Main

The most frequent cause of deficient Aldo biosynthesis is congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Signs of androgen excess appear in these patients. However, a small group of patients do have Aldo deficiency without any disturbances in cortisol and androgen biosynthesis. In the early 1960s, the diagnosis of a selective mineralocorticoid deficiency was established by urinary steroid metabolite determinations using improved laboratory methods such as gas chromatography(1, 2). In 1964, Visser and Cost(3) were the first to suggest a biosynthetic defect with autosomal recessive inheritance causing selective hypoaldosteronism due to deficient 18-hydroxylation of corticosterone(4). Subsequently, two patients with presumably deficient 18-OH-dehydrogenase were described by Ulick et al.(5) and Rappaport et al.(6). A decade later, Ulick(7) suggested the two biochemically different forms of selective Aldo deficiency be termed CMO deficiency types I and type II. In both CMO types, Aldo biosynthesis is impaired, whereas corticosterone of zona glomerulosa origin, under the primary control of the renin angiotensin system, is excessively produced. The two defects differ biochemically in that 18-OHB is deficient in CMO I, but overproduced in CMO II. Both disorders are characterized clinically by salt-wasting, failure to thrive, and growth retardation. The clinical spectrum of the defects has been summarized by Veldhuis and Melby(8), Rösler(9), and Ulick(10). The frequency of the two distinct forms of CMO deficiency has not yet been determined.

The molecular basis of terminal Aldo biosynthesis has been studied in detail in the last years. It is now clear that the terminal steps of Aldo biosynthesis in the zona glomerulosa are catalyzed by a single cytochrome P450 enzyme termed P450aldo(11). Humans have a distinct cytochrome P450 isoenzyme that catalyses hydroxylation at position 11β in the zona fasciculata termed P450c11(12). The genes encoding for P450c11 and P450aldo are termed CYP11B1 and CYP11B2, respectively(12). The human P450c11 and P450aldo enzymes have been predicted to be 93% identical in their amino acid sequence. Both genes are located on chromosome 8q22(13).

So far, the molecular basis of CMO I and CMO II has been explained in several families of Iranian Jewish origin(14, 15) and in a North American kindred(16), respectively, and, recently, in one of our own patients of Caucasian origin(17).

METHODS

Subjects. Our present series of 14 infants (8 female and 6 male patients) from 12 families diagnosed to have selective hypoaldosteronism was collected during the last 10 y. All babies were born at term. Fourteen patients presented with classical clinical features in the first weeks of life. The leading symptoms were recurrent vomiting, failure to thrive, and severe dehydration. Age at diagnosis ranged from 3 to 24 wk of life with a median of 8 wk. Birth weight was normal in all children ranging from 2900 to 4180 g with a median of 3440 g. Weight at time of diagnosis, before treatment was started, ranged from 2700 to 4680 g with a median of 3420 g which was not significantly higher than weight at birth. Some children presented with a weight even below their birth weight, whereas others showed insufficient weight gain. There was no genital ambiguity in any case. Laboratory signs in the untreated patients were hyponatremia, hyperkalemia, increased urinary sodium excretion, metabolic acidosis, elevated plasma renin activity, and decreased plasma Aldo. All patients had normal renal function indicated by normal serum creatinine. One patient was discovered by biochemical screening of asymptomatic family members. Of the 12 families, 5 came from Central Europe, 5 from Turkey, 1 from Lebanon, and 1 from Pakistan. The parents were consanguineous in two families. Clinical and initial laboratory data of all patients are summarized in Table 1.

