Genetic, Biochemical, and Clinical Studies of Patients With A328V or R213C Mutations in 11βHSD2 Causing Apparent Mineralocorticoid Excess
Abstract—Apparent mineralocorticoid excess is a recessively inherited hypertensive syndrome caused by mutations in the 11β-hydroxysteroid dehydrogenase type 2 gene, which encodes the enzyme normally responsible for converting cortisol to inactive cortisone. Failure to convert cortisol to cortisone in mineralocorticoid-sensitive tissues permits cortisol to bind to and activate mineralocorticoid receptors, causing hypervolemic hypertension. Typically, these patients have increased ratios of cortisol to cortisone and of 5α- to 5β-cortisol metabolites in serum and urine. We have studied 3 patients in 2 families with severe, apparent mineralocorticoid excess and other family members in terms of their genetic, biochemical, and clinical parameters, as well as normal controls. Two brothers were homozygous for an A328V mutation and the third patient was homozygous for an R213C mutation in the 11β-hydroxysteroid dehydrogenase type 2 gene; both mutations caused a marked reduction in the activity of the encoded enzymes in transfection assays. The steroid profiles of the 7 heterozygotes and 2 other family members studied were completely normal. The results of a novel assay used to distinguish 5α- and 5β-tetrahydrometabolites suggest that 5β-reductase activity is reduced or inhibited in apparent mineralocorticoid excess. In 1 patient undergoing renal dialysis for chronic renal insufficiency, direct control of salt and water balance completely corrected the hypertension, emphasizing the importance of mineralocorticoid action in this syndrome.
Apparent mineralocorticoid excess (AME) is a rare and severe form of hypertension characterized by an early age of onset and signs of excess mineralocorticoid activity.1 2 It has an autosomal recessive mode of inheritance3 and is caused by mutations in the 11β-hydroxysteroid dehydrogenase type 2 (11βHSD2) gene, which result in a deficiency in the activity of the encoded 11βHSD2 enzyme4 5 that catalyzes the conversion of cortisol to cortisone and of corticosterone to 11-dehydrocorticosterone.6 Cortisol and corticosterone are agonists with high affinity for the mineralocorticoid receptor, comparable to the principal mineralocorticoid hormone aldosterone, whereas cortisone and 11-dehydrocorticosterone have a low affinity for the receptor.7 The catalytic action of 11βHSD2 ensures, in part, that binding and activation of the mineralocorticoid receptor is effected only by aldosterone, since this steroid is not a substrate for the 11βHSD2 enzyme.
In patients with a hereditary defect of 11βHSD2 enzymatic activity, cortisol, which circulates at much greater concentrations than aldosterone, acts as a potent mineralocorticoid hormone, inducing hypervolemic hypertension. The AME syndrome has been biochemically characterized by increased cortisol-to-cortisone ratios in serum and urine, or of the urinary cortisol metabolites, 5β-tetrahydrocortisol (5βTHF) and 5α-tetrahydrocortisol (5αTHF), relative to the cortisone metabolite 5β-tetrahydrocortisone (5βTHE) (Figure 1⇓). An increase in the ratio of urinary 5αTHF to 5βTHF is also typically found in most of these patients,2 8 9 for which no explanation has yet been furnished.10 A second form of AME (AME type 2) has also been reported,11 12 and its molecular basis has recently been ascribed to mutations that partially inactivate the 11βHSD2 enzyme.10 Here we report genetic, biochemical, and clinical follow-up studies of 3 AME type 1 patients (2 of them siblings), as well as their relatives, and 91 normal individuals. The AME patients were investigated by a newly developed assay for determination of the urinary 5α-tetrahydroderivative of cortisone (5αTHE) and dehydrocorticosterone (5αTHA). Quantification of these steroids has enabled us to postulate a mechanism for the increased 5αTHF-to-5βTHF ratio encountered in AME. One of our patients was subjected to long-term hemodialysis, and she was evaluated both before and after dialysis.
Patients A and B
Patients A and B are brothers of Portuguese origin whose parents are first cousins. They were diagnosed as having AME in 1993 on the basis of hypertension, hypokalemia, hypoaldosteronemia, hyporeninemia, and low excretion of urinary 5βTHE compared with 5βTHF and 5αTHF.13 The 2 brothers have been repeatedly evaluated for their urinary corticoid excretion and blood cortisol, cortisone, renin, aldosterone, deoxycorticosterone, and corticosterone levels. Similar studies were performed on their mother, father, unaffected brother, and uncle.
