The 11beta hydroxysteroid dehydrogenase 2 exists as an inactive dimer.

The 11beta-hydroxysteroid dehydrogenase types 1 and 2 enzymes (11beta-HSD1 and 11beta-HSD2), modulate glucocorticoid occupation of the mineralocorticoid and glucocorticoid receptors by interconverting corticosterone and cortisol to the inactive metabolites 11-dehydrocorticosterone and cortisone within the target cells. The NAD(+)-dependent 11-HSD 2 in the kidney inactivates corticosterone and cortisol, allowing aldosterone, which is not metabolized, access to the receptor. Studies of the kinetics of 11-HSD 2 activity in the rat kidney have produced inconsistent results. Western blots done in the absence of the reducing agent beta-mercaptoethanol showed two bands with approximate MW of 40 and 80 kDa. When beta-mercaptoethanol was used, only the 40 kDa was detected, indicating that under non-denaturing conditions a significant proportion of the 11beta-HSD 2 exists as a dimer. NAD(+)-dependent conversion of 3H-corticosterone by 20 microg of microsomal protein increased approximately 10 fold with the addition of 5 mM DTT concentration. NADP(+)-dependent activity with 20 microg of microsomal protein was very low and did not change significantly when using DTT. In the presence of DTT, the predominant 11-HSD activity in the rat kidney is NAD(+)-dependent with a K(m) of 15.1 nM, similar to that of the cloned and expressed enzyme. These data suggest that dimerization and subsequent enzyme inactivation occur when protocols promoting oxidation of this protein are used.

Aldosterone esters and the heart.

There are clinical and experimental situations in which symptoms of mineralocorticoid excess are remediable with mineralocorticoid receptor antagonist treatment, in spite of paradoxically low levels of plasma renin and aldosterone. Several decades ago, a factor isolated from the heart was described that had mineralocorticoid properties like those of aldosterone, but much more potent. It was thought to be similar to aldosterone-18-monoacetate or -21-monoacetate, acetyl derivatives of aldosterone that are very rapidly hydrolyzed in the circulation. In our efforts to confirm and extend these observations, we extracted rat hearts and plasma harvested in a manner that would minimize hydrolysis. The product was subjected to several forms of TLC and HPLC and compared to several acetylated derivatives of aldosterone standards. We found that 68% of the aldosterone extracted from fresh myocardium corresponded to an aldosterone derivative that migrates at the same rate as aldosterone-20-monoacetate. The identity of this compound awaits definitive analysis. Tritiated aldosterone-21-monoacetate hydrolyzed to form aldosterone very rapidly; negligible monoacetate remained in blood left at 37 degrees C for 5 min or in hearts left at room temperature for 30 min. Regulation of aldosterone production serves the requirements of fluid and electrolyte homeostasis provided by transport epithelia, primarily that of the kidney. Nonepithelial actions of aldosterone would be freed of these regulatory constraints if the formation of a more potent derivative of the parent compound to which it is almost immediately hydrolyzed in the circulation were regulated within the nonepithelial target tissues.

Role of central mineralocorticoid receptors in cardiovascular disease.

Mineralocorticoids act directly through their receptors in specific centers in the central nervous system, kidneys, heart, and vascular smooth muscle to mediate hemodynamic homeostasis. These steroids also modulate renal and cardiovascular function indirectly through the autonomic nervous system. Complex homeostatic mechanisms under normal hormonal control become pathogenic when there is an excess of regulatory hormone. Experiments in which mineralocorticoid receptor antagonists or antisense oligodeoxynucleotides were administered centrally have clearly shown that centrally mediated effects on salt appetite, baroreceptor function, and autonomic drive to the renal and cardiovascular systems are crucial to the pathogenesis of hypertension and cardiovascular disease of hyperaldosteronism, and certain forms of genetic hypertension.

An alternatively spliced rat mineralocorticoid receptor mRNA causing truncation of the steroid binding domain.

