Identification of Heterozygotic Carriers of21-Hydroxylase Deficiency: Sensitivity of ACTHStimulation Tests

Selma F. Witchel* and Peter A. LeeDivision of Endocrinology, Diabetes, and Metabolism, Department of Pediatrics, Children’s Hospital of Pittsburgh,University of Pittsburgh, Pittsburgh, Pennsylvania

Congenital adrenal hyperplasia due to 21-hydroxylase deficiency is a common autoso-mal-recessive disorder. To ascertain carrierstatus, adrenocorticotropin (ACTH) stimu-lation tests are often used. To determine thesensitivity of ACTH stimulation to detectheterozygotes and to correlate stimulated17-hydroxyprogesterone responses with mo-lecular genotype, we compared moleculargenetic analysis of the 21-hydroxylase(CYP21) gene with 17-hydroxyprogesteroneresponses at 30 min in 51 individuals. Mo-lecular genotype analysis and ACTH stimu-lation tests were performed in healthy vol-unteers (n = 20) and relatives of patientswith congenital adrenal hyperplasia (n =31). Polymerase chain reaction (PCR) ampli-fication, single-strand conformational poly-morphism (SSCP) analysis, allele-specificoligonucleotide hybridization (ASOH)analysis, and restriction fragment lengthpolymorphism (RFLP) analysis were uti-lized to screen for 14 CYP21 mutationswhich account for >90% of the mutations as-sociated with 21-hydroxylase deficiency.Molecular genotype analysis classified 28individuals as heterozygotic carriers and23 individuals as normal for all muta-tions tested. As a group, the heterozygoteshad significantly greater stimulated17-hydroxyprogesterone responses at 10and 30 min (P < 0.0005). However, on an in-dividual basis, 14/28 (50%) genotyped het-erozygotic carriers had stimulated 17-hy-droxyprogesterone concentrat ions ,17-hydroxyprogesterone/cortisol ratios, and

17-hydroxyprogesterone incremental eleva-tions indistinguishable from the genotypednormal individuals. Thus, a normal 17-hydroxyprogesterone response to ACTHstimulation testing does not exclude carrierstatus for 21-hydroxylase deficiency. Mo-lecular genotype analysis is a more reliablemethod to determine 21-hydroxylase het-erozygotes. Am. J. Med. Genet. 76:337–342,1998. © 1998 Wiley-Liss, Inc.

KEY WORDS: congenital adrenal hyperpla-sia; 21-hydroxylase defi-ciency; ACTH stimulationtests; mutation detection;heterozygotic carriers

INTRODUCTIONCongenital adrenal hyperplasia due to 21-

hydroxylase deficiency is a common autosomal-recessive disorder due to mutations in the 21-hydroxylase gene (CYP21) [Speiser et al., 1985; Whiteet al., 1985, 1986; Amor et al., 1988; Higashi et al.,1988; Mornet et al., 1991; Tusie-Luna et al., 1991;Helmberg et al., 1992; Owerbach et al., 1992; Miller,1994]. Decreased adrenal 21-hydroxylase activity im-pairs cortisol and aldosterone biosynthesis, resulting inglucocorticoid and, often, mineralocorticoid deficien-cies. Diminished negative feedback inhibition by corti-sol leads to increased adrenocorticotropin (ACTH) se-cretion. Since the adrenal androgen biosynthetic path-way is unaffected, increased ACTH secretion provokesincreased adrenal androgen secretion.

In the more severe forms in which excessive adrenalandrogen secretion begins in utero, affected female in-fants are often identified at birth because of genitalambiguity. In some delivery rooms, virilized female in-fants may not be recognized and may be misassigned asboys. Undiagnosed salt-losing male infants typicallypresent in adrenal crisis, associated with unacceptablyhigh morbidity and mortality [Nass and Baker, 1991].Once an index case has been identified, relatives areoften eager to ascertain their carrier status and, hence,risk for affected offspring.

Contract grant sponsor: NIH; Contract grant numbers: HD-00965, 5MO1-RR-00084, 5-P30 HD28836-04; Contract grantsponsor: Renziehausen Trust Fund.