The third patient (C) was diagnosed at age 4 as having arterial hypertension (140/60 mm Hg), hypokalemic alkalosis, low plasma renin and aldosterone levels, microcalcifications, and pseudocystic right kidney with fibrosis of 7 of 22 glomeruli on renal biopsy.14 A positive diagnosis of AME was established at age 30, and the patient was reevaluated at the beginning of 1997, before and after institution of dexamethasone (DXM, 0.5 mg/d) therapy, and later in 1997 after hemodialysis (4 hours, 3 times a week) had been initiated. She was subjected to the same biochemical and genetic testing15 as patients A and B, together with her mother, sister, and 3 children.
Healthy, white, nonsmoking volunteers (51 women and 40 men), with ages ranging from 21 to 46 years and taking no medication, contraceptives, or licorice, provided blood samples at 8 am and 24-hour urine samples to establish the reference values of the parameters tested. All investigations conformed to the ethical standards of the Helsinki declaration (1975) as revised at Tokyo (1983).
Analysis of Genomic DNA
Genomic DNA was extracted from peripheral blood leukocytes, and the exons and intron-exon junctions of the 11βHSD2 gene were amplified by polymerase chain reaction (PCR). The samples were denatured at 94°C for 1 minute, followed by 30 amplification cycles (94°C for 1 minute, 60°C to 70°C for 1 minute, and 72°C for 2 minutes, with an additional 5 seconds in each cycle) and 1 cycle at 72°C for 7 minutes. Amplified DNA fragments were purified from low-melting-point agarose gels with Gelase (Epicentre Technologies), and the purified DNA was then sequenced on both strands by the Sanger method of dideoxynucleotide chain termination by using T7 Sequenase version 2 (Amersham).
In Vitro Mutagenesis
Mutations were introduced into the wild-type 11βHSD2 cDNA by site-directed PCR mutagenesis, as previously described.16 The PCR reaction was optimized to facilitate amplification of the GC-rich (62%) sequence by using a mix of Taq DNA polymerase (5 U/mL) and Pfu DNA polymerase (2.5 U/mL) in 200 mmol/L Tris, pH 8.8; 3 mmol/L MgCl2; 160 mmol/L NH4SO4; and 7.5% dimethyl sulfoxide in a single reaction.17 The mutated 11βHSD2 cDNAs were subcloned into the HindIII and XbaI restriction sites of the pcDNA1 expression vector (In Vitrogen).
Transfection and Expression Studies
Normal and mutated 11βHSD2 cDNAs (2 μg) were transfected into 70% confluent cells from a Chinese hamster ovary line (CHOP cells, kindly provided by Dr James Dennis) with Lipofectamine (GIBCO BRL) according to the manufacturer’s instructions. CHOP cells were grown in Ham’s F12 medium with 2 mmol/L glutamine and 10% decomplemented fetal bovine serum (Boehringer Mannheim). After 12 hours, the transfected cells were incubated with 2 nmol/L tritiated and 200 nmol/L unlabeled steroid (cortisol or corticosterone), and the resulting products were analyzed by high-performance liquid chromatography with an online β-detection system (Radiomatic, Flo-one, Packard Instruments). Apparent kinetic constants (Km and Vmax) were calculated from 2-hour incubations of 2 nmol/L tritiated steroid with 10, 50, 200, 600, 1500, and 5000 nmol/L unlabeled steroid and analysis by Lineweaver-Burk plot. The results presented are the mean of 2 independent experiments. The efficiency of the transfection was verified by Northern blot analysis (5 μg total RNA) and reverse transcription PCR (data not shown).
