Cellular Pathophysiology of Mutant Voltage-Dependent Ca²⁺ Channel CACNA1H in Primary Aldosteronism

Abstract

The physiological stimulation of aldosterone production in adrenocortical glomerulosa cells by angiotensin II and high plasma K⁺ depends on the depolarization of the cell membrane potential and the subsequent Ca²⁺ influx via voltage-activated Ca²⁺ channels. Germline mutations of the low-voltage activated T-type Ca²⁺ channel CACNA1H (Cav3.2) have been found in patients with primary aldosteronism. Here, we investigated the electrophysiology and Ca²⁺ signaling of adrenal NCI-H295R cells overexpressing CACNA1H wildtype and mutant M1549V in order to understand how mutant CACNA1H alters adrenal cell function. Whole-cell patch-clamp measurements revealed a strong activation of mutant CACNA1H at the resting membrane potential of adrenal cells. Both the expression of wildtype and mutant CACNA1H led to a depolarized membrane potential. In addition, cells expressing mutant CACNA1H developed pronounced action potential–like membrane voltage oscillations. Ca²⁺ measurements showed an increased basal Ca²⁺ activity, an altered K⁺ sensitivity, and abnormal oscillating Ca²⁺ changes in cells with mutant CACNA1H. In addition, removal of extracellular Na⁺ reduced CACNA1H current, voltage oscillations, and Ca²⁺ levels in mutant cells, suggesting a role of the partial Na⁺ conductance of CACNA1H in cellular pathology. In conclusion, the pathogenesis of stimulus-independent aldosterone production in patients with CACNA1H mutations involves several factors: i) a loss of normal control of the membrane potential, ii) an increased Ca²⁺ influx at basal conditions, and iii) alterations in sensitivity to extracellular K⁺ and Na⁺. Finally, our findings underline the importance of CACNA1H in the control of aldosterone production and support the concept of the glomerulosa cell as an electrical oscillator.

adrenal gland, primary aldosteronism, voltage-activated T-type calcium channel, CaV3.2, membrane voltage oscillation

Aldosterone is an important regulator of salt and fluid balance and thereby also controls arterial blood pressure. Aldosterone production in adrenocortical glomerulosa cells is mainly stimulated by angiotensin II (ang II), which increases due to volume depletion, and hyperkalemia (high plasma K⁺ concentrations). The stimulatory effect of ang II and high K⁺ in glomerulosa cells largely depends on a modulation of the cell membrane potential and the cytosolic Ca²⁺ activity. Knowledge about the electrophysiological properties and changes in Ca²⁺ activity in aldosterone-producing cells has largely been gained through in vitro studies using either isolated primary glomerulosa cells from different species (1-18) or adrenal cell lines (19-21). According to these studies, aldosterone-producing cells have a constantly hyperpolarized resting membrane potential due to high background K⁺ conductance. This conductance is based on the expression of different K⁺ channels (22), such as TASK K⁺ channels (23, 24). In addition, efficient Ca²⁺ export mechanisms and the Ca²⁺ retention in the endoplasmic reticulum (ER) maintain a very low cytosolic Ca²⁺ activity under resting conditions. Ang II and high K⁺ lead to a constant membrane depolarization by inhibiting the K⁺ channels or by shifting the K⁺ equilibrium potential. This triggers the opening of voltage-dependent Ca²⁺ channels. The resulting rapid Ca²⁺ influx, together with an ang II–induced Ca²⁺ release from the endoplasmic reticulum, greatly enhance the cytosolic Ca²⁺ activity. Finally, Ca²⁺-dependent signaling stimulates the activity and expression of factors required for aldosterone synthesis.

