A Novel Nonstop Mutation in the Stop Codon and a Novel Missense ...

0013-7227/02/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 87(6):2556 –2563 Copyright © 2002 by The Endocrine Society

A Novel Nonstop Mutation in the Stop Codon and a Novel Missense Mutation in the Type II 3␤Hydroxysteroid Dehydrogenase (3␤-HSD) Gene Causing, Respectively, Nonclassic and Classic 3␤-HSD Deficiency Congenital Adrenal Hyperplasia SONGYA PANG, WEIHUA WANG, BARRY RICH, RAPHAEL DAVID, YING TAI CHANG, GOLDY CARBUNARU, SUSAN E. MYERS, A. FORBES HOWIE, KAREN J. SMILLIE, AND J. IAN MASON Department of Pediatrics (S.P.,W.W., Y.T.C., G.C.), College of Medicine, University of Illinois, Chicago, Illinois 60612; Department of Pediatrics (B.R.), School of Medicine, University of Chicago, Chicago, Illinois 60637; New York University (R.D.), New York, New York 10016; Department of Pediatrics (S.E.M.), St. Louis University, St. Louis, Missouri 63104; Department of Reproductive and Developmental Sciences (Clinical Biochemistry Section) (A.F.H., K.J.S., J.I.M.), University of Edinburgh, EH3 9YW Edinburgh, Scotland, United Kingdom We investigated two novel point mutations in the human type II 3␤-hydroxysteroid dehydrogenase (3␤-HSD) gene causing a mild and a severe form of 3␤-HSD deficiency congenital adrenal hyperplasia. The first is a nonstop mutation in the normal stop codon 373 of the gene in exon IV [TGA (Stop) 3 TGC (Cys) ⴝ Stop373C) identified from one allele of a female child with premature pubarche whose second allele had an E142K mutation. The Stop373C mutation predictably results in an open reading frame and a mutant-type (MT) II 3␤-HSD protein containing 467 amino acid residues, compared with the 372 amino acid residues of wild-type (WT) protein. The second is a homozygous missense mutation in codon 222 [CCA (Pro) 3 ACT (Thr) ⴝ P222T] in the gene identified from a female neonate with salt-wasting disorder. The pcDNA vectors containing the constructs of WT II 3␤-HSD cDNA, WT cDNA with the open reading frame (WT cDNAⴙ), MT Stop373C with the open reading frame (Stop373Cⴙ) and MT P222T cDNA were transfected in COS-I and 293T cells and expressed a similar amount of 3␤-HSD mRNA. The enzyme activity in intact cells using pregnenolone and dehydroepiandrosterone as substrate in the medium (1 ␮mol/liter) was identical between the WT cDNA and the WT cDNAⴙ, but was decreased to 27% of the WT enzymes at 6 h by MT Stop373Cⴙ enzyme, and was undetectable

3

␤-HYDROXYSTEROID dehydrogenase/⌬5 3⌬4 isomerase (3␤-HSD) catalyzes the conversion of ⌬5-3␤-hydroxysteroids to ⌬4-3␤ keterosteroids and is essential for the formation of progesterone (P), the precursor hormone for aldosterone, 17-OH progesterone (17-OHP), the precursor hormone for cortisol (F) in the adrenal cortex, as well as androstenedione (⌬4-A), T, and estrogen in the adrenal cortex and gonads (1–3). In humans, 3␤-HSD expression in the adrenals and gonads is under the control of type II-3␤-HSD gene (4). The deficient activity of type II 3␤-HSD enzyme in the adrenal cortex is a well-known cause of congenital adAbbreviations: CAH, Congenital adrenal hyperplasia; DHEA, dehydroepiandrosterone; F, cortisol; HPO, hypothalamic-pituitary-ovarian; 3␤-HSD, 3␤-hydroxysteroid dehydrogenase; MT, mutant-type; 17-OHP, 17-OH progesterone; P, progesterone; WT, wild-type.

by P222T enzyme. In the homogenates of the cells, both MT Stop373Cⴙ and P222T enzyme activities and enzymes were undetectable despite clear detection of WT enzyme activities and WT enzymes. LH response to an LHRH analog stimulation in the pubertal female with the Stop373C/E142K genotypes and in a pubertal female with compound 273/318 frameshift genotypes were comparable to and higher than control females, respectively. In conclusion, a structurally lengthy MT II 3␤-HSD enzyme due to a nonstop mutation was relatively detrimental in intact cells causing the nonclassic phenotype of 3␤-HSD deficiency. A missense P222T mutation was seriously detrimental, causing the classic phenotype of 3␤-HSD deficiency. The undetectable Stop373C and P222T enzymes on Western blottings, together with the respective in vivo and in vitro data, suggest that a relative instability of Stop373C enzyme and a profound instability of the P222T enzyme are likely the detrimental molecular mechanisms. The increased LH in the female with the frameshift genotype and the appropriate LH response in the female with the nonstop genotype correlated with predictably severe and mild ovarian type II 3␤-HSD deficiency, respectively. (J Clin Endocrinol Metab 87: 2556 –2563, 2002)

renal hyperplasia (CAH) (5, 6) associated with salt-wasting disorder in the severely deficient form and nonsalt-wasting disorder in the less severely deficient form. Simultaneous fetal testicular type II 3␤-HSD deficiency results in incomplete virilization of the external genitalia in males. In females, mild virilization of external genitalia may be present in some by the effect of androgen metabolite of excess dehydroepiandrosterone (DHEA) secreted from the affected fetal adrenals. Mild variant of 3␤-HSD deficiency CAH manifests in subtle clinical symptoms of androgen excess in children and adults (6 –9). Classic salt-wasting and nonsalt-wasting 3␤-HSD deficiency CAH have been reported to result from either a deleterious homozygous or compound heterozygous mutations in the structure of type II-3␤-HSD gene such as frameshift,

2556

Pang et al. • Novel Nonstop Mutation in the Stop Codon

J Clin Endocrinol Metab, June 2002, 87(6):2556 –2563 2557

premature stop codon, or substitution of an amino acid residue (6, 10). However, information on the genetic basis of mild nonclassic variant of 3␤-HSD deficiency CAH is limited. Only a few subjects with clinically nonclassic presentation of the disorder were reported to have a missense mutation in coding region of the gene at least on one allele (6, 10, 11). Further molecular studies from 3␤-HSD deficiency patients with various spectra of the disorder would advance the understanding of the gene function in relation to the gene structure as well as the molecular basis for various phenotypes of the disorder. In this report, we characterized two additional novel point mutations in the type II 3␤-HSD gene resulting in a nonclassic variant and classic 3␤-HSD deficiency CAH and elaborated a possible mechanism for the detrimental 3␤-HSD activity due to the novel point mutation. In addition, clinical and hormonal phenotypes on the maturation of hypothalamic-pituitary-ovarian (HPO) axes were compared with type II 3␤-HSD genotypes in two pubertal females with 3␤-HSD deficiency CAH. Materials and Methods The study was approved by the Institutional Review Board of the University of Illinois at Chicago and was conducted after informed consent from the parents of minor subjects or from the subjects.