Serum was drawn between 8 and 9 am after 1 hour of upright posture. Renin and aldosterone were evaluated by radioimmunoassay with Renin III (ERIA Diagnostics Pasteur) and TKAL (DPC) kits. Cortisol and cortisone in blood and 24-hour urine were determined by radioimmunoassay after chromatography as previously described.18
Estimation of 17 hydroxycorticosteroids was carried out by gas chromatography with flame-ionization detection of the methyloxime trimethylsilyl ether derivatives after extraction, hydrolysis, and purification of 24-hour urinary steroids as previously described.13 To quantify 5αTHE and 5αTHA, which coelute with 5βTHF and 5αTHB, respectively, on OV1 columns, the gaseous phase was separated into 2 parts and passed simultaneously over an OV1 and an SPB20 capillary column before detection of the steroids; the SPB20 column allows separation and quantification of 5αTHE and 5αTHA. The remaining steroids were determined from the OV1 effluent, directly for 5βTHE, 5αTHF, 5βTHB, and 5βTHA and after subtraction of 5αTHE and 5αTHA for 5βTHF and 5αTHB, respectively.
Gas Chromatography–Mass Spectrometry
The identity of the trimethylsilyl ether derivatives of the steroids was assessed by mass fragmentography by using a Nermag R30-10 triple-quadrupole spectrometer and electron-impact ionization.
Gas Chromatography–Mass Spectrometry Analysis
Effective separation of 5αTHE from 5βTHF and of 5αTHA from 5αTHB was obtained on the SPB20 column (Figure 2⇓). Identification of 5αTHE and 5αTHA methyloxime trimethylsilyl ether derivatives by electron-impact ionization mass spectrometry produced mass spectra that were identical to those of 5βTHE and 5βTHA (Figure 2⇓). The detection limit of the method (zero+3SD) is 0.01 mg, with an interassay reproducibility (coefficient of variation for 10 repeats) between 11% and 14% for 5βTHE, 5βTHF, 5αTHF, and 5βTHB and of 12% to 17% for 5αTHE, 5αTHB, 5αTHA, and 5βTHA. Our normal values (see Table⇓) are in good agreement with those of other authors,19 20 but the normal ratios we observed are slightly wider than those usually reported.3 12 21
Direct DNA sequencing of 11βHSD2 exons from patients A and B revealed a homozygous mutation of codon 328 (exon 5) of the 11βHSD2 gene, from GCG to GTG (A328V). The father and mother were heterozygous for the same mutation, whereas the brother and uncle were homozygous for the normal sequence.
Patient C had a homozygous mutation in codon 213 (exon 3) of CGC to TGC (R213C). Her mother, sister, son, and 2 daughters were heterozygous for the same mutation. Expression of the mutated cDNA, followed by 16 hours of incubation with tritiated steroid precursors (Figure 3⇓), showed that the 2 mutant enzymes converted <5% of cortisol or corticosterone to cortisone or 11-dehydrocorticosterone, respectively. The normal 11βHSD2 cDNA resulted in 100% conversion of the substrates under the same conditions. Both reverse transcription PCR and Northern blot analysis of RNA from the transfected cells verified the efficiency of the transfection and transcription of the mRNAs (not shown). The apparent Km and Vmax for the normal 11βHSD2 enzyme in our system were, respectively, 117 nmol/L and 1.07 pmol · min−1 · well−1 for corticosterone and 254 nmol/L and 0.46 pmol · min−1 · well−1 for cortisol, consistent with reports from other authors using the same system.22 23 Similar constants could not be calculated for the mutant enzymes owing to their very low activity.
None of the unaffected members of the 2 families explored, including the heterozygous individuals and the homozygous normal individuals, showed any metabolic abnormality in blood or urine (kalemia, renin, aldosterone, cortisol, cortisone, or their metabolites), as is expected for a recessively inherited disease.
The 3 patients, on the contrary, displayed all of the features of AME (the Table⇑). Serum and urinary cortisone levels were low and cortisol-to-cortisone ratios (F-to-E) were high. Owing to the reduced conversion of cortisol to cortisone, urinary free cortisol was high in non–DXM-treated patients. A disturbed balance between cortisol and cortisone metabolites was observed in urinary samples, with an elevated ratio of (5βTHF+5αTHF) to 5βTHE, due principally to diminished secretion of 5βTHE. This was accompanied by an increase in the 5αTHF-to-5βTHF ratios and also the 5αTHE-to-5βTHE ratios, indicating a shift from 5β-reductase toward 5α-reductase metabolism. The serum and urinary F-to-E and (5βTHF+5αTHF)-to-5βTHE ratios remained elevated for patient A, irrespective of the efficacy of blood pressure control, whereas 5αTHF/5βTHF decreased to normal values after DXM treatment. His cortisol clearance (as reflected by the [5βTHF+5αTHF+5βTHE]-to–urinary free cortisol ratio) was generally diminished, but the A-ring reduction constant ([5βTHF+5αTHF]/urinary free cortisol) was normal.
