Minireview: Aldosterone Biosynthesis: Electrically Gated for Our Protection

Abstract

Aldosterone produced by adrenal zona glomerulosa (ZG) cells plays an important role in maintaining salt/water balance and, hence, blood pressure homeostasis. However, when dysregulated, aldosterone advances renal and cardiovascular disease states. Multiple steps in the steroidogenic pathway require Ca²⁺, and the sustained production of aldosterone depends on maintained Ca²⁺ entry into the ZG cell. Nevertheless, the recorded membrane potential of isolated ZG cells is extremely hyperpolarized, allowing the opening of only a small fraction of low-voltage-activated Ca²⁺ channels of the Ca_v3.x family, the major Ca²⁺ conductance on the ZG cell membrane. As a consequence, to activate sufficient Ca²⁺ channels to sustain the production of aldosterone, aldosterone secretagogs would be required to affect large decreases in membrane voltage, a requirement that is inconsistent with the exquisite sensitivity of aldosterone production in vivo to small changes (0.1 mm) in extracellular K⁺. In this review, we evaluate the contribution of membrane voltage and voltage-dependent Ca²⁺ channels to the control of aldosterone production and consider data highlighting the electrical excitability of the ZG cell. This intrinsic capacity of ZG cells to behave as electrical oscillators provides a platform from which to generate a recurring Ca²⁺ signal that is compatible with the lengthy time course of steroidogenesis and provides an alternative model for the physiological regulation of aldosterone production that permits both amplitude and temporal modulation of the Ca²⁺ signal.

Aldosterone is an important determinant of water and electrolyte balance and, hence, blood pressure homeostasis. Under conditions of restricted dietary sodium or hypovolemia, aldosterone acts via the nuclear mineralocorticoid receptor (MR) to retain sodium (and water) by increasing the capacity of the nephron for Na⁺ reabsorption. However, when levels of aldosterone are inappropriate for salt status (1), aldosterone advances disease processes, such as cardiac fibrosis, nephrosclerosis, and arteriosclerosis, that pervade the common chronic disease states of congestive heart failure, resistant arterial hypertension, and chronic kidney disease (2–5). Consequently, the addition of an MR antagonist to the standard of care of patients with a diverse disease spectrum [congestive heart failure (6), nonfatal myocardial infarction (7, 8), atrial fibrillation, and resistant hypertension (5, 9)] has been of great therapeutic benefit.

Several mechanisms may mediate the cardiovascular and reno-protective benefits of MR blockade that include: reduced cardiac electrical remodeling (10, 11), diminished myocardial and renal structural hypertrophy with reduced perivascular inflammation and fibrosis (12–14), improved endothelial and baroreceptor function (15), as well as the maintenance of serum K⁺ and magnesium levels that prevent arrhythmogenesis (16). Although cardiovascular and renal damage have been associated with MR activation, there remains an obligate requirement for dietary salt that is not understood (9, 17, 18). Also, poorly understood are the pathogenic contributions of aldosterone that are independent of MR activation (19), most notably, acute vasoconstrictor events that induce their own damage (19). Therefore, there is strong rationale for exploring therapies that are complementary to MR receptor blockade that directly target the regulation of aldosterone production.

In this minireview, we define and evaluate the contribution of membrane voltage and voltage-dependent Ca²⁺ channels to the control of aldosterone production. We synthesize old observations with new findings to advance an electrical model of the zona glomerulosa (ZG) cell that emphasizes amplitude and temporal modulation of the Ca²⁺ signal and that invites the consideration of a broader participation of ionic conductances in the control of aldosterone production (Fig. 1). For models of the ZG cell that emphasize important spatial organization of the Ca²⁺ signal, we direct the readership to previously published reviews (20, 21).

Fig. 1.

Schematic model of electrically excitable ZG cell. Ang II increases the frequency of membrane potential oscillations in ZG cells, resulting in enhanced Ca²⁺ entry that acts at multiple sites to promote the production if aldosterone. The ZG electrical response to Ang II stimulation is mediated by the coordinated activities of multiple channel conductances. Transient inhibition of K⁺ channels via AT₁R activation produces a membrane depolarization and the opening of voltage-dependent Ca²⁺ channels. The activities of high-voltage dependent and/or Ca²⁺-activated K⁺ conductances participate in the restoration of baseline Vm, returning Ca²⁺ channels to a closed state from which they can open during the next oscillatory cycle. Green arrows indicate aldosterone promoting action, whereas red lines indicate basal inhibitory activities.

