The brains of rats and humans express the enzymes required for the synthesis of aldosterone from cholesterol, including the 3β-steroid dehydrogenase that catalyzes the conversion of pregnenolone to progesterone in the pathway of adrenal steroid synthesis. Salt-induced hypertension in the Dahl inbred salt-sensitive (SS/jr) rat is associated with normal to low levels of circulating aldosterone, yet it is abrogated by the central infusion of mineralocorticoid receptor antagonists. To test the hypothesis that de novo synthesis of aldosterone in the brain has a pathophysiological role in the salt-induced hypertension of the SS rat, the 3β-steroid dehydrogenase antagonist trilostane was infused continuously intracerebroventricularly or subcutaneously in two different cohorts of Dahl SS/jr rats, one female, the other male, during and after the development of salt-induced hypertension. The doses of trilostane used had no effect on blood pressure when infused subcutaneously. Animals receiving vehicle intracerebroventricularly experienced a 30- to 45-mmHg increase in systolic blood pressure measured by tail cuff. The intracerebroventricular, but not subcutaneous, infusion of 0.3 μg/h trilostane effectively blocked the increase in systolic blood pressure and reversed the hypertension produced by drinking 0.9% saline. Trilostane was equally effective in female and male rats. Weight gain, serum aldosterone and corticosterone concentrations, and behavior assessed subjectively and by elevated plus maze were unchanged by the trilostane treatment. These studies suggest that the synthesis in the brain of a mineralocorticoid receptor agonist, probably aldosterone, is responsible in part for the salt-induced hypertension of the inbred Dahl SS/jr rat.
- aldosterone synthesis
all of the enzymes required for the synthesis of adrenal corticosteroids from cholesterol are expressed in specific anatomic areas of the rat brain, albeit in very small amounts (7, 12, 31, 32). The mRNAs for these enzymes have also been demonstrated in human brain (46). In addition, ancillary proteins required for de novo synthesis of adrenal steroids, StAR protein, and adrenodoxin and adrenodoxin reductase are also expressed in the brain (7, 8, 26, 33, 39). Aldosterone, corticosterone, and 18-hydroxycorticosterone are synthesized by minces of specific regions of brains from adrenalectomized rats from endogenous substrates and exogenous tritiated precursors (12, 13). The amount of corticosteroid biosynthesis by brain tissue in vitro is exceedingly small, even when exogenous substrates are provided, and circulating aldosterone and corticosterone in plasma are undetectable in the adrenalectomized individual. Therefore, the in vivo activity of these steroids, if any, must be local, either paracrine or autocrine.
Aldosterone synthesis within the brain may be important in the development of hypertension in the inbred Dahl salt-sensitive (SS/jr) rat. The SS/jr are spontaneously hypertensive on a normal 0.3% NaCl diet, but salt challenge greatly accelerates the development and increases the amplitude of the hypertension. We have confirmed the reports of Baba et al. (1), as well as others, that plasma levels of aldosterone of the SS/jr rats are normal and lowered further by a high-salt diet. Activation of mineralocorticoid receptors (MR) within the brain is responsible in great measure for the salt-induced increase in blood pressure, since the central infusions of MR antagonists at doses that have no effect systemically block the hypertension in the SS/jr rat (18). The central effect of MR activation, like that in epithelial cells of the kidney and colon, includes the increase in epithelial sodium channel (ENaC) activity. Salt-induced hypertension in the SS/jr rat is also blocked by the intracerebroventricular infusion of an ENaC-selective amiloride, benzamil, at doses lower than that required to lower the blood pressure systemically (21). This sensitivity to antagonists of aldosterone action, even though circulating aldosterone concentrations are not elevated, may be analogous to that seen in patients with low-renin essential hypertension who respond to antimineralocorticoid therapy with spironolactone or amiloride, even though their aldosterone levels are not elevated (4). The salt-induced hypertension in the SS/jr is partially inhibited by the chronic intracerebroventricular infusion of 19-ethinyl DOC, a suicide inhibitor of the aldosterone synthase, at a dose that had no effect when infused systemically (12). However, 19-ethinyl DOC is a partial agonist, and its toxicity limits studies of aldosterone synthase (8).