However, in vivo studies with freshly prepared adrenal slices revealed that the hyperpolarized baseline potential in glomerulosa cells is superimposed by oscillatory changes, similar to action potentials in excitable cells (25). Ang II and high plasma K⁺ control the frequency and amplitude of these oscillations as well as the baseline potential. Glomerulosa cells express different types of voltage-gated Ca²⁺ channels (26-28), which become activated depending on the degree of depolarization (29-31). T-type Ca²⁺ channels open at slight depolarization close to the hyperpolarized resting membrane potential (low-voltage activated) as induced by physiological increases in plasma K⁺ and ang II. Moreover, ang II has been shown to shift the activation of T-type Ca²⁺ channels to more hyperpolarized membrane potentials (32, 33). In contrast, L-type Ca²⁺ channels open only at stronger depolarization (high-voltage activated), as occurs during membrane voltage oscillations or when stimulating cells with supraphysiological high ang II or K⁺ levels. The modulation of voltage oscillations by ang II and high K⁺ probably determines the rate and specificity of repetitive opening of the different voltage-activated Ca²⁺ channels in vivo. Inhibition of either L-type or T-type calcium channels reduced aldosterone production from cultivated human adrenal slices, indicating that both types contribute to the control of aldosterone (34). On the other hand, these Ca²⁺ channels are themselves involved in the generation of membrane voltage oscillations (25). At the same time, Ca²⁺ export to the extracellular site and Ca²⁺ recycling between the cytoplasm and the ER, catalyzed by ion transporters (Ca²⁺-ATPases) and exchangers (eg, Na⁺/Ca²⁺-exchangers) takes place (35, 36). In addition, plasma membrane Ca²⁺-ATPases are involved in the generation of Ca²⁺ oscillations in different cell types (37, 38).

In concert, these mechanisms allow a complex regulation of cytosolic Ca²⁺ activity in glomerulosa cells, which can also lead to Ca²⁺ oscillations. So far, there are only a few studies in which Ca²⁺ measurements of glomerulosa cells have been carried out on adrenal slices (39-42). Despite the reported basal membrane voltage oscillations (25), Ca²⁺ peaks were missing (39, 42) or only rarely detected (40, 41) under control conditions. On the other hand, even a very low ang II concentration of 20 pM, as found in human plasma under healthy, unstimulated conditions (43), was sufficient to trigger more single Ca²⁺ spikes or Ca²⁺ bursts in adrenal slices (40). Further elucidation by Guagliardo et al revealed that ang II increases the number of Ca²⁺ bursts in a concentration-dependent manner, but not the frequency of oscillation within each burst (41). Similar to the membrane potential oscillations, Ca²⁺ oscillations were inhibited by T-type Ca²⁺ channel blockers (25, 41). This effect underlines the importance of these channels for the interaction between membrane voltage and Ca²⁺ activity.

Primary aldosteronism, characterized by stimulus-independent aldosterone synthesis, is the leading cause for secondary hypertension and is estimated to occur in about 3% to 12% of all hypertensive patients (44). Bilateral adrenal hyperplasia and unilateral adrenal adenoma (APA) are the major causes of primary aldosteronism (45). Recurrent somatic mutations of ion channels (KCNJ5, CACNA1D) and transporters (ATP1A1, ATP2B3) were found in up to 90% of APAs (46-48). In addition, de novo and familial germline mutations of some of these and other genes, including the chloride channel CLCN2 (49, 50) and the T-type voltage-activated Ca²⁺ channel CACNA1H (CaV3.2), have been identified in patients with primary aldosteronism (51, 52). Recently, a somatic mutation of CACNA1H was found in APA (53). Little is known about the cellular mechanism that induces autonomous aldosterone production in cells with mutant CACNA1H. The functional consequences of CACNA1H mutations on the electrophysiological properties of the channel have so far only been studied in nonadrenal cell models (52, 54). These measurements revealed a gain of channel activity in mutant CACNA1H, which was speculated to increase Ca²⁺ activity and thus aldosterone production in adrenal cells. The expression of mutant CACNA1H^M1549V in an adrenal cell line enhanced basal and stimulated aldosterone production (54). However, Ca²⁺ activity in adrenal cells with mutant CACNA1H has not yet been measured directly. As mentioned above, T-type Ca²⁺ channels influence the control of the membrane potential in adrenal cells in vivo. Therefore, mutations of CACNA1H may also increase Ca²⁺ activity indirectly through changes in the membrane potential.

Here, we examined the electrophysiology and the Ca²⁺ activity of adrenal NCI-H295R cells expressing CACNA1H wildtype and mutant M1549V. In the adrenal cell system, the mutant CACNA1H was activated at a more hyperpolarized membrane voltage in the region of the resting potential of adrenal cells. In addition, mutant CACNA1H enhanced the development of membrane potential oscillations in adrenal cells. Consequently, the mutant CACNA1H caused a severe disorder of the Ca²⁺ homeostasis characterized by increased basal Ca²⁺ activity, altered K⁺ sensitivity, and abnormal Ca²⁺ oscillations. Finally, we found a Na⁺ dependence of CACNA1H current, voltage oscillations, and Ca²⁺ levels in mutant cells, suggesting the involvement of Na⁺ conductance in the pathogenesis of primary aldosteronism in patients with CACNA1H mutations.