Subjects Subject 1. The patient of family 1 is a Caucasian female child born at term from a G1P1 mother with unremarkable prenatal history. There was no family history of consanguinity. At birth, the infant had normal female external genitalia and neonatal history was unremarkable except for a positive result from a newborn screening 17-OHP level obtained on d 3. The neonate had no symptoms and was thriving appropriately, and random serum electrolytes and CO2 levels were unremarkable (Table 1). Hormonal work-up at that time revealed appropriately normal basal and ACTH stimulated serum F levels (Table 1). The baseline serum 17-OHP level was slightly elevated for age (reference value: ⬍ 15 nmol/ liter) but did not rise excessively in response to ACTH stimulation excluding 21-hydroxylase deficiency CAH (Table 1). The baseline and ACTH stimulated serum 17-OH pregnenolone (⌬5-17P) levels were

thought to be inconclusive for diagnosing 3␤-HSD deficiency CAH due to the fact that control neonatal reference values were not available at that time. A bone age x-ray of the child obtained at chronological age 3 yr was reported to be delayed by 0.5–1 yr. The child’s medical history thereafter was unremarkable until age 5.5 yr when the child began to develop premature sexual hair growth. Bone age x-ray at that time was reported to be compatible with the chronological age. Between ages 5–7 yr, the child’s height percentile increased gradually from 25th to 60th in both height and weight. At age 7 yr, the child had no symptom of salt craving nor easy fatigability nor hyperpigmentation. The child developed a few facial acnes and body odor. At age 7.3 yr, the child’s physical examination revealed height 124.5 cm (60th percentile), weight 25 kg (60th percentile), blood pressure 97/61, pulse rate 72/min, Tanner stage III pubic hair with a few axillary hairs, some facial acne and no development of breast tissue. The remaining physical examination was unremarkable including external genitalia. Baseline hormonal evaluation at 0900 h revealed mildly elevated ⌬5-17P level (reference value: ⬍ 7 nmol/liter for pubic hair stage II–III development) and moderately elevated DHEA (reference value: ⬍14.5 nmol/liter), and DHEA-sulfate (DHEA-S) levels (ref value ⬍ 5.7 nmol/liter), whereas F level was in the low normal range (reference value: ⬎ 0.24 ␮mol/liter) (Table 1). Baseline serum17-OHP, ⌬4–A, and T levels were slightly elevated or appropriate for pubic hair stage III development. Simultaneous PRA and serum aldosterone levels were appropriately in the normal range (Table 1). Hormonal levels 1 h following ACTH stimulation (250 ␮g iv bolus of Cortrosyn) revealed unequivocally elevated ⌬5–17P (reference value: ⬍ 25.5 nmol/liter) and DHEA (ref value: ⬍26 nmol/liter) levels, whereas F response was slightly low (reference value ⬎ 0.57 ␮mol/liter) (Table 1). ACTH stimulated 17-OHP, ⌬4–A, and T rose appropriately or did not rise (Table 1). These data confirmed the diagnosis of nonsalt-wasting 3␤-HSD deficiency CAH in the patient. At age 71⁄2 yr, glucocorticoid replacement therapy was begun with 15 mg hydrocortisone/d in three divided doses, resulting in suppression of AM basal ⌬5-17P (0.99 – 6.4 nmol/liter), and DHEA-S (0.24 ␮mol/liter) levels. With the therapy, height percentile declined from 60th to 50th by age 9 yr. The patient subsequently developed thelarche at age 9 yr. At age 10 yr, the patient had an LHRH analog stimulation test by administering leuprolide acetate 20 ␮g/kg of weight sc to assess changes of HPO axes. Blood samples were obtained before and then 1, 3, 6, and 12 h following leuprolide administration for determination of LH, FSH, and/or E2 levels. For a type II 3␤-HSD gene study, blood samples were obtained from the patient and her parents. Subject 2. The patient of family 2 is a female infant born from a G1 P1 mother with unremarkable prenatal history. Her parents were of Eastern

TABLE 1. Biochemical data of patient from subjects 1 and 2‘ Subject 1

Age at study

3 wk Baseline level

Hormones

⌬5-17P (nmol/liter) DHEA (nmol/liter) DHEA-S (␮mol/liter) 17-OHP (nmol/liter) ⌬4-A (nmol/liter) T (nmol/liter) F (␮mol/liter) S (nmol/liter) Aldo (nmol/liter) Renin (ng/liter䡠sec) Na (meq/liter) K (meq/liter) CL (meq/liter) CO2 (mm/liter)

Subject 2

PT

180 37.3

Age reference range

(1.9 –31) (0.7–25)

19.1 1.75

(1.7– 4.2) (024 –2.5)

0.39 9.1

(0.06 – 0.51) –

134 5.5 100 25

7.3 yr

ACTH stimulated levels PT

760 72.4 – 35 2.33 – 0.75 16.6 – – – – – –

Age reference range

(20 –150) (6 –112) (2.8 – 8) (0.7– 4.6) (0.54 – 4.2) –

Baseline level PT

29.5 32.1 10.9 5 3.7 1.07 0.29 – 0.41 1.1– 4.4 – – – –

Age reference range

(0.4 –5.1) (0.18 – 4.9) (0.2–1.8) (0.21– 4.2) (0.4 –11) (0.1– 0.5) (0.15– 0.6) (0.1–7)

4 wk

ACTH stimulated levels PT

754 90.9 9.8 12.5 3.87 0.76 0.5 – – – – – – –

Age reference range

(0.81–21) (0.4 –11) (0.2– 4) (3.8 –17) (0.2–1.7) (0.1– 0.6) (0.54 –1.2)

