L’hyperaldosteronisme primaire est la forme la plus fréquente d’hypertension artérielle secondaire. Bien que des mutations somatiques récurrentes dans différents gènes codant pour des canaux ioniques et des ATPases, impliquées dans la régulation de la biosynthèse d’aldostérone, aient été identifiées dans les adénomes produisant de l’aldostérone (APA), leur rôle dans les phénomènes de tumorigenèse reste à démontrer. L’analyse transcriptomique de 48 échantillons d’APA et 11 surrénales contrôles ont permis d’identifier la voie de signalisation de l’acide rétinoïque comme un élément central dans le développement de l’APA.

Objective: Recurrent somatic mutations in KCNJ5, CACNA1D, ATP1A1 and ATP2B3 have been identified in aldosterone producing adenoma (APA). Although the role of these mutations in regulating aldosterone biosynthesis has been clearly established, the mechanisms involved in proliferation and APA formation still remains to be elucidated. The aim of our study was to identify pathways involved in adrenal cortex nodulation and APA formation. Design and method: We have analyzed transcriptome data generated from 123 APA and 11 control adrenals and correlated gene expression to the genetic, morphological and functional characteristics of the tumors. To evaluate the role of retinoic acid receptor RAR in the development of APA, we have investigated the consequences of modulation of RAR signaling in cell lines derived from human adrenal cortex and in rar&agr;-/- mice. Results: We have identified RAR signaling as a central molecular network involved in APA formation independently of the mutation status Treatment of H295R cells with all-trans retinoic acid and 9-cis retinoic acid reduced cell viability in a time- and dose-dependent manner. This effect was due to decreased cell proliferation and increased cell apoptosis. In contrast to the effects observed in vitro, 9-cis retinoic acid did not modify tumor progression in a mouse xenograft model. RAR&agr; invalidation by shRNA transduction in H295R cells did not affect their proliferative properties, but induced a major change in cellular phenotype with cells forming spheroid structures. Investigation of the adrenal phenotype of rar&agr;-/- mice demonstrated that in young (12 weeks) and old (52 weeks) rar&agr;-/- mice the characteristic cellular arrangement of the adrenal cortex was replaced by an enlarged zona glomerulosa and a disorganized zona fasciculata, this effect being more pronounced in old mice. RAR&agr; invalidation resulted in a significant decrease of CYP11B2 expression and/or aldosterone production, both in cell and mice models. Conclusions: Our results suggest that RAR signaling contributes to normal adrenal morphology and functional zonation and that its disruption could contribute to abnormal cell proliferation in the adrenal cortex, creating a propitious environment for the emergence of specific driver mutations in APA.

Nous avons récemment démontré la présence de mastocytes dans la zone définitive de la surrénale fœtale humaine à partir de 18–20 semaines d’aménorrhée (SA). Dans la surrénale adulte normale et dans les adénomes de Conn, il a été montré que les mastocytes contrôlent la sécrétion d’aldostérone. Nous avons donc émis l’hypothèse d’une implication similaire des mastocytes dans la surrénale fœtale. Dans cette étude, nous avons analysé la cinétique d’expression des acteurs de la stéroïdogenèse et de l’apparition des mastocytes par des études immunohistochimiques et RT-QPCR dans les surrénales fœtales humaines de 18 SA jusqu’au terme. Nous avons ainsi observé une expression de la 3β-hydroxystéroïde déshydrogénase et de l’aldostérone synthase (CYP11B2) dès 18 et 24 SA, respectivement, dans la zone définitive à proximité des mastocytes. En revanche, aucune corrélation temporo-spatiale n’a été observée avec l’expression de la 17α-hydroxylase ou de la 11β-hydroxylase. De plus, l’expression du récepteur au HDL cholestérol a été détectée précocement dans la zone fœtale, avec une extension dans la zone définitive à partir de 24 SA. Les résultats de QPCR ont corroboré les données immunohistochimiques. Au total, nous montrons pour la première fois la distribution des cellules sécrétrices de gluco- et minéralocorticoïdes dans la surrénale fœtale avec une corrélation temporo-spatiale entre les mastocytes et les enzymes nécessaires à la production d’aldostérone. Cette observation suggère une contribution des mastocytes dans le contrôle de la synthèse des minéralocorticoïdes durant le développement, processus essentiel pour le maintien de l’homéostasie hydrosodée en période périnatale.