In the urine of normal subjects, 5αTHE values were low compared with 5βTHF, 5αTHF, and 5βTHE values, indicating that cortisone is a less-efficient substrate for 5α-reductase than is cortisol, whereas cortisol and cortisone have similar affinities for the 5β-reductase enzyme.
In addition to the severe hypertension seen in patients A and B, their uncle exhibited sustained hypertension (160 to 180/120 to 140 mm Hg), whereas their father, mother, and brother were normotensive. When reevaluated in 1997, patient A was treated with nifedipine LD 20 mg/d (a calcium antagonist) and DXM 0.5 mg/d. Arterial blood pressure remained elevated (172/106 supine). Plasma potassium was low to normal (3.7 to 4.2 mmol/L), and urinary free cortisol was low, reflecting the DXM treatment. Ultrasonography revealed slight, diffuse, renal cortical atrophy, and the heart showed an eccentric left ventricular hypertrophy, with a high cardiac mass at 131 g/m2 (normal<125). Addition of an angiotensin converting enzyme inhibitor (benazepril 10 mg/d) to his therapy lowered blood pressure to 144/86 and the left cardiac hypertrophy regressed (cardiac mass 102 g/m2). However, within an additional 6 months, his blood pressure had again risen to 162/80. Additional treatment with furosemide (40 mg/d) and amiloride (5 mg/d) resulted in sustained lowering of blood pressure (140/90).
In 1997, patient B had followed no therapy other than potassium supplementation (24 mEq/d). Arterial blood pressure was 162/88 mm Hg supine. There was no sign of left ventricular hypertrophy. Renal ultrasonography was normal, but there was slight proteinuria (0.55 g/24 h) accompanied by high calcium and sodium excretion (9 mmol and 181 mmol/24 h, respectively). In spite of the potassium supplementation, serum potassium was low (3.1 to 3.3 mmol/L), as were renin and aldosterone, and free urinary cortisol was high (198 μg/24 h). A low-sodium diet and DXM 0.5 mg/d were recommended to the patient, which normalized blood pressure (145/85, 18 months later).
The mother of patient C had experienced 3 miscarriages before giving birth to 2 daughters and was diagnosed as having hypertension during her first pregnancy. The patient’s grandmother was also hypertensive, whereas her sister and her 3 children were normotensive. Her father had a history of hypertension and had died of a cerebrovascular accident.
Patient C had been prescribed several classes of antihypertensive agents over the last 20 years, including amlodipine (5 mg×2/d), rilmenidine (1 mg/d), and urapidil (60 mg×2/d), with poor blood pressure stabilization ≈160/100 mm Hg. She also received 32 meq K+/d to correct hypokalemia and calcium carbonate (3.62 g Ca2+/d) and alfacalcidol (1 μg/d) to correct hypocalcemia, which nonetheless remained at 1.75 mmol/L (normal 2.4 to 2.62). She had previously undergone subtotal parathyroidectomy for hyperparathyroidism that was secondary to the persistent hypocalcemia but that had attained a degree of autonomous regulation. Because analysis of her urinary corticoid metabolites (the Table⇑) was consistent with AME, her antihypertensive treatment was modified by the addition of DXM (0.5, then 1 mg/d), resulting in better control of blood pressure, and the associated therapy was reduced to atenolol (50 mg/d) and DXM (1 mg/d). An attempt to withdraw the atenolol completely was followed by a mild increase in blood pressure. At the beginning of 1997, she presented with severe renal insufficiency and was put on hemodialysis (4 hours, 3 times a week), enabling direct control of water and salt balance that, with accompanying weight loss, further improved her blood pressure. All antihypertensive therapy was then progressively removed until her blood pressure was equilibrated by dialysis alone.