Baseline level PT

1102 385.5 – 184 48 3.1

Age reference range

(1.9 –31) (0.7–25) (1.7– 4.2) (0.24 –2.5) (0.06 – 0.51)

– – 143 127 7

(0.1–9)

ACTH stimulation, 1 h after 0.25 mg Cortrosyn IV Bolus; ⌬5-17P, 17-hydroxypregnenolone (to ng/dl ⫻ 33.3); 17-OHP, 17-hydroxyprogesterone (to ng/dl ⫻ 33.3); F, Cortisol (to ␮g/dl ⫻ 36); PT, patient indicated above; DHEA, dehydroepiandrosterone (to ng/dl ⫻ 28.6); ⌬4-A, Androsteridione (to ng/dl ⫻ 28.6); Aldo, Aldosterone (to ng/dl ⫻ 37); DHEA-8, DHEA Sulfate (to ␮g/dl ⫻ 37); T, Testosterone (to ng/dl ⫻ 26); Renin (to ng/ml/hr ⫻ 3.61).

2558

J Clin Endocrinol Metab, June 2002, 87(6):2556 –2563

European Jewish ethnicity, and there was no family history of consanguinity. The infant had normal female external genitalia at birth but developed salt-wasting adrenal crises at age 4 wk (Table 1). Hormonal work-up at this time revealed highly elevated random basal serum ⌬5-17P (reference value: ⬍31 nmol/liter) and DHEA (reference value: ⬍24.6 nmol/liter) levels, whereas serum 17-OHP (reference value: ⬍3.8 nmol/liter) level was moderately elevated (Table 1). Serum ⌬4-A level was markedly elevated (reference value: ⬍2.55 nmol/liter), and serum T level was moderately elevated (reference value: ⬍0.52 nmol/liter) (Table 1). Initially the infant was diagnosed with 21-hydroxylase deficiency CAH and glucocorticoid and mineralocorticoid replacement therapy was initiated. However, because of the absence of genital virilization, 3␤-HSD deficiency CAH was subsequently suspected. Blood samples were obtained from the patient and her parents for type II 3␤-HSD gene study. Additional subject studied. A 14-yr-old female with salt-wasting 3␤-HSD deficiency CAH due to a compound heterozygous frameshift mutation resulting from a nucleotide deletion in codon 318 on one allele and two nucleotide deletions in codon 273 on the other allele of type II 3␤-HSD gene has been treated with glucocorticoid and mineralocorticoid replacement therapy since early infancy (12). At age 12 yr, breast development progressed to Tanner II and pubic hair to Tanner IV stage with axillary hair. Random LH and FSH levels were 6.4 and 5.5 mIU/ml, respectively and random E2 level was between Tanner I and II levels (94 pmol/liter). At age 13 yr, the patient was treated with 25 mg hydrocortisone/d in three divided doses and Florinef 0.15 mg taken orally daily and had her first vaginal spotting. Her menses were initially either spotting alone or menometrorrhagia lasting 3 d every 1–2 months, followed by secondary ammenorhea for the next 8 months. At age 14 yr, the patient’s examination revealed some acne on the face and upper back with no apparent hirsutism, Tanner V pubic and axillary hair, and Tanner II breasts. The patient had an LHRH analog stimulation test by administering 20 ␮g of leuprolide acetate per kg of weight sc. Blood samples were obtained before and 1, 3, 6, and 12 h after the injection for determination of LH, FSH, and or E2 levels.

Laboratory methods PCR and sequencing of the type II 3␤-HSD gene. PCR and sequencing of the gene were performed using primers previously reported to examine the sequence of the regions of a putative promoter from ⫺1053 base, exons I, II, III, and IV, and all exon and intron boundaries of the type II 3␤-HSD gene as reported previously (13, 14). Site-directed mutagenesis. Full-length wild-type (WT) type II 3␤-HSD cDNA cloned in a pcDNA-1/neo vector was cut by XbaI and XhoI and ligated in the pcDNA-3.1 vector (Stratagene, La Jolla, CA). The technique of site-directed mutagenesis by Quikchange site-directed mutagenesis kit (Stratagene) was employed to generate the P222T mutations using oligonucleotide primers of sense 5⬘-CTCTACAGTCAACACAGTCTATGTTGGCAACGTGGCC-3⬘ and antisense 5⬘-GGCCACGTTGCCAACATAGACTGTGTTGACTGTACTAG-3⬘, which contained a mutation on the target sequence site of the reported type II 3␤-HSD gene structure (15), as was previously reported (14). For the construct of the mutanttype (MT) Stop373C continuing WT exon IV sequence up to nucleotide no. 7756 of the gene (the open reading frame of 95 additional amino acid residues), ending with a stop codon sequence (MT Stop373C⫹), the no. 1 subject’s genomic type II 3␤-HSD DNA containing the Stop373C mutation was used as the template. The patient’s exon IV PCR product was digested with Xba-1 and BamH-1 enzymes to generate a PCR DNA piece containing the sequence from the 5⬘ end of exon IV to nucleotide no. 7756 of the gene with XbaI and BamHI stick ends. A BamHI restriction site in the native pcDNA vector was eliminated by a site-directed mutagenesis. The WT II 3␤-HSD cDNA in the pcDNA 3.1 vector was then digested with XbaI and BamHI enzymes to generate a plasmid piece with XbaI and BamHI stick ends. Both pieces of the patient’s PCR DNA and WT II plasmid DNA with XbaI and BamHI stick ends were combined for ligation. The ligated products were used for transformation to generate the plasmids. All colonies contained the mutant Stop373C⫹ DNA. The constructs of WT II 3␤-HSD cDNA⫹ containing the normal stop codon 373 followed by the sequence of exon 1V up to nucleotide no. 7756 was developed by a site-directed mutagenesis from the mutated Stop373C⫹


