Prevalence of CYP21 mutations and IRS1 variant among women with polycystic ovary syndrome and adrenal androgen excess

Prevalence of CYP21 mutations and IRS1 variant among women with polycystic ovary syndrome and adrenal androgen excess Selma F. Witchel, M.D.,a Melissa Kahsar-Miller, Ph.D.,b,c Christopher E. Aston, Ph.D.,d Carlie White, B.S.,a and Ricardo Azziz, M.D., M.P.H.b a Department of Pediatrics, Children’s Hospital of Pittsburgh, University of Pittsburgh, Pittsburgh, Pennsylvania; b Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, Alabama (R.A.’s current address: Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center, Los Angeles, California); c Medical and Molecular Genetics, University of Indiana–Purdue University, Indianapolis, Indiana; and d Genetic Epidemiology Unit, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma

Objective: To determine whether frequencies of the mutations in the 21-hydroxylase (CYP21) gene and the G972R variant of the insulin receptor substrate-1 (IRS1) gene are increased in women with polycystic ovary syndrome (PCOS) and adrenal androgen (AA) excess. Design: Prospective case– control study. Setting: University reproductive endocrinology laboratory and outpatient clinic. Patient(s): Consecutive patients of non-Hispanic white race diagnosed with PCOS (n ⫽ 114) and healthy controls (n ⫽ 95). Intervention(s): Blood and DNA sampling before hormonal therapy. Main Outcome Measure(s): Polycystic ovary syndrome patient and healthy control genotypes, with the CYP21 and IRS1 variants. Result(s): Fifty-four PCOS patients with (DHEAS ⬎3000 ng/mL) and 55 without (DHEAS ⬍2500 ng/mL) AA excess, respectively, were studied. Of 109 patients studied, 16 (14.7%) were found to be heterozygous carriers of mutations in the CYP21 gene. Of these 16, 10 (62.5%) had excessive AA secretion (i.e., excess DHEAS levels). Fifteen patients (13.8%) were found to be heterozygous carriers of the IRS1 variant; 9 (60.0%) of these 15 had excessive AA secretion. There were no significant differences in the allele frequency of CYP21 mutations or the IRS1 variant between PCOS patients with and without AA excess, and controls. None of the subjects were found to be homozygous carriers of CYP21 mutations or the IRS1 variant. Combined heterozygosity for CYP21 mutations and the IRS1 variant was limited to women with PCOS and excessive AA (n ⫽ 3). Conclusion(s): The G972R variant of the IRS1 gene might represent a modifier locus among women who are heterozygous carriers of CYP21 mutations, potentially increasing their risk of developing AA excess in PCOS. Nonetheless, this IRS1 variant and CYP21 mutations seem to play a limited role in the development of PCOS in the population studied. (Fertil Steril威 2005;83:371–5. ©2005 by American Society for Reproductive Medicine.) Key Words: Polycystic ovary syndrome, CYP21, insulin receptor substrate-1, adrenal, genetics, androgen

Polycystic ovary syndrome (PCOS) is a common endocrine disorder affecting approximately 4% of women (1, 2). Polycystic ovary syndrome is characterized by ovulatory dysfunction and hyperandrogenism, which can be clinically suspected by the presence of hirsutism, acne, or androgenic alopecia and confirmed by the finding of elevated androgen concentrations (3, 4). Although the underlying mechanisms responsible for PCOS remain unclear, familial aggregation indicates the importance of genetic factors (5). Despite intensive efforts, no single PCOS gene has been identified. Instead, mutation detection, linkage analysis, and case– Received October 15, 2003; revised and accepted October 14, 2004. Supported in part by grants RO1-HD29364 (R.A.) and R29-HD34808 (S.F.W.) from the National Institutes of Health, Bethesda, Maryland, and grants from the American Heart Association (S.F.W.). Reprint requests: Selma F. Witchel, M.D., Childrens Hospital of Pittsburgh, Department of Pediatrics, 3705 Fifth Avenue, Pittsburgh, Pennsylvania 15213 (FAX: 412-692-5834; E-mail: selma.witchel@ chp.edu).