We attempted to clone the putative 11-dehydrocorticosterone receptor by RT-PCR with two degenerate primers from highly homologous regions of the DNA and steroid binding domains of the receptor subfamily. In doing so, we have identified an alternatively spliced variant mRNA of the rat mineralocorticoid (MR) with a ten bp deletion in the C-terminal steroid binding domain. This deletion results in a truncated MR receptor of 807 amino acids in comparison to the wild type of 981 amino acids. The deletion variant was expressed in colon, kidney, heart, liver, aorta and brain tissues. The relative abundance of the deletion variant compared to the wild type MR was estimated to be 6% in rat kidney and 4% in hippocampus. This deletion was also detected in human kidney by RT-PCR. Site-directed mutagenesis was used to create the eukaryotic expression plasmid pCR3-rMRdel10 from the wild type for a transactivation assay using the luciferase reporter system in CV-1 cells. The deletion variant had the same baseline transactivation activity as the wild type MR, but did not respond to aldosterone or corticosterone stimulation. Co-transfection of MR with the deletion variant had no significant effect on transactivation activity of the MR, indicating that the deletion variant is unlikely to serve as a negative regulator of MR function.

Maternal hypertension and progeny blood pressure: role of aldosterone and 11beta-HSD.

Epidemiological and experimental evidence suggests that gestational events modulate the level of blood pressure that will be "normal" for the individual as an adult. Glucocorticoid excess during gestation is associated with low birth weight, a large placenta, and adult hypertension in humans and animals. It has been proposed that the deficiency in placental 11beta-hydroxysteroid dehydrogenase activity in humans produces a gestational hormonal milieu, notwithstanding normal circulating levels of glucocorticoids, that predisposes the adult progeny to hypertension. Animal studies indicate that maternal hypertension, excess glucocorticoids, and hydroxysteroid dehydrogenase inhibition program adult blood pressure. Blood pressures of Sprague-Dawley rat dams were manipulated during gestation with continuous intracerebroventricular infusions of vehicle, aldosterone, 11alpha-hydroxyprogesterone, or carbenoxolone at doses known to produce hypertension with no renal effects or with subcutaneous infusions of larger, equally hypertensinogenic doses that produce systemic effects. Blood pressures of all treated dams were significantly greater (P<0.01) during gestation than those of the vehicle ICV control rats but not significantly different from each other. The blood pressures of both male and female progeny (n>/=6 per group, comprising representatives from at least 4 litters) were measured after 6 weeks of age. No significant difference was found in the blood pressure of the pups regardless of the maternal gestational blood pressure or treatment with an enzyme inhibitor, even after high-salt diet challenge.

Embryo transfer in the rat as a tool to determine genetic components of the gestational environment.

Embryo transfer, with the recipient dam nursing the transferred progeny, was used to study the impact of the gestational environment on adult blood pressure (BP) in three inbred rat strains, the hypertensive Dahl salt-sensitive SS/JrCtr, the normotensive Dahl salt-hypertension resistant SR/Jr, and the normotensive Dark Agouti rat. Rats that had been cross-fostered within 6 h of birth were included as a control for lactational and nurturing factors. Systolic BP was measured by tail-cuff plethysmography twice a week in rats after the age of 7 weeks. Embryo transfer success, measured as the percentage of embryos transferred resulting in pups weaned at 4 weeks, was 27% between the SS/JrCtr and SR/Jr and 53% for the SS/JrCtr and Dark Agouti. This assessment included all failures, some of which probably were not associated with the transfer. If only the number of embryos transferred to dams with successful pregnancies was included, the success rate was 48% between the SS/JrCtr and SR/Jr and 82% between the SS/JrCtr and Dark Agouti strains. Anomalies in pups were not evident. In contrast to the lactational environment, the gestational milieu had a profound effect on basal blood pressure of the hypertensive SS/JrCtr progeny, less of an effect on that of the Dark Agouti, and no effect on that of the SR/Jr. Although the SS/JrCtr strain is significantly larger than the SR/Jr and Dark Agouti strains, neither embryo transfer nor cross-fostering altered body weight of rats at the age of 6 weeks. These data indicate that embryo transfer can be an easy and efficient method of isolating genetically determined factors of the gestational environment.

Modulation of blood pressure in the Dahl SS/jr rat by embryo transfer.