DNA. The plasmids containing WT cDNA, WT cDNA⫹, MT Stop373C⫹, and MT P222T cDNA were confirmed by double-strand DNA sequence for its authenticity. Expression of WT II 3␤-HSD cDNA, WT II cDNA⫹, MT Stop373C⫹, MT P222T cDNA. pcDNA 3.1 3␤-HSD expression vectors containing the entire type II 3␤-HSD coding region of WT cDNA, WT cDNA⫹, MT 373C⫹, and MT P222T cDNA were purified for transfection by QIAGEN plasmid Midi kit (QIAGEN). COS-1 or 293T cells were grown in DMEM (Life Technologies, Inc., Gaithersburg, MD) plus 10% FBS and 1% antibacterial antibiotic solution (Sigma, St. Louis, MO) and transfected with 1 ␮g plasmid DNA per 3 ␮l transfection reagent using Lipofectamine (Life Technologies, Inc.). Transfection efficiency tested by a ␤-galactosidase activity assay using a ␤-galactosidase vector (cytomegalovirus-␤) plasmid was 26% on average in two separate experiments. The presence of type II 3␤-HSD mRNA from the harvested cells transfected for 48 h with WT and MT II 3␤-HSD cDNAs and cDNA⫹ was determined by RT-PCR. The translated type II 3␤-HSD protein was determined by Western immunoblot analysis by harvesting the cells, as was reported (14). Assay of WT and MT II 3␤-HSD enzyme activity. 3␤-HSD enzyme activity in the intact transfected cells was determined by adding 1 ␮ mol/liter of pregnenolone (⌬5-P) or DHEA to the medium as was previously reported (14). 3␤-HSD activity for various concentrations of ⌬5-P and DHEA was performed at 37 C in a total reaction volume of 0.5 ml consisting of 100 ␮g total protein, 50 mmol/liter Tris, 1 mmol/liter NAD⫹, radiolabeled substrate (0.025 ␮Ci [3 H] ⌬5-P, 0.05 ␮Ci [3 H] ⌬5-P, or 0.05 ␮Ci [14C]DHEA), and the desired concentration of the relevant unlabeled substrate. The reaction was stopped by the addition of dichloromethane, and after extraction, steroids were separated by thin layer chromatography and quantified by liquid scintillation spectrometry as reported previously (14). Serum steroid concentrations were determined by a previously described RIA (16). Serum LH and FSH concentrations were determined by a previously described sensitive immunoradiometric assay (14).

Results Type II 3␤-HSD gene sequence (Figs. 1 and 2)

Auto-sequencing of the PCR products of subject no. 1 with nonclassic presentation of 3␤-HSD deficiency causing premature sexual hair growth and mild growth acceleration in childhood revealed a single nucleotide substitution mutation in one allele from A to C at nucleotide no. 7468 of the gene in exon IV, resulting in a nonstop mutation in the normally stop codon 373 [TGA (stop) 3 TGC (CYS) ⫽ Stop373C] of the gene and in the second allele, a missense mutation in codon 142 [CAA (Glu) 3 AAA (Lys) ⫽ E142K] in exon IV of the gene (Fig. 1). Remaining sequences of all exons, exon, and intron boundaries as well as sequences of the putative promoter region were normal. The possibility of PCR artifacts resulting in the mutation in exon IV of both alleles of the patient was excluded by confirmation of the same findings in the repeat PCR products and by proof of the Stop373C mutation in one allele of the mother and E142K mutation in one allele of the father (not shown on Fig. 1). These data confirmed that the compound heterozygous Stop373C and E142K mutations in the patient were inherited by a recessive trait. The Stop373C mutation would result predictably in an open reading frame of 95 additional amino acid residues, resulting in a mutant enzyme with 467 amino acids residues as compared with the WT II 3␤-HSD protein with 372 amino acid residues. Auto-sequencing of PCR products of the subject no. 2 with classic presentation of salt-wasting 3␤-HSD deficiency revealed a single nucleotide substitution at nucleotide no. 7013 of the gene in exon IV in both alleles, resulting in a homozygous missense mutation in the codon



FIG. 1. Partial nucleotide sequence of the sense strand (allele 1) and antisense strand (allele 2) of exon IV of type 3␤HSD gene from the subject 1 with nonclassic presentation of 3␤-HSD deficiency CAH, as determined by auto-sequencing of the PCR products and depicting the heterozygous point mutation involving codon 142 and stop codon 373. Allele 1 was inherited from the father and allele 2 was inherited from the mother. In allele 1 of the patient’s sequence (PT’s sequence) there was a nucleotide substitution from WT sequence G to A in codon 142 [indicated by an upward arrow (1)] resulting in the substitution of WT Glu (GAA) residue by Lys (AAA) ( ⫽ E142). In allele 2 of the PT’s sequence, there was a nucleotide substitution from the WT stop codon TGA to TGC [indicated by an upward arrow (1)] resulting in a nonstop codon by a Cys residue ( ⫽ Stop373C). The region of the codon 142 (allele 1) and stop codon mutations (allele 2) is indicated by an arrow on the schematic of the type II 3␤-HSD gene map. The sequenced region of the gene included the promoter (⫺1053 to ⫺1 bp), exons I, II, III, IV, and all exon and intron boundaries.

FIG. 2. Partial nucelotide sequence of the sense strand of exon IV from the subject 2 with classic salt-wasting presentation of 3␤-HSD deficiency CAH, as determined by auto-sequencing of the PCR products depicting a homozygous missense mutation in the codon 222. There was a nucleotide substitution indicated by upward arrow (1) from the WT sequence CCA (Pro) to ACA (Thr) ( ⫽ P222T). The region of the codon 222 mutation is indicated by an arrow on the autosequencing of exon IV (2) as well as on the schematic of the type II 3␤- HSD gene map (2). The sequenced region of the gene included the promoter (⫺1053 to ⫺1bp), exons I, II, III, IV, and all exon and intron boundaries.

