| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Endocrinology (R.R.S., L.G.G., N.H., W.L.M.), Department of Pediatrics, University of California, San Francisco, San Francisco, California 94143; and Endocrinology Service and Research Center (R.R.S., G.V.V.), Sainte-Justine Hospital, and Department of Pediatrics, University of Montreal, Montreal, Canada H3T 1C5
Address all correspondence and requests for reprints to: Prof. Walter L. Miller, Pediatric Endocrinology, 672-S, University of California, San Francisco, San Francisco, California 94143-0434. E-mail: wlmlab{at}ucsf.edu.
| Abstract |
|---|
|
|
|---|
Objective: The aim was to explore the manifestations of combined deficiencies of 21-hydroxylase and POR and to search for lesions in apparent manifesting POR heterozygotes.
Patients and Methods: A newborn female had craniosynostosis, severe salt wasting, minimal virilization, grossly elevated 17OH-progesterone, and minimally elevated androgens. DNA encoding 21-hydroxylase, POR, and fibroblast growth factor receptor 2 was sequenced. For POR, the first untranslated exon (exon 1U), 5' flanking DNA, and most introns were sequenced in five apparent manifesting POR heterozygotes.
Results: CYP21B mutations were found on both alleles, proving classical 21-hydroxylase deficiency. Fibroblast growth factor receptor 2 exons 8 and 10 were normal. A POR mutation, A287P, was found only on the maternal allele. Five previously reported patients had POR mutations found on only one allele, but their clinical characteristics were indistinguishable from patients with mutations on both alleles. Sequencing of exon 1U, 274 bp of POR 5' flanking DNA, and 12 of the 15 POR introns did not identify additional mutations affecting gene expression or splicing.
Conclusion: Manifesting heterozygosity is a possible feature of POR deficiency and may ameliorate the findings in coexisting 21-hydroxylase deficiency.
| Introduction |
|---|
|
|
|---|
-hydroxylase, 17,20 lyase, and 21-hydroxylase was first reported in 1985 (5), and similar patients were reported subsequently (reviewed in Refs. 3 and 6). Despite the early suggestion that this disorder might be caused by mutations in POR (7), the lethality of POR gene ablation in mice (8, 9) implied that human POR mutations would be disastrous. Nevertheless, Flück et al. (1) reported four patients with POR mutations in early 2004, and subsequent reports have documented at least 41 such patients (1, 2, 6, 10, 11, 12). The rapid discovery of so many patients and the observation that POR mutations may cause mild disease without skeletal anomalies (1, 6) suggest that POR deficiency may not be rare, but its diagnosis remains difficult because of the complex pattern of disordered plasma steroids and their urinary metabolites resulting from incomplete interruption of several steps in steroidogenesis. POR deficiency manifests with and without skeletal anomalies. Most patients described to date have had a congenital malformation disorder termed Antley-Bixler syndrome (ABS), which is characterized by craniosynostosis, radio-ulnar or radio-humeral synostosis, bowed femora, and other skeletal anomalies (13, 14). The craniosynostosis of ABS resembles that of Apert, Crouzon, Pfeiffer, and Jackson-Weiss syndromes. Work in the mid-1990s showed that these craniosynostosis syndromes were caused by dominant, gain-of-function mutations in exons 8 and 10 of the gene for fibroblast growth factor receptor 2 (FGFR2) (reviewed in Refs. 3 and 6). Initial work identified similar (or identical) FGFR2 mutations in some patients with ABS (15), but other patients with ABS had no mutations in FGFR2 (15). ABS patients harboring FGFR2 mutations have normal genitalia, whereas those who also have genital anomalies and/or an abnormal pattern of urinary steroids lack FGFR2 mutations (15). Analysis of the FGFR2 and POR genes in a large group of ABS patients established that the same skeletal malformation syndrome may result from either dominant FGFR2 mutations or recessive POR mutations, and that all ABS patients with genital anomalies or disordered steroidogenesis had POR mutations (6).