Dans le cortex surrénalien, la différenciation des cellules productrices d’aldostérone dans la zone glomérulée (ZG) est sous le contrôle de l’activation de la voie Wnt/β-caténine alors que l’activation de la voie AMPc/PKA par l’ACTH oriente la différenciation des cellules vers la synthèse de cortisol dans la zone fasciculée (ZF). La sur activation de la voie AMPc/PKA due à une hypersécrétion d’ACTH pourrait perturber l’organisation anatomique et fonctionnelle du cortex surrénalien. Nous avons étudié l’histologie surrénalienne et l’expression des enzymes de la corticostéroïdogénèse chez des patients présentant une hypersécrétion d’ACTH liée à une maladie de Cushing ou un bloc en 21-hydroxylase (21-OHD). Nos résultats révèlent une organisation des cellules en faisceaux dans la zone sous-capsulaire, et non en rosettes comme dans la surrénale saine, suggérant l’absence de ZG dans ces tissus pathologiques. Toutefois, la large distribution du marquage immunohistochimique Dab2, marqueur de ZG, dans les surrénales de patients avec 21-OHD, témoigne d’une ZG étendue. Chez ces patients, la ZG est fortement immunopositive à l’aldostérone synthase (CYP11B2) mais négative pour la 11β-hydroxylase (CYP11B1), enzyme retrouvée exclusivement dans la ZF. À contrario, le marquage Dab2 est restreint à quelques groupes de cellules chez les patients avec tumeur sécrétrice d’ACTH. La réduction de leur ZG est associée à une faible expression de CYP11B2 au niveau de l’ARNm et de la protéine, qui contraste avec une forte expression de CYP11B1. Nos résultats indiquent que la sur activation de la voie AMPc/PKA ne s’oppose pas à la différenciation fonctionnelle des cellules de la ZG des patients avec 21-OHD.

Primary aldosteronism is the most common form of secondary hypertension. Aldosterone producing adenoma (APA) and bilateral adrenal hyperplasia together account for ~95% of cases. Recurrent somatic mutations in KCNJ5, CACNA1D, ATP1A1 and ATP2B3 have been identified in APA. This has allowed us to establish a pathophysiological model involving modifications in intracellular ionic homeostasis and membrane potential leading to the activation of calcium signaling, the major trigger for aldosterone production. In parallel, we have identified retinoic acid receptor (RAR) signaling as a central molecular network involved in APA formation independently of the mutation status. The aim of our study was to evaluate the role of RAR signaling in the development of APA. To this purpose we have investigated the modulation of RAR signaling in two different cell lines derived from the human adrenal cortex as well as in different mouse models. Whereas treatment of H295R cells with 10–6M of all trans retinoic acid (ATRA, a RARα agonist) and 9-cis retinoic acid (9-cis RA, a RXRα agonist) did not affect cell viability, 10–5M of ATRA dramatically decreased cell number in a time-dependent manner. This effect was due to decreased cell proliferation and increased cell apoptosis. However, 9-cis RA did not modify tumor growth in a mouse xenograft model. RARα invalidation by shRNA transduction in H295R cells did not affect their proliferative properties, but induced a major change in cellular phenotype with cells forming spheroid structures able to grow in suspension. Investigation of the adrenal phenotype of rarα knock-out mice demonstrated that in young (12 weeks) and old (40 weeks) rarα-/- mice the characteristic cellular arrangement of the adrenal cortex was replaced by an enlarged zona glomerulosa and a disorganized zona fasciculata, this effect being more pronounced in old mice. Our results suggest that RAR signaling contributes to normal adrenal morphology and functional zonation. Disruption of RAR signaling could trigger abnormal proliferation of cells in the adrenal cortex, creating a propitious environment for the emergence of specific mutations affecting ionic channels and ATPases leading to increased aldosterone production.