222 (CAC (Pro) 3 ACA (Thr) ⫽ P222T) (Fig. 2). The remaining sequences of all exons and exon and intron boundaries as well as sequences of regions of the putative promoter were normal. A repeat PCR product from the subject revealed an identical sequence finding as well as PCR products of exon IV from the subject’s parents revealed P222T mutation in one allele of both parents (not shown of Fig. 2) confirming the homozygous inherited mutation in the patient. RT-PCR analysis of type II 3␤-HSD RNA from the cells transfected with WT and MT II 3␤-HSD cDNA and cDNA⫹ (Fig. 3)

An autoradiogram of the RT-PCR products revealed a similar band intensity of type II 3␤-HSD mRNA from the

cells transfected with WT II 3␤-HSD cDNA, WT II 3␤-HSD cDNA⫹, MT Stop373C⫹ DNA, and MT P222T cDNA. The control mock vector containing no type II 3␤-HSD cDNA or cDNA⫹ generated no mRNA. These findings confirmed the presence of transfected WT and MT II 3␤-HSD genes in the COS-I cells. Activities of the MT II 3␤-HSD Stop373C⫹ and P222T proteins compared with the activities of the WT II 3␤-HSD proteins in intact COS-I cells (Fig. 4)

The time course and amount of P formation from ⌬5 -P and of ⌬4-A formation from DHEA in intact COS-I cells are depicted in Fig. 4. The percentages of P formed after 1, 2, 3, 4, and 6 h of incubation with ⌬5-P substrate in cells transfected

2560



or cDNA⫹ (76 ⫾ 10 pmol/min䡠mg protein; n ⫽ 4; two homogenates of each plasmid) and no detectable activity distinguishable from basal activity in the homogenates of the cells transfected with the Stop373C⫹ and P222T DNAs (⬍3 pmol/min䡠mg protein). Western blot analysis of WT and MT II 3␤-HSD proteins (Fig. 5)

WT II 3␤-HSD proteins from the cells transfected with the WT cDNA and WT cDNA⫹ were clearly detectable. However, MT Stop373C⫹ and MT P222T as well as control unstable MT L6F protein were repeatedly undetectable. Comparison between phenotype maturation of hypothalamic-pituitary-ovarian axes and type II 3␤-HSD genotype in pubertal females with 3␤-HSD deficiency CAH (Fig. 6) FIG. 3. An autoradiogram of the RT-PCR analysis of the type II 3␤HSD mRNA from COS-I cells transfected with pcDNA 3.1 vectors containing no cDNA or no cDNA⫹ (Mock vector); WT II 3␤-HSD cDNA, WT cDNA continuing with additional WT genome sequences up to nucleotide no. 7756 of the gene ending with a stop codon sequence (WT II cDNA⫹), MT nonstop codon 373C sequence continuing with additional WT genomic sequence up to no. 7756 of the gene ending with a stop codon sequence (MT Stop373C⫹), and MT P222T cDNA. A similar amount of the type II 3␤-HSD mRNA via RT-PCR was detected in the lanes of the cells transfected with WT and MT II 3␤-HSD cDNAs and cDNA⫹s and absent mRNA in the lane of the control mock vector containing no cDNA or cDNA⫹.

with pcDNA3.1 WT II 3␤-HSD cDNA were 28%, 48%, 50%, 56%, and 62%, respectively, above the values of the control mock pcDNA vector. This finding was nearly identical to the percentages of P formed at the same incubation time points in cells transfected with pcDNA WT II cDNA⫹, which were 30%, 51%, 56%, 59%, and 61%, respectively, above the values of the control vector. In cells transfected with pcDNA Stop373C⫹, the percentages of P formed at the same incubation time points were 3%, 7%, 8%, 12% and 16%, respectively, above the values of the control vector. In cells transfected with pcDNA P222T and mock control vector, no enzyme activity was detected at any time point. The percentages of ⌬4-A formed at 2, 4, 5, and 6 h of incubation with DHEA substrate in cells transfected with pcDNA 3.1 WT II 3␤-HSD cDNA were 24%, 35%, 42% and 43%, respectively, above the values of the control mock pcDNA vector. In cells transfected with pcDNA WT II cDNA⫹, the percentages of ⌬4-A formed at the same time incubation points with DHEA substrate were 29%, 35%, 39%, and 45%, respectively, above the control mock vector. In cells transfected with pcDNA MT Stop373C⫹, the percentages of ⌬4-A formed at the same time incubation points with DHEA substrate were 4%, 9%, 10%, and 12%, respectively, above the control vector values. In cells transfected with pcDNA P222T and control mock vector, no enzyme activity was detected at any time point. Similar results were obtained when 3␤-HSD activities were assayed in intact 293T cells after transfection with the various plasmids as in the COS-I cells (data not shown). 3␤-HSD activity in the homogenates of COS-I cells in the presence of NAD⫹ revealed detectable activity only in the homogenates of cells transfected with WT II 3␤-HSD cDNA

The 10-yr-old pubertal female (subject 1) with the compound heterozygous Stop373C and E142K genotype had Tanner breast stage II and Tanner pubic hair stage III development. Baseline LH level and LH response to leuprolide stimulation in the patient were comparable to those of the control female with Tanner II-III breast development (17) (Fig. 6). Basal FSH level (1.87 mIU/ml) and peak FSH response to leuprolide stimulation (16 mIU/ml at 6 h) in this patient were also comparable to the control Tanner II-III females [Basal: 1.7 ⫾ 1.0; Peak: 25 ⫾ 5 (se) mIU/ml] (17). Baseline E2 level (73.4 pmol/liter) was in the Tanner II level and did not rise significantly by the 12 h after the leuprolide stimulation. The 14-yr-old pubertal female (additional subject) with the compound heterozygous frameshift genotype resulting in salt-wasting disorder (12) had Tanner breast II stage and Tanner pubic hair V stage development. Baseline LH level and 3, 6, and 12 h leuprolide stimulated LH levels were higher than those in the control females with Tanner breast III –V stage development (17) (Fig. 6). Basal FSH level (4.3 mIU/ml) and peak leuprolide stimulated FSH level at 12 h (15 mIU/ml) in this patient was comparable to the control Tanner III-V females [Basal: 3 ⫾ 5, Peak: 20 ⫾ 4 (se) mIU/ml] (17). The baseline serum E2 level (66 pmol/liter) was in Tanner II level and did not rise by 12 h after the leuprolide stimulation. Discussion

The present study demonstrates a novel molecular basis for the nonclassic and classic 3␤-HSD deficiency CAH due to a point mutation in the type II 3␤-HSD gene in both the female child with mild symptoms and in the female infants with severe symptoms of the disorder. The novel stop codon 373C mutation in the type II 3␤-HSD gene identified from one allele of the child with the nonclassic presentation of the disorder and a novel homozygous P222T mutation identified from the infant with classic salt-wasting disorder together with the in vitro expression of the mutant genes helped to further the understanding the relationship between the structure and function of human type II 3␤-HSD gene. The autosomal recessive inheritance of the mutant alleles was proven by the identification of the mutant genotype on one allele on each of the respective carrier parents.