Hormonal Synergy: Angiotensin II (Ang II) and K⁺

Aldosterone is produced from the subcapsular layer of the adrenal gland, the ZG, and is the product of the activities of several mixed function oxidases and hydroxylases that are distributed among the cytoplasmic, mitochondrial, and endoplasmic reticular compartments of the cell. Production of this steroid hormone reflects both synthesis and secretion. Aldosterone production is increased by circulating concentrations of its two major secretagogs, Ang II and plasma K⁺, and is acutely regulated by adrenocorticotropic hormone and locally produced endothelin-1. It is inhibited acutely by circulating atrial natriuretic peptide and tonically by locally produced dopamine (20, 22).

Although Ang II and K⁺ regulate aldosterone production independently, Ang II synergizes with serum K⁺ in the rich ionic milieu of plasma with noteworthy consequences. First, low concentrations of either agonist (≤25 pg/ml Ang II; ≤3.0 mm K⁺) preclude a robust aldosterone response to the other stimulator, with K⁺ exerting greater regulatory control over the responses elicited by Ang II than the converse. Secondly, when the concentration of Ang II or K⁺ is modest to high, small changes in the concentration of the other agonist elicit large increases in aldosterone production (23, 24). As a result of this multiplicative interplay, aldosterone production in vivo remains remarkably sensitive to small incremental changes in plasma K⁺ (0.1 mm) that can be affected by changes in the level of dietary K⁺ intake (25).

Aldosterone Production: Dependence on Ca²⁺ and Ca_v3.2 Ca²⁺ Currents

Regulation of aldosterone production by Ang II and K⁺ is Ca²⁺ dependent; inorganic cations and organic Ca²⁺ channel antagonists that block Ca²⁺ entry inhibit stimulated production (26–32). The sites of action of Ca²⁺ in steroidogenesis are many: 1) Ca²⁺ increases the activity of cholesterol ester hydrolase, which deesterifies cholesterol to release it from its cytoplasmic storage site (33); 2) Ca²⁺ increases the cytoskeletal delivery of cholesterol to the outer mitochondrial membrane (34, 35); 3) Ca²⁺ increases the intramitochondrial transfer of cholesterol to the inner membrane by promoting the transcription and translation of steroidogenic acute regulator protein (35, 36); 4) Ca²⁺ increases mitochondrial oxidative metabolism and the Ca²⁺-dependent formation of reduced nicotinamide adenine dinucleotide phosphate, a cofactor required for the activities of P450 cytochromes such as CYP11A1 and CYP11B2 (37, 38); and 5) Ca²⁺ stimulates the Ca²⁺/calmodulin kinase I-dependent transcription of CYP11B2, the aldosterone synthase gene product that catalyzes the three-step conversion of 11-deoxycorticosterone to aldosterone (39). Given the importance of Ca²⁺ to steroidogenesis and the requirement for sustained Ca²⁺ influx for stimulated production, it is therefore not surprising that modulation of Ca²⁺ entry into the ZG cell is a formidable way to control the production of aldosterone.

Voltage-gated Ca²⁺ channels are the major Ca²⁺-selective conductances that have been identified on the plasma membrane of the ZG cell. Early electrophysiological recordings uniformly described two components of Ca²⁺ current in rat and bovine ZG cells that could be ascribed to distinct populations of Ca²⁺ channel classes: low-voltage activated (LVA) (T-type, Ca_v3.x) and high-voltage activated (HVA) (subfamily L-type, Ca_v1.x) (40–43). These channel classes differ in their voltage dependence of opening and inactivation (T-type channels require smaller depolarizations from a hyperpolarized baseline to open and inactivate) in their kinetics for closing and inactivation (T-type channels close more slowly but inactivate more rapidly in response to strong depolarizations) and in their susceptibility to modification by dihydropryidines (selective for L-type at low dose) (44). Although there are reported quantitative differences in the expression of T-type and L-type currents among isolated ZG cell preparations (species and conditions of culture), current-voltage relationships derived from whole-cell patch recordings of Ca²⁺ current generally show current peaking at membrane potentials (V_m) that are approximately 30 mV negative to 0 mV, consistent with a dominance of T-type Ca²⁺ channel activity (41, 45). However, L-type currents are more susceptible to rundown and Ca²⁺-induced inactivation and, thus, may be underestimated.