Table 1 Clinical and laboratory data of 14 children with isolated terminal Aldo biosynthesis defect
Full size table

Procedure. The diagnosis of corticosterone methyl oxidase deficiency was confirmed by measurement of adrenal plasma steroids before and after short term ACTH stimulation. The ACTH stimulation test was performed to exclude other defects of adrenal steroidogenesis. The ACTH test was performed in the untreated patients with an i.v. bolus injection of 125 μg 1,24-ACTH(Synacthen, CIBA-Geigy, Wehr, Germany) between 0800 and 1000 h. Blood samples were drawn immediately before and 60 min after ACTH injection, collected in prechilled heparinized tubes, and immediately centrifuged at 4°C. Plasma was kept frozen at -20°C until assayed.

Plasma steroids were measured using a previously described method for the simultaneous determination of multiple adrenal steroids in a small plasma volume of 1 to 2 mL developed in our laboratory(18). After methylene chloride extraction of the unconjugated steroids from plasma to which tritiated steroid tracers had been added, Aldo, 18-OHB, corticosterone, 18-OHDOC, DOC, progesterone, 17-OHP, 11-deoxycortisol, and cortisol were separated by means of Sephadex LH-20 automated multicolumn chromatography(19). Each of the isolated steroids was then quantified by specific RIA. The results are expressed in nanomoles/L; to convert to nanograms/mL, divide by the following factors: Aldo 2.774, 18-OHB 2.759, corticosterone 2.886, 18-OHDOC 2.886, DOC 3.026, 17-OHP 3.026, 11-deoxycortisol 2.887, and cortisol 2.759.

Normal ranges for different pediatric age groups using our method of multisteroid analysis(17, 18) have been reported previously(20, 21).

RESULTS

Two groups within all children with congenital hypoaldosteronism could be separated on the basis of their 18-OHB plasma levels.

CMO I Patients

Aldo. Aldo plasma levels were near the lower limit of detection of the assay (0.055 nmol/L). Basal Aldo ranged from 0.055 to 0.11 nmol/L(normal, 0.11-0.61 nmol/L), whereas PRA (Table 1) was excessively elevated in all patients ranging from 45 to >300 ng/mL/h(normal range for age, 2.2-10.2 ng/mL × h). After short-term ACTH stimulation, there was a small rise of Aldo plasma levels in some patients(range 0.055-0.31 nmol/L), which, however, did not reach the normal range of 0.43 to 1.43 nmol/L (Fig. 1, lower left panel).

Figure 1
figure 1

(Lower panel) Aldo plasma levels before(basal) and after short-term ACTH stimulation in normal infants below age 12 mo (boxes) and in Aldo-deficient infants with CMO I(left panels) and CMO II (right panels). (Middle panel) 18-OHB plasma levels (logarithmic scale). (Upper panel) Corticosterone (B) plasma levels.

Full size image

18-OHB. Although Aldo was very low in all affected children, there was one group with decreased or low normal 18-OHB plasma levels (range, 0.063-0.44; normal, 0.56-1.46 nmol/L). After ACTH, 18-OHB continued to be decreased or low normal (range, 0.38-1.96; normal 0.83-5.5 nmol/L)(Fig. 1, middle left panel).

Corticosterone. Corticosterone plasma levels were already excessively high in the basal state (range, 19-154; normal, 3.5-13.2 nmol/L) with a moderate rise (range, 60.3-206; normal, 32-104 nmol/L) after ACTH stimulation (Fig. 1, upper left panel).

18-OHDOC and DOC. Plasma levels of 18-OHDOC were elevated already in the basal state (range, 0.98-3.1; normal, 0.13-0.72 nmol/L) with a further increase (range, 3.1-11; normal, 0.58-6.1 nmol/L) after ACTH stimulation. Plasma DOC was also markedly elevated in all patients before(range, 0.79-4.7; normal, 0.15-0.67) and after (range, 3.4-49.3; normal, 0.76-3.12 nmol/L) ACTH stimulation.