Materials and Methods

CACNA1H expression vectors

Plasmid (pcDNA3 [ampR]) containing the human wildtype CACNA1H (Cav3.2a) cDNA sequence (GenBank accession number AF051946), a gift from Dr. Edward Perez-Reyes (University of Virginia, Charlottesville, VA; Addgene plasmid ID45809 (55)), was used as template for subcloning the coding sequence of CACNA1H into pIRES2-DsRed-Express (Clontech Laboratories, Mountain View, CA). Following that, the point mutation c.A4645G (p.Met1549Val) was introduced by site-directed mutagenesis. Finally, coding sequences of wildtype CACNA1H and mutant CACNA1H^M1549V were subcloned into a variant (kindly provided by Dr. Anselm Zdebik, University College London, UK) of the bicistronic expression vector pIRES-CD8 (56), here modified with the multiple cloning site of pIRES2-DsRed-Express.

Cell culture and transfection

Adrenocortical carcinoma NCI-H295R cells (CLS, Eppelheim, Germany) were cultured in DMEM/F12 medium (MG-42, CLS, Eppelheim, Germany) containing: 15 mM HEPES, 6.25 µg/mL insulin, 6.25 µg/mL transferrin, 6.25 ng/mL selenium, 1.25 mg/mL bovine serum albumin, 5.35 µg/mL linoleic acid, and 2.5% Nu-Serum I, supplemented with antibiotics (100 units/mL penicillin, 100 µg/mL streptomycin; Life Technologies GmbH, Darmstadt, Germany). The cells were maintained at 37 °C under a humid atmosphere of 95% air and 5% CO₂. Before the experiments, 2 × 10⁶ NCI-H295R cells were transfected with 5 µg of wildtype CACNA1H or mutant CACNA1H^M1549V containing plasmids or with 3 µg of empty vector using an electroporation system (NEON, Life Technologies GmbH, Darmstadt, Germany). Electroporation was done according to the manufacturer using 1 pulse of 40 ms at 1100 V. After electroporation, cells were cultured in the serum containing medium (1.25 mg/mL bovine serum albumin and 2.5% Nu-Serum I) but without antibiotics, on fibronectin/collagen-coated glass cover slips or without cover slips on 35 mm Falcon dishes (Corning Life Sciences, Amsterdam, the Netherlands). Next, the cells were analyzed by Fura-2 Ca²⁺ measurements (24 hours after electroporation) and patch clamp measurements (24 and 48 hours after electroporation).

Chinese hamster ovary (CHO) cells (CLS, Eppelheim, Germany) were cultured in MEM-alpha medium (Life Technologies GmbH, Darmstadt, Germany) containing 10% FCS, supplemented with antibiotics (50 units/mL penicillin, 50 µg/mL streptomycin; Life Technologies GmbH, Darmstadt, Germany). The cells were maintained at 37 °C under a humid atmosphere of 95% air and 5% CO₂. Before the experiments, 5 × 10⁵ CHO cells were transfected similar to the protocol described for NCI-H295R cells but using 2 pulses of 20 ms at 1400 V for electroporation. After electroporation, cells were cultured in the serum containing antibiotic-free medium on fibronectin/collagen-coated glass cover slips (Corning Life Sciences, Amsterdam, the Netherlands). Following that, cells were analyzed by patch clamp measurements 24 hours after electroporation.