Gestational hypertension and malnutrition are associated with hypertension and ischemic heart disease in the adult human. The impact of the gestational environment on the adult blood pressure in two well-characterized genetically homogeneous rat strains, the hypertensive SS/jr and normotensive SR/jr, was studied by cross-fostering within 6 hours of birth and by embryo transplantation with the recipient dam nursing the transplanted pups. Systolic blood pressure (BP) was measured by tail-cuff plethysmography twice a week after the age of 7 weeks. The lactational environment (cross-fostering) had no effect on blood pressure. Embryo transfer between like strains had no effect on the development of hypertension, nor did the BP of R transferred to S (RetS) differ from that of normal R or RetR. At 7 weeks of age, the BP of SetR was significantly lower than that of S or SetS (P<.01) and was similar to that of RetR and R. With age, the blood pressures of the S, SetS and SetR increased at approximately the same rate but from a significantly different baseline. Salt-sensitivity in the S and resistance in the R were not altered. The protective effect of the R gestational environment on SetR female BP was abrogated during whelping and lactation. Embryo transfer and cross-fostering did not alter the weight of rats older than 7 weeks. Because the BP of the R dams were significantly lower than that of the S dams, these studies do not distinguish between the effects of the R dams' lower blood pressure per se and hormonal influences of the R uterus on the S blood pressure phenotype.

Central hypertensive effects of aldosterone.

The soluble mineralocorticoid receptor bound to an agonist acts as a transcription factor for several genes relevant to ion transport by kidney and colon epithelial cells and is a major regulator of electrolyte and fluid homeostasis. Mineralocorticoids, the most prominent of which is aldosterone, also influence the activity of nonepithelial target cells, including vascular smooth muscle cells, by altering intracellular ion transport and content. Evidence is summarized for mineralocorticoid modulation of neuronal activity in a center or centers within the brain, probably in the periventricular area of the anterior hypothalamus, where information on electrolyte, fluid, and cardiovascular status is received and integrated, resulting in alterations in central sympathetic efferent activity. These functions are distinct from central aldosterone effects on salt appetite and peripheral trophic effects on cardiovascular tissue. The isolated mineralocorticoid receptor binds several adrenal steroids, including aldosterone and the major glucocorticoids, with equal affinity. Ligand specificity for the mineralocorticoid receptor differs between tissues, including different organs in the brain. Specificity is conferred extrinsically by the 11-beta-hydroxysteroid dehydrogenase enzymes in transport epithelia, but mechanisms for mineralocorticoid ligand specificity have not been completely defined in the brain. The functional interaction between the mineralocorticoid receptor bound to different ligands and between the mineralocorticoid and glucocorticoid receptors is complex and as yet unresolved. Evidence is presented for the de novo synthesis of adrenal corticosteroids in the brain which may, by paracrine regulation of central control mechanisms, be relevant for certain clinical and experimental forms of hypertension characterized by low circulating levels of mineralocorticoids which respond to mineralocorticoid receptor antagonists.

Aldosterone biosynthesis in the rat brain.

Messenger RNA (mRNA) for enzymes involved in adrenal steroid biosynthesis are expressed in the brain, and the coded enzymes have been shown to be active. The expression of mRNA for the cytochrome P-450 enzyme aldosterone synthase, crucial for the final step in the synthesis of aldosterone and the synthesis of aldosterone was studied in several anatomic areas of the rat brain. Expression of the mRNA for the aldosterone synthase was demonstrated by RT-PCR/Southern blot in adrenal, aorta, hypothalamus, hippocampus, amygdala, cerebrum, and cerebellum. Incubation of brain minces from intact and adrenalectomized rats demonstrated the synthesis of corticosterone and aldosterone from endogenous precursors. Incubations of brain minces with [1,2(3)H]-deoxycorticosterone, followed by extraction and three different successive TLCs, demonstrated the presence of labeled aldosterone, corticosterone, and 18-hydroxy-deoxycorticosterone. Incubation, in the presence of 10 microM cortisol or metyrapone, inhibited the synthesis of aldosterone or both aldosterone and corticosterone, respectively. These studies indicate that the rat brain has the enzymatic machinery for the synthesis of adrenal corticosteroids and is capable of synthesizing aldosterone. Aldosterone synthesized in the brain might play a paracrine role in the regulation of blood pressure.

Cloning of two alternatively spliced 21-hydroxylase CDNAs from rat adrenal.

