-hydroxyprogesterone, and 16α-hydroxyprogesterone located by UV light. TLC was also run in two different solvent systems—chloroform/ethylacetate 3:1, and methylene chloride/methanol/H2O 200:20:1—to confirm the steroidal identities (25). The 21-hydroxylation of [C]progesterone to DOC and [H]17OHP to 11-deoxycortisol was quantified by phosphorimaging. All experiments were performed three times, each in duplicate. The Michaelis constant (Km), maximum velocity (Vmax), and the Vmax/Km (an estimate of enzymatic efficiency) were calculated by Lineweaver-Burk plots using GraphPad Prism 3 software (GraphPad Software, San Diego, CA).

CYP2C19 sequencing

The nine coding exons of the CYP2C19 gene, including at least 50 bp of the flaking introns, were sequenced in the five patients, and 3.8 kb of the 5′ flanking DNA (promoter region) was sequenced in the five patients (Table 11)) and 40 controls. Primers flaking 3.8 kb of the promoter region and exons 1 to 9, including the exon-intron junctions, were designed using the University of California, Santa Cruz genome browser (http://genome.ucsc.edu/index.html), and Primer 3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) (Table 22).). Regions containing known single nucleotide polymorphisms were not used for primers.

Table 1

Clinical, hormonal, and genetic data

Patient no.Sex (social/genetic)Age (chron/bone) at diagnosis (yr)Clinical findingsBasal 17OHP (nmol/liter)Basal Δ4 and T (nmol/liter)Basal PRA (ng/ml · h)Genotype21OHD phenotype (predicted/ observed)
1M/F4.5/11Phallic enlargement915.159/25a5.6R365W/R365WSW/SV
2M/M4.5/12.5Phallic enlargement, pubarche163.625.8/25.56.5Del 21A2/Q318XSW/SV
3M/M5.3/11Phallic enlargement, pubarche581.812.9/17.72.8Del 21A2/Del 21A2bSW/SV
4M/M1/−Phallic enlargement721.1/469c126/40.5c>25R365W/Ins T, V281LSW/SV
5M/M5/11.5Phallic enlargement, pubarche42434.9/5.918Q318X/R408CSW/SV

M, Male; F, female; chron, chronological; Δ4, androstenedione; T, testosterone.

Patient 1 hormonal data at 34 yr without treatment.
The breakpoint of the CYP21A2 hybrid gene was analyzed by allele-specific PCR. Alleles present the following point mutations: P30L, I2 splice, Del 8nt, I172N and Cluster. Parents of Patient 3 are first cousins.
Patient 4 hormonal data before re-starting treatment at 13 yr.

Table 2

Oligonucleotide primers for PCR and sequencing of CYP2C19

Sequence nameProduct size (bp)Location UCSC Genome BrowserForward primer (5′-3′)Reverse primer (5′-3′)
Promoter 1 (P1)86696508784–96509649GACCTTGATCTGGCAATGGTTGCAGTGTTGGCATAGTTTTG
Promoter 2 (P2)84796509374–96510220CACACAAAAATCTGCACATGGTGAGCAATTGTTGACTCAGTG
Promoter 3 (P3)72396510057–96510779TGTGGAGGGCTTAATGTTGATGCTGGTGCTAGAGCTGAGA
Promoter 4 (P4)75596510730–96511484CCTTCAAGATGCAGGGCTTAAGGAAAACAGCCCCAGAGAT
Promoter 5 (P5)80496511356–96512159TCAAGCCCTTAGCACCAAATCCAATGCACCGTCATAATTG
Promoter 6/exon 1 (P6-E1)103896512038–96513075TGGCCATTTCCGTTAAATCACTAAACCCACAGCTGCTTCC
Exon 2/exon 3 (E2-E3)83396524597–96525429TTGTCTGACCATTGCCTTGATCTCAGCTTCAAACCCTGCT
Exon 4 (E4)64196529991–96530631AAGACAAATAGGCCGGGAATTCAGGGAGCTAATGGGCTTA
Exon 5 (E5)66496531188–96531851TCAGGTTGTGCAAACTCTTTTCAAGCATTACTCCTTGACCTG
Exon 6 (E6)61696570009–96570624TGACAAACCCACAGCCAATATCCTAGCCTCAATGGTCCAC
Exon 7 (E7)60496592319–96592922TTCATTTCTTCCTGCCTTCCCCCAAACTGGAATCAACAGAA
Exon 8 (E8)60296599397–96599998GTCCCCGAAGTGTGATGTTCTCTCCAAAACCCACTAATCTGG
Exon 9 (E9)61596602271–96602885TTGCCTATCCATCCATTCATCTCAGCATTATGTGGCACTCA

The amplification of the 5′ flanking DNA segments P1, P2, and P4 (Table 22)) was performed under touchdown cycling conditions: 95 C for 4 min, then 15 touchdown cycles of 95 C for 30 sec, 62 C for 30 sec (decreasing by 0.5 C with each cycle), and 72 C for 45 sec, followed by 35 cycles of 95 C for 30 sec, 55 C for 30 sec, and 72 C for 45 sec. The final extension was held at 72 C for 7 min, and then the reaction was stopped at 4 C. The amplifications of the 5′ flanking DNA segments P3, P5 (Table 22)) and exons 2, 3, 4, 5, 6, 7, and 9 was performed as described above, except that the annealing temperature started at 60 C and finished at 53 C after the 15 touchdown cycles, and then continued at 53 C for the remaining 35 cycles. The amplifications of 5′ flanking DNA segment P6 plus exon 1 (Table 22)) were performed as described, except that the annealing started at 59 C and finished at 51 C, and for exon 8 the annealing temperature started at 64 C and finished at 57 C. All cycles for PCR were performed on Bio-Rad MyCycler and i-Cycler (Bio-Rad Laboratories, Hercules, CA). The sequencing was done with ABI BigDye terminator version 3.1 (Applied Biosystems, Foster City, CA) and displayed on ABI 3730 X 1 DNA Analyzer. Sequence variations were analyzed with DNA Sequencher 5.2 (Gene Codes, Ann Arbor, MI). All sequence variations were confirmed by repeating the PCR amplification and sequencing the opposite strand.