POR is required for the activity of three steroidogenic enzymes: P450c17, which catalyzes 17
-hydroxylase and 17,20 lyase activities; P450c21, the adrenal 21-hydroxylase; and P450aro, which aromatizes androgens to estrogens. Thus, POR deficiency affects multiple steps in steroidogenesis (Fig. 1
). Furthermore, most POR lesions retain partial activity, so that these enzymatic steps may be incompletely blocked. Therefore, the array of disordered steroids seen in blood and urine can be variable, making the diagnosis difficult. POR deficiency can lead to genital ambiguity in both sexes; impairment of testicular P450c17 in severely affected male fetuses leads to incomplete masculinization, whereas impairment of placental aromatase and/or activation of the "backdoor pathway" of androgen biosynthesis (2, 16, 17) leads to partial virilization of female fetuses. The impairment in 17,20 lyase activity is greater than the impairment in 17
-hydroxylase activity, consistent with the established enzymology of P450c17 (18, 19, 20); hence, affected individuals typically have moderately elevated concentrations of 17OH-progesterone (17OHP) with an exaggerated response to acute stimulation with ACTH. The values are generally comparable to those seen in nonclassic 21-hydroxylase deficiency (1, 6, 11, 12). Maternal urinary estriol is very low due to impaired feto/placental steroidogenesis (21), and 21-deoxycortisol has been elevated in all patients in whom it has been measured. Basal cortisol is typically normal but poorly responsive to ACTH, indicating compensated adrenal insufficiency, and dehydroepiandrosterone (DHEA), DHEA sulfate (DHEAS), and androstenedione are typically low. Thus, the hormonal data may suggest POR deficiency but may not provide an unambiguous diagnosis; hence DNA sequencing has assumed a more important role in this disease than in other forms of adrenal hyperplasia.
|
| Subject and Methods |
|---|
|
|
|---|
DNA was collected in accordance with local institutional review board guidelines. CYP21B genetic analysis was done at the Alberta Childrens Hospital molecular diagnostic laboratory. The 15 coding exons of POR, including at least 100 bp of the flanking introns, were sequenced as described (6). We characterized POR exon 1U and the 5' flanking DNA by identifying a region of high homology with the known rat sequence using BLAST from the Ensembl Genome Browser (http://www.ensembl.org/index.html). Primers flanking exon 1U, 274 bp of the 5' flanking DNA, and introns 3, 4, 5, 6, 7, 9, and 11 were designed using the University of California, Santa Cruz genome browser (http://genome.ucsc.edu; accession no. NM_000941) and using Primer 3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) (Table 1
). Regions containing known single nucleotide polymorphisms (SNPs) were not used for primers. We performed PCR with a 20:1 mix of Taq (Promega, Madison, WI) and pfu DNA polymerases (Stratagene, Cedar Creek, TX) (22). We amplified introns 3, 4, 5, 6, and 7 (downstream fragment) 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 1 min, followed by 25 cycles of 95 C for 30 sec, 55 C for 30 sec, and 72 C for 1 min. The final extension was held at 72 C for 7 min, and then the reaction was stopped at 4 C. We amplified introns 7 (upstream fragment), 9, and 11 as described above, except that the annealing temperature started at 63 C and finished at 56 C after the 15 touchdown cycles and then continued at 56 C for the remaining 25 cycles. Amplification of exon 1U and the 5' flanking DNA was performed under the same conditions except that annealing started at 64 C for 15 touchdown cycles and then continued at 57 C for the remaining 25 cycles. All cycling for PCR was performed on the Bio-Rad MyCycler and i-Cycler (Bio-Rad Laboratories, Hercules, CA); PCR for sequencing was performed on the ABI GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA). Automated direct sequencing of PCR products employed the primers in Table 1
and ABI BigDye terminator, version 3.1 (Applied Biosystems). Data were displayed with ABI 3730#1 DNA Analyzer and analyzed using Sequencher demo version 4.5 (http://www.genecodes.com).
|
| Case Report |
|---|
|
|
|---|
| Results |
|---|
|
|
|---|
Sequencing of the CYP21B gene from the patient and her parents identified a deletion of the paternal allele and two mutations on the maternal allele: the severe IVS213A/C>G splice site mutation and the mild P30L mutation. Because of the mild, nonprogressive virilization, the craniosynostosis, and the unusual steroid pattern, we also sequenced the POR gene, revealing the common A287P mutation on the maternal allele. No lesions were found on the paternal allele. No mutations were found in exons 8 and 10 of FGFR2, a common genetic cause of craniosynostosis.