Verhovez, A.; Williams, T. A.; Crudo, V.; Burrello, J.; Covella, M.; Leone, D.; Berra, E.; Fallo, F.; Fabris, B.; Amenta, F.; Gomez-Sanchez, C. E.; Veglio, F.; Mulatero, P. Author Information

The enzymes required for aldosterone synthesis from cholesterol are expressed in rat and human brains. The hypertension of Dahl salt-sensitive (SS) rats is mitigated by the intracerebroventricular (i.c.v.) infusion of antagonists of the mineralocorticoid receptor (MR) and downstream effectors of mineralocorticoid action, as well as ablations of brain areas that also abrogate mineralocorticoid-salt excess hypertension in normotensive rats. We used real time RT-PCR to measure mRNA of aldosterone synthase and 11 beta-hydroxylase, the requisite enzymes for the last step in the synthesis of aldosterone and corticosterone, respectively, MR and the determinants of MR ligand specificity, 11 beta-hydroxysteroid dehydrogenase types 1 and 2 (11 beta-HSD1&2) and hexose-6-phosphate dehydrogenase (H6PDH). A combination of extraction and ELISA was used to measure aldosterone concentrations in tissue and urine of SS and Sprague-Dawley (SD) rats. Aldosterone synthase mRNA expression was higher in the brains and lower in the adrenal glands of SS compared with SD rats. The amounts of mRNA for MR, 11 beta-hydroxylase, 11 beta-HSD1&2 and H6PD were similar. Aldosterone concentrations were greater in brains of SS than SD rats, yet, in keeping with the literature, the circulating and total aldosterone production of aldosterone in SS rats were not. The selective inhibitor of aldosterone synthase, FAD286, was infused i.c.v. or subcutaneously in a cross-over blood pressure study in hypertensive SS rats further challenged by a high-salt diet. The i.c.v. infusion of FAD286, at a dose that had no effect systemically, significantly and reversibly lowered blood pressure in SS rats. Aldosterone synthesis in brains of SS rats is greater than in SD rats and is important in the genesis of their salt-sensitive hypertension.

The central nervous system produces many of the enzymes responsible for corticosteroid synthesis. A model system to study the regulation of this local system would be valuable. Previously, we have shown that primary cultures of hippocampal neurons isolated from the fetal rat can perform the biochemical reactions associated with the enzymes 11beta-hydroxylase and aldosterone synthase. Here, we demonstrate directly that these enzymes are present within primary cultures of fetal rat hippocampal neurons.

A simple mathematical model for studying mechanism-based inhibitors (MBIs) is presented. The mathematical equations are deduced for an experimental protocol consisting of a first incubation of the enzyme in the presence of MBI followed by a washing protocol to eliminate free MBI. Finally enzyme activity (initial velocity) is measured with specific substrate. The representation of the final equation obtained is a straight line, and the MBI-specific association constant of velocity (k) can be calculated from its slope. The mathematical model was then challenged with the effect of 18-ethynyl-11-deoxycorticosterone (18-EtDOC) as an MBI on aldosterone biosynthesis from 11-deoxycorticosterone (DOC) in rat adrenal mitochondria. The last step of the mitochondrial biosynthesis of aldosterone consists of the conversion of DOC into corticosterone (B) or 18-hydroxy-11-deoxycorticosterone (18-OHDOC), and both steroids can then be transformed into aldosterone. The k (mM(-1) x min(-1)) values obtained for 18-EtDOC were: 451 +/- 36 for DOC to aldosterone; 177 +/- 16 for B to aldosterone; 175 +/- 15 for 18-OHDOC to aldosterone; and 2.7 +/- 0.2 for DOC to B. These results show that this MBI practically does not affect the metabolism of DOC to B in our enzyme preparation and that conversions of B and 18-OHDOC into aldosterone are catalyzed by the same enzyme.

TA cloning is one of the simplest and most efficient methods for the cloning of PCR products. The procedure exploits the terminal transferase activity of certain thermophilic DNA polymerases, including Thermus aquaticus (Taq) polymerase. Taq polymerase has non-template dependent activity which preferentially adds a single adenosine to the 3'-ends of a double stranded DNA molecule, and thus most of the molecules PCR amplified by Taq polymerase possess single 3'-A overhangs. The use of a linearized "T-vector" which has single 3'-T overhangs on both ends allows direct, high-efficiency cloning of PCR products, facilitated by complementarity between the PCR product 3'-A overhangs and vector 3'-T overhangs. The TA cloning method can be easily modified so that the same T-vector can be used to clone any double-stranded DNA fragment, including PCR products amplified by any DNA polymerase, as well as all blunt- and sticky-ended DNA species. This technique is especially useful when compatible restriction sites are not available for the subcloning of DNA fragments from one vector to another. Directional cloning is made possible by appropriate hemi-phosphorylation of both the T-vectors and the inserts. With a single T-vector at hand, any DNA fragment can be cloned without compromising the cloning efficiency. The universal TA cloning method is thus both convenient and labor-saving.