Among the three Ca_v3.x family members, Ca_v3.2 channels are the predominant subtype expressed in ZG cells across species as determined by quantitative RT-PCR, in situ hybridization (46, 47), and high-affinity inhibition of ZG Ca²⁺ current by nickel (IC₅₀, 5–20 μm for Ca_v3.2 vs. 200–300 μm for Ca_v3.1 or 3.3) (48). Among the members of the Ca_v1.x family, mRNA for Ca_v1.2 (α1C) and Ca_v1.3 (α1D) channels is expressed in ZG cells (47, 49) in agreement with high-affinity inhibition of ZG HVA current by dihydropyridines.

The relative contribution of T-type vs. L-type Ca²⁺ channels to the sustained production of aldosterone has been debated. Several lines of evidence support a primary role for T-type channels as the principal Ca²⁺ influx pathway activated by physiological concentrations of Ang II and K⁺. First, because the voltage-range over which Ca_v3.2 channels activate and inactivate overlaps, Ca_v3.2 channels can carry steady-state (window) currents within a voltage range that parallels and can account for the biphasic dependence of aldosterone secretion on extracellular K⁺ (42). Second, in perforated patch-clamp recordings in which the intracellular milieu of the cell is not disrupted, low physiological concentrations of Ang II (0.1 to 1 nm) that markedly increase aldosterone production elicit slow depolarizing responses (6–12 mV) that may be sufficient to activate a small proportion of T-type but not L-type Ca²⁺ channels. In contrast, supraphysiologic concentrations of Ang II or K⁺ that induce large maintained depolarizations of adequate magnitude (30–40 mV) to drive T-type Ca²⁺ channel inactivation to completion elicit only modest increases in aldosterone production (45, 50). Third, block of L-type Ca²⁺ channel currents by ω-agatoxin IIIA or their inhibition by Ang II fails to decrease the steroidogenic response to physiological concentrations of K⁺ or Ang II (45, 51), whereas the activation of these currents and the ensuing large rise in cytosolic Ca²⁺ that is produced actually inhibits aldosterone biosynthesis (37, 51, 52).

ZG Cell: A Model Nonexcitable Cell?

The ZG cell does not express fast Na⁺ channels and thus does not fire Na⁺-dependent action potentials. Rather, in physiological, asymmetrical K⁺ solutions, the ZG V_m follows closely the predicted equilibrium potential for K⁺ over a broad range of extracellular K⁺ concentrations (4–28 mm). Nevertheless, because the V_m of ZG cells is more depolarized than E_K (dependence less than 58.8 mV/10-fold change in K and especially so <4 mm K⁺) (45, 53), the ZG membrane at rest is not exclusively K⁺ permeable; rather a persistent depolarizing conductance(s) must also be active. A likely candidate for this conductance is a Ca²⁺-permeable nonselective cation channel previously identified in rat ZG cells (54). An unlikely candidate is Ca_v3.2 channels, because there is uniform agreement that the V_m of unstimulated ZG cells (at 3.5 mm K⁺) is negative to −78 mV (−78 to −90 mV) (50, 53, 55), at which the open probability of LVA Ca²⁺ channels is extremely low (50, 56). Therefore, Ca_v3.2 channels in ZG cells are predicted to be mainly closed at rest but available for opening upon membrane depolarization. Hence, sustained inhibition of a dominant background K⁺ conductance is required of aldosterone secretagogs to move the V_m of ZG cells into a voltage range where steady-state Ca²⁺ entry through Ca_v3.2 channels is probable.

Several voltage-gated K⁺ conductances have been identified in ZG cells, although their expression varies across species and with the conditions of cell culture (20, 55). These include a delayed rectifier (57, 58), a transient A-type current [dominant in bovine (57, 58) and human (43) ZG], a slow delayed rectifier (43, 59, 60), an outward rectifier (55), and a Ca²⁺-dependent maxi-K⁺-type current (Table 1) (43, 55, 58, 61). However, collective analyses of these voltage-dependent conductances suggest that they are not likely to be active at the hyperpolarized resting V_m of ZG cells, and thus, although inhibited by Ang II, they are not suitable to initiate the depolarizing action that is required of aldosterone secretagogs. By contrast, two background K⁺ currents have been identified in ZG cells: inward rectifier (58, 62) and leak K⁺ currents (61, 63–65). However, at negative resting membrane voltages, only leak K⁺ currents have been recorded at both the single channel and macroscopic levels (55, 64). This contrasts with the early recording of “inwardly rectifying” single-channel currents (58) that originally were attributed to an inward rectifier, but for which no macroscopic correlate of current could be identified, raising the possibility that the subsequently identified nonselective cation channel (54, 66) may have been responsible for these currents.