FIG. 4. The comparison between WT and MT II 3␤-HSD enzyme activities in intact COS-I cells, based on the percent conversion of 1 ␮mol/liter of ⌬5-P (pregnenolone) to P (progesterone) and 1 ␮mol/liter DHEA to ⌬4-A (Androstenedione) in the medium. The cells were transfected with pcDNA 3.1 vectors carrying the respective WT and MT II 3␤-HSD cDNA or cDNA⫹ via site-directed mutagenesis. The results represent the mean of the two determinations. The control mock vector carrying no type II 3␤-HSD cDNA or cDNA⫹ exhibited no enzyme activity (᭙-᭙). WT cDNA refers to the vector carrying the plasmid of WT II 3␤-HSD cDNA (E–E). WTcDNA⫹ contains WT cDNA continuing with additional WT exon IV sequence up to nucleotide no. 7756 of the gene ending with a stop codon sequence (䡺-䡺). MT Stop373C⫹ refers to the nonstop mutation in what would normally be the stop codon 373 substituted by a cystine residue, continuing with additional WT exon IV sequence up to nucleotide no. 7756 of the gene ending with a stop codon (f)-f). P222T cDNA has an amino acid residue substitution in codon 222 from proline to threonine (F-F).

The functional significance of the normal stop codon 373 in the nucleotide no. 7468 –7471 in the type II 3␤-HSD gene is obvious and result predictably in the WT II 3␤-HSD protein containing 372 amino acid residues in humans (4, 15). It is also evident that the stop codon in the family of 3␤-HSD genes is conserved throughout all mammalian and other vertebrate species. The stop codon structural integrity is essential in translation of WT 3␤-HSD protein containing either 372 or 371 amino acid residues throughout the species (2, 3). The Stop373C identified from the child with the nonclassic 3␤-HSD deficiency disorder would result predictably in an open reading frame of the coding sequence in the exon IV of the gene up to nucleotide no. 7753 followed by a subsequent stop codon sequence at nucleotide nos. 7754 –7756 of the gene (15). This would result predictably in an additional 95 amino acid residue sequence, in turn resulting in a longer MT II 3␤-HSD protein containing a 467 amino acid residues (Stop373C⫹) in contrast to the WT protein containing a 372 amino acid residues (4, 15). In vitro, WT enzymes from the cells transfected with vectors carrying either WT II cDNA or the construct of WT II cDNA⫹ demonstrated identical activity because only WT II 3␤-HSD enzyme would be translated from both WT cDNA and WT cDNA⫹. The MT Stop373C enzyme activity at the 6 h point in intact COS-I cells or 293T cells was detectable but decreased to 27% of WT enzyme activity. Nonetheless, the substantially measurable mutant Stop373C enzyme activity for both substrates of ⌬517P and DHEA in intact cells corroborated with the mild nonclassic clinical and hormonal phenotypes in the affected patient who had a E142K mutation in the second allele. E142K mutation has been previously reported to be a severely detrimental mutation resulting in no enzyme activity in vitro (18). Thus, the nonclassic manifestation of 3␤-HSD deficiency in this patient with the compound heterozygous mutation of the Stop373C/E142K genotypes is due to the

FIG. 5. An auto-radiogram of the Western blot analysis of the type II 3␤-HSD protein (15 ␮g protein per lane) from the 293T cells transfected with pcDNA 3.1 vectors carrying WT II cDNA and WT II cDNA⫹, MT II Stop373C⫹, MT II P222T cDNA, a control reported unstable MT L6F cDNA (16) and a control cell carrying no pcDNA vector. A 43-kDa 3␤-HSD protein was detected only from the cells transfected with WT II 3␤-HSD cDNA and WT cDNA⫹. The 3␤-HSD protein was not detectable from the cells transfected with MT P222T and MT Stop373C⫹ as well as the control MT L6F cDNA and the control cells carrying no pcDNA vector. The MT L6F was previously proven to be relatively unstable (Ref. 16). For details of mutant DNA construct, see Figs. 3 and 4.

effect of the partial enzyme activity of the Stop373C⫹ mutant protein. The nonclassic presentation of the patient was evident by normal genitalia, premature sexual hair growth and mild growth acceleration between the ages of 5–7 yr, normal PRA and aldosterone levels, low normal cortisol secretion, and only mildly elevated basal ⌬5-17P level. It is of interest that the in vitro Stop373C enzyme exhibited a similar degree

2562


FIG. 6. LH response to leuprolide stimulation (20 ␮g/kg sc) in the pubertal female with the compound heterozygous Stop373C/E142K genotypes and nonclassic presentation of 3␤-HSD deficiency CAH (nonclassic patient, f⫽f) and in the pubertal female with the compound heterozygous 273/318 frameshift genotypes and classic SW presentation of 3␤-HSD deficiency CAH (classic patient, Œ-Œ) LH was assayed by a highly specific immunoradiometric assay and the LH response of the control females for development of breast stage of Tanner I (䡺-䡺), Tanner II–III (‚-‚), and Tanner III–IV (E-E) were previously reported (17).

of catalytic activity for ⌬5-P and DHEA substrates. However, in vivo hormonal data of this subject revealed a greater elevation of ⌬5-17P than DHEA. It is possible that intraadrenal 17–20 lyase activity might have played a role for the greater production of ⌬5-17P than DHEA in the in vivo setting by a local regulatory mechanism. It is also possible that variable kinetics of this mutant enzyme for ⌬5-17P and DHEA substrate may be the factor but it was not possible to prove this due to the instability of the enzyme. Despite detectable MT Stop373C⫹ enzyme activity in intact cells, the enzyme activity in the homogenates of the cells was undetectable in the presence of NAD⫹ as cofactor. Furthermore the Stop373C⫹ MT protein was not detectable on Western blottings when WT II 3␤-HSD enzymes were clearly detected from the cells transfected with the WT cDNA and cDNA⫹. Hence, it was not possible to determine the kinetics of this mutated enzyme. These findings suggest that a relative instability of the Stop373C⫹ mutant protein with 467 amino acid residues is likely the detrimental mechanism for the compromised enzyme activity. We hypothesize that the extra tail of such an enzyme renders the protein unstable. Thus, the protein is made but quickly falls apart. This would explain the presence of detectable mRNA but decreased enzyme activity in intact cells, but no detectable protein and enzyme activity in the homogenates. A similar finding has been reported on the adenine phosphoryltransferase gene due to a mutation in the stop codon resulting in an extended open reading frame of the gene (19). This also suggests an insight to the detrimental mechanism of the mutated structure by a lengthy protein.