Patch-clamp measurements

Electrophysiological measurements by whole-cell patch-clamp were performed at room temperature on transfected NCI-H295R cells or CHO cells using an EPC 10 amplifier (Heka, Lambrecht, Germany) coupled to a personal computer and a Powerlab Data Acquisition System (ADInstruments GmbH, Spechbach, Germany). The PatchMaster v2x50 software (Heka, Lambrecht, Germany) was used for pulse generation and data acquisition; LabChartPro v7 software (ADInstruments GmbH, Spechbach, Germany) was used for additional data acquisition. Patch pipettes with 5 to 10 MΩ were used for the recordings. The capacitance of NCI-H295R cells averaged 14 picofarads (pF). Transfected cells were identified using anti-CD8 coated Dynabeads (Life Technologies GmbH, Darmstadt, Germany). The patch pipette solution contained: 95 mM K-gluconate, 30 mM KCl, 4.8 mM Na₂HPO₄, 1.2 mM NaH₂PO₄, 5 mM glucose, 2.38 mM MgCl₂, 0.726 mM CaCl₂, 1 mM EGTA, 3 mM ATP, pH 7.2. The extracellular Ringer-type control solution contained (mM): 145 NaCl, 0.4 KH₂PO₄, 1.6 K₂HPO₄, 5 glucose, 1 MgCl₂, 1.3 CaCl₂, 5 HEPES, pH 7.4. For some measurements, the extracellular K⁺ concentration was increased, while Na⁺ was reduced to the same extent. In the Ca²⁺-free solution, Ca²⁺ was replaced by Na⁺. For Na⁺-free solution, Na⁺ was replaced by N-methyl-D-glucamine (NMDG⁺). In the Na⁺- and Ca²⁺-free solution, both ions were replaced by NMDG⁺. The cellular resting membrane potential (V_m) was measured under current clamp (CC0) condition in Ringer-type control solution immediately after breaking into the cell. Voltage-dependent Ca²⁺ currents in NCI-H295R cells transfected with either wildtype or mutant CACNA1H were activated in voltage clamp mode by a series of depolarizing steps (from −90 to −20 mV incremented in 5 mV steps) from a holding potential of −100 mV. Peak inward currents (I) at each voltage clamp step (V_c) normalized to maximal peak current (I_max) were used to obtain the voltage dependence of activation for each cell. Activation and inactivation kinetics as a function of potential were calculated using double-exponential fits to the current traces. Data were analyzed using Fit-Master v2x90.5 software (Heka, Lambrecht, Germany). For evaluation of membrane potential oscillations, baseline membrane potential was hyperpolarized to about −80 mV by injecting a constant negative current (in the range of −3 up to −15 pA), while measuring spontaneous changes of membrane voltage (V_mc) in current clamp mode. This protocol has been chosen, because oscillations in native adrenal glomerulosa cells were only present near the normal resting potential (–80 mV) (25), but most NCI-H295R cells overexpressing wildtype or mutant CACNA1H^M1549V had a more depolarized baseline membrane potential.

Ca²⁺ measurements

Cytosolic free Ca²⁺ activity was measured using the ratiometric fluorescent Ca²⁺ sensitive dye Fura-2-AM (Life Technologies GmbH, Darmstadt, Germany). NCI-H295R cells were loaded at room temperature for 60 minutes with 2 µM Fura-2-AM in the presence of 1X Power Load permeabilizing reagent (Life Technologies GmbH, Darmstadt, Germany). The extracellular Ringer-type control solution contained: 145 mM NaCl, 0.4 mM KH₂PO₄, 1.6 mM K₂HPO₄, 5 mM glucose, 1 mM MgCl₂, 1.3 mM CaCl₂, 5 mM HEPES, pH 7.4. For some measurements, the extracellular K⁺ concentration was increased while Na⁺ was reduced to the same extent. For Na⁺-free Ringer solution, Na⁺ was replaced by NMDG⁺. Mean fluorescence ratios of emission at 490 to 530 nm after excitation at 340 nm and 380 nm were calculated for single transfected cells after subtraction of the background signal using the Axiovision software (Zeiss, Jena, Germany). Transfected cells were identified using anti-CD8 coated Dynabeads (Life Technologies GmbH, Darmstadt, Germany). In some sets of experiments, untransfected cells (without Dynabeads) from each dish with wildtype or mutant CACNA1H–expressing cells were used as an appropriate internal control instead of empty vector–expressing cells as control. Ca²⁺ measurements were performed at 37 °C.

Statistics

The numbers of independent experiments (n) refer to the number of cells or dishes studied to calculate mean values ± SEM or medians. The measurements were carried out at different days and from different cell preparations using different cell passages to ensure the reproducibility of the experiments. In Ca²⁺ measurements, the averaged signals per dish represent the number of experiments since several cells per dish were stimulated at the same time. For patch-clamp experiments, each cell represents an independent experiment because the stimulation protocol for individual cells was started separately. Data were analyzed in Prism software (GraphPad, San Diego, CA) using the appropriate statistical tests as indicated in the text. Unpaired 2-tailed Student t tests or Mann-Whitney tests were used as appropriate to calculate the level of significance for single comparisons. One-way ANOVA plus Tukey post-test was used for multiple comparisons of 1 variable in more than 2 groups. Two-way ANOVA plus Bonferroni post-test was used as appropriate to calculate the level of significance for multiple comparisons of more than one variable. Differences between the groups were considered significant for P < 0.05 or in the case of a paired t test with Bonferroni correction for a lower P value as calculated appropriately.