Our results highlight the severe complications that may arise in inadequately treated AME patients and the clinical differences that can be observed in individuals with identical mutations. The A328V mutation has been independently reported in a 7-year-old Brazilian girl with AME24 whose heterozygous father was found to be hypertensive, with suppressed plasma renin activity and aldosterone and a persistently elevated (5αTHF+5βTHF)/5βTHE ratio of 2.79 (normal 0.8 to 1.4), but with normokalemia. This was the first such suggestion of hypertension with an AME pattern of steroid metabolites in a heterozygote for an 11βHSD2 mutation.
In the family with an A328V mutation that we studied, there was little evidence for an effect of the mutation in heterozygous individuals. The heterozygous mother and brother were normotensive, and the uncle, who was homozygous for the normal sequence, had established hypertension. Furthermore, the biological exploration and steroid metabolites were completely normal for all heterozygotes in this family as well as in the other family investigated.
The 2 affected brothers displayed some clinical differences in the expression of the disease. Although untreated, patient B showed no sign of left ventricular hypertrophy, while his brother (patient A), who was receiving DXM therapy, had significant left ventricular hypertrophy. Clearly, there is variability of the clinical and biological phenotype independent of the 11βHSD2 genotype.
The R213C mutation has been reported in 2 homozygous sisters with AME5 and in an unrelated male.25 We found the activity of the mutated enzyme to be <5% of the wild-type enzyme for both cortisol and corticosterone, which compares well with the 3.6% and 2.2%, respectively, previously reported.23
The most significant diagnostic feature for AME appears to be the decrease of cortisone in blood and urine and the increased ratio of 11-hydroxymetabolites to 11-oxometabolites of cortisol. The (5βTHF+5αTHF)-to-5βTHE ratio and the cortisol-metabolizing quotient (5βTHF+ 5αTHF+5βTHE/urinary free cortisol) were also a useful diagnostic feature. The A-ring reduction constant ([5βTHF+ 5αTHF]/urinary free cortisol) was always the same in our patients as in normal subjects and thus seems to be a less useful diagnostic indicator of AME.
The method of steroid determination used in this study allowed separate quantification of 5αTHE. The amounts of 5αTHE secreted are low compared with 5βTHF, 5αTHF, and 5βTHE, consistent with a previous report26 that found <2% metabolism of [4-14C]cortisol into 5αTHE and with the values reported27 in cases of congenital adrenal hyperplasia. Separate quantification of this metabolite confirms that cortisone has a low affinity for the 5α-reductase enzyme, whereas cortisol has an approximately equal affinity for the 5α- and 5β-reductase enzymes. In normal subjects, both 5α-reductase and 5β-reductase metabolize cortisol, leading to approximately equal amounts of 5αTHF and 5βTHF, whereas cortisone is reduced mainly to 5βTHE by 5β-reductase. It appears that in AME patients, 5β-reductase activity is reduced, as reflected by the generally low ratios of 5βTHF/F and 5βTHE/E (the Table⇑). In contrast, the ratios of 5αTHF/F are in the normal range. These results suggest that the increased ratio of 5αTHF/5βTHF, often observed in AME patients, is principally due to downregulation of 5β-reductase activity of undetermined origin. Although frequent in AME, the high 5αTHF-to-5βTHF ratio is not universal.3 28 For example, this ratio decreased in our patient A after DXM treatment for reasons that are not clear at the present time. Further study of steroid metabolism in AME patients by the methods presented here is clearly warranted.
As expected in a patient with hypermineralocorticism, the blood pressure of our AME patient with chronic renal insufficiency was completely normalized by hemodialysis, enabling discontinuation of her antihypertensive therapy. This can be attributed to a lowering of serum sodium levels and direct control of water balance. In most AME cases reported, a low-sodium diet has had a beneficial effect on the control of hypertension. This suggests that studies of 11βHSD2 function in hypertension should be centered on salt-sensitive and low-renin forms of hypertension.
Several studies in addition to ours have noted the occurrence of multiple miscarriages in mothers of AME patients.24 29 Because 11βHSD2 is strongly expressed in the placenta and may play a protective role for the developing fetus,30 a more systematic study of this phenomenon may be warranted.