Interest in extra-adrenal corticosteroid synthesis has been revived by technological advances and the quest for answers to clinical problems. The cytochrome P450 21-hydroxylase converts progesterone to deoxycorticosterone, the obligatory substrate for the production of the main adrenal steroids aldosterone, cortisol and corticosterone. The rat P450 21-hydroxylase was cloned and two constructs, 21OH-5 and 21OH-6, sequenced. The constructs are similar, except that 21OH-6 has three additional major insertions of 64, 70 and 84 bp, a 3 bp deletion, and four extra base pairs immediately before the poly-A sequence. The entire coding region of 21OH-5 has 87 and 71% homology with the mouse and human 21-hydroxylase cDNA, respectively, whereas the encoded protein has 84 and 65% homology. Reverse transcriptase-polymerase chain reaction (RT-PCR) combined with Southern blot demonstrated expression of both transcripts in the kidney, aorta, liver, cerebellum, hypothalamus and brain stem, heart and cerebrum, but not the hippocampus, in addition to the adrenal. The entire coding region of 21OH-5 and the corresponding region of 21OH-6 including the three introns were cloned into pCR3 and the plasmids transiently transfected into COS-7 cells. Only 21OH-5 was translated into active protein, converting approximately 64% of 3H-progesterone to deoxycorticosterone in 2 h.

Inhibition of steroidogenesis in rat adrenal cells by 18-ethynyldeoxycorticosterone: evidence for an alternative pathway of aldosterone biosynthesis.

The effect of the mechanism-based inhibitor 18-ethynyldeoxycorticosterone (18-E-DOC) on the late steps of the aldosterone biosynthetic pathway was examined in freshly isolated cells of the zona glomerulosa (ZG) and fasciculata (ZF) from rat adrenal glands. ZG synthesis of aldosterone was inhibited by 18-E-DOC in a time- and concentration-dependent manner with a Ki of approximately 0.05 microM. The maximal degree of inhibition of ZG production of aldosterone and 18-hydroxycorticosterone (18-OH-B) was approximately 80%. ZF cells, perhaps surprisingly, were found to secrete 18-OH-B at levels approximately one-third to one-fourth those of ZG cells and the Ki of 18-E-DOC inhibition of 18-OH-B secretion was approximately 10 microM for ZF cells, 200-fold higher than for ZG cells. The inhibitor had no effect on the secretion of corticosterone by either ZG or ZF, and the secretion of 18-hydroxydeoxycorticosterone (18-OH-DOC) by both the ZG and ZF was inhibited only to a minor degree. 18-E-DOC inhibited the biosynthesis of aldosterone by ZG cells incubated with 10 microM added DOC or 18-OH-DOC by approximately 75%, similar to the degree of inhibition of aldosterone biosynthesis from endogenous substrate, whereas ZF biosynthesis of 18-OH-B from either substrate was inhibited by less than 40%. ZF cells do not express aldosterone synthase, the only enzyme known to convert 18-OH-DOC into 18-OH-B. Incubation of MA-10 cells stably transfected with the cDNA of the rat aldosterone synthase with 18-E-DOC resulted in a complete inhibition of the conversion of DOC to aldosterone with a Ki of approximately 0.02 microM. In addition, transfected cells expressing 11beta-hydroxylase convert DOC to 18-OH-B in very small quantities only and cannot convert 18-OH-DOC to 18-OH-B. These data suggest that neither 11beta-hydroxylase nor aldosterone synthase are responsible for the biosynthesis of 18-OH-B by ZF cells from DOC or 18-OH-DOC, that 20% of aldosterone synthesis appears not to be attributable to the actions of aldosterone synthase and that an unknown CYP11B enzyme is also involved in the biosynthesis of 18-OH-B.

The sheep kidney contains a novel unidirectional, high affinity NADP(+)-dependent 11 beta-hydroxysteroid dehydrogenase (11 beta-HSD-3).