0015-0282/05/$30.00 doi:10.1016/j.fertnstert.2004.10.027

control studies have affirmed the complexity of the molecular genetics of this disorder (6). Hence, PCOS seems to be yet another example of a complex multifactorial genetic disorder in which the phenotype is influenced by genetic heterogeneity, gene– gene interactions, gene– environment interactions, and environmental factors (7, 8). Abnormalities of the hypothalamic–pituitary– ovarian axis, ovarian and adrenal steroidogenesis, and/or insulin action have been demonstrated among women with PCOS. Although the ovaries seem to be the principal source of androgens in the majority of patients, studies have shown that approximately 50%– 60% exhibit elevated circulating adrenal androgens (AAs), particularly DHEAS and 11hydroxyandrostenedione (9 –11). Although adrenal androgen excess might arise, in part, in response to ovarian or hyperinsulinemic effects, most data suggest that this abnormality is independent of these influences (12). Heritability of DHEAS concentrations suggests that in a subset of women

Fertility and Sterility姞 Vol. 83, No. 2, February 2005 Copyright ©2005 American Society for Reproductive Medicine, Published by Elsevier Inc.

371

with PCOS, excessive AA secretion plays a major role in the pathophysiology (13, 14). Because phenotypic heterogeneity has plagued the search for the PCOS genes, segregation of women with PCOS by DHEAS concentrations will lessen the hormonal phenotypic heterogeneity and increase the likelihood of identifying genes associated with PCOS. An increased frequency of heterozygosity for mutations in the 21-hydroxylase gene (CYP21) has been reported among women with PCOS, adolescent girls with hyperandrogenism, and children with premature pubarche (15–17). Yet, although heterozygosity for CYP21 mutations is one potential explanation for the excessive AA secretion, obligate heterozygotic CYP21 mutation carriers (mothers of children with classic congenital adrenal hyperplasia) tend to be asymptomatic (18). To resolve this paradox, we hypothesized that variants in genes involved in insulin signal transduction might represent modifier loci that increase the risk of developing PCOS. This hypothesis is based on our in vitro observation that insulin might play a role in directly stimulating AA secretion by the adrenal cortex (19). The binding of insulin to its receptor initiates a series of events, including phosphorylation of the insulin receptor substrate proteins (protein: IRS; gene: IRS). Several IRS proteins, encoded by different genes, have been characterized. The IRS-1 protein plays an important role in skeletal muscle, whereas IRS-2 is important in liver, ovary, and pancreatic ␤ cells. After phosphorylation, IRS-1 and IRS-2 bind and activate downstream effectors, such as phosphoinositide 3-kinase (PI 3-kinase), to promote the metabolic and mitogenic actions of insulin. One variant of IRS1, G972R, has been associated with PCOS and type 2 diabetes mellitus (20, 21). In addition to being associated with type 2 diabetes mellitus in some populations, this variant has been associated with obesity and insulin resistance among nondiabetic Italian adults and decreased insulin secretion in glucose-tolerant adults (22, 23). Pancreatic islets cells obtained from G972R carriers show decreased insulin secretion and decreased IRS1-stimulated PI 3-kinase activity (24). Accumulating evidence indicates that this variant impairs both insulin signal transduction in skeletal muscle and insulin secretion by the ␤ cells. Among adolescent girls with hyperinsulinemia, hyperandrogenemia, and a history of premature pubarche, this variant seems to be a modifying locus (25). We hypothesized that the G972R variant might play a role in the pathogenesis of the PCOS, particularly among patients who are heterozygous carriers of CYP21 mutations. To test this hypothesis, we studied 114 consecutive patients with PCOS and 95 nonhirsute eumenorrheic controls. MATERIALS AND METHODS Subjects A total of 114 consecutive patients seeking treatment for PCOS were recruited from the reproductive endocrinology 372

Witchel et al.