Cholesterol is successively converted to pregnenolone, then progesterone and 11-DOC by the sequential actions of the cytochrome P-450 side chain cleavage, 3β-hydroxysteroid dehydrogenase/isomerase (3β-HSD), and cytochrome P-45021 enzymes, respectively. DOC is converted to corticosterone by cytochrome P-450–11β-hydroxylase (CYP11B1) and by a series of hydroxylations to aldosterone by cytochrome P-450-aldosterone synthase (CYP11B2). The predominant neurosteroids are GABAA receptor-modulating derivatives produced from pregnenolone and progesterone through the action of the 5α-reductase-3α-hydroxysteroid enzyme complex (30, 38). The CYP-21 hydroxylase, which produces DOC from progesterone, and 11β-hydroxylase and aldosterone synthase, which use DOC as substrate, are expressed in very low amounts in the brain (46, 47). The expression of the CYP-21 hydroxylase was in question before methods to detect very low levels were developed (12, 34), and it appears that the cytochrome P-450–2D4 is the most likely enzyme to catalyze the 21-hydroxylation in the brain (27). Trilostane is an inhibitor of 3β-HSD that was developed as an adrenal steroid synthesis inhibitor and has been used for almost 30 years to study adrenal, gonadal, and neurosteroid synthesis and metabolism and physiological activity, as well as the treatment of Cushing’s syndrome, adrenal adenomas, and breast cancer (6, 25, 41). Trilostane does not have MR antagonistic activity (24, 41).
Some neurosteroids derived from progesterone modulate GABAA receptors and decrease seizure threshold. Extrapolating from toxicology data, we selected doses of trilostane that would decrease progesterone levels and its derivatives, but not to levels that caused seizures (28, 37, 45). The intent was to reduce the amount of progesterone available for the formation of DOC, the precursor for aldosterone and corticosterone and a mineralocorticoid itself, to a level below that which allowed the production of enough aldosterone to inappropriately activate the MR in the brain that are responsible for the hypertension in the Dahl SS rat. We are reporting the effect of the chronic intracerebroventricular infusion of trilostane, an inhibitor of the 3β-HSD, on the blood pressure of SS rats challenged with a high-salt diet.
MATERIALS AND METHODS
These studies were carried out in young adult female and male inbred Dahl SS rats, SS/jrctr. The original SS/jrctr were kindly given to us by John Rapp, Medical College of Ohio, in 1984 and have been maintained as a closed colony since. Husbandry and all procedures since 1984 have followed the National Research Council Guide for the Care and Use of Laboratory Animals and were performed in American Association for the Accreditation of Laboratory Animal Care-accredited facilities. The animal care and use protocol for the current studies was approved by the Jackson Veterans Affairs Institutional Animal Care and Use Committee. Rats born within 7 days of each other were used in each experiment. Littermates were randomly assigned to all groups, with no more than two littermates in the same group. Males and females were studied in separate experiments.
Trilostane (2α-cyano-4α,5α-epoxyandrostan-17β-ol-3-one; Sanofi, Malverne, PA) was dissolved in 45% 2-hydroxypropyl β-cyclodextrin in a proportion of 1:2 molar ratio (RBI, Natick, MA) and then diluted to 5% β-cyclodextrin with saline. The control solution was 5% β-cyclodextrin. Placement of the intracerebroventricular cannulas and pumps was as described previously (14). The doses were calculated from previous reports to partially block the 3β-HSD (45).
For the experiment with the female rats, 14-day miniosmotic pumps (Alza, Palo Alto, CA) delivering 0.48 μl/h were used to deliver either 0.1 or 0.3 μg/h trilostane intracerebroventricularly and changed every 13–14 days. At 6 wk, the time of the third pump change, the pumps of the animals receiving 0.1 μg/h were switched to deliver 0.3 μg/h. The pumps in the group receiving 0.3 μg/h were switched to vehicle. For the experiment with the males, 28-day pumps delivering 0.25 μl/h were used. Two groups received vehicle intracerebroventricularly, one group received 0.3 μg/h icv, and one group received 0.3 μg/h sc. At 3.5 wk, the pumps were changed. The pump content of one of the vehicle intracerebroventricular groups was changed to deliver 0.3 μg/h icv; the rest of animals continued to receive the original infusions.
Systolic blood pressure.
Systolic blood pressure was measured by tail-cuff plethysmography in trained unheated animals between 8:00 and 12:00 AM two to three times a week. To minimize stress, no animal was restrained for >10 min at a time. If a rat did not become settled immediately after entering the tube, it was allowed cage time before trying to measure the pressure again. The time that it took the rat to settle in its tube and relax its tail was noted and used as a subjective measure of anxiety. Body weight was recorded after each blood pressure measurement.