*Correspondence to: Selma F. Witchel, M.D., Division of Endo-crinology, Diabetes, and Metabolism, Department of Pediatrics,Children’s Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh,PA 15213. E-mail: sfs@med.pitt.edu

Received 3 November 1997; Accepted 15 December 1997

American Journal of Medical Genetics 76:337–342 (1998)

© 1998 Wiley-Liss, Inc.

As a group, obligate heterozygotes show greaterACTH-stimulated 17-hydroxyprogesterone (17-OHP)responses than control subjects [Lee and Gareis, 1975].In another study, obligate heterozygotes demonstratedgreater mean 17-OHP and 21-deoxycortisol responsesto ACTH stimulation than nongenotyped controls, butshowed much overlap in hormone concentrations[Krensky et al., 1977]. To determine the sensitivity ofACTH stimulation tests and to determine whether themagnitude of ACTH-stimulated steroid hormone re-sponses correlated with molecular genotype, we com-pared the 21-hydroxylase molecular genotype with ste-roid hormone responses in 51 individuals.

MATERIALS AND METHODSSubjects

Relatives of patients with 21-hydroxylase deficiency(n 4 31; 15 men and 16 women) participated in thisstudy. Relatives included 8 mothers, 7 fathers, 6 grand-fathers, 4 grandmothers, 3 aunts, 2 uncles, and 1 adultsister. Healthy volunteers (n 4 20; 9 men and 11women) were recruited. None of the women partici-pants (relatives or volunteers) were hirsute; all hadregular menses. This protocol was approved by the Hu-man Rights Committee of the Children’s Hospital ofPittsburgh. Informed consent was obtained from alladult participants.

Molecular Genotype Analysis

Blood samples were obtained for HLA haplotypeanalysis and for DNA extraction from peripheral bloodlymphocytes. Standard serologic methods were used todetermine the extended HLA haplotypes for the A, B,C, DR, and DQ loci [Tong et al., 1993; Hsia et al., 1993].DNA was extracted from peripheral blood lymphocytes[Miller et al., 1988].

For each polymerase chain reaction (PCR) amplifica-tion performed, at least one primer specific for thefunctional 21-hydroxylase gene was used. Portions ofthe 21-hydroxylase gene were amplified as previouslydescribed [Siegel et al., 1995]. Allele-specific oligo-nucleotide hybridization (ASOH) was performed aspreviously described [Siegel et al., 1995]. Additionaloligonucleotide probes utilized were: Arg339-WT, 58-TACAAGGACCGTGCACGGCT-38; His339-MUT, 58-TACAAGGACCATGCACGGCT-38; exon 6-WT, 58-CTCAGCTGCATCTCCACGA-38; exon 6-MUT, 58-CTCAGCTGCTTCTCCTCGT-38; GT1769-WT, 58-ACCCTGAGGTGCGTCCTG-38; CT1769-MUT, 58-ACCCTGAGCTGCGTCCTG-38 ; Pro453-WT, 58-C G C T G C T G C C C T C C G G G G - 3 8 ; S e r 4 5 3 , 5 8 -CGCTGCTGTCCTCCGGGG-38 ; Arg484-WT, 58-ATCCCCCGGGGCTGCAG-38 ; Pro484-MUT, 58-ATCCCCGGGGGCTGCAG-38; STOP484-MUT, 58-A T C C C G G G G G C T G C A G - 3 8 ; 1 7 6 1 - W T , 5 8 -CGTGAAGCAAAAAAACCACGG-38; and i1761-MUT,58-CGTGAAGCAAAAAAAACCACGG-38.

Single-strand conformational polymorphism (SSCP)analysis was used to detect the splicing mutation inintron 2, V281L in exon 7, Q318X in exon 8, andR356W in exon 8, as previously described [Siegelet al., 1994; Witchel et al., 1996]. Primers 694F

(58-ACCTGTCCTTGGGAGACTAC-38) and 1122R (58-TCGTCCTGCCAGAAAAGGAG-38) were used to detectthe I172N mutation on a 5% acrylamide gel preparedwith 10% glycerol, which was electrophoresed at 40 wat 4°C for 11 hr, as previously described [Witchel et al.,1997]. Following electrophoresis, the gels were driedand autoradiography was performed at −80°C.