In operant conditioning experiments, two methods are commonly used to motivate laboratory rats to perform designated tasks. The first is restricting food so that rats are forced to lose 20% of body weight within one week, followed by maintenance at 80% of the baseline weight for the remainder of the experiment. The second is restricting access to water to 15 min in each 24 h period. These methods are effective in motivating the animals. There is, however, little information available on the effects on performance in tests of behaviour that are not related to operant conditioning. In addition, it is not clear if these commonly used methods of food and water restriction will lead to physiological stress as indicated by an elevation of serum corticosterone. Male rats were either food-restricted to reduce and maintain their weight at 80% of baseline weight, or were restricted to 15 min access to water every 24 h. Activity in the open field was significantly greater in food-restricted rats than in water-restricted or control rats, but freezing behaviour was similar in all experimental groups. Food-restricted rats had a higher mean serum corticosterone level than water-restricted and control rats 37 days after the start of the experimental period. These data suggested that chronically restricting food and maintenance of body weight at 80% of baseline body weight led to significant behavioural changes and physiological stress. In contrast, water restriction did not lead to changes in behaviour or corticosterone levels. A second experiment was conducted to compare the effects of food restriction to 80% of baseline body weight, as described above, with a less stringent protocol in which test rats were initially reduced to 80% of baseline weight, but were then maintained at 80% of an ad libitum fed control rat's weight. Serum corticosterone levels and adrenal gland weights were measured after the initial week of forced weight loss and after maintenance for 21 days. Forced loss of 20% of body weight in the first week led to significantly increased serum corticosterone levels and adrenal gland weights compared to ad libitum fed controls. Serum corticosterone levels and adrenal gland weights in rats maintained at 80% of their initial body weight for 21 days remained higher than ad libitum fed control rats. However, rats maintained at 80% of an ad libitum fed control rat's weight did not differ from control rats in serum corticosterone levels or adrenal gland weights at the end of the 21-day study period. Adjustment of the feeding regimen in this manner eliminated physiological evidence of chronic stress.

Objective To investigate phenotypic consequences of renin gene polymorphism between Lyon hypertensive (LH) and normotensive (LN) rats because previously we demonstrated cosegregation of the LH allele with increased blood pressure in a cross of LH with LN rats. Design Two studies were conducted. Study 1 used a cohort of male F2 rats from a LH × LN cross. Eighty-two rats homozygous for the hypertensive (HH) renin gene allele were compared with 82 rats homozygous for the normotensive (NN) allele. Urinary steroid excretion was measured in 24 h urine samples collected from rats aged 6 weeks. The direct aortic blood pressure was recorded in 30-week-old rats and, after they had been killed, their kidney renin concentration (KRC) was measured. In study 2, renin, angiotensinogen and angiotensin converting enzyme plasma concentrations and renin messenger RNA (mRNA) levels were measured in renal and extra-renal tissues from 6- and 25-week-old LH and LN parental and HH and NN F2 male rats. Methods Urinary steroids and plasma components of the renin–angiotensin system (RAS) were measured using specific radioimmunoassays. mRNA levels were quantified by northern blotting. Results In study 1, HH F2 rats had a higher blood pressure (151.5 ± 8.2 versus 146.0 ± 7.4 mmHg, P <0.001) and a lower KRC (514 ± 203 versus 666 ± 304 μg A1/h per g cortex, P <0.01) than did NN rats aged 30 weeks. In covariate analysis the decrease in KRC in HH rats was attributable to their increased blood pressure rather than to the renin genotype. The renin genotype of rats aged 6 weeks was not associated with a change in the urinary excretion of aldosterone, desoxycorticosterone, corticosterone or 18-hydroxy desoxycorticosterone. In study 2, we found no difference either in plasma levels of RAS components or in renal or extra-renal renin mRNA levels either between parental LH and LN rats or between HH and NN F2 rats apart from a higher plasma renin concentration in LH rats aged 6 weeks. Renal, but not extra-renal, renin mRNA levels declined with age. Conclusions We found no evidence of a renin genotypedependent phenotypic difference in the RAS that could account for the effect of the renin locus on blood pressure in Lyon rats. Our findings suggest that the effect of the locus on blood pressure might be due to an as yet unidentified gene linked to renin.