The 11 beta-hydroxysteroid dehydrogenase (11 beta-HSD) enzymes convert corticosterone and cortisol to 11-dehydrocorticosterone and cortisone, and are thought to convey extrinsic specificity to the mineralocorticoid receptor by limiting access of the relatively more abundant glucocorticoids to it. Two different 11 beta-hydroxysteroid dehydrogenases (11 beta-HSD) have been described and cloned. The liver-type, NADP(+)-dependent 11 beta-HSD-1, has an affinity in the micromolar range and bidirectional activity. The NAD(+)-dependent 11 beta-HSD-2 has a higher affinity, in the nanomolar range, and exhibits only oxidase activity. 11 beta-HSD-2, because of its affinity and co-localization with the mineralocorticoid receptor, is likely to serve as the "gatekeeper" for the mineralocorticoid receptor in the kidney. Although the rat kidney expresses both isoforms, only the high-affinity, NAD(+)-dependent 11 beta-HSD-2 has been reported in the sheep kidney. We found both 11 beta-HSD NAD(+)- and NADP(+)-dependent activities in sheep kidney to be present. The NAD(+)-dependent activity exhibited a Km similar to that reported in the literature, 3.85 +/- 1.28 nM for corticosterone and 21.3 +/- 5.8 for cortisol, was distributed in approximately equal amounts between microsomes and nuclei, and was unidirectional, converting corticosterone to 11-dehydrocorticosterone. The enzyme exhibited prominent substrate inhibition. The NADP(+)-dependent activity had a Km for corticosterone of 4 +/- 1.3 nM for a Km for cortisol of 35.2 +/- 2 nM, 100-fold lower than that described for the 11 beta-HSD-1 in the liver of sheep and other species, and was more prevalent in the microsomes than the nuclei. This enzyme was not inhibited by its substrate. The NAD(+)-dependent activity was approximately 3-10 times greater than the NADP(+)-dependent activity when incubated with 5 nM corticosterone substrate, but had similar activity when incubated with 100 nM substrate concentrations. CHOP cells (a modified Chinese hamster ovary cell line) transiently transfected with the sheep 11 beta-HSD-2 plasmid exhibited a marked preference for NAD+ as co-factor. Oxidation of corticosterone by transfected cells in the presence of NADP+ was present, but minimal; NADP+ did not support the metabolism of cortisol, the primary glucocorticoid of sheep. These data suggest the existence of another NADP(+)-dependent enzyme, 11 beta-HSD-3, which, because of its high affinity and unidirectional oxidase activity, may play a physiological role in the modulation of glucocorticoid binding to both the mineralocorticoid and glucocorticoid receptors.

Regulation of the 11 beta-hydroxysteroid dehydrogenase in the rat adrenal. Decrease enzymatic activity induced by ACTH.

Patients with ectopic ACTH syndrome often develop hypertension and hypokalemic alkalosis with an abnormal increase in the ratio of plasma cortisol to cortisone, indicating that 11 beta-hydroxysteroid dehydrogenase (11 beta HSD) activity is inhibited. Inhibition of 11 beta HSD allows access of cortisol or corticosterone to the mineralocorticoid receptor where it act as a mineralocorticoid. Two isozymes, 11 beta HSD-1 and 11 beta HSD-2, have been cloned and characterized. The rat adrenal expresses the mRNAs for 11 beta HSD-2 and, in lesser amounts, 11 beta HSD-1. We investigated the effect of ACTH on the 11 11 beta HSD-2 activity in the rat adrenal. Rat adrenal cells zone fasciculata (ZF) were dispersed and incubated separately with increasing concentrations of ACTH for 90 min, and secretion of corticosterone (B) and 11-dehydrocorticosterone (A) in the media was measured by enzyme-linked immunoabsorbent assays (ELISA). The conversion of [3H]B to [3H]A in the presence of 0.5 mM NAD+ was evaluated in microsomes prepared from dispersed cells preincubated for 30 min with cyanoketone and metyrapone followed by incubation for 30 min with the same inhibitors, with and without 10 nM ACTH. The dispersed cells of the ZF produced significant amounts of A which increased with ACTH. The basal B/A ratio was 0.97 +/- 0.05. ACTH caused a concentration-dependent increase in the ratio of B/A with a maximum ratio of 9.58 +/- 0.20. ACTH also inhibited the conversion of [3H]B to [3H]A in microsomes in which endogenous B production was inhibited by cyanoketone and metyrapone. ACTH did not change the K(m) for B conversion, but the Vmax was reduced significantly (1.73 +/- 0.43 pmol/min. mg protein), indicating that ACTH suppressed the 11 beta HSD-2 in a noncompetitive fashion. Dibutyryl cyclic AMP (dcAMP) also produced a concentration-dependent increase in the B/A ratio, but various concentrations of calcium did not affect the enzyme activity. In summary, adrenal cells treated with ACTH results in a significant increase in the ratio of B/A in the ZF owing a noncompetitive inhibition of the 11 beta HSD-2 via the ACTH receptor.

Corticosteroid synthesis in the central nervous system.