Table 1.

K⁺ conductances in ZG cells

Channel type/current	Properties	K+ channel family	Species	Inhibition
Delayed rectifier (IK)	Slowly voltage activated/slowly voltage inactivating or noninactivating	?	Rat (57, 58)	Ang ll
			Human (57)	Ang II
Slow delayed rectifier (IKs)	Very slowly voltage activated/noninactivating	Kv7.1	Mouse (60)
		?	Human (43)
		?	Rat (59)
Outward rectifying	Rapidly voltage activated/noninactivating	?	Rat (55)
A-type K+ channel (IA)	Rapidly voltage activated/rapidly voltage inactivating	?	Bovine (57, 58)	Ang II
		?	Human (43)	Ang II
Maxi K (BK)	Ca2+ and voltage activated	KCa1.1	Rat (55, 58, 61)	Ang II
Small K (SKCa)	Ca2+/CaM activated		Bovine (58)
			Human (43)	Ang II
		KCa2.2	Rat (77)
Leak K (Ileak)	Weakly voltage activated/noninactivating	TASK-1/3	Rat (61, 63, 64)	Ang II/[H+]o
			Mouse (65)	Ang II/[H+]o
		TREK-1	Bovine (76)	Ang II
Inward rectifier (IIRK)	Gβγ activated/voltage inactivating	Kir3.4	Human (62)
		?	Rat (58)	Ang ll
		?	Bovine (58)	Ang II

Summary of reported K⁺ conductances in ZG cells of various species. ? indicates the molecular correlate is not identified.

Nevertheless, it is of note that mutations in or close to the selectivity filter of the G protein-activated inward rectifier K⁺ channel (K_ir3.4) are found in a subset (36–41%) of aldosterone-producing adenomas (62, 67). Heterologous expression of these mutant channels in model cells results in loss of K⁺ selectivity, commensurate with an increase in Na⁺ conductance to produce membrane depolarization, and in model adrenal cells (HAC-15) an elevation in aldosterone production (62, 68). Taken together, these studies provide proof of principal for the importance of inward rectifier currents in the modulation of aldosterone production in tumorigenic ZG cell function.

The KCNK background (“leak”) K⁺ channel family is encoded by multiple genes that are grouped into seven subfamilies. These channels show little or no voltage dependence and thus generate background K⁺ currents that play a major role in setting negative resting membrane voltages. Functional channels are formed by the dimeric association of subunits that contain two-pore and four transmembrane spanning domains (2P/4TM). Two family members, TWIK-related acid-sensitive K⁺ (TASK) channel TASK-1 and TASK-3 subunits, are abundantly expressed in the ZG layer (rat, mouse, and human) (63–65), allowing the potential for homo- and heteromeric TASK channel conformations (69). The combined deletion of TASK-1 and TASK-3 subunits from mouse ZG cells results in an approximately 20-mV baseline membrane depolarization and markedly increases (400%) aldosterone production (65, 70). Although the pharmacological characterization of gK_leak in rat recapitulates properties of TASK-3 homomeric channels, the IC₅₀ for pH inhibition (pH 7.5) is not in agreement with that of TASK-3 channels (pH 6.8) (64), nor does the sole deletion of TASK-3 subunits from mouse ZG cells result in a baseline membrane depolarization (70). Therefore, the subunit composition of native ZG channels remains unresolved. Nevertheless, in either homo- or heteromeric configuration, the activation of Gq-coupled receptors induces an inhibition of TASK channel activity. Thus, the inhibition of ZG TASK channels by Ang II is predicted to induce a passive analog shift in voltage that removes the restraint on aldosterone production. However, under physiological conditions, whether at rest or during stimulation with Ang II or K⁺, the elicited depolarization is small (4–10 mV from baseline) and the V_m remains in a range where steady-state Ca²⁺ entry is very low. As a consequence, mechanisms which modify Ca_v3.2 channel gating by shifting the voltage dependence of channel opening will increase (hyperpolarized shift) or decrease (depolarized shift) steady-state Ca²⁺ entry and hence contribute to the control of aldosterone production. Of note, phosphorylation of the II–III linker of Ca_v3.2 channels by Ca²⁺/calmodulin (CaM)-dependent kinase II induces a 10-mV hyperpolarizing shift, or by Rho-associated kinase, a 10-mV depolarizing shift, in the voltage dependence for half-activation of Ca_v3.2 channels to elicit opposing changes in Ca²⁺ entry.

The ZG Cell: A Model Excitable Cell?