Hormonal Analysis

ACTH stimulation tests were performed on 31 rela-tives and 20 healthy adults (11 women and 9 men).Four obligate heterozygotic female carriers were testedapproximately 6 weeks after the birth of their affectedchild. All women had regular menstrual cycles. Theother 23 women were tested during the follicular phaseof the menstrual cycle (basal plasma progesterone <45ng/dl). None were hirsute or taking oral contraceptives.Samples were obtained for progesterone, 17-hydroxypregnenolone, 17-hydroxyprogesterone, dehy-droepiandrosterone (DHA), androstenedione (D4), andcortisol prior to administration of Cortrosynt (Orga-non, Inc., West Orange, NJ), 0.25 mg, intravenous in-fusion over 1 min. Subsequent blood samples were ob-tained 30 min postinfusion (n 4 50). Plasma proges-terone, 17-hydroxyprogesterone, and androstenedionewere assayed using double-antibody radioimmunoas-say kits with 125I-labeled steroids (ICN Biomedicals,Inc., Carson, CA). Respective interassay coefficients ofvariation (n 4 20) were 12.6%, 11.0%, and 9.2%. Cor-tisol was assayed using a solid-phase 125I radioimmu-noassay (Diagnostic Products Corporation, Los Ange-les, CA) with interassay coefficient of variation (n 420) of 6.1% [Siegel et al., 1993]. For all subjects, basalcortisol was <20 mg/dl.

Statistical Analysis

Statistical analysis was performed using AbSTATstatistical software (Anderson-Bell, Arvada, CO). The99% confidence intervals (mean + 2.57 SD) of hormoneconcentrations measured in the healthy individualswere used to define the upper limit of normal hormoneconcentrations. Student’s independent t-test was usedto compare hormone concentrations between the twogroups.

RESULTSMolecular Genotype Analysis

Molecular genotype analysis was performed usingASOH, SSCP, and RFLP techniques to assess for 14distinct CYP21 mutations which account for >90% ofmutations identified to date [Wedell et al., 1994]. Fourrelatives carried no deleterious mutations by both mu-tation studies and HLA linkage data; these 4 are in-cluded with the healthy controls for statistical analysisof ACTH-stimulated hormone responses.

Twenty-seven of the 31 relatives studied were foundto be carriers of 21-hydroxylase deficiency. Mutationsdetected were: gene conversion/gene deletion (n 4 8),intron 2 splicing mutation (n 4 7), I172N (n 4 2), exon6 triple-codon mutation (n 4 1), V281L (n 4 1), i1761(n 4 2), Q318X (n 4 4), and R356W (n 4 2). One

338 Witchel and Lee

normal allele and one mutant allele were identified foreach heterozygotic carrier.

Nineteen of the 20 healthy volunteers who had ge-notype analysis and ACTH stimulation tests showednormal results for all mutations investigated. One vol-unteer carried one normal allele and the intron 2 splic-ing mutation on the second allele. The ACTH-stimulated hormone responses obtained in this volun-teer were, thus, included with the carrier group forstatistical analysis. Accordingly, results of ACTHstimulation tests obtained in 28 carriers and 23 geno-typed normal subjects were compared.

Hormonal Responses in GenotypedNormal Subjects

The mean basal 17-OHP concentration was 103 ± 65ng/dl in the healthy controls (Table I). At 30 min, themean 17-OHP concentration was 201 ± 91 ng/dl. Wedefined the 99% confidence interval (mean + 2.57 SD)as the upper limit of normal, i.e., 434 ng/dl. The mean17-OHP incremental elevation at 30 min was 98 ± 76ng/dl (difference between stimulated and basal concen-tration).

For androstenedione, the mean basal concentrationwas 201 ± 98 ng/dl. At 30 min, the mean concentrationwas 300 ± 132 ng/dl. The mean incremental elevationat 30 min was 102 ± 65 ng/dl.