The importance of Pro 222 in the type II 3␤-HSD enzyme may be predicted by the adjacent flanking substrate binding domain from Asn 176 to Arg186 and from Gly 251 to Cys 274 region according to affinity labeling study of purified human type 1 3␤-HSD (20). In addition, Pro 222 region was suggested to be in a membrane-spanning domain in a family of 3␤-HSD isoenzyme in various species (2, 21). Pro 222 is also a conserved amino acid residue throughout the mammalian and vertebrate species (2). Thus, the mutation involving this region of the gene is likely severely detrimental to the enzyme activity. A homozygous P222T mutation identified from the infant with classic salt-wasting disorder was proven to be severely detrimental to the enzyme activity in vitro. Furthermore P222T MT protein from the cells was not detectable on repeated Western blottings despite unequivocal detection of WT enzymes. These findings strongly suggest a profound instability of P222T protein, which is likely the detrimental mechanism for the absent 3␤-HSD activity both in vivo and in vitro. Proline residues may be important to the folding pattern of the enzyme and substitution of proline residues in codon 222 may contribute to the severe instability of the enzyme, as similarly detrimental and unstable P222Q and P222H mutant proteins have been reported (10). We hypothesize that the P222T protein may be folded to an unstable protein. Hence, the protein is made but falls apart severely. This would explain the detectable mRNA but no detectable enzyme activity in the intact cells and homogenates, and the absence of enzyme in the homogenates. To date, various degrees of unstable MT II 3␤-HSD proteins resulting from a missense mutation involving codons 6, 10, 15, 82, 253, and 259 have been reported (10, 14). Western blotting data also suggest that a missense mutation involving codons 108, 167, 173, 186, 245, and 294 may also be unstable (10). The Stop373C⫹ mutant protein is a structurally lengthy protein and adds to the list of MT enzymes that are relatively unstable, while P222T mutant protein adds to the list of severely unstable proteins. The study to compare the phenotype maturation in the HPO axes to the type II 3␤-HSD genotypes in the pubertal females with 3␤-HSD deficiency CAH revealed that in the patient with the compound heterozygous Stop373C/E142K genotypes, basal and LHRH analog stimulated LH and FSH levels were comparable to Tanner II-III control females. The clinical and hormonal findings corroborated with a predictably milder ovarian type II 3␤-HSD deficiency in this patient. In the patient with the compound heterozygous 273/318 frameshift genotypes (12), the minimal breast development and elevated basal and LHRH analog-stimulated LH levels corroborated with predictably insufficient ovarian estrogen synthesis due to severely deficient ovarian type II 3␤-HSD activity related to the frameshift mutation (12). Nonetheless, spontaneous thelarche and menarche in the patient with the predictably severely detrimental type II 3␤-HSD genotype is of interest. The gonads express very low level of type I 3␤-HSD (4), exhibiting an approximately 4 – 6 times greater enzyme efficiency than that of the type II 3␤-HSD enzyme (22–24). Thus it is possible that spontaneous thelarche and menarche in the patient with the frameshift genotypes might have been influenced by the ovarian type I 3␤-HSD activity, as was elaborated in a patient with classic 3␤-HSD deficiency



due to an A10E homozygous mutation in the type II 3␤-HSD gene (25). In summary, this work demonstrated that a structurally lengthy MT II 3␤-HSD enzyme due to a nonstop mutation in what would normally be the stop codon 373 of the type II 3␤-HSD gene was relatively detrimental whereas a missense mutation in codon 222 was severely detrimental to enzyme activity, thereby causing nonclassic and classic 3␤-HSD deficiency CAH, respectively. The in vitro and in vivo data further suggest that the mutant Stop373C and P222T proteins were relatively and severely unstable, respectively. The profound instability of P222T mutant enzyme may be the detrimental mechanism for the severely deleterious missense mutations involving this region of the gene. The findings of the HPO axis in 2 pubertal females with 3␤- HSD deficiency CAH corroborated with the MT II 3␤-HSD genotypes although type I ovarian 3␤-HSD activity appears to contribute to spontaneous thelarche and menarche in the pubertal female with a compound heterozygous frameshift genotype associated with salt-wasting 3␤-HSD deficiency CAH.

9. 10.

11.

12.

13.

14.

15.

Acknowledgments 16.

Received November 19, 2001. Accepted February 26, 2002. Address all correspondence and requests for reprints to: Dr. Songya Pang, Department of Pediatrics (M/C 856), University of Illinois, 840 South Wood Street, Chicago, Illinois 60612. This work was supported by a United States Public Health Service (USPHS) Grant R01-HD-36399 (to S.P.) and in part by a USPHS General Clinical Research Center grant to the University of Illinois at Chicago, College of Medicine.

17.

18.

19.

References 5

1. Goldman AS, Bongiovanni AM, Yakovac WC, Prader A 1964 Study of c 3␤-hydroxysteroid dehydrogenase in normal hyperplastic and neoplastic adrenal cortical tissue. J Clin Endocrinol Metab 24:894 –909 2. Mason JI 1993 The 3␤-hydroxysteroid dehydrogenase gene family of enzymes. Trends Endocrinol Metab 4:199 –202 3. Labrie F, Simard J, Luu-The V, Belanger A, Peletier G 1992 Structure, function and tissue-specific gene expression of 3␤-hydroxysteroid dehydrogenase/5ene-4-ene isomerase enzymes in classical and peripheral intracrine steroidogenic tissues. J Steroid Biochem Mol Biol 43:805– 826 4. Rheaume E, Lachance Y, Zhao H, Breton N, Dumont M, de Launoit Y, Trudel C, Luu-The V, Simard J, Labrie F 1991 Structure and expression of new complementary DNA encoding the almost exclusive 3␤-hydroxysteroid dehydrogenase/⌬5-⌬4-isomerase in human adrenals and gonads. Mol Endocrinol 5:1147–1157 5. Bongiovanni AM, Kellebenz G 1962 The androgenital syndrome with deficiency of 3␤-hydroxysteroid dehydrogenase. J Clin Invest 41:2086 –2092 6. Pang S 2001 Congenital adrenal hyperplasia owing to 3␤-hydroxysteroid dehydrogenase deficiency. Endocrinol Metab Clin North Am 30:81–100 7. Pang S, Levine LS, Stoner E, Opitz JM, New MI 1983 Non salt-losing congenital adrenal hyperplasia due to 3␤-hydroxysteroid dehydrogenase deficiency with normal glomerulosa function. J Clin Endocrinol Metab 56:808 – 818 8. Chang YT, Kulin HE, Garibalsi L, Suriano MJ, Bracki K, Pang S 1992 Hypothalamic-pituitary-gonadal axis function in pubertal male and female siblings with glucocorticoid-treated non salt-wasting 3␤-hydroxysteroid dehy-

20.