Elevated plus maze.
Elevated plus maze was originally designed to assess anxiety but also allows easy observation of ataxia, tremors, or other motor or behavioral abnormalities (2, 43). The apparatus comprises four connecting limbs, each 50 cm long × 10 cm wide, in the shape of a “+,” two limbs with, and two without, 10-cm side walls. The rat was placed in the center and allowed to explore for 5 min. It was scored by recording the total distance in centimeters that it traveled, with travel in the sideless arms weighted by multiplying by two. The more anxious and fearful, the less distance will be traveled, especially in the sideless arms of the apparatus. The number of times that the rat urinates or defecates also correlates positively with anxiety (2, 43). The females were observed on the maze three times, one time before the experiment started, just before the crossover in treatment, and at the end of the experiment. The males were tested one time, near the end of the experiment, to avoid a training bias.
The week before the end of the experiment, the male rats were placed in stainless steel rat metabolism cages for 3 days of an acclimatization period, followed by three consecutive 24-h urine collections. Urine albumin, aldosterone, and corticosterone were measured and normalized as a ratio to creatinine.
The rats were mask induced with isoflurane delivered in oxygen with a standard anesthesia machine, 4–5 ml blood were drawn from the left ventricle in an EDTA vacutainer, the aorta was cut, draining blood from the brain, and the brain was removed and frozen in liquid nitrogen within 4–6 min of the rat being taken from its cage.
Aldosterone and corticosterone assays in plasma, urine, and brain.
The brain was weighed, and 5 ml of water were added containing 2,000 counts/min of tritiated aldosterone for estimation of recoveries. It was then homogenized and extracted with 25 ml dichloromethane. The sample was cleaned by passing the dichloromethane through a silica gel column (Silica Gel Grade 62; Sigma-Aldrich, St. Louis, MO) prewashed with 5 ml dichloromethane, where the steroids were adsorbed. The steroids were then eluted with 5 ml dichloromethane containing 7% methanol. The organic extract was then evaporated in a Vortex Evaporator, reconstituted, and assayed using an ELISA buffer (20 mM sodium phosphate, 100 mM sodium chloride, 0.01% thimerosal, and 0.05% Tween 20). The sensitivity of the ELISA for aldosterone was 1 pg/well and for corticosterone 10 pg/well. Results are expressed as picograms per gram (aldosterone) or nanograms per gram (corticosterone) of tissue (pg/ml and ng/ml, respectively, in plasma).
Urine albumin was measured by ELISA using a sheep antibody raised against rat albumin produced in-house. Urine creatinine was measured by the standard picric acid method.
Differences between groups were evaluated by ANOVA, followed by a Fisher least-significant difference or Tukey contrast where appropriate (STATISTICA 6.0; StatSoft package). Results are expressed as means ± SE.
The experiment with female rats comprised the following three groups, with n = 5 animals in each group: vehicle administered intracerebroventricularly, 0.1 μg/h icv trilostane, and 0.3 μg/h icv trilostane. The dose of 0.3 μg/h, but not 0.1 μg/h icv, blocked the rise in blood pressure produced by drinking 0.9% saline (Fig. 1). At 4 wk of treatment, the pressures of rats receiving 0.1 μg/h trilostane intracerebroventricularly were no different from those receiving the vehicle intracerebroventricularly, an increase of ∼30 mmHg from a pretreatment baseline average systolic blood pressure of 120 mmHg. After 4 wk, the pump solutions were changed so that the rats that had been receiving 0.1 μg/h trilostane intracerebroventricularly received 0.3 μg/h icv, and those receiving 0.3 μg/h trilostane intracerebroventricularly received vehicle. The pressures of the former decreased significantly within 1 wk and stayed low until the pumps ran out 4 wk later, 8 wk into the experiment, at which time they again increased to the level of the controls. Blood pressures of the female rats started on 0.3 μg/h trilostane and then switched to vehicle intracerebroventricularly increased to control levels within a week of changing their treatment to the vehicle.
There was no difference in weight gain between the groups of females throughout the experiment, nor was there a difference in tibia length; the weights of the kidneys, adrenal glands, or hearts; or the ratio of the heart to body weight at the end of the experiment. Their weight increased from 200.1 ± 2.1 g at 11 wk to 255.3 ± 3.5 at 19 wk of age. Similarly, there were no differences in weight gain between the groups of male rats. Behavior assessed subjectively, and by elevated plus maze, was not altered by these doses of trilostane.
