Cardiac Steroidogenesis—New Sites of Synthesis, or Much Ado About Nothing?

The primary adrenocortical steroids are aldosterone, synthesized in the outermost layer of cells of the adrenal cortex, the zona glomerulosa; and cortisol and corticosterone, synthesized in the next layer, or zona fasciculata. Young et al. (1) have presented data in this journal complementing other suggestions that the heart is capable of synthesizing adrenocortical steroids that have physiological or pathophysiological importance through paracrine or autocrine effects (2–6). Aldosterone produced in the zona glomerulosa is released into the circulation, to be carried to various target organs throughout the body, where it binds the mineralocorticoid receptor. This receptor, like others in the steroid receptor superfamily, acts as a transcription factor modulating the transcription of message for several proteins, many of which have not been completely characterized. Aldosterone enhances the vectorial transfer of sodium in transport epithelia (7), increases the blood pressure through its action in the brain via central sympathetic neurons, some of which alter renal function (8, 9), and promotes hypertrophy and fibrosis through direct effects on the heart and vessels (10–13). Left ventricular mass has been found to correlate with plasma aldosterone in patients with both primary aldosteronism and essential hypertension, as well as in a population-based sample (14–16), suggesting that aldosterone plays an important role in cardiac remodeling. Administration of low doses of the aldosterone receptor antagonist spironololactone in the Randomized Aldactone Evaluation Study trial in patients with congestive heart failure decreased cardiovascular related mortality by 30% and morbidity by 35% (17). This low dose of spironolactone did not affect blood pressure, suggesting a direct effect within the heart. In addition to suffering from the effects of excessive circulating aldosterone, the heart has the enzymatic machinery and the synthetic ability to produce aldosterone and other corticosteroids, which might then act in a paracrine and/or autocrine manner (1, 2–6, 18). The synthesis of steroids from cholesterol in tissues other than the adrenal or gonadal glands was first demonstrated by Baulieu and collaborators (19, 20), who demonstrated steroid synthesis within the central nervous system. They called the products of this de novo synthesis “neurosteroids” (19, 20). Subsequent studies by Takeda and colleagues (21–25) have demonstrated that human vascular endothelial and smooth muscle cells in vitro and rat mesenteric artery ex vivo release aldosterone and corticosterone into the culture and perfusion media and that cardiovascular cells express and regulate the expression of the enzymes in the late pathway in the biosynthesis of corticosteroids including the 11β-hydroxylase (CYP11B1) and aldosterone synthase (CYP11B2). Studies by Delcayre’s group (3) demonstrated that the ex vivo perfused rat heart released corticosterone and aldosterone into the perfusate. As in the adrenal gland, angiotensin II or ACTH increased both the release of these steroids into the perfusate and their concentration in the heart at the end of the perfusion (3). They also demonstrated that aldosterone synthase mRNA expression in the heart was regulated by the infusion of angiotensin II and by chronic sodium depletion. Aldosterone synthase activity was demonstrated by incubating tritiated deoxycorticosterone with heart homogenates and measure of aldosterone generation (3). Experimental myocardial infarction resulted in a 2-fold increase in aldosterone synthase mRNA and a 3.7-fold increase in aldosterone content in adjacent noninfarcted myocardium (26). The increase was blocked by the administration of an angiotensin II receptor blocker (26). These studies suggested that aldosterone biosynthesis participates in cardiac remodeling after myocardial infaction. Cardiac aldosterone production, aldosterone synthase enzymatic activity, and the expression of aldosterone synthase mRNA were also increased in the stroke-prone spontaneously hypertensive rat in comparison to Wistar Kyoto (WKY) control rats. Aldosterone production and aldosterone synthase mRNA expression were also increased by adrenalectomy (4). In contrast to what occurs in the adrenal gland, chronic sodium loading increased aldosterone synthase mRNA, aldosterone synthase activity, and aldosterone release from perfused WKY rat hearts (5). Expression of the aldosterone synthase gene differs depending on the strain of rat. The heart of the Sprague Dawley rat expresses the aldosterone synthase gene after chronic stimulation with angiotensin II, but not under basal conditions as is reported for the WKY or Spontaneously Hypertensive Rat (18). Two different strains of mouse heart do not express the aldosterone synthase (1), indicating that the expression of steroidogenic enzymes varies according to the species and strains studied. The mRNA for the P-450scc, 3β-ol-dehydrogenase, 21-hydroxylase, and 11β-hydroxylase were detected in the human heart using RT-PCR and Southern blotting at levels 100- to 10,000-fold lower than those in the adrenal. The aldosterone synthase mRNA was only detected in fetal human heart and aortic samples. Young et al. (1) did not find 11β-hydroxylase and aldosterone synthase mRNA expression in normal human heart samples, but did find it in samples of hearts in congestive heart failure. This suggests that under the chronic stress of congestive heart failure, the heart is stimulated to synthesize aldosterone with possible deleterious paracrine or autocrine effects (1). Simultaneously, plasma levels of aldosterone in the anterior interventricular vein, coronary sinus, and aortic root of control people and patients with left ventricular systolic dysfunction or left ventricular diastolic dysfunction were collected by Mizuno et al. (6). Plasma levels of aldosterone in these patients were not different