clinic at the University of Alabama at Birmingham (UAB). Inclusion criteria were evidence of ovulatory dysfunction in conjunction with hirsutism and/or hyperandrogenemia, as well as exclusion of other disorders, such as nonclassic 21-hydroxylase-deficient adrenal hyperplasia (NCAH), Cushing’s syndrome, hyperprolactinemia, and hypothyroidism. Ovulatory dysfunction was defined as menstrual cycles ⬎45 days in length, or a P level ⬍2 ng/mL on days 22–24 of the menstrual cycle, in conjunction with a monophasic basal body temperature chart. Hirsutism was defined as a modified Ferriman-Gallwey score of ⱖ6 (2). Hyperandrogenemia was defined as serum levels of free T above the 95th percentile of the controls, as previously reported (2). Twenty-one hydroxylase-deficient NCAH was excluded by a basal follicular phase 17-hydroxyprogesterone (17-OHP) level ⬍2 ng/mL, or a 17-OHP level after acute ACTH stimulation ⬍10 ng/mL (26). Normal levels of prolactin and TSH excluded hyperprolactinemia and hypothyroidism, respectively; and Cushing’s syndrome was excluded by a 24-hour urine free cortisol level ⬍100 ␮g/day, when screening was clinically indicated. In addition, 95 healthy, nonhirsute women with regular menstrual cycles and not taking hormonal medications were recruited as controls. All subjects were non-Hispanic white. The UAB Institutional Review Board and the Human Rights Committee of the Children’s Hospital of Pittsburgh approved this study, and written informed consent was obtained from all subjects. Hormonal Assays Baseline samples for T, DHEAS, and sex hormone-binding globulin (SHBG) were frozen and assayed later in larger batches to minimize the impact of interassay variability. Prolactin, TSH, and 17-OHP were measured at the time of the initial clinical evaluation, and P was measured as needed. Total T was measured by an in-house RIA method after serum extraction, as previously described (26, 27), with the exception that before the RIA the extracted serum sample was separated with the use of methyl chloride:isopropyl alcohol (97:3) over a Sephadex LH-20 column (Sigma Chemical Co., St. Louis, MO). Sex hormone-binding globulin was measured by diffusion equilibrium dialysis with Sephadex G-25 and [3H] T as the ligand, and the free T was calculated as previously described (28). Dehydroepiandrosterone sulfate, P, prolactin, TSH, and 17-OHP were measured by direct RIA with commercially available kits (DHEAS and P from Diagnostic Products, Los Angeles, CA; TSH and prolactin from Nichols Institute Diagnostics, San Juan Capistrano, CA, as previously described) (2). Genotyping Genomic DNA was extracted from the peripheral blood lymphocytes by a nonorganic, nonenzymatic method (29). Mutation detection analyses of CYP21 were performed as previously described. Specifically, CYP21 mutations (gene

CYP21 mutations and IRS1 variant in PCOS

Vol. 83, No. 2, February 2005

TABLE 1 Results of CYP21 genotype analyses by specific mutation. Genotype CYP21-P30L CYP21-V281L CYP21-P453S CYP21-intron 2 CYP21-Q318X CYP21 (all mutations)

PCOS with AA (n ⴝ 54)

PCOS without AA (n ⴝ 55)

Control (n ⴝ 95)

5 1 2 2 0 10 (18.5%)

5 1 0 0 0 6 (10.9%)

4 1 4 0 2 11 (11.5%)

Note: AA ⫽ adrenal androgen; PCOS ⫽ polycystic ovary syndrome. Witchel. CYP21 mutations and IRS1 variant in PCOS. Fertil Steril 2005.