This work was supported in part by a grant from the IPSEN Foundation for Therapeutic Research to L.P. We thank Dr Zygmund Krozowski for his generous gift of 11βHSD cDNA; Drs Pierre-François Plouin, Christine Massien-Simon, and Marie-Christine Jubert for monitoring patients; Denis Lesage and Jean-Claude Tabet for mass spectrometry expertise; Florent Soubrier and Hany Soliman for helpful advice; and Marlyn Cohen, Françoise Quilin, Carole Ouabdelkader, Eliane Rolle, Laurence Homyrda, and François Rocher for expert technical assistance.
Reprint requests to Dr Gilles Morineau, Biologie Hormonale, Hôpital Saint-Louis, 1 Ave Claude Vellefaux, 75010 Paris, France.
- Received March 8, 1999.
- Revision received March 25, 1999.
- Accepted May 24, 1999.
Krozowski ZS, Funder JW. Renal mineralocorticoid receptors and hippocampal corticosterone-binding species have identical intrinsic steroid specificity. Proc Natl Acad Sci U S A.. 1983;80:6056–6060.
Monder C, Shackleton CH, Bradlow HL, New MI, Stoner E, Iohan F, Lakshmi V. The syndrome of apparent mineralocorticoid excess: its association with 11β-dehydrogenase and 5β-reductase deficiency and some consequences for corticosteroid metabolism. J Clin Endocrinol Metab.. 1986;63:550–557.
Gourmelen M, Saint-Jacques I, Morineau G, Soliman H, Julien R, Fiet J. 11β-Hydroxysteroid dehydrogenase deficit: a rare cause of arterial hypertension: diagnosis and therapeutic approach in two young brothers. Eur J Endocrinol.. 1996;135:238–244.
Morineau G, Pascoe L, Krozowski ZS. Mutations R213C et A328V du gène de la 11β-HSD2 responsables du syndrome d’excès apparent de minéralocorticoïdes. Ann Endocrinol.. 1997;58:2S-128. Abstract.
Curnow KM, Slutsker L, Vitek J, Cole T, Speiser PW, New MI, White PC, Pascoe L. Mutations in the CYP11B1 gene causing congenital adrenal hyperplasia and hypertension cluster in exons 6, 7, and 8. Proc Natl Acad Sci U S A. 1993;90:4552–4556.
Cheng S, Fockler C, Barnes WM, Higuchi R. Effective amplification of long targets from cloned inserts and human genomic DNA. Proc Natl Acad Sci U S A.. 1994;91:5695–5699.
Morineau G, Boudi A, Barka A, Gourmelen M, Degeilh F, Hardy N, al-Halnak A, Soliman H, Gosling JP, Julien R, Brerault JL, Boudou P, Aubert P, Villette JM, Pruna A, Galons H, Fiet J. Radioimmunoassay of cortisone in serum, urine, and saliva to assess the status of the cortisol-cortisone shuttle. Clin Chem.. 1997;43:1397–1407.
Weykamp CW, Penders TJ, Schmidt NA, Borburgh AJ, van de Calseyde JF, Wolthers BJ. Steroid profile for urine: reference values. Clin Chem.. 1989;35:2281–2284.
Soro A, Ingram MC, Tonolo G, Glorioso N, Fraser R. Evidence of coexisting changes in 11β-hydroxysteroid dehydrogenase and 5β-reductase activity in subjects with untreated essential hypertension. Hypertension. 1995;25:67–70.
Mune T, White PC. Apparent mineralocorticoid excess: genotype is correlated with biochemical phenotype. Hypertension. 1996;27:1193–1199.
Flood C, Layne DS, Ramcharan SE, Tait JF, Tait SAS. An investigation of the urinary metabolites and secretion rates of aldosterone and cortisol in man and a description of methods for their measurement. Acta Endocrinol (Copenh). 1961;36:237–264.
Dave-Sharma S, Wilson RC, Harbison MD, Newfield R, Azar MR, Krozowski ZS, Funder JW, Shackleton CH, Bradlow HL, Wei JQ, Hertecant J, Moran A, Neiberger RE, Balfe JW, Fattah A, Daneman D, Akkurt HI, De Santis C, New MI. Examination of genotype and phenotype relationships in 14 patients with apparent mineralocorticoid excess. J Clin Endocrinol Metab.. 1998;83:2244–2254.