Characterization of exon 1U
All previous analyses of the POR gene in patients with apparent POR deficiency have described the POR gene as containing 15 exons and 14 introns (1, 2, 3, 6, 10, 11, 12). However the NCBI Entrez Nucleotide database entry for the human POR cDNA and mRNA included 5' untranslated sequences not found in "exon 1" (www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=Nucleotide; accession no. NM_000941.2), and previous studies showed that the rat POR gene contains an untranslated exon 30.5 kb upstream from the first coding exon (23, 24, 25). To determine whether the human gene is organized similarly, we probed the human genome database with the 56-nucleotide sequence of the rat upstream untranslated exon and identified a similar human DNA region about 38.8 kb upstream from what has previously been called exon 1; we term this exon "1U" so as to preserve the numbering from previous studies (1, 2, 3, 6, 10, 11, 12). The 3' end of exon 1U is defined by the mRNA/cDNA sequence (accession no. NM_000941.2) and is immediately followed by a canonical GTGAG splice donor site. The 5' end of the exon has not been defined experimentally, but a putative transcriptional start site has been chosen by analogy with the rat gene, whose transcriptional start site has been identified experimentally (23). Putative promoter regions important for transcription of the rat gene lie immediately upstream and are highly conserved in the human sequence (Fig. 2
).
|
Our previous reports included five patients in whom we had found only one POR mutation; these included subject 2 (1) and subjects CF1, 259, 337, and 14683 (6). We sought cryptic mutations in our present patient and in those five previous patients by sequencing a 478-bp fragment encompassing exon 1U and 274 bp of 5' flanking, putative promoter DNA, as well as sequencing all of introns 3, 4, 5, 6, 7, 9, and 11. The introns between exons 1U and 1 (38,809 bp), exons 1 and 2 (18,231 bp) and exons 2 and 3 (6,988 bp) were too large for PCR amplification, so that only 100 bp at each end, adjacent to the exons, was sequenced.
Patients 259 and 337 (6) were homozygous for known SNP rs3823884 in the 5' untranslated region of exon 1U in position 47A>C. Our patient had two POR intronic substitutions: IVS6 + 499C>G and IVS11 + 31C>T. Patient CF1 (6) had substitutions IVS6 + 418G>A and IVS11 + 32G>A; no other intronic sequence variants and no promoter sequence variants were found. These intronic nucleotide substitutions are not known SNPs, but we have observed the same intron 11 substitutions in our sequence analysis of 851 normal controls (Huang, N., V. Agrawal, K. M. Giacomini, and W. L. Miller, unpublished data). The location of the intron 6 substitutions was not covered by the sequencing strategy used for normal controls, so we do not know whether these SNPs are present in healthy individuals. Analysis of these nucleotides with NetGene2 (http://www.cbs.dtu.dk/services/NetGene2) indicates that these changes would not disrupt or create a cryptic splice acceptor or donor site.
| Discussion |
|---|
|
|
|---|
-hydroxylase activity (1, 3, 4, 6), so that 17OHP accumulates, but androgens do not.
POR impairs both the 17
-hydroxylase and 17,20 lyase activities of CYP17, partially blocking the formation of androgen via the classical pathway. Virilization in females with POR deficiency appears to arise both from impaired placental aromatase activity and activation of the backdoor pathway to dihydrotestosterone (16, 17, 30). The role of the backdoor pathway is supported by finding increased amounts of androsterone in the urine of a woman carrying a POR-deficient fetus (30) and the observation that CYP17 has greater affinity for 5
reduced C21 steroids than for 17OHP or 17OH-pregnenelone (31). However, the kinetics of the other enzymes in the backdoor pathway are unknown, and the contribution of this pathway to human androgen production is not yet defined. The lack of severe virilization and low C19 steroids both suggest a relative deficiency of androgen synthesis at the level of 17,20 lyase activity secondary to POR deficiency.