conversion events, P30L, intron 2 splicing mutation, I172N, V281L, Q318X, R356W, and P453S) were detected with CYP21-specific polymerase chain reaction (PCR) followed by mutation-specific methods, including allele-specific amplification, allele-specific restriction fragment length polymorphism (RFLP), or single-strand conformational polymorphism analyses (15). The G972R variant of IRS1 was assayed by gene-specific PCR followed by allele-specific RFLP. The PCR primers were forward (IRS-1F: 5=-CTTCTGTCAGGTGTCCATCC-3=) and reverse (IRS-1R: 5=-TGGCGAGGTGTCCACGTAGC-3=). Expected size for the PCR product was 262 base pairs (bp). After digestion of the PCR product with BstNI, the fragments were analyzed on a 4.5% agarose gel. The wild-type allele showed three bands with sizes of 158 bp, 81 bp, and 23 bp. The point mutation creates a BstNI site, such that the mutant allele showed four bands with sizes of 107 bp, 81 bp, 51 bp, and 23 bp (30). Positive control genomic DNA samples were included in all assays. Statistical Analysis Data were tabulated in 2 ⫻ 2 categoric tables [case/control ⫻ presence/absence of variant(s)] and analyzed with Fisher’s exact test because of the low frequencies of the mutations. Calculations were performed with commercial software (AbSTAT; Anderson-Bell, Boulder, CO). RESULTS Patients with PCOS were classified according to their baseline serum DHEAS levels, a biochemical marker for AA excess (9 –11). The upper normal limit of serum DHEAS levels was determined by the 95th percentile concentration of our healthy controls. As such, patients with a DHEAS level ⬎3,000 ng/mL were considered to have excessive AA secretion (n ⫽ 54), whereas those with DHEA levels ⬍2,500 ng/mL (n ⫽ 55) were considered to have normal AA secretion. To reduce possible misclassification of patients, those with intermediate serum DHEAS (i.e., 2,500 –2,999 ng/mL) Fertility and Sterility姞

were excluded (n ⫽ 5), with 109 PCOS patients remaining for analysis. Of 109 patients, 16 (14.5%) with PCOS were found to be heterozygous carriers of CYP21 mutations (Table 1). Mutations typically associated with NCAH (P30L, V281L, and P453S) were the most common mutations detected among all three groups. Of these 16, 10 (62.5%) had AA excess (i.e., elevated DHEAS levels). The frequency of the CYP21 mutations among PCOS patients with excessive AA secretion was 18.5%, and that of PCOS without AA excess was 10.9%. Of 95 healthy controls, 11 (11.5%) were heterozygous carriers of CYP21 mutations, for an allele frequency of 5.8%. None of the subjects was found to be homozygous for CYP21 mutations. There was no statistical difference in the heterozygote or variant allele frequencies between PCOS patients with AA excess vs. those without AA excess or healthy controls. None of the women with PCOS or the control subjects was homozygous for the G972R variant of the IRS1 gene. Nine women (16.7%) with PCOS and AA excess, six women (10.9%) with PCOS without AA excess, and six control women (6.3%) were heterozygous carriers of this variant (Table 2). The frequency of heterozygosity for the IRS1 variant among women with PCOS and AA excess compared with the healthy control women approached statistical significance (P⫽.052). Three women with PCOS and AA excess were heterozygous carriers for both a mutation in CYP21 and the IRS-1 variant. None of the women with PCOS without AA excess and none of the control women was a heterozygous carrier of both variants. The frequency of combined heterozygosity for CYP21 and IRS1 variants was significantly greater among women with PCOS and AA excess compared with both other groups (P⫽.046). DISCUSSION Dehydroepiandrosterone sulfate is a sulfated metabolite of an adrenal androgen precursor that seems to have minimal intrinsic biological activity. We used DHEAS concentrations 373

does not differ, what is the biological basis for the phenotypic differences between the manifesting and obligate carriers?

TABLE 2 Results of IRS1 and CYP21 genotype analyses.

Gene CYP21 IRS1 CYP21 and IRS1

PCOS with AA

PCOS without AA

Control

10 9 3a

6 6 0a

11 6 0a

Note: AA ⫽ adrenal androgen; PCOS ⫽ polycystic ovary syndrome. a The frequency of combined heterozygosity was significantly greater among women with PCOS and AA (P⫽.046). Witchel. CYP21 mutations and IRS1 variant in PCOS. Fertil Steril 2005.