CMO II Patients

Aldo. Aldo plasma levels were near the lower limit of detection of the assay (0.055 nmol/L). Basal Aldo ranged from 0.055 to 0.097 nmol/L(normal, 0.11-0.61 nmol/L), whereas PRA (Table 1) was excessively elevated in all patients ranging from 46 to >300 ng/mL/h(normal range for age, 2.2-10.2 ng/mL × h). After short-term ACTH stimulation, there was a small rise of Aldo plasma levels in some patients(range 0.055-0.36 nmol/L) which, however, did not reach the normal range of 0.43 to 1.43 nmol/L (Fig. 1, lower right panel).

18-OHB. Although Aldo was very low in all affected children, there was one group with clearly elevated basal 18-OHB plasma levels (range, 12.1-52.1; normal, 0.56-1.46 nmol/L) which remained elevated (17.1-121.4 nmol/L) above the normal range (0.83-5.5 nmol/L) (Fig. 1, middle right panel).

Corticosterone. Corticosterone plasma levels were already excessively high in the basal state (range, 32.2-77.9; normal, 3.5-13.2 nmol/L) with a moderate rise (range, 98.6-134; normal, 32-104 nmol/L) after ACTH stimulation (Fig. 1, upper right panel).

18-OHDOC and DOC. Plasma levels of 18-OHDOC were elevated already in the basal state (range, 0.81-7.8; normal, 0.13-0.72 nmol/L) with a further increase (range, 6.7-16.2; normal, 0.58-6.1 nmol/L) after ACTH stimulation. Plasma DOC was also markedly elevated in all patients before(range, 0.70-7.05; normal, 0.15-0.67) and after (range, 2.78-47.2; normal, 0.76-3.12 nmol/L) ACTH stimulation.

CMO I and II Patients

Cortisol and its precursor steroids. Plasma cortisol and its direct precursors 11-deoxycortisol and 17-OHP were all within the normal range for age before and after ACTH stimulation excluding other adrenal disorders with salt-wasting such as 21-hydroxylase deficiency or congenital adrenal hypoplasia. Ranges found in the whole group of 14 CMO patients were as follows: 17-OHP basal 0.12-3.29, after ACTH 3.09-9.08 nmol/L; 11-deoxycortisol basal 0.09-5.64, and after ACTH 5.48-13.92 nmol/L; cortisol basal 262-772, after ACTH 508-1424 nmol/L.

18-OHB/Aldo ratio. Calculating the ratio of plasma 18-OHB to Aldo differentiated both types of disorders of terminal Aldo biosynthesis. Patients with CMO I invariably had a ratio of less than 10, whereas in all CMO II patients the ratio was above 100 (Table 2). The disadvantage of this ratio is that it cannot be precisely calculated when Aldo plasma levels are below the lower limit of detection of the assay.

Table 2 Ratios of 18-OHB/Aldo and B/18-OHB of 14 children with isolated terminal Aldo biosynthesis defect
Full size table

Corticosterone/18-OHB ratio. An even better parameter for the differentiation between CMO types I and II deficiency is the ratio of plasma corticosterone to 18-OHB (Table 2). All patients with CMO I had a ratio of more than 40, whereas in all CMO II patients the ratio was less than 10.

DISCUSSION

Thirty years after the first description by Visser and Cost(3) of a biosynthetic defect of Aldo production causing isolated hypoaldosteronism, the clinical features, biochemical findings, and diagnostic criteria, and, finally, etiology and pathogenesis of this rare, autosomal recessive inherited inborn error of metabolism appear much more precisely defined.

Clinical presentation and treatment. In accordance with the published clinical case reports(22–30), we did not observe any differences in the severity of clinical signs between young infants with CMO I and CMO II, respectively. All children presented with frequent vomiting, failure to thrive, and severe, life-threatening salt loss in the first weeks of life. They fully recovered under treatment with an oral mineralocorticoid. Replacement therapy with 9α-fluorocortisol in doses of 100-250 μg/m2/d is recommended at least during early infancy and childhood. In addition, generous sodium supplementation may be given. Rösler(9) has shown that in each affected individual the clinical severity of the disease decreases with age. Adolescents and adults may show only the abnormal steroid pattern which persists throughout life. Continued mineralocorticoid replacement therapy after childhood is not always necessary as in the clinical observation compensatory extra-adrenal salt-conserving mechanisms mature with age. The same natural course of the disease is observed in many patients with pseudohypoaldosteronism(9).