Results

Electrophysiological properties of wildtype and mutant CACNA1H in adrenal cells

Recently, electrophysiological studies of mutants of the voltage-activated Ca²⁺ channel CACNA1H (Cav3.2) in nonadrenal cell models revealed a higher activity of the mutant channels (51, 52). Here, we examined the properties of mutant CACNA1H and the effects on cellular electrophysiology using adrenal NCI-H295R cells transfected with either human CACNA1H^WT or CACNA1H^M1549V mutant. First, the voltage-dependent activation and inactivation of the CACNA1H-induced currents in adrenal cells was investigated by whole-cell patch-clamp recordings. Similar to the experiments in HEK cells, published by Scholl et al, mutant CACNA1H Ca²⁺ channels displayed a delayed activation and inactivation when expressed in the adrenal cell model (typical trace Fig. 1A and summary of activation and inactivation kinetics in Fig. 1D). However, wildtype and mutant CACNA1H were activated at more negative membrane potentials (median of V_c for I_max: WT at −50 mV, MUT at −60 mV, P < 0.05 applying the Mann-Whitney test) and the difference in activation between the wildtype and mutant channels was more pronounced in adrenal cells (Fig.1B; for comparison with data from HEK cells see (51)). Accordingly, a large part of the mutant CACNA1H channels was already activated in the range of the resting membrane potential of native adrenal cells, which normally lies between −90 and −70 mV at physiological K⁺ concentrations (2, 25, 57). In contrast, wildtype CACNA1H was activated only at membrane potentials higher than −70 mV. Overexpression of wildtype or mutant CACNA1H resulted in depolarization of the membrane potential compared to vector control cells. Depolarization in mutant cells was slightly but not significantly more pronounced than in wildtype cells (Fig. 1C). In order to test possible cell line–specific effects on the electrophysiological properties of CACNA1H, we carried out patch-clamp experiments with CHO cells that overexpressed wildtype or mutant CACNA1H. In the nonadrenal CHO cells expressing wildtype or mutant CACNA1H, the maximum of the transient inward currents was present at a more depolarized membrane potential compared with adrenal NCI-H295R cells (Fig. 1E). Measurements of CHO cells were carried out using the same solutions and patch-clamp protocols as described for NCI-H295R cells. Therefore, cell-specific factors could be the reason for the difference in the voltage-dependent activation of CACNA1H when comparing adrenal and nonadrenal cells. Nevertheless, mutant CACNA1H was also activated at more hyperpolarized potentials than wildtype CACNA1H when it was expressed in CHO cells (median of V_c for I_max: WT at −40 mV, MUT at −53 mV, P < 0.05 applying the Mann-Whitney test).

Figure 1.

Voltage-activated currents (a, b), membrane potential (c), and kinetics of current activation and inactivation (d) in adrenal NCI-H295R cells transfected with human wildtype CACNA1H (black) or mutant CACNA1H^M1549V (grey), compared to vector control cells (white). Voltage-activated currents in nonadrenal Chinese hamster ovary (CHO) cells transfected with human wildtype CACNA1H (black) or mutant CACNA1H^M1549V (grey) (e). (a) Typical traces of transiently activated inward currents in wildtype or mutant CACNA1H–expressing NCI-H295R cells as measured by whole-cell patch-clamp in Ringer-type extracellular solution with 3.6 mM K⁺. Currents were activated by a series of depolarizing steps from −90 to −20 mV (in 5 mV increments) from a holding potential of −100 mV. The dotted line shows the level of zero current. Representative recordings show delayed activation and inactivation of currents in mutant CACNA1H–expressing cells (summary of activation and inactivation kinetics is shown in (d), * indicates P < 0.05 comparing groups at each voltage, tested by 2-way ANOVA and Bonferroni post-test.). Transiently activated inward currents were not detectable under these conditions in vector control cells (data not shown). (b) Voltage dependence of normalized peak currents in adrenal NCI-H295R cells shown as mean values (± SEM) revealed a shift in activation of current in mutant-expressing cells towards the normal negative resting membrane potential (normally about −80 mV) of adrenal cells. Numbers in brackets represent independent experiments per group. * indicates P < 0.05 comparing groups at each voltage, tested by 2-way ANOVA and Bonferroni post-test. (c) Membrane potential under control conditions was depolarized in cells overexpressing wildtype or mutant CACNA1H compared to vector control cells and is shown as mean values (± SEM). The depolarization was more pronounced in mutant cells. Numbers in brackets represent independent measurements per group. * indicates P < 0.05 comparing vector control cells (CD8) with wildtype (WT) or mutant (MUT) cells (tested by 1-way ANOVA and Tukey post-test). Experiments with NCI-H295R cells were carried out at 17 days using cells from different passages. (e) In CHO cells, the maximum of the transient inward currents was present at a more depolarized membrane potential compared to adrenal NCI-H295R cells. However, the mutant CACNA1H was also activated in CHO cells with more hyperpolarized potential than wildtype CACNA1H. * indicates P < 0.05 comparing groups at each voltage, tested by 2-way ANOVA and Bonferroni post-test. Experiments with CHO cells were carried out at 8 days using cells from different passages.