There were four groups (n = 8) in the experiment with males (Fig. 2) : these groups received vehicle intracerebroventricularly; vehicle intracerebroventricularly 3.5 wk followed by 0.3 μg/h trilostane intracerebroventricularly; 0.3 μg/h trilostane intracerebroventricularly; and vehicle intracerebroventricularly + 0.3 μg/h trilostane subcutaneously. The pretreatment mean systolic blood pressures for all rats was 139.7 ± 3.8 mmHg. Within 3.5 wk of a high-salt intake, the two groups receiving vehicle intracerebroventricularly and the group receiving vehicle intracerebroventricularly + 0.3 μg/h trilostane subcutaneously had similar increases in blood pressure to 164 ± 3.6, 159 ± 8.6, and 158 ± 3.1, respectively, whereas the pressures of the rats receiving 0.3 μg/h trilostane intracerebroventricularly remained at baseline. The pumps were replaced with pumps delivering the original solution except for those in the second intracerebroventricular vehicle group; these were switched to 0.3 μg/h trilostane intracerebroventricularly. Within 1 wk, the pressures of these rats decreased to the level of those receiving 0.3 μg/h trilostane intracerebroventricularly from the beginning, whereas the pressures of the vehicle and intracerebroventricular vehicle + 0.3 μg/h sc trilostane groups continued to increase slowly. After the second pumps were implanted (3.5 wk), the systolic blood pressures were 176 ± 1.1 and 174.5 ± 2.3 for the intracerebroventricular vehicle and subcutaneous trilostane groups and 136.3 ± 2.6 and 140.3 ± 4.7 for the two groups receiving trilostane intracerebroventricularly. All rats gained weight at the same constant rate, starting at 309.2 ± 5.25 at 11 wk and reaching 418.8 ± 5.8 g at 19 wk of age.
There were no significant differences between groups in or plasma or brain aldosterone or corticosterone concentrations or in urine albumin or aldosterone normalized to creatinine excretion. Behavior assessed subjectively, and by elevated plus maze, was not altered by treatment with trilostane.
Trilostane is an inhibitor of the 3β-HSD responsible for the conversion of pregnenolone to progesterone. The doses chosen for this study were three orders of magnitude below the reported seizure ED50 for mice, with the postulate that enough progesterone would still be formed to maintain normal neuronal excitability (45), but the formation of DOC and the subsequent steroids in the cascade would be below the amount required to elevate the blood pressure upon salt challenge in these Dahl SS rats. The central infusion of 0.3 μg/h trilostane intracerebroventricularly consistently decreased systolic blood pressure in both female and male rats in separate experiments without altering the rate of growth, behavior, or gross motor ability. Because growth and behavior were not altered by the dose of trilostane that prevented the salt-induced increase in blood pressure in the Dahl SS rats, it is assumed that enough of the GABAergic modulating neurosteroid derivatives of progesterone remained to sustain a normal life and that the effects seen on the blood pressure were not the result of generalized weakness or poor health.
The elevated plus maze has been used successfully to follow corticosterone and hippocampal MR-induced behavioral responses (2) and was used in these experiments to assess effects of the trilostane infusions on activity, anxiety, and motor abilities. Our results suggest that, although trilostane reached the circumventricular organs in effective inhibitory amounts, at the intracerebroventricular infusion rate of 0.3 μg/h, it either did not diffuse in relevant amounts to the hippocampus, or aldosterone or corticosterone synthesis in the hippocampus is not very important for those parameters assessed.
The MR is abundantly and widely expressed within the brain. In normotensive Sprague-Dawley or Wistar rats, the chronic intracerebroventricular infusion of as little as 5 ng/h aldosterone, but not corticosterone, produces hypertension of the same rate of onset and amplitude as 500 ng/h aldosterone infused subcutaneously (23). The intracerebroventricular infusion of an MR antagonist prevents the hypertension produced by the systemic infusion of aldosterone in genetically normotensive rats and by salt challenge in the Dahl SS/jr rat, even though circulating aldosterone levels in the latter are not very elevated (15, 19). These studies and ablation studies suggest that MR in the circumventricular organs, particularly those of the hypothalamus, mediate MR effects on the blood pressure (3, 9, 15, 16). Absolute plasma levels of aldosterone must be interpreted considering physiological need. A high-salt diet suppresses adrenal production of aldosterone. In the present experiments, the rats consumed a high-salt diet, and aldosterone was suppressed compared with rats on a low- or normal-salt diet (17); however, their response to MR antagonists when salt challenged indicates that suppression of aldosterone synthesis is not adequate. Little in known about the regulation of aldosterone synthesis in the brain, but the brain renin-angiotensin system, rather than the renal or systemic renin-angiotensin-aldosterone system, and sodium sensors in the central nervous system are the most promising candidates (42, 44).