Frequency. Precise data regarding the frequency of selective hypoaldosteronism and its two biochemical types in the general population are not available. It is generally assumed that CMO deficiency type II is more frequent than CMO I. This is mainly due to the observation of Rösler(9) who published data of 21 patients of Iranian Jewish origin with selective Aldo deficiency due to CMO II. This preponderance of CMO type II deficiency is also reflected in the overall published case reports(22–30). In contrast to this, our material, which is the first large series of unrelated patients diagnosed by strictly the same methodology, indicates a comparable frequency of both CMO types.

Biochemical findings and diagnostic recommendations. The biochemical criteria for the differentiation of the two CMO variants are based on specific steroid determinations in urine and/or plasma. The criteria for differentiation of CMO types I and II have been proposed by Ulick(7). The same author published revised and more precise biochemical criteria for the diagnosis of the two CMO variants based of urinary steroid measurements using gas chromatography/mass fragmentography(31). According to this nomenclature, the historic Visser and Cost family(3) would be classified as CMO type I deficiency, whereas the family described by Ulick et al.(5), would be termed CMO type II deficiency.

Using specific steroid determinations, the plasma level of 18-OHB distinguishes between CMO I (where it is decreased or low normal) and CMO II(where it is markedly elevated). The clearest discriminative parameter between the two CMO types reflecting impaired 18-hydroxylation is that the ratio of plasma corticosterone/18-OHB which is elevated in CMO I and decreased in CMO II. Moreover, the ratio of plasma 18-OHB/Aldo can also discriminate between the two CMO variants. The disadvantage of this ratio is that plasma Aldo levels in these patients often are at or even below the lower limit of detection.

The plasma levels of DOC and 18-OHDOC were elevated in all Aldo-deficient children. There was no difference between the two CMO types. Both plasma levels increase with ACTH in normal children as well as in Aldo-deficient children. This suggests, at least in part, a zona fasciculata origin as a source of both steroids. This is in agreement with data from animal studies which have shown that, at least in the rat, most of the 18-OHDOC is derived from the zona fasciculata, and its biosynthesis appears to be mainly under ACTH control(32). However, 18-OHDOC is also produced by the zona glomerulosa(32). These data are transferable to humans because both rats and humans have distinct steroid 11β-hydroxylase isoenzymes active in the zona fasciculata and glomerulosa(33, 34). P450c11 can in vitro hydroxylate DOC at the C18 position(35). Elevated DOC and 18-OHDOC levels in both CMO types can be well explained by substrate accumulation proximal to the deficient enzyme activities.

For practical reasons, we recommend the following procedure for the diagnosis of CMO deficiency in young infants. All salt-wasting neonates should be examined carefully. First of all, 21-hydroxylase deficiency should be excluded by measurement of basal 17-OHP (RIA after extraction). Second, basal cortisol, Aldo, and PRA should be determined. Highly specific methods (RIA after extraction and chromatography) for the determination of plasma steroids(particularly for Aldo) are necessary in the early life period. In our own experience, many direct RIA methods yield far too high results in newborns and young infants due to vast amounts of interfering steroids from the fetoplacental unit and/or fetal adrenal cortex. In children with low or undetectably low plasma Aldo and elevated PRA, corticosterone should be determined as next. A high ratio of corticosterone over Aldo leads to the diagnosis of CMO deficiency. The ratios of 18-OHB/Aldo and corticosterone/18-OHB is necessary for further differentiation of CMO type I and II deficiency. An ACTH test is not necessary for the diagnosis. The diagnosis can also be made from 24-h urine sample(31) which, however, is difficult to collect in this age group.