Subjects and genetic analysis

Under a study protocol approved by the Ethical Committee of São Paulo University, informed consent was obtained from the patients or caretakers. We selected 109 Brazilian patients with classical 21OHD bearing severe CYP21A2 gene mutations on both alleles that completely abolish enzyme activity (5,19). These patients were classified clinically as having simple virilizing (SV) CAH, characterized by ambiguous genitalia in girls and signs of early postnatal virilization in both sexes, or SW CAH, characterized by adrenal crisis or hyponatremia with high plasma renin activity (PRA) in the first month of life.

The CYP21A2 genes of all 109 patients were previously analyzed by Southern blotting to determine large rearrangements (26), and by allele-specific PCR to determine 15 common microconversions resembling point mutations (27,28). To rule out the presence of additional rare mutations, the entire CYP21A2 gene, including approximately 700 bp of 5′ flanking DNA, was sequenced in patients who had a phenotype that was discordant with that predicted by the genotype. After the complete CYP21A2 genetic analysis, five patients remained who had a discordance between phenotype and genotype (Table 11).). None of these patients experienced SW or needed mineralocorticoid replacement; all were treated with cortisone acetate ≤18 mg/m · d (see Case reports).

We sequenced the CYP2C19 gene in these five patients and in a control group of 40 21OHD patients having severe CYP21A2 mutations on both alleles, but who experienced the expected SW crisis. We excluded 64 21OHD patients bearing the common intronic mutation IVS2-13 (A/C to G) from the study because this mutation is associated with phenotypic variability, possibly due to variations in RNA splicing (29).

Case reports

Patient 1, a 46,XX individual, had Prader V external genitalia at birth and was raised as a male. At 4 equation M1yr the diagnosis of 21OHD was made based on phallic enlargement, sexual precocity (Tanner III), tall stature, and advanced bone age (11 yr). The patient has abandoned follow-up, takes no medication, is married, and lives as a male.

Patient 2 was diagnosed at 4 equation M2yr with penile enlargement (11 × 2.0 cm), sexual precocity (Tanner III), tall stature (+2.5 sd), and advanced bone age (12.5 yr old). At diagnosis, his sodium was 141 mEq/liter, potassium 5.2 mEq/liter, 17OHP 163.6 nmol/liter, androstenedione 25.8 nmol/liter, testosterone 25.5 nmol/liter, and PRA 6.5 ng/ml · h (elevated), with normal blood pressure. He abandoned follow-up at age 11 yr.

Patient 3 was diagnosed at 5 equation M3yr with penile enlargement (9.0 × 2.7 cm), sexual precocity (Tanner III), tall stature (+2.8 sd), advanced bone age (11 yr old), and normal blood pressure. At diagnosis, his sodium was 136 mEq/liter, potassium 4.3 mEq/liter, 17OHP 581.8 nmol/liter, androstenedione 12.9 nmol/liter, testosterone 17.7 nmol/liter, and PRA 2.8 ng/ml · h.

Patient 4 was diagnosed at 1 yr with penile enlargement, but his parents delayed treatment with cortisone acetate until he was 2 equation M4yr old. His bone age was advanced (5 yr), and his 17OHP was 721.1 nmol/ml. Treatment was started elsewhere without baseline steroid measurements. He came to our clinic at 4 equation M5yr with sexual precocity (Tanner II), penile enlargement (10.5 × 2.5 cm), normal height (+0.5 sd), and advanced bone age (11 yr). He stopped cortisone acetate intake from 11 to 13 yr of age, and did not have SW crisis. At age 13, he had 17OHP 469 nmol/liter, androstenedione 126 nmol/liter, testosterone 40.5 nmol/liter, and PRA above 25 ng/ml · h, but his blood pressure was normal, with sodium 136 mEq/liter and potassium 4.1 mEq/liter.

Patient 5 was diagnosed at 5 yr with penile enlargement (6.0 × 2.0 cm), sexual precocity (Tanner II), tall stature (+2 sd), and advanced bone age (11.5 yr). His 17OHP was 424 nmol/liter, androstenedione 34.9 nmol/liter, and testosterone 5.9 nmol/liter. He had normal blood pressure, normal sodium of 137 mEq/liter, and potassium 4.7mEq/liter, and high PRA level of 18 ng/ml · h.

Cytochrome P450 enzymes

Purified, bacterially expressed human CYP2C19 and CYP3A4 were obtained from Invitrogen (Madison, WI). P450c21 was expressed in bacteria and purified as described (19). Human P450c21 cDNA, with the N-terminal region replaced by the sequence MALLLAVFL (20) and with a C-terminal 6-His-tag, was cloned in pCWori (construct built by Dr. Christa E. Flück). This construct was expressed in Escherichia coli C41(DE3)pLysS, and bacterial membranes were prepared as described (21). For protein purification, membranes were applied to an Ni-NTA agarose column (Sigma, St. Louis, MO), then to a DEAE-Sepharose column (Sigma), and finally to an Sp-Sepharose column (Sigma) (22). The purification was assessed by Coomasie Blue-staining of an SDS-PAGE gel and confirmed by Western blotting using our rabbit antiserum against bacterially expressed human P450c21.

Expression of POR and cytochrome b5

POR lacking 27 N-terminal residues was cloned in pET22b, and the A503V POR variant was generated by site-directed mutagenesis (23). Expression vectors for wild-type and A503V POR were expressed in E. coli C41(DE3)pLysS, and the bacterial membranes were prepared as described (23). Wild-type and A503V POR were quantified by Western blotting, with comparison to a standard curve of purified wild-type POR (24). The POR proteins were separated by SDS-PAGE, transferred to Immobilon-FL transfer membrane (Millipore, Bedford, MA), and incubated with rabbit polyclonal antibody against rat POR (Stressgene, Ann Arbor, MI). The blot was analyzed on an Odyssey Infrared Imaging System (LI-COR Bioscience, Lincoln, NE) (24). Human cytochrome b5 was expressed in E. coli and purified as described (23).

Enzyme assays

For CYP2C19 assays, 5 pmol of purified CYP2C19 was incubated with bacterial membranes containing 10 pmol of wild-type or A503V POR, 10 pmol of cytochrome b5, 10 μg of 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC), and 10 μg of 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) in a buffer mix containing 50 mm HEPES/KOH (pH 7.4), 3 mm reduced glutathione, and 30 mm MgCl2. The assays were performed with 3.0, 8.0, 15, and 30 μm [C]progesterone and [H]17OHP as substrates in a total volume of 200 μl.