The earliest sharp microelectrode recordings of “peripheral” cells (ZG) in cat adrenal slices showed rapid spontaneous fluctuations in V_m in a percentage of impaled cells and “spike potentials” that were evoked by Ang II or K⁺ in most others (71). This hint of electrical excitability was reanalyzed in dissociated/cultured ZG cell preparations in the mid-1980s. Although isolated ZG cells showed no spontaneous spiking activity, Ca²⁺-dependent V_m fluctuations could be evoked by depolarizing current injection and/or pharmacological blockade of outward K⁺ currents (72). Active responses also have been measured in acutely dissociated cells using the patch-clamp technique. Superimposed on larger analog shifts in membrane voltage, Ang II evoked small depolarizing membrane potential fluctuations. These active responses were almost certainly curtailed by the recording protocol that repetitively delivered hyperpolarizing current to permit the concurrent measurement of membrane resistance (45).

Taken together, these data are inconsistent with the accepted dogma of ZG cell nonexcitability and provided the impetus for our laboratory to determine the membrane properties of ZG cells retained within an acute adrenal slice (47). We find that mouse ZG cells in adrenal slices act as electrical oscillators, showing both a voltage-threshold and a voltage dependence to their oscillatory activity. ZG V_m oscillations are spontaneous, large (deviating ∼70 mV from baseline V_m), of low periodicity (∼2 sec), and are modulated in frequency by Ang II and K⁺. Notably, these oscillations provide a platform from which to generate a significant and recurring steady-state Ca²⁺ signal that is compatible with the lengthy time course of steroid hormone production (minutes-hours) and that can be fine-tuned by aldosterone secretagogs. This innate capacity of the ZG cell to behave as an electrical oscillator may provide a novel and/or alternative molecular explanation for: 1) the oscillatory character of cytosolic Ca²⁺ signaling in ZG cells evoked by physiological concentrations of Ang II; 2) the requirement for a large depolarization from a baseline voltage of approximately −90 mV to activate even LVA Ca²⁺ channels; 3) the generation of high-Ca²⁺ microdomains that are necessary for the efficient transfer of Ca²⁺ into the mitochondria by the low-Ca²⁺ affinity mitochondrial uniporter; and 4) the hyperaldosteronism (60) that is produced by the deletion of the KCNE1 subunit in mice (the regulatory subunit of a voltage activated KCNQ1 K⁺ channel, with a voltage of half-activation equal to 5 mV).

Recognition of the intrinsic capacity of ZG cells to generate rhythmic electrical signals provides not only a novel/alternative way to understand the diverse set of previously mentioned signaling incongruities but also may serve to broaden our perspective of cellular response elements that contribute to the regulation of aldosterone production. For example, oscillations in membrane voltage allow for the full complement of K⁺ conductances previously identified in ZG cells to participate in the regulation of steroidogenesis. As a consequence, when viewed in the context of an excitable cell, TASK channels will retain their singular importance during the depolarization phase of the oscillatory cycle when membrane voltage is highly negative by affecting the rate of depolarization between spikes, whereas delayed rectifiers and BK channels will play a more dominant role during the later phases of the oscillatory cycle when membrane voltage is depolarized by controlling the rate of repolarization. Notably, this shared importance is supported by mouse models of human disease, in which the combined genetic deletion of TASK-1 and TASK-3 subunits (65, 70) or KCNE1 (60) subunits results in hyperaldosteroinsm. By analogy, although substantial data support a role for Ca_v3.2 channels in the control of steroid hormone production, in the excitable ZG cell, HVA Ca²⁺ channels of the Ca_v1.x family unquestionably are activated, raising the possibility that as in subsets of neurons, these channels may support an alternate function, that of modulating gene transcription (73).

Finally, an excitable ZG cell is a cell that allows for nuanced steroid hormone regulation: a change in intracellular response can be achieved by either amplitude or temporal modulation of the Ca²⁺ signal, the former involving a change in the concentration of Ca²⁺, the latter involving a change in pulsatility or oscillation frequency. It is noteworthy, therefore, that frequency decoding of the Ca²⁺ signal is a property of a subclass of CaM-dependent effectors, Ca²⁺/CaM-dependent protein kinase I and II, that can accumulate, and thus integrate, activity in response to pulsatile Ca²⁺ signals (74, 75). Given the importance of these kinases at multiple steps in aldosterone biosynthesis, they may be at the very heart of information transfer in the excitable ZG cell coordinating the flow of information from the cell surface to the cell interior to control both the short- and long-term production of aldosterone.