Glucocorticoid-suppressible hyperaldosteronism is a dominantly inherited form of hypertension believed to be caused by the presence of a hybrid CYP11B1/CYP11B2 gene which has arisen from an unequal crossing over between the two CYP11B genes in a previous meiosis. We have studied a French pedigree with seven affected individuals in which two affected individuals also have adrenal tumors and two others have micronodular adrenal hyperplasia. One of the adrenal tumors and the surrounding adrenal tissue has been removed, giving a rare opportunity to study the regulation and action of the hybrid gene causing the disease. The hybrid CYP11B gene was demonstrated to be expressed at higher levels than either CYP11B1 or CYP11B2 in the cortex of the adrenal by RT-PCR and Northern blot analysis. In situ hybridization showed that both CYP11B1 and the hybrid gene were expressed in all three zones of the cortex. In cell culture experiments hybrid gene expression was stimulated by ACTH leading to increased production of aldosterone and the hybrid steroids characteristic of glucocorticoid-suppressible hyperaldosteronism. The genetic basis of the adrenal pathologies in this family is not known but may be related to the duplication causing the hyperaldosteronism.

A variant cytochrome P450 11β-hydroxylase/aldosterone synthase cDNA (CYP11B2) clone, pALDO-4, was isolated and sequenced from a rat zona glomerulosa cDNA library. This clone contains a 1521 bp open reading frame coding for a 506 amino acid residue protein which has a 30 amino acid extension peptide which is 6 residues longer than that reported by others. Seven nucleotide mutations in the coding region of the mature protein were found in comparison to the reported cDNAs. These mutations resulted in three amino acid changes in the mature protein which were identical to those present in the cytochrome P450 11β-hydroxylase (CYP11B1). Two nucleotide substitutions and two nucleotide insertions were also identified in the 3′-noncoding region. The expression product of this cDNA transfected into COS7 cells converted DOC to aldosterone, corticosterone and 18-hydroxycorticosterone less efficiently than the plasmid containing the cDNA with normal 24 amino acid extension peptide.

The effects of specific active immunization against renin were compared with those of chronic angiotensin converting enzyme (ACE) inhibition. Male spontaneously hypertensive rats (SHR) were immunized (SHR-I) (n = 10) against pure murine renin (four injections of 30 micrograms/kg s.c.) or received (SHR-P) (n = 11) a converting enzyme inhibitor (perindopril, 2 mg/kg/day per os for 4 weeks). Sham-immunized SHR (SHR-S) (n = 12) and normotensive Wistar-Kyoto (WKY-S) (n = 12) rats served as controls. At 15 weeks of age, 24-hour average blood pressure was obtained in freely moving rats using intra-aortic pressure recording with computer analysis. Antirenin immunization induced high circulating titers of antibodies, a fall in plasma renin activity (-95%), and urinary excretion of mineralocorticoids. Perindopril abolished the pressor response to angiotensin I, whereas plasma ACE was only partly (-56%) decreased. It also increased plasma renin activity and did not alter the urinary excretion of steroids. Both immunization and perindopril allowed the blood pressure of SHR to return to the level of WKY-S rats and reduced the left ventricular weight. These decreases were associated with an elevated sympathetic nervous system activity as indicated by increases in the urinary excretion of catecholamines and their metabolites. It is conclude that, apart from an unaltered steroid synthesis, most of the cardiovascular effects of chronic ACE inhibition are similar to those of antirenin immunization, thus indicating that blockade of the circulating and renal renin-angiotensin system accounts for most of the effects of ACE inhibitors.