Comparison of ACTH Stimulation Tests inHeterozygotic 21-Hydroxylase Carriers and

Healthy Controls

Comparison of ACTH-stimulated hormone responsesbetween the heterozygotes and the healthy controlsshowed comparable mean basal 17-hydroxyprogester-one levels (106 ± 63 vs. 103 ± 65 ng/dl, respectively).Individual basal 17-OHP concentrations showed muchoverlap (Fig. 1). As expected, when considered as agroup, the carriers had significantly greater mean 17-hydroxyprogesterone responses at 30 min than thehealthy controls (Table I; Fig. 1) (P < 0.0001). Mean17-hydroxyprogesterone concentration among the car-riers was 468 ± 243 ng/dl. Mean 17-OHP incrementalelevation for the carriers was 362 ± 225 ng/dl. How-ever, the stimulated 17-OHP concentrations of indi-

viduals from both groups overlapped for values be-tween 200–400 ng/dl. Only carriers showed stimulated17-OHP responses >400 ng/dl (Figs. 1, 2).

Mean basal androstenedione concentrations wereslightly greater for the healthy controls than for theheterozygotes (Table I). Mean basal 17-hydroxy-pregnenolone and DHEA concentrations were signifi-cantly greater in the healthy controls than the hetero-zygotic carriers (P < 0.05 and P < 0.005, respectively).Following ACTH stimulation, the 17-hydroxypreg-nenolone, DHEA, and androstenedione concentrationswere comparable between the two groups (Table I).

We compared hormone ratios and the calculatedcombined rate of rise of progesterone and 17-OHP be-tween the two groups. The mean basal 17-OHP/cortisolratio (Table I) was similar between the genotyped nor-mal subjects and genotyped heterozygotes. At 30 min,the mean 17-OHP/cortisol ratio (Table I) was signifi-cantly greater in the genotyped heterozygotes than inthe genotyped normal subjects (P < 0.0001). The meancalculated rate of rise for progesterone and 17-OHPwas significantly greater among the heterozygotic car-riers, i.e., 13.7 ± 9.1 vs. 3.6 ± 3.2 ng/dl/min (P < 0.0001).

Fourteen of 28 heterozygotic carriers (50%) and noneof the healthy controls (0%) had 30-min 17-OHP con-centrations >434 ng/dl. The CYP21 mutations (Fig. 3)identified in the 14 heterozygotic carriers were geneconversions/deletions (n 4 5), splicing mutations (n 44), I172N (n 4 1), i1761 (n 4 1), Q318X (n 4 1), andR356W (n 4 2). The other 14 genotyped heterozygoticcarriers had 30-min 17-OHP concentrations <434 ng/dl. Among this group, CYP21 mutations (Fig. 3) de-tected were gene conversions/deletions (n 4 3), splicingmutations (n 4 4), I172N (n 4 1), an exon 6 triple-codon mutation (n 4 1), i1761 (n 4 1), Q318X (n 4 3),and V281L (n 4 1).

Thus, ACTH stimulation testing with sampling at 30min failed to characterize 14 of 28 genotyped heterozy-gotes as carriers of 21-hydroxylase deficiency; 13 indi-viduals had values for ACTH-stimulated 17-OHP con-centrations, 17-OHP incremental elevations, 17-OHP/F ratios, and combined 17-OHP and progesteronerates of rise which were indistinguishable from thehealthy genotyped normal subjects. Mutations identi-

TABLE I. Basal and ACTH-Stimulated Hormone Concentrations and Ratios in Controlsand Carriers*