For CYP3A4 assays, 5 pmol of purified CYP3A4 was incubated with bacterial membranes containing 10 pmol POR, 10 pmol cytochrome b5, 10 μg of a lipid mixture containing three synthetic phospholipids [DLPC, 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), and 1,2-dilauroyl-sn-glycero-3-phosphoserine (PS)] at a ratio of 1:1:1, 10 μg CHAPS, and the same HEPES buffer mix used for CYP2C19. The assays were performed with 30, 60, 90, and 150 μm [C]progesterone and [H]17OHP as substrates, in a total volume of 200 μl.

For P450c21 assays, 5 pmol of purified P450c21 was incubated with bacterial membranes containing 10 pmol POR, 10 μg DLPC, 10 μg CHAPS, 50 mm potassium phosphate (pH 7.4), 6 mm potassium acetate, 1 mm reduced glutathione, and 10 mm MgCl2. The assays were performed with 0.3, 1, 3, and 5 μm [C]progesterone and [H]17OHP as substrates, in a total volume of 200 μl.

The reactions were started by adding 2 mm nicotinamide adenine dinucleotide phosphate and were stopped after 1 h of incubation at 37 C, by adding 450 μl ethylacetate/isooctane 1:1. Steroids were analyzed by thin layer chromatography (TLC), and steroids were identified by cochromatography, with the labeled progesterone and 17OHP located by autoradiography, and with unlabeled DOC, 11-deoxycortisol, 6β-hydroxyprogesterone, and 16α-hydroxyprogesterone located by UV light. TLC was also run in two different solvent systems—chloroform/ethylacetate 3:1, and methylene chloride/methanol/H2O 200:20:1—to confirm the steroidal identities (25). The 21-hydroxylation of [C]progesterone to DOC and [H]17OHP to 11-deoxycortisol was quantified by phosphorimaging. All experiments were performed three times, each in duplicate. The Michaelis constant (Km), maximum velocity (Vmax), and the Vmax/Km (an estimate of enzymatic efficiency) were calculated by Lineweaver-Burk plots using GraphPad Prism 3 software (GraphPad Software, San Diego, CA).

CYP2C19 sequencing

The nine coding exons of the CYP2C19 gene, including at least 50 bp of the flaking introns, were sequenced in the five patients, and 3.8 kb of the 5′ flanking DNA (promoter region) was sequenced in the five patients (Table 11)) and 40 controls. Primers flaking 3.8 kb of the promoter region and exons 1 to 9, including the exon-intron junctions, were designed using the University of California, Santa Cruz genome browser (http://genome.ucsc.edu/index.html), and Primer 3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) (Table 22).). Regions containing known single nucleotide polymorphisms were not used for primers.

Table 1

Clinical, hormonal, and genetic data

Patient no.Sex (social/genetic)Age (chron/bone) at diagnosis (yr)Clinical findingsBasal 17OHP (nmol/liter)Basal Δ4 and T (nmol/liter)Basal PRA (ng/ml · h)Genotype21OHD phenotype (predicted/ observed)
1M/F4.5/11Phallic enlargement915.159/25a5.6R365W/R365WSW/SV
2M/M4.5/12.5Phallic enlargement, pubarche163.625.8/25.56.5Del 21A2/Q318XSW/SV
3M/M5.3/11Phallic enlargement, pubarche581.812.9/17.72.8Del 21A2/Del 21A2bSW/SV
4M/M1/−Phallic enlargement721.1/469c126/40.5c>25R365W/Ins T, V281LSW/SV
5M/M5/11.5Phallic enlargement, pubarche42434.9/5.918Q318X/R408CSW/SV

M, Male; F, female; chron, chronological; Δ4, androstenedione; T, testosterone.

Patient 1 hormonal data at 34 yr without treatment.
The breakpoint of the CYP21A2 hybrid gene was analyzed by allele-specific PCR. Alleles present the following point mutations: P30L, I2 splice, Del 8nt, I172N and Cluster. Parents of Patient 3 are first cousins.
Patient 4 hormonal data before re-starting treatment at 13 yr.

Table 2

Oligonucleotide primers for PCR and sequencing of CYP2C19

Sequence nameProduct size (bp)Location UCSC Genome BrowserForward primer (5′-3′)Reverse primer (5′-3′)
Promoter 1 (P1)86696508784–96509649GACCTTGATCTGGCAATGGTTGCAGTGTTGGCATAGTTTTG
Promoter 2 (P2)84796509374–96510220CACACAAAAATCTGCACATGGTGAGCAATTGTTGACTCAGTG
Promoter 3 (P3)72396510057–96510779TGTGGAGGGCTTAATGTTGATGCTGGTGCTAGAGCTGAGA
Promoter 4 (P4)75596510730–96511484CCTTCAAGATGCAGGGCTTAAGGAAAACAGCCCCAGAGAT
Promoter 5 (P5)80496511356–96512159TCAAGCCCTTAGCACCAAATCCAATGCACCGTCATAATTG
Promoter 6/exon 1 (P6-E1)103896512038–96513075TGGCCATTTCCGTTAAATCACTAAACCCACAGCTGCTTCC
Exon 2/exon 3 (E2-E3)83396524597–96525429TTGTCTGACCATTGCCTTGATCTCAGCTTCAAACCCTGCT
Exon 4 (E4)64196529991–96530631AAGACAAATAGGCCGGGAATTCAGGGAGCTAATGGGCTTA
Exon 5 (E5)66496531188–96531851TCAGGTTGTGCAAACTCTTTTCAAGCATTACTCCTTGACCTG
Exon 6 (E6)61696570009–96570624TGACAAACCCACAGCCAATATCCTAGCCTCAATGGTCCAC
Exon 7 (E7)60496592319–96592922TTCATTTCTTCCTGCCTTCCCCCAAACTGGAATCAACAGAA
Exon 8 (E8)60296599397–96599998GTCCCCGAAGTGTGATGTTCTCTCCAAAACCCACTAATCTGG
Exon 9 (E9)61596602271–96602885TTGCCTATCCATCCATTCATCTCAGCATTATGTGGCACTCA