Molecular genetics. So far, the molecular basis of Aldo deficiency has been elucidated in three different pedigrees(Table 3). Among Jews of Iranian origin CMO type II is observed exclusively(9). All affected individuals are homozygous for two missense mutations in CYP11B2, R181W and V386A. Individuals homozygous for either mutation alone have no clinical symptoms(14, 15). Three patients of a consanguineous Amish kindred affected with CMO type I deficiency have a deletion of 5 nucleotides in exon 1 of CYP11B2, resulting in a frameshift to form a stop codon in the same exon(16). In a boy with proven CMO I deficiency of a nonconsanguineous north German family, we identified a R384P amino acid substitution in CYP11B2(17). This family has lived for generations in the province of East Frisia, not too far from the Dutch area of Friesland (Groningen) where the historic Visser and Cost family lived. To date, it can only be speculated whether both families might have the same disease-causing mutation.

Table 3 Mutations within the CYP11B2 gene causing aldosterone synthase deficiency
Full size table

In both CMO type I-deficient alleles a total loss of 11-hydroxylase as well as 18-hydroxylase activity of P450aldo has been demonstrated(16, 17). In CMO II, the R181W mutation in CYP11B2 does not impair 11-hydroxylase activity, markedly decreases 18-hydroxylase activity, and abolishes 18-oxidase activity. The V386A mutation in CYP11B2 causes a small but consistent reduction in 18-hydroxylase activity(15).

Based on the biochemical findings, the existence of two different enzymes catalyzing the terminal Aldo biosynthesis (CMO type I = 18-hydroxylase and CMO type II = 18-dehydrogenase) has been postulated(7). However, the existence of these two different enzymes has not so far been proven. Investigations on cultured cells and in Aldo-secreting tumors have shown that a single enzyme with 11β-hydroxylase as well as 18-hydroxylase/18-dehydrogenase activity is expressed in the adrenal zona glomerulosa(11, 36). This enzyme, termed P450aldo, is identical to CMO I and II activity, and appears to catalyze exclusively the terminal Aldo biosynthesis in humans(11, 35). It shows considerable homology with the enzyme P450c11 which catalyzes the conversion of 11-deoxycortisol to cortisol in the zona fasciculata of the adrenal cortex(12). P450c11 can also convert DOC to 18-OHDOC, but it 18-hydroxylates corticosterone poorly. It cannot convert corticosterone into Aldo(35). Both enzymes are encoded by different genes in man, i.e. CYP11B1 for P450c11 and CYP11B2 for P450aldo. In both types of CMO defects, corticosterone plasma levels are elevated, possibly due to enhanced 11β-hydroxylation of DOC, whereas methyl oxidation at C-18 is impaired(31).

The two variants of CMO deficiency are caused by different mutations within the same gene(14–17). In their recent review, White et al.(35) discussed the molecular genetic differences between CMO types I and II deficiency. They stated that CMO type I might be caused by a more severe reduction of enzyme activity of P450aldo. The high levels of corticosterone in both CMO types can be explained by up-regulation of P450c11 in the zona glomerulosa under the control of elevated PRA(36). It is, however, difficult to explain the different 18-OHB levels in both CMO types. In vitro, 11β-hydroxylase has 18-hydroxylase activity at a level of about 1/10th of that of P450aldo. If 11β-hydroxylase activity in the zona glomerulosa is increased in both CMO types, then 18-OHB should likewise be elevated in both. Another explanation for increased 18-OHB levels might be an overexpression of CYP11B2 in CMO type II. However, the R181W/V386A mutant P450aldo from Iranian patients with CMO type II deficiency has normal 11β-hydroxylase activityin vitro but low 18-hydroxylase activity(15). To clarify these discrepancies, further molecular studies on a greater number of Aldo-deficient patients from various genetic backgrounds are necessary to elucidate the structure-function relationship of the enzyme P450aldo.