The MR binds corticosterone, cortisol, aldosterone, and progesterone with similar affinity. Coexpression of 11-HSD enzyme, which inactivates corticosterone or cortisol, with the MR allows aldosterone to access the MR despite a ratio of <1:100 circulating aldosterone-corticosterone. Where there is little or no 11-HSD2, for example, the hippocampus, the MR is occupied by glucocorticoids. The amount and distribution in the brain of 11-HSD2 enzyme is also reported to be inadequate to protect MR in the circumventricular organs of the anterior hypothalamus (40). Aldosterone synthase immunoreactivity is found in these areas (unpublished observation) and in the hippocampus and cerebellum (29). Synthesis of aldosterone in or near cells expressing MR would increase the ratio of aldosterone to corticosterone concentrations just as destruction of corticosterone would (11, 22).
We speculate that de novo synthesis of aldosterone within circumventricular organs thought to be involved in salt-sensitive and mineralocorticoid hypertension (9, 15) is responsible for mediating the blood pressure increase in the Dahl SS rat and that trilostane infused intracerebroventricularly at 0.3 μg/h partially inhibited the 3β-HSD in circumventricular organs, decreasing the amount of progesterone available for the production of DOC and aldosterone to levels below that which elevate the pressure in the salt-challenged Dahl SS rats. Trilostane is not an antagonist of the MR (24). A decrease in plasma and brain aldosterone and corticosterone content in the rats receiving trilostane was not detected. The dose of trilostane was well below the effective systemic dose, so a decrease in adrenal aldosterone and corticosterone synthesis was not expected. The concentration of aldosterone and corticosterone in the brain of intact rats derives primarily from the circulation (unpublished observation). Adrenalectomy results in an undetectable amount of aldosterone and corticosterone in plasma and low but consistently detectable amounts in the brain (unpublished observation). The large amounts of aldosterone and corticosterone entering the brain from the circulation mask the small amounts produced in the brain, and the effect of local inhibition by trilostane would not be detected by the methods available (12, 13, and unpublished observation). An assay sensitive enough to measure differences in aldosterone synthesis in circumventricular organs by the intracerebroventricular trilostane treatment is not yet available. It is assumed from the physiological data available that the ratio of agonists (DOC and aldosterone) to antagonists (progesterone and corticosterone) at the level to the relevant brain MR was too low to sustain the salt-induced hypertension in these Dahl SS rats.
Progesterone treatment increases the expression of MR in hippocampal neurons, but only after estrogen pretreatment (5). An alternative explanation, that the trilostane-induced decrease in progesterone may have decreased the density of MR in the circumventricular areas, is less plausible than that of a decrease in MR agonists, since the increase in MR synthesis produced by progesterone requires previous exposure to estrogens and the antihypertensive effect of trilostane was seen in both males and females (5).
In conclusion, the brain is capable of the de novo synthesis of a small amount of adrenal corticosteroids from cholesterol. Unlike the adrenal, the brain produces very small amounts of DOC, the substrate for the aldosterone synthase and 11β-hydroxylase enzymes, in favor of other neurosteroids. The intracerebroventricular infusion of the 3β-HSD antagonist trilostane at a rate that is too low to completely inhibit the conversion of pregnenolone to progesterone prevented the increase in blood pressure and normalized the hypertension of two different cohorts of Dahl SS rats challenged with a high-salt diet. These data support the hypothesis that synthesis of aldosterone in the circumventricular areas of the brain is part of the etiology of the hypertension in the Dahl SS/jr rat and that such synthesis may be part of normal blood pressure homeostasis.
This work was supported by medical research funds from the Department of Veterans Affairs and by National Heart, Lung, and Blood Institute Grants HL-27737 and HL-075321.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society