Endothelin, a hormone secreted by endothelial cells, has potent vasoconstrictive properties. Due to its potential paracrine nature, we examined the effect of endothelin-I on atrial natriuretic peptide (ANP) secretion in vitro. Isolated superfused rat left atria, paced at 2 Hz, were used for study. Endothelin (1-100 nM) increased ANP secretion in a dose-dependent manner from 1.6- to 6.7-fold above baseline. Spontaneously beating right atria increased ANP secretion by 2.3-fold in response to 10 nM endothelin without a change in beat frequency. However, the right atrial ANP secretory response was less than the 3.8-fold increase seen by left atria, and the time to peak response was slower. The calcium dependency of endothelin-stimulated ANP secretion was examined using paced left atria. The dependency of endothelin-stimulated secretion on calcium influx was examined by lowering the superfusate calcium from 1.8 to 0.2 mM. The ANP secretory response to 10 nM endothelin was reduced by 65% with 0.2 mM calcium. Influx of calcium through voltage-dependent calcium channels was examined by superfusion with 50 microM nitrendipine. Nitrendipine decreased endothelin-stimulated ANP secretion by 51% without affecting endothelin binding. The role of intracellular calcium release from the sarcoplasmic reticulum (SR) was examined by superfusion with 1 microM ryanodine, an inhibitor of SR calcium release. Ryanodine had no effect on endothelin-stimulated ANP secretion. We conclude: 1) Endothelin is a potent stimulus of ANP secretion in vitro. 2) The relative secretory response of right atria to endothelin expressed as a function of basal secretion is less and the time to peak secretion delayed relative to left atria. 3) Enhanced calcium influx, primarily through voltage-dependent calcium channels, plays a significant role in endothelin-stimulated secretion. 4) Release of intracellular calcium from the SR does not participate in the secretory response. 5) Part of the stimulatory signal appears to be independent of calcium influx or intracellular calcium release. Thus, endothelin may be an important secretagogue or modulator of ANP secretion in vivo; however, its physiological role in regulating ANP secretion in vivo remains to be determined.

SUMMARY 1. In order to determine whether the antihypertensive effect of neonatal thymectomy in genetically hypertensive rats could be mediated through altered adrenal function, systolic blood pressure (SBP) and urinary excretion of deoxycorticosterone (DOC), corticosterone (B) and aldosterone were measured in thymectomized hypertensive (LH), normotensive (LN) and low‐blood pressure (LL) rats of the Lyon strain. Sham‐operated animals served as controls. 2. Neonatal thymectomy prevented the spontaneous increase of SBP in LH rats while it slightly decreased the SBP of LN and did not change that of LL rats. 3. Five week old sham‐operated LH rats exhibited an increased urinary excretion of DOC and a decreased excretion of B compared with both LN and LL controls. Thymectomy did not alter the urinary excretion of adrenal steroids in LN and LL rats. The urinary excretion of B was markedly enhanced in thymectomized LH rats whereas that of DOC remained unmodified. 4. These data suggested that the thymus could be involved in the development of hypertension in LH rats. 5. The antihypertensive effect of thymectomy did not seem to be mediated by a decreased mineralocorticoid production in the genetically hypertensive rat of the Lyon strain.

Glucocorticoid-suppressible aldosteronism (GSA) is a rare form of hyperaldosteronism in which the increased secretion of aldosterone and the elevation of blood pressure are corrected when ACTH secretion is suppressed. Two 17-hydroxylated analogs of 18-hydroxycorticosterone and aldosterone, 18-hydroxycortisol and 18-oxocortisol, which had been identified in the urine of patients with hyperaldosteronism due to an adrenal adenoma and in bullfrog adrenal tissue incubated with cortisol, are produced in greater than normal quantities in patients with GSA. The excretion of 18-hydroxycortisol and 18-oxocortisol in nine patients with GSA was 2914 +/- 923 (+/- SD) nmol/day [1108 +/- 351 micrograms/day; normal, 165 +/- 94 nmol/day (63 +/- 36 micrograms/day)] and 141 +/- 77 nmol/day [53 +/- 29 micrograms/day; normal, 3.2 +/- 2.4 nmol/day (1.2 +/- 0.9 micrograms/day)], respectively. The excretion of aldosterone 18-oxoglucuronide was 53 +/- 18 nmold/day [19.4 +/- 6.8 micrograms/day; normal, 16.9 +/- 7.5 nmol/day (6.1 +/- 2.7 micrograms/day)]. Aldosterone excretion was elevated in six patients and within the normal range in three patients. The degree of abnormality in 18-hydroxycortisol and 18-oxocortisol excretion was significantly greater than that in aldosterone. Dexamethasone administration decreased excretion of the three steroids to the normal range. ACTH administration for 3 days resulted in an exaggerated increase in the excretion of these steroids, suggesting ACTH dependence of these steroids in patients with GSA. The excessive production of these steroids, which are 17- and 18-hydroxylated, indicate that they are produced by hybrid-type cells which we have called transitional cells, which are also capable of producing aldosterone. These findings are consistent with the postulate that cytochrome P-450-corticosterone methyl oxidase fails to disappear normally in the zona glomerulosa cells as they migrate to the zona fasciculata and acquire 17-hydroxylase activity. This abnormality explains the supernormal conversion of cortisol to 18-hydroxycortisol and 18-oxocortisol and the ACTH dependence of aldosterone secretion in GSA.