The amplification of the 5′ flanking DNA segments P1, P2, and P4 (Table 22)) was performed under touchdown cycling conditions: 95 C for 4 min, then 15 touchdown cycles of 95 C for 30 sec, 62 C for 30 sec (decreasing by 0.5 C with each cycle), and 72 C for 45 sec, followed by 35 cycles of 95 C for 30 sec, 55 C for 30 sec, and 72 C for 45 sec. The final extension was held at 72 C for 7 min, and then the reaction was stopped at 4 C. The amplifications of the 5′ flanking DNA segments P3, P5 (Table 22)) and exons 2, 3, 4, 5, 6, 7, and 9 was performed as described above, except that the annealing temperature started at 60 C and finished at 53 C after the 15 touchdown cycles, and then continued at 53 C for the remaining 35 cycles. The amplifications of 5′ flanking DNA segment P6 plus exon 1 (Table 22)) were performed as described, except that the annealing started at 59 C and finished at 51 C, and for exon 8 the annealing temperature started at 64 C and finished at 57 C. All cycles for PCR were performed on Bio-Rad MyCycler and i-Cycler (Bio-Rad Laboratories, Hercules, CA). The sequencing was done with ABI BigDye terminator version 3.1 (Applied Biosystems, Foster City, CA) and displayed on ABI 3730 X 1 DNA Analyzer. Sequence variations were analyzed with DNA Sequencher 5.2 (Gene Codes, Ann Arbor, MI). All sequence variations were confirmed by repeating the PCR amplification and sequencing the opposite strand.

Subjects and genetic analysis

Under a study protocol approved by the Ethical Committee of São Paulo University, informed consent was obtained from the patients or caretakers. We selected 109 Brazilian patients with classical 21OHD bearing severe CYP21A2 gene mutations on both alleles that completely abolish enzyme activity (5,19). These patients were classified clinically as having simple virilizing (SV) CAH, characterized by ambiguous genitalia in girls and signs of early postnatal virilization in both sexes, or SW CAH, characterized by adrenal crisis or hyponatremia with high plasma renin activity (PRA) in the first month of life.

The CYP21A2 genes of all 109 patients were previously analyzed by Southern blotting to determine large rearrangements (26), and by allele-specific PCR to determine 15 common microconversions resembling point mutations (27,28). To rule out the presence of additional rare mutations, the entire CYP21A2 gene, including approximately 700 bp of 5′ flanking DNA, was sequenced in patients who had a phenotype that was discordant with that predicted by the genotype. After the complete CYP21A2 genetic analysis, five patients remained who had a discordance between phenotype and genotype (Table 11).). None of these patients experienced SW or needed mineralocorticoid replacement; all were treated with cortisone acetate ≤18 mg/m · d (see Case reports).

We sequenced the CYP2C19 gene in these five patients and in a control group of 40 21OHD patients having severe CYP21A2 mutations on both alleles, but who experienced the expected SW crisis. We excluded 64 21OHD patients bearing the common intronic mutation IVS2-13 (A/C to G) from the study because this mutation is associated with phenotypic variability, possibly due to variations in RNA splicing (29).

Case reports

Patient 1, a 46,XX individual, had Prader V external genitalia at birth and was raised as a male. At 4 equation M1yr the diagnosis of 21OHD was made based on phallic enlargement, sexual precocity (Tanner III), tall stature, and advanced bone age (11 yr). The patient has abandoned follow-up, takes no medication, is married, and lives as a male.

Patient 2 was diagnosed at 4 equation M2yr with penile enlargement (11 × 2.0 cm), sexual precocity (Tanner III), tall stature (+2.5 sd), and advanced bone age (12.5 yr old). At diagnosis, his sodium was 141 mEq/liter, potassium 5.2 mEq/liter, 17OHP 163.6 nmol/liter, androstenedione 25.8 nmol/liter, testosterone 25.5 nmol/liter, and PRA 6.5 ng/ml · h (elevated), with normal blood pressure. He abandoned follow-up at age 11 yr.

Patient 3 was diagnosed at 5 equation M3yr with penile enlargement (9.0 × 2.7 cm), sexual precocity (Tanner III), tall stature (+2.8 sd), advanced bone age (11 yr old), and normal blood pressure. At diagnosis, his sodium was 136 mEq/liter, potassium 4.3 mEq/liter, 17OHP 581.8 nmol/liter, androstenedione 12.9 nmol/liter, testosterone 17.7 nmol/liter, and PRA 2.8 ng/ml · h.

Patient 4 was diagnosed at 1 yr with penile enlargement, but his parents delayed treatment with cortisone acetate until he was 2 equation M4yr old. His bone age was advanced (5 yr), and his 17OHP was 721.1 nmol/ml. Treatment was started elsewhere without baseline steroid measurements. He came to our clinic at 4 equation M5yr with sexual precocity (Tanner II), penile enlargement (10.5 × 2.5 cm), normal height (+0.5 sd), and advanced bone age (11 yr). He stopped cortisone acetate intake from 11 to 13 yr of age, and did not have SW crisis. At age 13, he had 17OHP 469 nmol/liter, androstenedione 126 nmol/liter, testosterone 40.5 nmol/liter, and PRA above 25 ng/ml · h, but his blood pressure was normal, with sodium 136 mEq/liter and potassium 4.1 mEq/liter.

Patient 5 was diagnosed at 5 yr with penile enlargement (6.0 × 2.0 cm), sexual precocity (Tanner II), tall stature (+2 sd), and advanced bone age (11.5 yr). His 17OHP was 424 nmol/liter, androstenedione 34.9 nmol/liter, and testosterone 5.9 nmol/liter. He had normal blood pressure, normal sodium of 137 mEq/liter, and potassium 4.7mEq/liter, and high PRA level of 18 ng/ml · h.