Normal Carriers P value

Prog-08 (ng/dl) 21 ± 5 (22) 24 ± 16 (25) NSProg-308 (ng/dl) 32 ± 20 (23) 78 ± 61 (25) 0.000617-Preg-08 (ng/dl) 209 ± 166 (23) 123 ± 101 (24) 0.042717-Preg-308 (ng/dl) 1004 ± 518 (23) 988 ± 470 (24) NS17-OHP-08 (ng/dl) 103 ± 65 (23) 106 ± 63 (28) NS17-OHP-308 (ng/dl) 201 ± 91 (23) 468 ± 243 (28) 0.0001DHA-08 (ng/dl) 586 ± 335 (23) 343 ± 232 (27) 0.0040DHA-308 (ng/dl) 1125 ± 436 (23) 946 ± 663 (27) NSD4-08 (ng/dl) 201 ± 98 (23) 158 ± 67 (28) 0.0642D4-308 (ng/dl) 300 ± 132 (23) 287 ± 140 (28) NS17-Preg/17-OHP-08 2.17 ± 1.6 (23) 1.44 ± 0.86 (24) 0.064117-Preg/17-OHP-308 5.31 ± 2.07 (23) 2.68 ± 1.30 (24) 0.000017-OHP/F-08 9.13 ± 4.97 (23) 10.91 ± 7.93 (28) 0.356017-OHP/F-308 8.35 ± 3.38 (23) 10.62 ± 8.09 (28) 0.0000

*Number in parentheses indicates number of subjects.

ACTH Stim Tests in 21-Hydroxylase Carriers 339

fied in these individuals were gene deletion/conversion(n 4 3), splicing mutations (n 4 4), I172N (n 4 1),exon 6 (n 4 1), i1761 (n 4 1), L281 (n 4 1), and Q318X(n 4 2). Use of the 95% confidence interval would haveascertained one additional heterozygotic carrier(Fig. 3).

Correlation of Hormone ResponseWith Genotype

Correlation of clinical findings with molecular geno-type has shown specific mutations to be consistentlyassociated with classical salt-losing 21-hydroxylase de-ficiency. Of the 17 subjects with these mutations, i.e.,gene deletion/conversion, Q318X, R356W, exon 6, andi1761, only 9 (53%) subjects showed elevated 17-OHPresponses. Four of 8 patients with the splicing muta-tion showed elevated 17-OHP responses. The 2 relatedcarriers of the I172N mutation showed vastly different17-OHP responses; the grandfather’s stimulated 17-OHP concentration was 1,078 ng/dl whereas themother, tested in the postpartum period, had a stimu-lated 17-OHP concentration of 219 ng/dl. No additionalmutations were observed in either individual. The ex-tended HLA haplotypes were the same for these 2 in-dividuals, signifying that the CYP21 allele was identi-cal. The heterozygotic carrier for the V281L mutationcommonly associated with nonclassical or late-onsetcongenital adrenal hyperplasia showed a stimulated17-OHP concentration of 353 ng/dl.

DISCUSSION

We compared CYP21 molecular genotypes withACTH-stimulated steroid hormone responses in 51 in-dividuals. Four relatives (aunts and uncles of the pro-positi) showed wild-type CYP21 sequences for all mu-tations tested; extended HLA haplotypes showed thatthese 4 individuals carried unaffected CYP21 alleleswith no evidence of recombination events. One of 19healthy controls was found to be a carrier on genotypeanalysis, which was not unexpected given the high fre-quency of CYP21 mutations [Speiser et al., 1985].

Basal and stimulated steroid hormone responsesmeasured in our genotyped normal subjects were usedto establish normal values. Our genotypically normalsubjects showed stimulated 17-OHP concentrationsranging from 106–376 ng/dl. We compared the 17-OHPresponses of our genotyped normal subjects to thosereported in a nomogram created by New et al. [1983].Although the nomogram utilizes hormone concentra-

Fig. 1. Basal and stimulated 17-hydroxyprogesterone (17-OHP) concen-trations. Basal and ACTH-stimulated 17-OHP concentrations (mean ±SEM) are indicated for healthy controls (circles) and heterozygotic car-riers (squares). Basal 17-OHP concentrations are not different. Stim-ulated 17-OHP concentrations are significantly greater in the carriers(P < 0.0005), but there is much overlap.

Fig. 2. Stimulated 17-OHP re-sponses. Bars indicate stimulated17-OHP concentrations in all sub-jects. Solid bars indicate healthy con-trols, and hatched bars indicate car-riers. Solid line indicates 435 ng/dl(99% confidence interval). Dotted lineindicates 383 ng/dl (95% confidenceinterval).