21. 22. 23. 24. 25.

drogenase deficiency congenital adrenal hyperplasia. J Clin Endocrinol Metab 77:1251–1257 Rosenfield RL, Rich BH, Wolsdorf JI, Cassorla F, Parks JS, Bongiovanni AM, Wu CH, Shackelton CH 1980 Pubertal presentation of congenital ⌬5-3␤ hydroxysteroid dehydrogenase deficiency. J Clin Endocrinol Metab 51:345–353 Moison AM, Ricketts MK, Veronique T, Desrochers M, Me´barki F, Chaussain J-L, Cabrol S, Ravx-Demay MC, Forest MG, Sippell WG, Peter M, Morel Y, Simard J 1999 New Insight in mutations in the HSD3B2 gene in eleven patients from seven new families and comparison of the functional properties of twenty-five mutant enzymes. J Clin Endocrinol Metab 84:4410 – 4425 Mauri S, Castro M, Latronico AC, Elias LLK, Arnhold JJP, Moreira AC, Mendonca BD 2000 Mutations in the type II 3␤-hydroxysteroid dehydrogenase (HSD3B2) gene can cause premature pubarche in girls. Clin Endocrinol 52:67–75 Zhang L, Sakkal-Alkaddour HS, Chang YT, Yang X, Pang S 1995 A new compound heterozygous frameshift mutation in the type II 3␤-hydroxysteroid dehydrogenase (3␤-HSD) gene causing salt-wasting 3␤-HSD deficiency congenital adrenal hyperplasia. J Clin Endocrinol Metab 81:291–295 Sakkal-Alkaddour H, Zhang L, Xiaojiang Y, Chang YT, Kappy M, Slover RS, Jorgensesn Y, Pang S 1996 Studies of 3␤-hydroxysteroid dehydrogenase gene in infants and children manifesting premature pubarche and increased ACTH stimulated ⌬5 steroid levels. J Clin Endocrinol Metab 81:3961–3965 Zhang L, Mason I, Naiki Y, Copeland KS, Castro-Magana M, Cordon Walker TA, Chang YT, Pang S 2000 Characterization of two novel missense mutations involving codon 6 and 259 of type II 3␤-hydroxysteroid dehydrogenase (3␤HSD) gene causing respectively, non salt-wasting and salt-wasting 3␤ HSD deficiency disorder. J Clin Endocrinol Metab 85:1678 –1685 Lachance Y, Luu-The V, Verreault H, Dumont M, Rheaume E, Leblanc G, Labrie F 1991 Structure of human type II 3␤-(hydroxysteroid dehydrogenase/ ⌬5-⌬4 isomerase (3␤-HSD) gene: adrenal and gonadal specificity. DNA Cell Biol 10:701–711 Sakkal-Alkadour H, Surain MJ, Riddick L, Chang YT, Ziai F, Pang S 1998 The effect of three menstrual cycle phases on adrenal hormonal dynamics and etiologies of hirsutism in five ethnic/racial female populations. Clin Pediatr Endocrinol 7:23–34 Garibaldi LR, Aceto T, Weber C, Pang S 1993 The relationship between luteinizing hormone and estradiol secretion in female precocious puberty: evaluation by sensitive gonadotropin assays and the leuprolide stimulation test. J Clin Endocrinol Metab 76:851– 856 Simard J, Rheaume E, Sanchez R, Laflamme N, de Launoit Y, Luu-The V, van Seters AP, Gordon RD, Bettendorf M, Heinrich U, Moshang T, New MI, Labrie F 1993 Molecular basis of congenital adrenal hyperplasia due to 3␤hydroxysteroid dehydrogenase deficiency. Mol Endocrinol 7:716 –728 Taniguchi A, Hakoda M, Yamanaka H, Terai C, Hikiji K, Kawaguch R, Konishi N, Kashiwazaki S, Kamatani N 1998 A germline mutation abolishing the original stop codon of the human adenine phosphoribosyltransferase (APRT) gene leads to complete loss of the enzyme protein. Hum Genet 102: 197–202 Thomas JL, Nash WE, Myers RP, Crankshaw MW, Strickler RC 1993 Affinity radiolabeling identifies peptides and amino acids associated with substrate binding in human-placental 3␤-hydroxysteroid dehydrogenase. J Biol Chem 268:18507–18512 Eisenberg D, Schwartz E, Kamoromy M, Wall R 1984 Analysis of membrane and surface protein sequences with the hydrophobic moment plot. J Mol Biol 179:125–142 Dumont M, Luu-The V, Dupont E, et al. 1992 Characterization, expression, and immunohistochemical localization of 3␤-hydroxysteroid dehydrogenase/ ⌬5-⌬4 isomerase in human skin. J Invest Dermatol 99:415– 421 Lorence MC, Corbin CJ, Kamimura N, Mahendroo MS, Mason JI 1990 Structural analysis of the gene encoding 3␤-hydroxysteroid dehydrogenase ⌬5-⌬4 isomerase. Mol Endocrinocrinol 4:1850 –1855 Rheaume E, Lachance Y, Zhao H-F 1991 Structure and expression of a new complementary DNA coding the almost exclusive 3␤-hydroxysteroid dehydrogenase. Mol Endocrinol 5:1147–1151 Alos N, Moisan A, Ward L, Desrochers M, Legault L, Leboeuf G, Van Vliet G, Simard J 2000 A novel A10E homozygous mutation in the HSD3B2 gene causing severe salt-wasting 3␤-hydroxysteroid dehydrogenase deficiency in 46,XX and 46,XY French-Canadians: evaluation of gonadal function after puberty. J Clin Endocrinol Metab 85:1968 –1974