Results

21-Hydroxylation of progesterone

To determine whether extraadrenal 21-hydroxylation performed by hepatic P450 enzymes modulates mineralocorticoid deficiency in 21OHD, we evaluated the ability of CYP2C19 and CYP3A4 to 21-hydroxylate progesterone, compared with P450c21. We reconstituted bacterially expressed P450s and bacterially expressed wild-type and A503V POR in a synthetic lipid environment, and we measured the conversion of radiolabeled progesterone to DOC by TLC (Fig. 11).). We determined the Km and Vmax of CYP2C19, CYP3A4, and P450c21. The data are linear, showing noncooperative kinetics (Fig. 22).). The Km of human P450c21 with progesterone was 2.7 μm. This value is in very good agreement with the value of 2.8 μm obtained in microsomes of COS-1 cells expressing human P450c21 (30), validating the biochemistry of our in vitro system. The Km for CYP2C19 with progesterone was 10.8 μm, and DOC was the only product detected by TLC (Fig. 11).). By contrast, the Km for CYP3A4 with progesterone was 112 μm, and the products were predominantly 6β-hydroxyprogesterone and 16α-hydroxyprogesterone, rather than DOC (Fig. 11).). Compared with P450c21, the efficiency (Vmax/Km) of 21-hydroxylation of progesterone by CYP2C19 was 17%, and the efficiency of 21-hydroxylation by 3A4 was 10% (Table 33).

An external file that holds a picture, illustration, etc.
Object name is zeg0010962740001.jpg

TLC analysis of the steroid 21-hydroxylase activities of CYP2C19 and CYP3A4. The four indicated concentrations of [C]progesterone (Prog) were used as substrate for each enzyme supported by wild-type POR. The identities of the radioactive spots corresponding to DOC and other metabolites were determined by cochromatography with labeled progesterone and with unlabeled DOC, 6β-OHProg, and 16α-OHProg, located by UV light. DOC is the only product resulting from metabolism of Prog by CYP2C19. The major products resulting from metabolism of Prog by 3A4 are 6β-OHProg and 16α-OHProg; DOC is a minor product.

An external file that holds a picture, illustration, etc.
Object name is zeg0010962740002.jpg

Kinetics of progesterone 21-hydroxylation by P450c21 (A), CYP2C19 (B), and CYP3A4 (C), as supported by wild-type (WT) or A503V POR. In each panel, the conversion of progesterone to DOC was quantitated by phosphorimage analysis of thin layer chromatograms, such as that in Fig 11,, and displayed as Lineweaver-Burk plots. Data are mean ± sem of three experiments, each performed in duplicate.

Table 3

21-Hydroxylation of progesterone by CYP2C19 and CYP3A4 supported by wild-type or A503 POR

Wild type
A503V
Km (μm)VmaxVmax/Km (% P450c21)Km (μm)VmaxVmax/Km (% PORwt)
P450c212.65 ± 0.370.26 ± 0.060.1 ± 0.03 (100%)2.47 ± 0.490.21 ± 0.070.08 ± 0.02 (80%)
CYP2C1910.4 ± 1.10.18 ± 0.020.017 ± 0.002 (17%)15.4 ± 3.50.21 ± 0.060.013 ± 0.003 (76%)
CYP3A4111.8 ± 8.31.16 ± 0.10.01 ± 0.001 (10%)107.6 ± 5.51.25 ± 0.40.012 ± 0.004 (120%)

Vmax is measured in picomoles per minute per picomoles P450.

21-Hydroxylation of 17OHP

To determine whether extraadrenal 21-hydroxylation of 17OHP performed by hepatic P450 enzymes modulates glucocorticoid deficiency in 21OHD, we evaluate the ability of CYP2C19 and CYP3A4 to catalyze the 21-hydroxylation of 17OHP. We reconstituted bacterially expressed P450s and POR as described for progesterone. The Km for the 21-hydroxylation of 17OHP by P450c21 was 1.4 μm, and the Vmax was 0.28 ± 0.08 pmol/min · pmol P450, consistent with previous data (30). Neither CYP2C19 nor CYP3A4 was able to 21-hydroxylate 17OHP.

Influence of the A503V POR variant

All catalysis by P450c21, CYP2C19, and CYP3A4 requires electron donation from POR. Because 28% of normal human POR alleles express the variant A503V (18), and because A503V reduces the 17α-hydroxylase and 17,20 lyase activities (Vmax/Km) of P450c17 to approximately 60% compared with wild-type POR (23), we assessed the potential impact of A503V POR on the 21-hydroxylation of progesterone by P450c21, CYP2C19, and CYP3A4. Compared with wild-type POR, there were no differences in the enzymatic parameters of any of these enzymes using A503V POR (Table 33).

Patients studied

To determine the effect of extraadrenal 21-hydroxylation in vivo, we selected five patients who had a phenotype discordant for their genotype. All patients first showed clinical signs of hyperandrogenism at 2–5 yr of age (Table 11),), and none had salt loss; thus, they presented clinically as SV 21OHD. However, their genotypes predicted SW 21OHD. Of the 10 CYP21A2 alleles in these five patients, three are deleted, one carries a frameshift (F306 + 1nt), and two carry a nonsense mutation (Q318X). None of these mutations encodes a full-length protein, and all are associated with SW 21OHD (31). Three of the remaining four alleles carry the missense mutation R356W and one carries R408C. Both mutations are associated with SW 21OHD (31), and, when assayed in vitro, both lack 21-hydroxylase activity using either progesterone or 17OHP as substrates (29,32). However, some cases of SV 21OHD have been associated with the R356W mutation (5).

CYP2C19 Genetic analysis

Relatively little 21-hydroxylase activity is needed to produce adequate amounts of aldosterone, as evidenced by patients with SV CAH. Hence, 21-hydroxylation of progesterone by CYP2C19 may be sufficient to modulate mineralocorticoid production in vivo, especially in individuals who overexpress this enzyme (33). Therefore, we sequenced the CYP2C19 gene in five 21OHD patients with discordant genotype/phenotype (Table 11)) and found polymorphisms only in the 5′ flanking DNA. This region was compared with 40 control CAH patients who had a good correlation between genotype and phenotype. Among these five individuals, patient 3 was heterozygous and patient 4 was homozygous for the 2C19 polymorphisms −807C>T and −3402C>T, which characterize the CYP2C19*17 ultrametabolizer allele (34). Among the 40 SW CAH controls, six were heterozygous for CYP2C19*17. As expected from their severe CYP21A2 genotypes, these six individuals required mineralocorticoid replacement beginning in the neonatal period, and all are now adolescents or adults.