The possibility that adrenocorticosteroids might be synthesized in the central nervous system was assessed by RT-PCR using primers for the CYP11B1 gene which codes for 11 beta-hydroxylase, the enzyme responsible for corticosterone and cortisol formation in the zona fasciculata, incubation of minces of several areas of the brain with 3H-DOC and measuring steroid metabolites, and determining the effect of the intracerebroventricular infusion of the 11 beta-hydroxylase mechanism-based inhibitor 19-ethynyldeoxycorticosterone upon the salt-induced increase in blood pressure in SS/jr rats. Significant, though small relative to the adrenal, amounts of mRNA for 11 beta-hydroxylase was found in the aorta, cerebrum, cerebellum, hippocampus, hypothalamus and amygdala, but not in the heart. Brain minces converted 3H-DOC to corticosterone and 11-dehydrocorticosterone to a greater degree than to 18-OH-DOC. The effect of 19-ethynyldeoxycorticosterone was dose dependent, with the lower doses preventing salt-induced hypertension and the higher doses having no effect or increasing the blood pressure.

11 beta-Hydroxysteroid dehydrogenase type 2 complementary deoxyribonucleic acid stably transfected into Chinese hamster ovary cells: specific inhibition by 11 alpha-hydroxyprogesterone.

The 11 beta-hydroxysteroid dehydrogenase type 2 (11 beta HSD-2) enzyme is thought to confer aldosterone specificity upon mineralocorticoid target tissues by protecting the mineralocorticoid receptor from binding by the more abundant glucocorticoids, corticosterone and cortisol. We have developed a Chinese hamster ovary cell line stably transfected with a plasmid containing the rat 11 beta HSD-2 complementary DNA. This cell line has expressed the enzyme consistently for many generations. The 11 beta HSD-2 was located primarily in the microsomes, but significant amounts also existed in the nuclei and mitochondria. The enzymatic reaction was unidirectional, oxidative, and inhibited by the product, 11-dehydrocorticosterone, with an IC50 of approximately 200 nM. The K(m) for corticosterone was 9.6 +/- 3.1 nM, and that for NAD+ was approximately 8 microM. The enzyme did not convert dexamethasone to 11-dehydrodexamethasone. Tunicamycin, an N-glycosylation inhibitor, had no effect on enzyme activity. 11 alpha-Hydroxyprogesterone (11 alpha OH-P) was an order of magnitude more potent a competitive inhibitor of the 11 beta HSD-2 than was glycyrrhetinic acid (GA) (approximate IC50 = 0.9 vs. 15 nM). 11 beta OH-P, progesterone, and GA were almost equipotent (IC50 = 10 and 6 nM, respectively), and 5 alpha-pregnandione and 5 beta-pregnandione were less potent (IC50 = 100 and 500 nM, respectively) inhibitors of the enzyme. When the inhibitory activities were examined with intact transfected cells, 11 alpha OH-P was more potent than GA (IC50 = 5 and 150 nM, respectively). 11 alpha OH-P was not metabolized by 11 beta HSD-2. We were unable to demonstrate the presence of 11 alpha OH-P in human urine. In conclusion, a cell line stably transfected with the rat 11 beta HSD-2 was created, and the enzyme kinetics, including inhibition, were characterized. 11 alpha OH-P was found to be a potent relatively specific inhibitor of the 11 beta HSD-2 enzyme. Its potential importance is that it is the most specific inhibitor of the 11 beta HSD-2 so far encountered and would aid in the study of the physiological importance of the isoenzyme.

Mineralocorticoids, salt and high blood pressure.

Essential hypertensive patients often respond to treatments mitigating mineralocorticoid action, even though circulating levels of these steroids are within normal ranges. In addition to the kidney, mineralocorticoid or Type I receptors are found in the brain and vascular smooth muscle where they mediate effects associated with several forms of experimental hypertension. Studies in which discrete anatomic or functional areas of the brain have been ablated demonstrate that the periventricular areas of the hypothalamus and the central sympathetic and baroreceptor systems are crucial for the development of hypertension in the renoprival, DOCA salt, and Dahl salt-sensitive rat. Intracerebroventricular (i.c.v.) infusion of aldosterone in both rats and dogs at doses that do not raise serum levels above normal produce hypertension. The hypertension produced by systemic mineralocorticoid excess, adrenal regeneration, and i.c.v. or oral administration of glycyrrhetinic acid or carbenoxolone in genetically normotensive rats and by dietary salt in the Dahl salt-sensitive rat is inhibited by the i.c.v. infusion of a mineralocorticoid receptor antagonist or a Na+ channel-selective amiloride analog. Recent data demonstrate the extraadrenal synthesis of steroids in aortic endothelial cells, smooth muscle cells and the brain. The role of the extraadrenal synthesis of steroids raises new avenues for research into the causes of hypertension.