from control subjects. However, aldosterone levels were significantly higher in the anterior interventricular vein and coronary sinus than in the aortic root from patients with left ventricular systolic and diastolic dysfunction, but not from control patients (6). These studies support the idea that steroidogenic enzymes expression and aldosterone synthesis are activated in the hearts of patients with congestive heart failure and that the favorable response to spironolactone in the Randomized Aldactone Evaluation Study (17) might have been through antagonizing the effects of locally synthesized, as well as circulating, aldosterone. The studies described above support the hypothesis that cardiovascular synthesis of aldosterone is relevant under pathophysiological conditions. However tempting, there are several caveats to the experimental details from these studies that need to be clarified before proving that cardiovascular tissues synthesize meaningful amounts of aldosterone. Human pulmonary artery endothelial cells and smooth muscle cells express aldosterone synthase enzyme, but do not express the P-450scc (cholesterol side chain cleavage enzyme) or the 11β-hydroxylase mRNAs, even when the RT-PCR is done for 50 cycles (27). The absence of the P-450scc in these cells suggests that the production of aldosterone in these cells would have to be from precursors from the circulation (27). However, cultured umbilical vein endothelial cells and vascular smooth muscle cells generate basal, as well as angiotensin II- and ACTH-regulated, aldosterone synthesis (22, 24) in the absence of added precursors to the culture. Either there is an as yet unknown alternative pathway to generate the pregnenolone necessary for the synthesis of aldosterone (22, 24) in umbilical vein endothelial cells and vascular smooth cells or they are different from pulmonary artery in the expression of the enzymes in the early pathway of steroidogenesis. In the adrenal, the rate limiting reactions in the synthesis of aldosterone are those of the P-450scc and the aldosterone synthase (28). In the human heart the P-450scc mRNA is between 1,000- and 10,000-fold less abundant than that of the adrenal (2). Even if the rate of protein turnover for these enzymes in the heart were significantly less than that of the adrenal, the Km for all of these enzymes are in the micromolar range (29). Because the expression of the rest of the enzymes required for aldosterone synthesis is also extremely low, all would become rate limiting. It is unknown if the mRNA are translated, because the aldosterone synthase or 11β-hydroxylase proteins have never been demonstrated in heart tissues. If adrenocortical steroidogenic enzymes were expressed in heart and vascular tissue, even at concentrations very much lower than those in the adrenal, it is difficult to explain why, given the combined mass of the heart, endothelial cells and vascular smooth muscle cells compared with that of the adrenal cortex, plasma concentrations of corticosterone and aldosterone in adrenalectomized animals are undetectable (4). There is another equally bothersome finding. The reported conversion rates of radiolabeled deoxycorticosterone to aldosterone by cultured endothelial and vascular smooth muscle cells and heart homogenates are extremely high, much higher than conversion rates known to occur in cells of the zona glomerulosa that express more than 1000-fold greater amounts of the aldosterone synthase than the vascular tissue (3–5, 22, 23, 30–32). In the brain where the expression of the mRNA for the aldosterone synthase is also similarly very low, the enzyme has been demonstrated using immunocytochemistry using tyramine-amplification techniques and synthesis of aldosterone has been demonstrated, the conversion rates of tritiated deoxycorticosterone to aldosterone are at least two to three orders of magnitude lower than those reported for vascular tissue (33, 34). Synthesis of aldosterone by the failing heart is not the only possible explanation for the clinical and experimental data suggesting that aldosterone plays an important role in congestive heart failure. Tsutamoto et al. (35) found that levels of aldosterone in the coronary sinus of patients with congestive heart failure and normal controls were lower than the levels in the aortic root, suggesting that the heart extracted aldosterone from the circulation by an unknown mechanism. Administration of spironolactone reversed the extraction of aldosterone from plasma (35). These results are exactly the opposite of those found by Mizuno et al. (6), as described above. Aldosterone concentrations in heart tissue are 10 times greater than those of plasma (3, 26), whether due to extraction of aldosterone from the circulation, its local synthesis, or both. Three decades ago, Lockett and co-workers (36, 37) demonstrated that the heart contained and produced a sodium-retaining factor that was released into the coronary sinus and that they identified as an aldosterone derivative similar to aldosterone-18-monoacetate. Recent studies suggest that the heart has a high concentrations of an aldosterone ester that might be aldosterone-20-monoacetate (38). Both the 21- and 18-monoacetates are ephemeral in circulation; they are converted to aldosterone within minutes at room temperature and even faster at body temperature. Both are more potent than aldosterone in diverse bioassays (37). In conclusion, our understanding of the role of aldosterone role in cardiovascular morbidity and mortality has regained importance in our therapeutic approaches to patients with congestive heart failure. Although there are several studies suggesting that the heart and vascular tissues have the enzymatic machinery to synthesize aldosterone, there are enough doubts about the evidence so far presented to question the importance of extra-adrenally synthesized aldosterone in cardiovascular pathophysiology. This important issue should be clarified in the next few years.