340 Witchel and Lee

tions obtained 60 min following stimulation and thevalues reported herein were obtained at 30 min, stimu-lated 17-OHP responses at 30 min are not significantlydifferent from those at 60 min. The nomogram, basedon basal and stimulated 17-OHP responses, was re-ported in 1983 prior to the identification and sequenc-ing of the CYP21 and CYP21P genes [New et al., 1983].Thus, genotype analysis was not available for the 51general-population control subjects in this prior study[New et al., 1983]. In the nomogram, the 5 individualsidentified as being unaffected based on HLA haplotypeinformation had ACTH-stimulated 17-OHP concentra-tions (60 min after stimulation) ranging from 120–350ng/dl, which are comparable to those of our genotypednormal subjects [New et al., 1983].

We then compared ACTH-stimulated hormone re-sponses between the 28 genotyped heterozygotic carri-ers and the 23 genotyped normal subjects. Not surpris-ingly, as a group, obligate heterozygotes had a signifi-cantly elevated 30-min 17-OHP response [Lee andGareis, 1975; Child et al., 1979]. However, when con-sidered on an individual basis, 50% of the carriers hadstimulated 17-OHP responses at 30 min within thenormal range (<434 ng/dl). Further, no consistent as-sociations were apparent between stimulated 17-OHPconcentrations and specific CYP21 mutations. For ex-ample, 17-OHP responses in 2 related carriers of theI172N mutation (father and daughter), typically asso-ciated with simple virilizing 21-hydroxylase deficiency,were 1,078 ng/dl and 219 ng/dl. The sensitivity ofACTH stimulation testing to detect heterozygotes was50%, whereas specificity was 100%. Thus, ACTHstimulation tests failed to identify half the carrierstested. Most importantly, if normal 17-hydroxy-progesterone responses to ACTH stimulation testingwere interpreted to indicate noncarrier status, inaccu-rate genetic counseling would have ensued.

Four women underwent ACTH stimulation testingpostpartum. Three women had given birth to girls withclassical 21-hydroxylase deficiency recognized by geni-tal ambiguity. The fourth infant was identified through

neonatal screening at age 6 weeks. All 4 women showed17-hydroxyprogesterone responses within the normalrange. None had resumed menses at the time of ACTHstimulation testing. Thus, ACTH stimulation testing isnot reliable postpartum.

To improve the predictive value of ACTH stimulationtesting, calculated sums and ratios of steroid hormoneswere utilized. For example, Peter et al. [1990] reportedthat the ACTH-stimulated ratio of 17-hydroxyproges-terone to 11-deoxycorticosterone showed no overlap be-tween carriers and controls as defined by HLA haplo-type analysis, but their radioimmunoassay requiredextraction and chromatography. Comparable to ourfindings, the calculated rate of rise of progesterone and17-OHP identified 50% of carriers [Gutai et al., 1977].The use of HLA haplotypes improves the ability toidentify carriers [Mauseth et al., 1980], although ge-netic recombinations can occur between the HLA lociand CYP21.

In summary, we have characterized 17-OHP re-sponses to ACTH stimulation in healthy individualswho showed wild-type nucleotide sequences for allCYP21 mutations assayed. Hence, whereas some pre-viously reported studies found a lower cutoff range fornormals, we would concur (Fig. 2) that 17-OHP re-sponses between 500–1,000 ng/dl are consistent withheterozygosity for 21-hydroxylase deficiency [Wedell etal., 1994; Dewailly et al., 1986; Azziz et al., 1994; Tardyet al., 1996]. Most importantly, a 17-OHP response toACTH stimulation <435 ng/dl does not exclude hetero-zygosity for 21-hydroxylase deficiency. Molecular ge-netic analysis is now feasible and provides greater ac-curacy regarding carrier status, with the caveat thatrelatives may carry mutations that differ from the in-dex case [Witchel et al., 1996].