to sort our patient population for several reasons: [1] ovarian steroidogenesis does not significantly contribute to circulating DHEAS concentrations, [2] DHEAS concentrations show heritability, and [3] higher DHEAS concentrations were found in brothers of women with PCOS (31). Use of this marker of excessive adrenal androgen secretion increased the homogeneity of our patient subgroups. Among women with PCOS and without AA excess, the frequency of heterozygosity (10.9%) was comparable to that among healthy control women (11.5%). These results are not surprising because women without AA excess are presumed to have increased ovarian androgen secretion. Although not statistically different, presumably owing to small sample size, the frequency of heterozygosity for CYP21 mutations was 18.5% among the women with PCOS and AA excess. Among this subset, the frequency of heterozygosity for CYP21 mutations was almost twice the frequency found in the group without AA excess and comparable to the frequency (20%) found in Spanish women with hirsutism (32). The frequency of heterozygosity for CYP21 mutations was significantly higher among these Spanish women with hirsutism compared with population controls (32). The frequency of heterozygosity for CYP21 mutations has been reported to be significantly greater in both American and Greek children with premature pubarche due to premature adrenarche and in American adolescent girls with hyperandrogenism (15, 33). Thus, there seems to be a subset of children with premature pubarche and women with PCOS in whom excessive adrenal androgen secretion contributes to the elevated androgen concentrations. Among this subset, the frequency of heterozygosity for CYP21 mutations is greater than among those without AA excess. The term “manifesting heterozygotic carrier” has been used to distinguish this subset of symptomatic CYP21 mutation carriers from asymptomatic carriers. Because the spectrum of specific CYP21 mutations 374

Witchel et al.

Insulin resistance/hyperinsulinemia plays a major role in the pathophysiology of PCOS. The G972R variant of the IRS1 gene has been associated with higher fasting insulin concentrations among women with PCOS (20) and lower SHBG concentrations in Spanish girls with a history of premature pubarche (25). Among offspring of patients with type 2 diabetes mellitus, G972R variant carriers manifest features of the metabolic syndrome, including decreased insulin sensitivity and greater serum concentrations of free fatty acids and triglycerides (34). Therefore, we ascertained the frequency of the G972R variant of the IRS1 gene in our patient population. Fifteen women (13.8%) with PCOS and six control women (6.3%) were heterozygous for the G972R variant of the IRS1 gene. Among our western Pennsylvania population, a trend toward increased frequency of this variant in children with premature pubarche and adolescent girls with hyperandrogenism compared with control subjects has been found (35). Three women with PCOS and AA were heterozygous carriers of CYP21 mutations and the IRS1 variant. Two women were heterozygous carriers of the P30L mutation, and the third carried the P453S mutation. In these three women, the molecular basis responsible for “manifesting heterozygosity” might be the combined heterozygosity for mild loss of function mutations in CYP21 and IRS1 (35–37). We speculate that the hyperinsulinemia/insulin resistance associated with the G972R variant stimulates adrenal steroidogenesis. In the presence of the CYP21 mutation with haploinsufficiency, adrenal androgen biosynthesis is increased. The combination of variants in genes involved in adrenal steroidogenesis and in genes involved in insulin signaling might thus promote excessive adrenal androgen biosynthesis (37). Our findings provide another example of genetic variations that influence the spectrum of both normal diversity and disease manifestations (38). The G972R variant of the IRS1 gene might represent a modifier locus among women who are heterozygous carriers of CYP21 mutations, potentially increasing their risk of developing AA excess in PCOS. Nonetheless, overall, the IRS1 variant and CYP21 mutations studied seem to play a limited role in the development of PCOS in the population studied. REFERENCES 1. Azziz R, Woods KS, Reyna R, Key TJ, Knochenhauer ES, Yildiz BO. The prevalence and features of the polycystic ovary syndrome in an unselected population. J Clin Endocrinol Metab 2004;89:2745–9. 2. Knochenhauer ES, Key TJ, Kahsar-Miller M, Waggoner W, Boots LR, Azziz R. Prevalence of the polycystic ovary syndrome in unselected

CYP21 mutations and IRS1 variant in PCOS

Vol. 83, No. 2, February 2005

3.

4. 5. 6.