21-Hydroxylation of progesterone

To determine whether extraadrenal 21-hydroxylation performed by hepatic P450 enzymes modulates mineralocorticoid deficiency in 21OHD, we evaluated the ability of CYP2C19 and CYP3A4 to 21-hydroxylate progesterone, compared with P450c21. We reconstituted bacterially expressed P450s and bacterially expressed wild-type and A503V POR in a synthetic lipid environment, and we measured the conversion of radiolabeled progesterone to DOC by TLC (Fig. 11).). We determined the Km and Vmax of CYP2C19, CYP3A4, and P450c21. The data are linear, showing noncooperative kinetics (Fig. 22).). The Km of human P450c21 with progesterone was 2.7 μm. This value is in very good agreement with the value of 2.8 μm obtained in microsomes of COS-1 cells expressing human P450c21 (30), validating the biochemistry of our in vitro system. The Km for CYP2C19 with progesterone was 10.8 μm, and DOC was the only product detected by TLC (Fig. 11).). By contrast, the Km for CYP3A4 with progesterone was 112 μm, and the products were predominantly 6β-hydroxyprogesterone and 16α-hydroxyprogesterone, rather than DOC (Fig. 11).). Compared with P450c21, the efficiency (Vmax/Km) of 21-hydroxylation of progesterone by CYP2C19 was 17%, and the efficiency of 21-hydroxylation by 3A4 was 10% (Table 33).

An external file that holds a picture, illustration, etc.
Object name is zeg0010962740001.jpg

TLC analysis of the steroid 21-hydroxylase activities of CYP2C19 and CYP3A4. The four indicated concentrations of [C]progesterone (Prog) were used as substrate for each enzyme supported by wild-type POR. The identities of the radioactive spots corresponding to DOC and other metabolites were determined by cochromatography with labeled progesterone and with unlabeled DOC, 6β-OHProg, and 16α-OHProg, located by UV light. DOC is the only product resulting from metabolism of Prog by CYP2C19. The major products resulting from metabolism of Prog by 3A4 are 6β-OHProg and 16α-OHProg; DOC is a minor product.

An external file that holds a picture, illustration, etc.
Object name is zeg0010962740002.jpg

Kinetics of progesterone 21-hydroxylation by P450c21 (A), CYP2C19 (B), and CYP3A4 (C), as supported by wild-type (WT) or A503V POR. In each panel, the conversion of progesterone to DOC was quantitated by phosphorimage analysis of thin layer chromatograms, such as that in Fig 11,, and displayed as Lineweaver-Burk plots. Data are mean ± sem of three experiments, each performed in duplicate.

Table 3

21-Hydroxylation of progesterone by CYP2C19 and CYP3A4 supported by wild-type or A503 POR

Wild type
A503V
Km (μm)VmaxVmax/Km (% P450c21)Km (μm)VmaxVmax/Km (% PORwt)
P450c212.65 ± 0.370.26 ± 0.060.1 ± 0.03 (100%)2.47 ± 0.490.21 ± 0.070.08 ± 0.02 (80%)
CYP2C1910.4 ± 1.10.18 ± 0.020.017 ± 0.002 (17%)15.4 ± 3.50.21 ± 0.060.013 ± 0.003 (76%)
CYP3A4111.8 ± 8.31.16 ± 0.10.01 ± 0.001 (10%)107.6 ± 5.51.25 ± 0.40.012 ± 0.004 (120%)

Vmax is measured in picomoles per minute per picomoles P450.

21-Hydroxylation of 17OHP

To determine whether extraadrenal 21-hydroxylation of 17OHP performed by hepatic P450 enzymes modulates glucocorticoid deficiency in 21OHD, we evaluate the ability of CYP2C19 and CYP3A4 to catalyze the 21-hydroxylation of 17OHP. We reconstituted bacterially expressed P450s and POR as described for progesterone. The Km for the 21-hydroxylation of 17OHP by P450c21 was 1.4 μm, and the Vmax was 0.28 ± 0.08 pmol/min · pmol P450, consistent with previous data (30). Neither CYP2C19 nor CYP3A4 was able to 21-hydroxylate 17OHP.

Influence of the A503V POR variant

All catalysis by P450c21, CYP2C19, and CYP3A4 requires electron donation from POR. Because 28% of normal human POR alleles express the variant A503V (18), and because A503V reduces the 17α-hydroxylase and 17,20 lyase activities (Vmax/Km) of P450c17 to approximately 60% compared with wild-type POR (23), we assessed the potential impact of A503V POR on the 21-hydroxylation of progesterone by P450c21, CYP2C19, and CYP3A4. Compared with wild-type POR, there were no differences in the enzymatic parameters of any of these enzymes using A503V POR (Table 33).

Patients studied

To determine the effect of extraadrenal 21-hydroxylation in vivo, we selected five patients who had a phenotype discordant for their genotype. All patients first showed clinical signs of hyperandrogenism at 2–5 yr of age (Table 11),), and none had salt loss; thus, they presented clinically as SV 21OHD. However, their genotypes predicted SW 21OHD. Of the 10 CYP21A2 alleles in these five patients, three are deleted, one carries a frameshift (F306 + 1nt), and two carry a nonsense mutation (Q318X). None of these mutations encodes a full-length protein, and all are associated with SW 21OHD (31). Three of the remaining four alleles carry the missense mutation R356W and one carries R408C. Both mutations are associated with SW 21OHD (31), and, when assayed in vitro, both lack 21-hydroxylase activity using either progesterone or 17OHP as substrates (29,32). However, some cases of SV 21OHD have been associated with the R356W mutation (5).

CYP2C19 Genetic analysis

Relatively little 21-hydroxylase activity is needed to produce adequate amounts of aldosterone, as evidenced by patients with SV CAH. Hence, 21-hydroxylation of progesterone by CYP2C19 may be sufficient to modulate mineralocorticoid production in vivo, especially in individuals who overexpress this enzyme (33). Therefore, we sequenced the CYP2C19 gene in five 21OHD patients with discordant genotype/phenotype (Table 11)) and found polymorphisms only in the 5′ flanking DNA. This region was compared with 40 control CAH patients who had a good correlation between genotype and phenotype. Among these five individuals, patient 3 was heterozygous and patient 4 was homozygous for the 2C19 polymorphisms −807C>T and −3402C>T, which characterize the CYP2C19*17 ultrametabolizer allele (34). Among the 40 SW CAH controls, six were heterozygous for CYP2C19*17. As expected from their severe CYP21A2 genotypes, these six individuals required mineralocorticoid replacement beginning in the neonatal period, and all are now adolescents or adults.