ACKNOWLEDGMENTS

We are indebted to Makiko Suda-Hartman for herexpert technical assistance. We also appreciate the as-sistance of Amy Jones, R.N., Janet Bell, R.N., Tamara

Fig. 3. Comparison of 17-OHP re-sponses with specific mutation. Mu-tations are grouped in order of sever-ity with most severe (del/gene conver-sion) at far left and most mildmutation (V281L) at far right. Spe-cific mutations are listed beneath thebars: X318, Gln3 1 8→X; W356,Arg356→Trp; AAA, Ile236→Asn,Val237→Glu, and Met239→Lys; iT, Tinsertion in exon 7; splicing, splicingmutation at nucleotide 655 in intron2; N172, I le→Asn; and L281,Val281→Leu.

ACTH Stim Tests in 21-Hydroxylase Carriers 341

Johnston, R.N., and Debbie Cleary. We gratefully ac-knowledge Eric P. Hoffman, Ph.D., for helpful discus-sions. This work was supported in part by NIH grantsHD-00965 (to S.F.W.), 5MO1-RR-00084 (to the GeneralClinical Research Center), and 5-P30 HD28836-04 (tothe Child Health Research Center). This work was alsosupported in part by the Renziehausen Trust Fund (tothe Children’s Hospital of Pittsburgh).

REFERENCESAmor M, Parker KL, Globerman H, New MI, White PC (1988): Mutation in

the CYP21B gene (Ile-172→Asn) causes steroid 21-hydroxylase defi-ciency. Proc Natl Acad Sci USA 85:1600–1604.

Azziz R, Dewailly D, Owerbach D (1994): Clinical review 56. Nonclassicadrenal hyperplasia: Current concepts. J Clin Endocrinol Metab 78:810–815.

Child DF, Bu’lock DE, Anderson DC (1979): Adrenal steroidogenesis inheterozygotes for 21-hydroxylase deficiency. Clin Endocrinol (Oxf) 11:391–398.

Dewailly D, Vantyghem-Haudiquet M-C, Sainsard C, Buvat J, CappoenJP, Ardaens K, Racadot A, Lefebvre J, Fossati P (1986): Clinical andbiological phenotypes in late-onset 21-hydroxylase deficiency. J ClinEndocrinol Metab 63:418–423.

Gutai JP, Kowarski AA, Migeon CJ (1977): The detection of the heterozy-gous carrier for congenital adrenal hyperplasia. J Pediatr 90:924–929.

Helmberg A, Tusie-Luna M-T, Tabarelli M, Kofler R, White PC (1992):R339H and P453S: CYP21 mutations associated with nonclassic ste-roid 21-hydroxylase deficiency that are not apparent gene conversions.Mol Endocrinol 6:1318–1322.

Higashi Y, Tanae A, Inoue H, Hiromasa T, Fujii-Kuriyama Y (1988): Ab-errant splicing and missense mutations cause steroid 21-hydroxylase[P-450(C21)] deficiency in humans: Possible gene conversion product.Proc Natl Acad Sci USA 85:7486–7490.

Hsia S, Tong JY, Parris GL, Nghiem DD, Cottington EM, Ruderf WA,Trucco M (1993): Molecular compatibility and renal graft survival—The HLA DRB1 genotyping. Transplantation 55:395–399.

Krensky AM, Bongiovanni AM, Marino J, Parks J, Tenore A (1977): Iden-tification of heterozygote carriers of congenital adrenal hyperplasia byradioimmunoassay of serum 17-OH progesterone. J Pediatr 90:930–933.

Lee PA, Gareis FJ (1975): Evidence for partial 21-hydroxylase deficiencyamong heterozygote carriers of congenital adrenal hyperplasia. J ClinEndocrinol Metab 41:415–418.

Mauseth RS, Hansen JA, Smith EK, Giblett ER, Kelley VC (1980): Detec-tion of heterozygotes for congenital adrenal hyperplasia: 21-hydroxylase deficiency—A comparison of HLA typing and 17-OH pro-gesterone response to ACTH infusion. J Pediatr 97:749–753.

Miller SA, Dykes DD, Polesky HF (1988): A simple salting out procedurefor extracting DNA from human nucleated cells. Nucleic Acids Res16:1215.