Mutant CACNA1H amplified membrane voltage oscillations in adrenal cells

In addition to the Ca²⁺ influx, T-type Ca²⁺ channels are involved in the generation of oscillatory changes in membrane potential in native glomerulosa cells. It is believed that these oscillations are important for controlling the aldosterone production (25). Under control conditions, adrenal NCI-H295R cells expressing wildtype CACNA1H and most cells with mutant CACNA1H did not show oscillations. Possibly, depolarization induced by overexpression of CACNA1H prevented the formation of membrane potential oscillations. Therefore, cells were hyperpolarized by injecting constant current to set a resting potential normally present in native adrenal cells (approximately −80 mV (25)). In addition to the induced hyperpolarized baseline potential, spontaneous changes in membrane voltage occurred over time (V_mc in Fig. 2). Under these conditions, cells expressing mutant CACNA1H developed marked membrane potential oscillations. These oscillatory changes were characterized by peak-like depolarizations up to a range of −25 to −10 mV. The peaks resulted from a rapid depolarization phase, followed by a fast repolarization back to the baseline, where the membrane potential then remained for a few seconds (Fig. 2B, upper trace). In other mutant cells, the repolarization was delayed, resulting in the following depolarization peak immediately after reaching the baseline potential (Fig. 2B, middle trace). Accordingly, the width of the peaks at the baseline potential of those mutant cells varied between 1.6 and 4.6 seconds (median 2.4 seconds) and the frequency of the oscillations was in a range of 0.05 to 0.55 Hz (median 0.22 Hz). Some cells displayed rather irregular oscillations with peaks of different amplitude and duration (Fig. 2B, lower trace). In contrast, membrane potential oscillations in wildtype CACNA1H–expressing cells (Fig. 2A) were rarely detected, and if so, they consisted of narrower peaks (peak width of 0.84-2.96 seconds, median 1.49 seconds) at a lower frequency (0.02-0.14 Hz, median 0.03 Hz) than in mutant cells (P < 0.05). Vector control cells were devoid of membrane voltage oscillations (data not shown). To determine the possible effects of temperature on the electrophysiological phenotype of mutant CACNA1H in adrenal cells, we repeated a series of patch clamp experiments at 37 °C. However, the membrane potential as well as the voltage dependence of the inward current and the development of membrane potential oscillations in cells with mutant CACNA1H were not significantly affected by the increase in temperature (data not shown).

Figure 2.

Oscillatory changes of the membrane potential in adrenal NCI-H295R cells transfected with human wildtype CACNA1H (a) or mutant CACNA1H^M1549V (b). Changes in membrane potential (V_mc) in wildtype or mutant CACNA1H–expressing cells were measured by whole-cell patch-clamp in Ringer-type extracellular solution containing 3.6 mM K⁺. Measurements were done at constant current injection (in the range of −3 to −15 pA) in order to set a hyperpolarized baseline potential of about −80 mV as found in resting native adrenal cells. Representative data showing membrane potential oscillations were obtained from independent measurements carried out on 17 different days in 39 mutant and 27 wildtype CACNA1H–expressing cells. 95% of mutant cells exhibited membrane potential oscillations. In contrast, oscillations could only be detected in 22% of the wildtype CACNA1H–expressing cells and were characterized by narrower and less frequent peaks. Oscillations were defined as more than 1 depolarizing peak over a period of 3 minutes during the control phase (only 30 seconds are shown here for better visualization of individual peaks).