A role for POR in our patient was initially suggested by her craniosynostosis in the presence of impaired 17,20 lyase activity and normal FGFR2 gene sequences. Craniosynostosis can be an isolated anomaly or part of a craniosynostosis syndrome, most of which are caused by FGFR mutations. The incidence of craniosynostosis is about 35 per 10,000 births, of which 86% is nonsyndromic (32, 33); most syndromic craniosynostosis is caused by gain-of-function mutations in exons 8 or 10 of FGFR2. Craniosynostosis has not been linked to defects of steroidogenesis other than POR deficiency.
POR deficiency is autosomal recessive (1, 2, 3, 6, 10, 11, 12). However, we were only able to find one mutation, A287P, in our patient. Consistent with recessive inheritance, the unaffected mother also carried A287P on one allele. Six of the 41 reported patients (15%) with POR deficiency have only one identified mutation, despite being clinically indistinguishable from patients with two identified mutations (1, 6). This is an unusually high percentage for an autosomal recessive disease. Therefore, we extended the search for POR mutations to include 12 of the 15 introns, exon 1U and the 5' flanking region, but were unable to identify nucleotide variations that appeared to cause abnormal splicing of the transcription product or abnormal regulation of gene expression. A recent brief report mentions that only one mutation was found in two of five patients, but, because clinical descriptions of all of these patients had been reported previously, it is not clear how many additional patients are described (34). Nevertheless, this report strengthens the observation that apparent manifesting heterozygotes of POR are common.
Manifesting heterozygosity occurs frequently in X-linked disorders such as Duchenne muscular dystrophy due to skewed X inactivation (35), but is also described in autosomal recessive disorders; for example, about 17% of thyroid peroxidase deficiency represents manifesting heterozygosity in which only the mutant allele is transcribed into RNA (36). Our six patients may represent a new example of manifesting heterozygosity in an autosomal recessive disease. Because RNA was not available for study from our patients, we cannot be certain whether they represent true manifesting heterozygosity or only apparent heterozygosity, with unidentified mutations or gene silencing via abnormal methylation on the seemingly unaffected allele. Thus, these patients have apparent, but unproven, manifesting heterozygosity.
| Acknowledgments |
|---|
| Footnotes |
|---|
Author Disclosure Summary: All authors have nothing to declare.
First Published Online March 27, 2007
1 R.R.S. and L.G.G. contributed equally to this work. ![]()
Abbreviations: ABS, Antley-Bixler syndrome; DHEA, dehydroepiandrosterone; DHEAS, DHEA sulfate; FGFR2, fibroblast growth factor receptor 2; 17OHP, 17OH-progesterone; POR, P450 oxidoreductase; SNP, single nucleotide polymorphism.
Received October 26, 2006.
Accepted March 19, 2007.
| References |
|---|
|
|
|---|
-hydroxylase/17,20-lyase): insights into reaction mechanisms and effects of mutations. Mol Endocrinol 13:11691182
-hydroxylase deficiency in 17 families. Am J Hum Genet 63:16941702[CrossRef][Medline]
-Reduced C21 steroids are substrates for human cytochrome P450c17. Arch Biochem Biophys 418:151160[CrossRef][Medline]This article has been cited by other articles:
![]() |
L. G. Gomes, N. Huang, V. Agrawal, B. B. Mendonca, T. A. S. S. Bachega, and W. L. Miller The Common P450 Oxidoreductase Variant A503V Is Not a Modifier Gene for 21-Hydroxylase Deficiency J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2913 - 2916. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. V. Pandey, P. Kempna, G. Hofer, P. E. Mullis, and C. E. Fluck Modulation of Human CYP19A1 Activity by Mutant NADPH P450 Oxidoreductase Mol. Endocrinol., October 1, 2007; 21(10): 2579 - 2595. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |