Aging and Adrenal Aldosterone Production

Aldosterone, the primary mineralocorticoid, is synthesized in the outer zone of the adrenal cortex called the zona glomerulosa (ZG). The production of aldosterone is tightly regulated by angiotensin II (Ang II) and circulating potassium levels¹. Physiologically, aldosterone plays a key role in the maintenance of intravascular volume and blood pressure through sodium retention in the kidney. Excess aldosterone causes hypertension and induces cardiovascular complications^2–5. The autonomous secretion of aldosterone, independent of Ang II and sodium status, is known as primary aldosteronism (PA). PA is the most common cause of endocrine-related hypertension with a prevalence of 5-10% in hypertensive population^6–10 and approximately 20% in resistant hypertension^11–13. Due to the increased risk for cardiovascular complications in patients with PA, early detection of the disease and targeted treatment is recommended¹⁴.

The molecular pathogenesis of PA was largely unknown until recently. The development of specific antibodies against aldosterone synthase (CYP11B2), which is required for the final steps of aldosterone production, has allowed the detection of aldosterone producing cells in resected adrenals^{15, 16}. Using these antibodies, non-neoplastic foci of CYP11B2-expressing cells called aldosterone-producing cell clusters (APCC)¹⁵ have been identified in adrenal tissues adjacent to aldosterone-producing adenomas (APA) as well as in normal human adrenal glands without tumor or hyperplasia^15–20. Studies have shown APCC are a common occurrence in normal human adrenals^21–23. Since circulating renin and Ang II levels are suppressed in patients with PA, the observation of APCC adjacent to APA suggested that APCC may represent a source of autonomous and renin-independent aldosterone secretion, perhaps even a precursor to APA²⁴.

Over the last six years, somatic and germline mutations that cause inappropriate aldosterone production have been identified in patients with PA. Most of the mutations cause increased intracellular calcium (Ca²⁺) levels, resulting in activated CYP11B2 transcription and elevation of aldosterone production^24–26. These somatic mutations have also been observed in APCC in normal adrenal glands^{23, 27}. Recent histopathological studies demonstrated that older age is associated with greater adrenal APCC content^21–23 and a concomitant biochemical phenotype that supports progressive autonomous aldosteronism²¹. These findings provide a possible link between age-related histological changes in adrenal CYP11B2 expression and age-related physiological changes in aldosterone secretion. In this review, we cover recent topics on physiology and dysregulation of aldosterone production and discuss potential age-related effects on the aldosterone and adrenal physiology.

Human adrenal zonation and steroid production

Human adrenal glands are composed of an outer cortex and an inner medulla. The adrenal cortex is further divided into three functionally distinct zones; ZG, zona fasciculata (ZF), and zona reticularis (ZR) (Figure 1A). In the ZG, the mineralocorticoid aldosterone is produced with the physiologic role of maintaining fluid and electrolyte balance. The production of aldosterone is mainly regulated by Ang II, potassium (K⁺), and adrenocorticotropic hormone (ACTH)¹. Glucocorticoids are synthesized in the ZF under the regulation of ACTH. The ZR is also mainly regulated by ACTH and produces adrenal androgens, including dehydroepiandrosterone (DHEA) and DHEA sulfate. There is considerable evidence that aging is associated with a decline in circulating levels of DHEA and DHEA sulfate²⁸. The decrease in these adrenal steroids results from intra-adrenal changes that include a decrease in ZR size and expression of steroidogenic enzymes required for DHEA synthesis^{29, 30}. In contrast to DHEA, little is known regarding the age-related adrenal changes related to aldosterone synthesis.

Figure 1. Adrenal zonation and steroidogenic pathway.

A. Adrenal cortex zonation with the primary regulators of steroid production. B. Steroidogenic pathway for the production of aldosterone and cortisol. Aldosterone and cortisol production occur in the zona glomerulosa (ZG) and zona fasciculata (ZF), respectively

ZG synthesis of aldosterone occurs through the action of a series of enzymes, including cholesterol side-chain cleavage (CYP11A1), type 2 3β-hydroxysteroid dehydrogenase (HSD3B2), 21-hydroxylase (CYP21A2), and CYP11B2¹ (Figure 1B). Of these enzymes, CYP11B2 is expressed specifically in the ZG, while expression of 11β-hydroxylase (CYP11B1) which is required for the final step in the biosynthesis of cortisol, is limited to the ZF and ZR³¹. This functional zonation acts to compartmentalize aldosterone production to the ZG^{32, 33}. Based on the critical nature of CYP11B2 in aldosterone production, we and others have studied the cell signaling mechanisms regulating this enzyme in normal and pathological conditions within the adrenal.

Physiological and pathological regulation of adrenal cell aldosterone production

The primary physiological regulators of aldosterone production are Ang II, K⁺, and ACTH. Other factors that have been proposed to modulate aldosterone production include serotonin^{34, 35}, estrogen³⁶, parathyroid hormone^{37, 38}, vasopressin^{39, 40}, endothelin-1^{41, 42}, and very low-density lipoprotein (VLDL)^43–45. A recent study demonstrated that leptin induces CYP11B2 expression and aldosterone production via Ca²⁺ dependent mechanisms⁴⁶. Several signaling pathways are involved in adrenal cell aldosterone production. The key intracellular pathways include Ca^{2+ 47, 48}, cyclic AMP ^{49, 50} and phospholipase D (PLD) signaling^51–53. Of these, the Ca²⁺ signaling pathway appears to be the most important in both physiological and pathological conditions. Both Ang II and K⁺ utilize Ca²⁺ for physiologic activation while inappropriate aldosterone production, as seen in PA, is primarily regulated by somatic and germline mutations that dysregulate and raise intracellular Ca²⁺.

Mutations in KCNJ5⁵⁴, ATP1A1⁵⁵, ATP2B3⁵⁵, CACNA1D^{56, 57}, and CACNA1H⁵⁸ (aldosterone-driver genes) have been identified in APA and/or familial hyperaldosteronism. In vitro studies have shown that these mutations cause increased aldosterone production^59–64. The KCNJ5 gene encodes the inward-rectifying potassium channel GIRK4 and somatic mutations in this gene are considered to be the most prevalent genetic alteration in APA^65–67. Mutations in KCNJ5 gene cause a loss of ion selectivity, increased Na⁺ conductance, cell membrane depolarization, and voltage-gated calcium channel opening, leading to increased intracellular Ca²⁺ concentration^{54, 59, 68}. The ATP1A1 gene encodes the α1-subunit of the Na⁺/K⁺ ATPase which acts to maintain the resting cell membrane potential and electrical excitability. ATP1A1 mutations disrupt this function and cause cell membrane depolarization and/or cellular acidification due to a pathological H⁺ leak, resulting in aldosterone overproduction^{55, 61, 69}. The ATP2B3 gene encodes the plasma membrane Ca²⁺ ATPase isoform 3, which regulates intracellular Ca²⁺ homeostasis by transporting cytoplasmic Ca²⁺ out of the cells. Mutations in ATP2B3 gene appear to increase intracellular Ca²⁺ levels through a reduced Ca²⁺ export as a result of a loss of pump function and/or non-physiological Na⁺ and possible Ca²⁺ permeability to the cells⁶². The CACNA1D gene encodes the voltage-dependent L-type calcium channel subunit alpha-1D (Ca_v1.3). Mutations in CACNA1D gene cause channel activation at less depolarized potentials, leading to increased Ca²⁺ influx and aldosterone overproduction^{56, 57, 63, 70}. The CACNA1H gene encodes the voltage-dependent T-type calcium channel subunit alpha-1H (Ca_v3.2). In a recent study, a recurrent germline mutation in CACNA1H gene (p.Met1549Val) was identified in early-onset PA patients⁵⁸. In vitro studies demonstrated that the CACNA1H mutation caused impaired channel inactivation and a shift of activation to less depolarized potentials, leading to increased Ca²⁺ entry and elevated aldosterone production^{58, 64}.

The cellular origins of APA with somatic mutations remain an area of active research. If the adrenal accumulation of somatic mutations follows other tissues, it would be expected that there would be an age-dependence. We and others have approached this research by focusing on CYP11B2, as it is required for aldosterone synthesis and its expression is highly cell specific within the adrenal. Development of specific antibodies against human CYP11B2 and CYP11B1 has provided tools to better study adrenal functional histopathology^{15, 16}. Nishimoto et al.¹⁵ reported that there are two patterns of CYP11B2 expression within the adrenal: 1) conventional zonation with ZG cells expressing CYP11B2; 2) variegated zonation consisting of a subcapsular non-neoplastic cell foci expressing CYP11B2 named APCC. The expression pattern of steroidogenic enzymes in APCC, i.e. high expression of HSD3B2 and CYP11B2 and low expression of 17α-hydroxylase (CYP17A1) and CYP11B1 (needed for cortisol biosynthesis) supports the ability of APCC to produce aldosterone¹⁵. APCC have also been identified in adrenal tissues adjacent to APA^{15, 17–20}; given that renin and Ang II are suppressed in PA caused by APA, these observations supported APCC as non-neoplastic sources of autonomous and renin-independent aldosterone secretion. Of note, before the development of CYP11B2 antibodies, in situ hybridization methods allowed the identification of cell foci expressing CYP11B2 mRNA but no CYP11B1 or CYP17 mRNA⁷¹.

Adrenal zona glomerulosa and aldosterone-producing cell clusters in aging

Age-related histological changes in adrenal CYP11B2 expression have been investigated. There is growing evidence for an age-dependent accumulation of APCC in adrenal glands^21–23. Further, there is age-related loss of the classic continuous CYP11B2 expression pattern within the ZG that is frequently observed in young adrenal glands (Figure 2)^{21, 72}. Collectively, older adrenal glands have less normal CYP11B2 expression in the ZG and greater amount of APCC. Although there is no direct evidence to explain why such age-related changes in CYP11B2 expression patterns might occur, potential considerations may include genomic, epigenetic, or other environmental factors. The histological findings suggest a transition from continuous CYP11B2 expression in the ZG to APCC predominance with advancing age²¹ and raise the hypothesis that with aging there may be a progressive transition from normal physiological aldosterone regulation to autonomous and renin-independent aldosterone secretion. If true, this clinical phenotype associated with these age-related histopathological changes should involve greater autonomous aldosteronism and higher risk for aldosterone-mediated cardiovascular disease.