Next, we tested the influence of increasing extracellular K⁺ concentrations on the membrane potential oscillations (Fig. 3). Under constant current injection, the membrane potential in wildtype CACNA1H–expressing cells as well as the baseline potential between peaks in oscillating mutant cells depolarized upon a stepwise increase of extracellular K⁺ (up to −56 ± 2.3 mV [WT] and −57 ± 1.5 mV [MUT] at 15 mM K⁺). Some wildtype cells, which showed no oscillations at the control K⁺ concentration, developed isolated peaks of depolarization in the range of 4.5 to 6 mM K⁺ (Fig. 3A). Strong basal membrane potential oscillations in mutant CACNA1H–expressing cells ceased at K⁺ concentrations greater than 6 mM, which corresponded to a membrane potential more positive than −70 mV (Fig. 3B). Some cells displayed a higher frequency and smaller amplitude of oscillations at 6 mM K⁺ before the oscillations stopped.

Figure 3.

Impact of increasing extracellular K⁺ concentrations on membrane voltage oscillations in adrenal NCI-H295R cells transfected with human wildtype CACNA1H (a) or mutant CACNA1H^M1549V (b). Typical traces of membrane potential (V_mc as measured in Fig. 2 with constant current injection) upon stepwise increase of extracellular K⁺ (3.6-15 mM). Strong basal oscillations in mutant-expressing cells (b) or rare basal oscillations in wildtype cells (a, upper trace) disappeared at concentrations greater than 6 mM K⁺. In wildtype cells, which showed no basal oscillations, isolated depolarizing peaks occasionally occurred at increased K⁺ concentrations (a, lower trace). The data shown are representative for independent experiments on 10 wildtype and 12 mutant CACNA1H–expressing cells measured on 12 different days.

Na⁺ dependence of CACNA1H current and membrane potential oscillations

In patients with primary hyperaldosteronism due to mutations of other ion channels and pumps, such as KCNJ5 and ATP2B3, autonomous aldosterone production was driven by abnormal Na⁺ conductance and subsequent depolarization (20, 58). In contrast to L-type Ca²⁺ channels, T-type Ca²⁺ channels are not strictly Ca²⁺ selective and allow Na⁺ to pass through to a various extent (59). Therefore, we investigated whether reduced Ca²⁺ selectivity and increased Na⁺ permeability could also play a role in the pathophysiology of mutant CACNA1H. The removal of extracellular Na⁺, which was replaced by impermeable NMDG⁺, equally reduced the voltage-activated transient inward currents in both wildtype and mutant CACNA1H–expressing adrenal NCI-H295R cells (Fig. 4A and 4B, % reduction at voltage with maximal control current: WT 24.5 ± 2.9%, MUT 26.6 ± 3.4%). In addition, a slight shift in the activation curve towards less hyperpolarized potential was seen in mutant cells upon removal of Na⁺ (delta of V_c at I_max: MUT 4.7 ± 1.2 mV, WT 0.8 ± 3.1 mV, P < 0.05, Mann-Whitney test). The simultaneous removal of extracellular Na⁺ and Ca²⁺ led to a complete disappearance of the transient inward current in both wildtype and mutant cells. With a solution free of Ca²⁺ but containing Na⁺, however, a significant inward current returned in mutant cells (Fig. 4B). The maximum of this Na⁺-dependent current in mutant cells was reached at a more negative membrane potential than the control current in the presence of both Ca²⁺ and Na⁺. In contrast, the current in wildtype cells was only slightly and not significantly increased when Na⁺ was added to the Ca²⁺-free solution (Fig. 4A).

Figure 4.