Discussion

Extraadrenal 21-hydroxylation is well described (8) and may modify the clinical picture of CAH (7), but is not catalyzed by the adrenal 21-hydroxylase, P450c21 (9). If extraadrenal 21-hydroxylation is catalyzed by one or more of the hepatic, drug-metabolizing enzymes, it might be possible to induce such an enzyme pharmacologically, thus ameliorating the clinical severity and possibly reducing the amount of hormonal replacement therapy needed. Thus we have focused on hepatic enzymes. Previous studies suggested that hepatic P450 enzymes of the 2C and 3A families may be involved (10,11,12,13). Using a baculovirus system coexpressing human P450 and rabbit POR, Yamazaki and Shimada (13) showed that CYP2C19 and CYP3A4 can 21-hydroxylate progesterone in vitro, but they did not compare these data with 21-hydroxylation by P450c21 or assess 17OHP as a substrate. Our data show that CYP2C19 and CYP3A4 can, respectively, 21-hydroxylase progesterone to DOC with about 17% and 10% of the efficiency of P450c21, but that neither CYP2C19 nor CYP3A4 could 21-hydroxylate 17OHP. Because the adrenal normally produces much less aldosterone than cortisol, low rates of progesterone 21-hydroxylation can suffice for salt balance. For example, the CYP21A2 mutation I172N is a common cause of SV 21OHD that retains only 2–5% of wild-type activity; thus very little progesterone 21-hydroxylase activity is needed to prevent salt loss (29,32). Therefore, extraadrenal 21-hydroxylation of progesterone by these two hepatic enzymes may be able to ameliorate the mineralocorticoid deficiency, but not the glucocorticoid deficiency of CAH. However, there is little hepatic expression of CYP2C19 before 5 months of age (35), and CYP3A4 expression is low throughout childhood (36). Hence, extraadrenal 21-hydroxylation by these enzymes does not normally prevent the typical SW syndrome seen in newborns with SW 21OHD.

The CYP2C19 gene is highly polymorphic, and some promoter polymorphisms can increase enzyme expression. Among five CAH patients whose phenotypes were less severe than predicted by their CYP21A2 genotypes, one was homozygous for the CYP2C19*17 ultrametabolizer allele. This allele carries two changes (−806C>T and −3402C>T) in the 5′-flanking region that create binding sites for hepatic transcription factors and increase gene expression; the net result is an increased amount of enzyme with the same Km and Vmax, so that this variant increases metabolism by CYP2C19 about 2- to 4-fold (34). Thus the CYP2C19*17 allele might increase the 21-hydroxylation of progesterone and synthesis of aldosterone. Because CYP2C19*17 represents a gain-of-function, one would expect it to exert a dominant effect. However, we found heterozygosity for this allele in six of 80 CYP2C19 alleles among 40 patients with salt-losing CAH whose phenotypes and genotypes were concordant, indicating that heterozygosity for CYP2C19*17 is insufficient to modulate the salt loss of CAH. Thus, if CYP2C19*17 is clinically important in CAH, it probably must be homozygous.

CYP3A4 is the most abundant P450 in the liver and metabolizes 40–45% of currently used drugs (15). In assaying the activity of CYP3A4 as a 21-hydroxylase, we found that CYP3A4 has a low affinity for progesterone (Km ∼110 μm), but a higher Vmax than P450c21, so that the efficiency of its 21-hydroxylation of progesterone (Vmax/Km) was approximately 10% of that for P450c21. Thus, this most abundant of all hepatic P450 enzymes may also contribute to extraadrenal 21-hydroxylation. Although CYP3A4 does not have common polymorphisms (37), there is substantial genetic variation in the metabolic clearance of its substrates (38). Therefore, we considered whether variations in POR might affect its 21-hydroxylase activity, but the common A503V variant of POR did not affect the 21-hydroxylase activity of CYP3A4.

The POR gene is very polymorphic, with the A503V variant polymorphism found in 28% of normal alleles (18). Although A503V decreases the activity of 17α-hydroxylase and 17,20-lyase activities of P450c17 (23), it does not affect 21-hydroxylation by P450c21 (19), and we now show that POR A503V does not affect 21-hydroxylation by CYP2C19 and CYP3A4. This variability is consistent with other studies showing that the activity of a POR variant with one P450 enzyme will not predict its activity with another P450 (24,39).

Thus, CYP2C19 and CYP3A4 can 21-hydroxylate progesterone in vitro and thus may modulate mineralocorticoid deficiency in vivo. However, we found a homozygous CYP2C19 ultrametabolizer allele in only one of five CAH patients whose phenotype was inappropriately mild for the CYP21 genotype. We propose that multiple enzymes, not necessarily confined to the liver, may exert clinically significant extraadrenal 21-hydroxylation in different individuals, depending on that individual’s array of polymorphic variants of several genes. Thus, extraadrenal 21-hydroxylation should not be viewed as the activity of only one or two enzymes, but instead as a genetic quantitative trait influenced by multiple genetic loci. The identification of these additional loci encoding factors that modify the 21OHD phenotype will contribute to a better understanding of this disease and its treatment.

Department of Pediatrics (L.G.G., N.H., V.A., W.L.M.), University of California San Francisco, San Francisco, California 94143-0978; and Department of Endocrinology-LIM42 (L.G.G., B.B.M., T.A.S.S.B.), Hospital das Clinicas da Universidade de São Paulo, 05403-900 São Paulo, Brazil
Address all correspondence and requests for reprints to: Prof. Walter L. Miller, Department of Pediatrics, HSE-1401, 513 Parnassus Avenue, University of California, San Francisco, San Francisco, California 94143-0978. E-mail: ude.fscu@balmlw.
Address all correspondence and requests for reprints to: Prof. Walter L. Miller, Department of Pediatrics, HSE-1401, 513 Parnassus Avenue, University of California, San Francisco, San Francisco, California 94143-0978. E-mail: ude.fscu@balmlw.
Received 2008 May 30; Accepted 2008 Oct 22.