Figure 2. Age-related histological changes in human adrenal glands.

Continuous CYP11B2 expression in the zona glomerulosa (ZG) is often seen in young adrenal glands, whereas older adrenals have less normal ZG CYP11B2 expression and more APCC. Potential effects that might cause the age-associated increase in CYP11B2-positive APCC include genomic, epigenetic events or environmental factors.

To better define the cellular characteristics of APCC, we performed transcriptome analysis and targeted next-generation sequencing (NGS) analysis on APCC isolated from normal adrenal glands from kidney donors²⁷. Transcriptome analysis demonstrated that APCC mRNA profile had similar characteristics to those of ZG but with higher CYP11B2 expression in APCC, indicating increased capacity to produce aldosterone in APCC²⁷. Mutation analysis revealed that 8 out of 23 APCC (35%) had known aldosterone-driver somatic mutations in genes including CACNA1D and ATP1A1²⁷. A recent study using a modified targeted NGS protocol identified somatic mutations in 21 of 61 APCC (34%) in adrenals from normotensive Japanese autopsy cases²³. No KCNJ5 mutation, which is the most prevalent somatic mutation in APA, was identified in either American or Japanese cohorts^{23, 27}. In contrast to APA, both studies suggest that APCC are most likely to have aldosterone-driver mutations in the CACNA1D gene^{23, 27}. These findings suggest that L-type calcium channel blockers should be considered as a targeted treatment option that could reduce inappropriate autonomous aldosterone production from APCC. Interestingly, in one early-onset PA patient with germline de novo CACNA1D mutation (p.Gly403Asp), treatment with a calcium channel blocker, amlodipine, resulted in normalization of blood pressure and resolution of biventricular hypertrophy that had been present since birth⁵⁷.

Although somatic mutations in APCC support the concept of renin-independent aldosterone production in APCC, genetic characteristics of majority of the APCC are still unknown. Further, since prior research on human APCC has been conducted from either post-mortem adrenal specimen, or adrenals containing APA, the biochemical phenotype of APCC has not been specifically quantified. Future clinical studies are needed to directly assess whether APCC autonomously secrete aldosterone.

The renin-angiotensin-aldosterone system in aging

With aging there is a decline in normal function of several physiological systems. Consistent with the histological findings of age-related changes in adrenal CYP11B2 expression patterns, physiological changes in the renin-angiotensin-aldosterone system seem to occur with advancing age. There have been several studies with relatively small sample size suggesting that older individuals may secrete less aldosterone and have lower plasma renin activity (PRA)^73–77. Specifically, studies have shown that older age is associated with a blunted ability to secrete aldosterone despite stimulation by its major regulators: Ang II with sodium restriction⁷⁴ and potassium during intravenous infusion⁷⁸. We recently demonstrated the dynamics of aldosterone physiology in a large cross-sectional study that included a broad age continuum²¹. The most notable observations were a strong and significant association between older age with higher aldosterone-to-renin ratio and a blunted ability to secrete aldosterone with sodium restriction²¹. Taken together, older age was associated with greater autonomous aldosterone secretion and less physiologic aldosterone secretion. Given the aforementioned age-related histopathological changes in CYP11B2 expression, we hypothesized that this clinical phenotype may have been the consequence of parallel age-related increases in adrenal APCC content and concomitant decreases in normal adrenal CYP11B2 expression.

Clinical perspectives

The concept of age-related autonomous aldosteronism may provide new avenues to investigate the pathogenesis and treatments for hypertension and cardiovascular diseases. Recent studies have demonstrated that autonomous aldosterone production and even overt PA can be detected in normotensive subjects^{79, 80}; in parallel with the finding that a subset of adrenal glands from normotensives have APCC harboring somatic mutations in aldosterone-driver genes²³, this may suggest that the spectrum of PA can range from mild and subclinical in normotension (possibly as a result of autonomous aldosterone production from APCC) to more severe in hypertension (possibly as a result of a combination of APCC and neoplastic and hyperplastic processes such as APA and bilateral adrenal hyperplasia)⁸¹. Consistent with this, normotensive patients with evidence for subclinical PA have a significantly higher risk for developing incident hypertension^{79, 81, 82}. A recent population-based cohort study demonstrated that older normotensives were more likely to have a suppressed renin phenotype, and that higher aldosterone levels in the context of this renin suppression were associated with a substantially higher risk for developing hypertension⁸¹. Further, physiologic investigations have also shown age-related declines in 11β-hydroxysteroid dehydrogenase type 2 activity that result in a phenotype of renin suppression and cortisol-mediated mineralocorticoid receptor (MR) activation⁸³. Given the accumulating evidence suggesting age-related histopathological and biochemical changes consistent with renin-independent aldosteronism and MR activation, future studies may investigate to what extent the “essential” hypertension of aging may be mediated by mineralocorticoids, such as aldosterone and/or cortisol.