Miller WL (1994): Clinical review 54. Genetics, diagnosis, and manage-ment of 21-hydroxylase deficiency. J Clin Endocrinol Metab 78:241–246.

Mornet E, Crete P, Kuttenn F, Raux-Demay M-C, Boue J, White PC, BoueA (1991): Distribution of deletions and seven point mutations onCYP21B genes in three clinical forms of steroid 21-hydroxylase defi-ciency. Am J Hum Genet 48:79–88.

Nass R, Baker S (1991): Learning disabilities in children with congenitaladrenal hyperplasia. J Child Neurol 6:306–312.

New MI, Lorenzen F, Lerner AJ, Kohn B, Oberfield SE, Pollack MS, Du-pont B, Stoner E, Levy DJ, Pang S, Levine LS (1983): Genotypingsteroid 21-hydroxylase deficiency: Hormonal reference data. J Clin En-docrinol Metab 57:320–326.

Owerbach D, Sherman L, Ballard A-L, Azziz R (1992): Pro-453 to Sermutation in CYP21 is associated with nonclassic steroid 21-hydroxylase deficiency. Mol Endocrinol 6:1211–1215.

Peter M, Sippell WG, Lorenzen F, Willig RP, Westphal E, Grosse-Wilde H(1990): Improved test to identify heterozygotes for congenital adrenalhyperplasia without index case examination. Lancet 335:1296–1299.

S iege l SF , F inego ld DN, Murray PJ , Lee PA (1993) : 17 -hydroxypregnenolone responses in adolescent girls with acanthosis ni-gricans. Adolesc Pediatr Gynecol 6:9–14.

Siegel SF, Hoffman EP, Trucco M (1994): Molecular diagnosis of 21-hydroxylase deficiency: Detection of four mutations on a single gel.Biochem Med Metab Biol 51:66–73.

Siegel SF, Lee PA, Rudert WA, Swinyard M, Trucco M (1995): Phenotype/genotype correlations in 21-hydroxylase deficiency. Adolesc Pediatr Gy-necol 8:9–16.

Speiser PW, Dupont B, Rubinstein P, Piazza A, Kastelan A, New MI(1985): High frequency of nonclassical steroid 21-hydroxylase defi-ciency. Am J Hum Genet 37:650–667.

Tardy V, Carel JC, Forest MG, et al. (1996): Nonclassical forms of 21-hydroxylase deficiency revisited by molecular biology. EndocrinologySociety Annual Meeting, Abstract P2-737, San Francisco, CA.

Tong JY, Hammad A, Rudert WA, Trucco M, Hsia S (1993): Heterodu-plexes for HLA DQB1 identity of family members and kidney donor-recipient pairs. Transplantation 55:741–745.

Tusie-Luna M-T, Speiser PW, Dumic M, New MI, White PC (1991): Amutation (Pro-30 to Leu) in CYP21 represents a potential nonclassicsteroid 21-hydroxylase deficiency allele. Mol Endocrinol 5:685–692.

Wedell A, Thilen A, Ritzen EM, Stengler B, Luthman H (1994): Mutationalspectrum of the steroid 21-hydroxylase gene in Sweden: Implicationsfor genetic diagnosis and association with disease manifestations. JClin Endocrinol Metab 78:1145–1152.

White PC, Grossberger D, Onufer BJ, Chaplin DD, New MI, Dupont B,Strominger JL (1985): Two genes encoding steroid 21-hydroxylase arelocated near the genes encoding the fourth component of complement inman. Proc Natl Acad Sci USA 82:1089–1093.

White PC, New MI, Dupont B (1986): Structure of human steroid 21-hydroxylase gene. Proc Natl Acad Sci USA 83:5111–5115.

Witchel SF, Nayak S, Suda-Hartman M, Lee PA (1997): Newborn screen-ing for 21-hydroxylase deficiency: Results of CYP21 molecular geneticanalysis. J Pediatr 131:328–331.

Witchel SS, Lee PA, Trucco M (1996): Who is a carrier? Detection of un-suspected mutations in 21-hydroxylase deficiency. Am J Med Genet61:2–9.

342 Witchel and Lee