7. 8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

black and white women of the southeastern United States: a prospective study. J Clin Endocrinol Metab 1998;83:3078 – 82. Zawadski J, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Dunaif A, Givens JR, Haseltine FP, Merriam GR, eds. Polycystic ovary syndrome. Oxford: Blackwell Science, 1992:377– 84. Azziz R. Androgen excess is the key element in polycystic ovary syndrome. Fertil Steril 2003;80:252– 4. Kahsar-Miller M, Azziz R. Heritability and the risk of developing androgen excess. J Steroid Biochem Mol Biol 1999;69:261– 8. Urbanek M, Legro RS, Driscoll DA, Azziz R, Ehrmann DA, Norman RJ, et al. Thirty-seven candidate genes for polycystic ovary syndrome: strongest evidence for linkage is with follistatin. Proc Natl Acad Sci U S A 1999;96:8573– 8. Legro RS, Strauss JF. Molecular progress in infertility: polycystic ovary syndrome. Fertil Steril 2002;78:569 –76. Brüning JC, Winnay J, Bonner-Weir S, Taylor SI, Accili D, Kahn CR. Development of a novel polygenic model of NIDDM in mice heterozygous for IR and IRS-1 null alleles. Cell 1997;88:561–72. Wild RA, Umstot ES, Andersen RN, Ranney GB, Givens JR. Androgen parameters and their correlation with body weight in one hundred thirty-eight women thought to have hyperandrogenism. Am J Obstet Gynecol 1983;146:602– 6. Hoffman DI, Klove K, Lobo RA. The prevalence and significance of elevated dehydroepiandrosterone sulfate levels in anovulatory women. Fertil Steril 1984;42:76 – 81. Carmina E, Koyama T, Chang L, Stanczyk FZ, Lobo RA. Does ethnicity influence the prevalence of adrenal hyperandrogenism and insulin resistance in polycystic ovary syndrome? Am J Obstet Gynecol. 1992;167:1807–12. Azziz R. Abnormalities of adrenocortical steroidogenesis in the polycystic ovary syndrome. In: Azziz R, Nestler JE, Dewailly D, eds. Androgen excess disorders in women. Philadelphia: Lippincott-Raven, 1997:403–14. Jaquish CE, Blangero J, Haffner SM, Stern MP, Maccluer JW. Quantitative genetics of dehydroepiandrosterone sulfate and its relation to possible cardiovascular disease risk factors in Mexican Americans. Hum Hered 1996;46:301–9. An P, Rice T, Gagnon J, Hong Y, Leon AS, Skinner JS, et al. Race differences in the pattern of familial aggregation for dehydroepiandrosterone sulfate and its responsiveness to training in the HERITAGE family study. Metabolism 2001;50:916 –20. Witchel SF, Lee PA, Suda-Hartman M, Hoffman EP. Hyperandrogenism and manifesting heterozygotes for 21-hydroxylase deficiency. Biochem Mol Med 1997;62:151– 8. Ostlere LS, Rumsby G, Holownia P, Jacobs HS, Rustin MH, Honour JW. Carrier status for steroid 21-hydroxylase deficiency is only one factor in the variable phenotype of acne. Clin Endocrinol 1998;48: 209 –15. Blanche H, Vexiau P, Clauin S, Le Gall I, Fiet J, Mornet E, et al. Exhaustive screening of the 21-hydroxylase gene in a population of hyperandrogenic women. Hum Genet 1997;101:56 – 60. Knochenhauer ES, Cortet-Rudelli C, Cunnigham RD, Conway-Myers BA, Dewailly D, Azziz R. Carriers of 21-hydroxylase deficiency are not at increased risk for hyperandrogenism. J Clin Endocrinol Metab. 1997;82:479 – 85. Hines GA, Smith ER, Azziz R. Influence of insulin and testosterone on adrenocortical steroidogenesis in vitro: preliminary studies. Fertil Steril 2001;76:730 –5. El Mkadem SA, Lautier C, Macari F, Molinari N, Lefebvre P, Renard E, et al. Role of allelic variants Gly972Arg of IRS-1 and Gly1057Asp of IRS-2 in moderate-to-severe insulin resistance of women with polycystic ovary syndrome. Diabetes 2001;50:2164 – 8.