Abstract

Context: 21-Hydroxylase deficiency (21OHD) is caused by CYP21A2 gene mutations disrupting the adrenal 21-hydroxylase, P450c21. CYP21A2 mutations generally correlate well with the 21OHD phenotype, but some children with severe CYP21A2 mutations have residual 21-hydroxylase activity. Some hepatic P450 enzymes can 21-hydroxylate progesterone, but their physiological relevance in modifying 21OHD is not known.

Objective: We determined the ability of CYP2C19 and CYP3A4 to 21-hydroxylate progesterone and 17-hydroxyprogesterone (17OHP), determined the impact of the common P450 oxidoreductase (POR) variant A503V on these activities, and examined correlations between CYP2C19 variants and phenotype in patients with 21OHD.

Methods: Bacterially expressed, N-terminally modified, C-His-tagged human P450c21, CYP2C19, and CYP3A4 were combined with bacterially expressed wild-type and A503V POR. The 21-hydroxylation of radiolabeled progesterone and 17OHP was assessed, and the Michaelis constant (Km) and maximum velocity (Vmax) of the reactions were measured. CYP2C19 was genotyped in 21OHD patients with genotypes predicting severe congenital adrenal hyperplasia.

Results: Compared to P450c21, the Vmax/Km for 21-hydroxylation of progesterone by CYP2C19 and CYP3A4 were 17 and 10%, respectively. With both forms of POR, the Km for P450c21 was approximately 2.6 μm, the Km for CYP2C19 was approximately 11 μm, and the Km for CYP3A4 was approximately 110 μm. Neither CYP2C19 nor CYP3A4 could 21-hydroxylate 17OHP. The CYP2C19 ultrametabolizer allele CYP2C19*17 was homozygous in one of five patients with a 21OHD phenotype that was milder than predicted by the CYP21A2 genotype.

Conclusions: CYP2C19 and CYP3A4 can 21-hydroxylate progesterone but not 17OHP, possibly ameliorating mineralocorticoid deficiency, but not glucocorticoid deficiency. Multiple enzymes probably contribute to extraadrenal 21-hydroxylation.

Abstract

Steroid 21-hydroxylase deficiency (21OHD) has an incidence of approximately 1 in 15,000 newborns (1) and accounts for 90–95% of cases of congenital adrenal hyperplasia (CAH). 21OHD is caused by mutations in the CYP21A2 gene, which encodes the adrenal 21-hydroxylase, P450c21 (2,3). P450c21 converts progesterone to deoxycorticosterone (DOC) and converts 17-hydroxyprogesterone (17OHP) to 11-deoxycortisol in the biosynthesis of aldosterone and cortisol. Patients with 21OHD have impaired synthesis of aldosterone and cortisol; the compensatory overproduction of ACTH leads to accumulation of androgen precursors.

In severe salt-wasting (SW) 21OHD, both sexes can experience a SW crisis in the first month of life, and girls are born with virilized external genitalia (2,3). Patients with SW-21OHD typically have severe CYP21A2 mutations that abolish P450c21 activity, so that they cannot synthesize aldosterone. However, some patients with severe mutations do not have clinically significant salt loss (4,5,6), and other patients appear to regain their ability to retain salt over time (7). Such recovery from salt loss may reflect increased dietary sodium, increased mineralocorticoid sensitivity, and increased activity of other enzymes that can 21-hydroxylate progesterone.

Many human extraadrenal tissues can 21-hydroxylate progesterone (8), but this activity is not due to P450c21 because its mRNA is not detected in these tissues (9). Hepatic P450 enzymes of the 2C subfamily can catalyze 21-hydroxylation of progesterone in rats, rabbits, and sheep (10,11,12). Recombinant human CYP3A4 and CYP2C19, which are hepatic, drug-metabolizing P450s enzymes, could 21-hydroxylate progesterone in vitro (13), but that study did not determine whether CYP3A4 and CYP2C19 could 21-hydroxylate 17OHP (13).

The enzymatic activities of CYP3A4 and CYP2C19 can vary enormously between individuals, depending on polymorphic variants that affect gene expression, classified as poor, intermediate, extensive, and ultra-metabolizers (14,15,16). We hypothesized that genetic variations in CYP3A4 and CYP2C19 might account for some of the differences in the 21-hydroxylation of progesterone and 17OHP, accounting for some of the phenotypic variation in 21OHD.

All microsomal P450 enzymes, including P450c21 and hepatic CYP3A4 and CYP2C19, receive electrons from nicotinamide adenine dinucleotide phosphate via the electron-transport flavoprotein, P450 oxidoreductase (POR) (17). The gene for human POR is also very polymorphic (18), and POR variants might also cause interindividual variability in steroid metabolism.

To determine whether hepatic CYP2C19 and CYP3A4 influence the 21OHD phenotype, we characterized their capacity to 21-hydroxylate progesterone and 17OHP using either wild-type POR or its common variant A503V as an electron donor; and we determined whether CYP2C19 polymorphisms modulate salt balance in patients with genotypes predicting severe 21OHD.

M, Male; F, female; chron, chronological; Δ4, androstenedione; T, testosterone.

Vmax is measured in picomoles per minute per picomoles P450.

Acknowledgments

We thank Ms. Izabella Damm for excellent technical assistance.

Acknowledgments

Footnotes

This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) Grant BEX1516/060 (to L.G.G.), by Fundacao de Amparo a Pesquisa do Estado de Sao Paulo Grants 05/55364-0 (to L.G.G.) and 05/04726-0 (to B.B.M.), and by National Institutes of Health Grant R01 GM073020 (to W.L.M.).

Disclosure Statement: The authors have nothing to disclose.

First Published Online October 28, 2008

Abbreviations: CAH, Congenital adrenal hyperplasia; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate; DLPC, 1,2-dilauroyl-sn-glycero-3-phosphocholine; DOC, deoxycorticosterone; Km, Michaelis constant; 17OHP, 17-hydroxyprogesterone; 21OHD, 21-hydroxylase deficiency; POR, P450 oxidoreductase; PRA, plasma renin activity; SV, simple virilizing; SW, salt-wasting; TLC, thin layer chromatography; Vmax, maximum velocity.

Footnotes
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.