Fertility and Sterility姞

21. Jellema A, Zeegers MP, Feskens EJ, Dagnelie PC, Mensink RP. Gly972Arg variant in the insulin receptor substrate-1 gene and association with type 2 diabetes: a meta-analysis of 27 studies. Diabetologia 2003;46:990 –5. 22. Baroni MG, Arca M, Sentinelli F, Buzzetti R, Capici F, Lovari S, et al. The G972R variant of the insulin receptor substrate-1 (IRS-1) gene, body fat distribution and insulin-resistance. Diabetologia 2001;44:367–72. 23. Stumvoll M, Fritsche A, Volk A, Stefan N, Madaus A, Maerker E, et al. The Gly972Arg polymorphism in the insulin receptor substrate-1 gene contributes to variation in insulin secretion in normal glucose-tolerant humans. Diabetes 2001;50:882–5. 24. Federici M, Hribal ML, Ranalli M, Marselli L, Porzio O, Lauro D, et al. The common Arg972 polymorphism in insulin receptor substrate-1 causes apoptosis of human pancreatic islets. FASEB J 2001;15:22– 4. 25. Ibáñez L, Marcos MV, Potau N, White C, Aston CE, Witchel SF. Increased frequency of the G972R variant of the insulin receptor substrate-1 (IRS-1) gene among girls with a history of precocious pubarche. Fertil Steril 2002;78:1288 –93. 26. Azziz R, Bradley EL Jr, Potter HD, Parker CR Jr, Boots LR. Chronic hyperinsulinemia and the adrenal androgen response to acute corticotropin-(1-24) stimulation in hyperandrogenic women. Am J Obstet Gynecol 1995;172:1251– 6. 27. Boots LR, Potter S, Potter D, Azziz R. Measurement of total serum testosterone levels using commercially available kits: high degree of between-kit variability. Fertil Steril 1998;69:286 –92. 28. Pearlman WH, Crepy O. Steroid-protein interaction with particular reference to testosterone binding by human serum. J Biol Chem 1967; 242:182–9. 29. Lahiri DK, Schnabel B. DNA isolation by a rapid method from human blood samples: effects of MgCl2, EDTA, storage time, and temperature on DNA yield and quality. Biochem Genet 1993;31:321– 8. 30. Hitman GA, Hawrami K, McCarthy MI, Viswanathan M, Snehalatha C, Ramachandran A, et al. Insulin receptor substrate-1 gene mutations in NIDDM: implications for the study of polygenic disease. Diabetologia 1995;38:481– 6. 31. Legro RS, Kunselman AR, Demers L, Wang SC, Bentley-Lewis R, Dunaif A. Elevated dehydroepiandrosterone sulfate levels as the reproductive phenotype in the brothers of women with polycystic ovary syndrome. J Clin Endocrinol Metab 2002; 87:2134 – 8. 32. Escobar-Morreale HF, San Millan JL, Smith RR, Sancho J, Witchel SF. The presence of the 21-hydroxylase deficiency carrier status in hirsute women: phenotype-genotype correlations. Fertil Steril 1999; 72:629 –38. 33. Dacou-Voutetakis C, Dracopoulou M. High incidence of molecular defects of the CYP21 gene in patients with premature adrenarche. J Clin Endocrinol Metab 1999;84:1570 – 4. 34. Marini MA, Frontoni S, Mineo D, Bracaglia D, Cardellini M, De Nicolais P, et al. The Arg972 variant in insulin receptor substrate-1 is associated with an atherogenic profile in offspring of type 2 diabetic patients. J Clin Endocrinol Metab 2003;88:3368 –71. 35. Witchel SF, Smith R, Tomboc M, Aston CE. Candidate gene analysis in premature pubarche and adolescent hyperandrogenism. Fertil Steril 2001;75:724 –30. 36. Witchel SF, Lee PA. Identification of heterozygotic carriers of 21hydroxylase deficiency: sensitivity of ACTH stimulation tests. Am J Med Genet 1998;76:337– 42. 37. Witchel SF, Aston CE. The role of heterozygosity for CYP21 in PCOS. J Pediatr Endocrinol Metab 2000;85:4013– 8. 38. Weiss KM. Genetic variation and human disease. Principles and evolutionary approaches. Cambridge, United Kingdom: Cambridge University Press, 1995.

375