Influence of single or combined removal of Ca²⁺ and Na⁺ from the extracellular solution on voltage-activated whole-cell currents in adrenal NCI-H295R cells transfected with (a) wildtype human CACNA1H or (b) mutant CACNA1H^M1549V. Impact of Na⁺ removal on membrane potential oscillations (c and d). Peak inward currents were measured by a series of depolarizing steps from −90 to −20 mV from a holding potential of −100 mV. Currents were normalized to the maximum peak current under control conditions and are shown as mean values (± SEM). Subsequently, Na⁺ and Ca²⁺ were stepwise removed and replaced by impermeable NMDG⁺. Finally, Na⁺ was reintroduced in a Ca²⁺-free solution (a, b). Removal of extracellular Na⁺ hyperpolarized the mean membrane potential (V_mc as measured in Fig. 2 with constant current injection) in cells with mutant CACNA1H and less pronounced in wildtype cells (c). Typical traces showing the effect of Na⁺ removal on membrane potential oscillations are given in (d). Numbers in brackets represent individual cells measured in independent experiments. * indicates significant difference comparing control vs Na⁺-free solution, ° indicates significant difference comparing Ca²⁺ free vs Ca²⁺/Na⁺-free solution (paired t test, level of significance was corrected by the number of groups and voltage steps, accordingly P < 0.0004 was considered significant).

Next, we tested the impact of the partial Na⁺ conductance of CACNA1H on the membrane potential oscillations observed in patch-clamp measurements with constant current injection. Under these conditions, replacement of extracellular Na⁺ by NMDG⁺ resulted in lower frequency or even disappearance of oscillations in cells with mutant CACNA1H (Fig. 4C and D). Accordingly, the mean membrane potential with constant current injection was hyperpolarized in the Na⁺-free solution (Fig. 4C). Wildtype CACNA1H–expressing cells that rarely showed membrane potential oscillations were also hyperpolarized by Na⁺ replacement, albeit less pronounced than mutant cells (Fig. 4C and D). In contrast, the membrane potential of vector control cells did not change in Na⁺-free solution (data not shown).

Disturbed Ca²⁺ signaling in adrenal cells expressing mutant CACNA1H

The altered electrophysiological properties of the mutant CACNA1H protein were predicted to increase the cytosolic Ca²⁺ activity and thereby stimulate autonomous aldosterone production (51, 52, 54). Here, we directly analyzed the Ca²⁺ signaling using ratiometric Fura-2 Ca²⁺ measurements of adrenal NCI-H295R cells transfected with either human CACNA1H^WT or the CACNA1H^M1549V mutant. Expression of the mutant CACNA1H led to a significant increase in mean cytosolic Ca²⁺ activity under control conditions (3.6 mM extracellular K⁺, Fig. 5A). This increased mean value resulted from a variable change in Ca²⁺ levels in individual mutant cells (Fig. 5C). About two-thirds of the mutant cells (10/16 analyzed cells) showed spontaneous oscillatory changes of the Ca²⁺ activity, while one-third of the cells had increased basal Ca²⁺ levels without oscillations. The Ca²⁺ oscillations were characterized by very high amplitudes with moderately elevated baseline levels (Fig. 5C upper trace) or by smaller amplitudes superimposed on greatly increased baseline Ca²⁺ levels (Fig. 5C middle trace). In contrast, Ca²⁺ levels in wildtype CACNA1H–expressing cells only showed a slight and statistically insignificant increase. Oscillatory Ca²⁺ changes at control conditions were rarely detectable in wildtype cells (4/17 analyzed cells) and if so, their frequency and amplitude were lower than those measured in mutant cells. Overall, the frequency of Ca²⁺ oscillations appeared to be lower than those originating from of membrane voltage oscillations.

Figure 5.

Basal Ca²⁺ activity (a) and impact of increasing extracellular K⁺ concentrations (b, c) or Na⁺ removal (d) on cytosolic Ca²⁺ activity in adrenal NCI-H295R cells transfected with human wildtype CACNA1H (blue) or mutant CACNA1H^M1549V (red), compared to control cells (black). Data represent the mean of the ratio (± SEM) of the Fura-2 fluorescence intensity after excitation at 340 and 380 nm as a measure of cytosolic Ca²⁺. The time course of the Ca²⁺ signal was measured at a basal K⁺ concentration of 3.6 mM followed by a stepwise increase up to 15 mM K⁺. Expression of mutant CACNA1H increased basal Ca²⁺ activity as shown in the bar graph of mean values (± SEM) calculated from the entire control phase of each individual cell (a). Mean Ca²⁺ level in mutant cells increased in the range of 4.5-6 mM K⁺, however oscillatory Ca²⁺ changes stopped at 7.5 mM K⁺ and higher K⁺ concentrations (b, typical traces of single mutant cells are shown in c). Expression of wildtype CACNA1H had no significant effect on basal Ca²⁺ activity (a) but led to a shift in K⁺ sensitivity reflected in a Ca²⁺ response to lower K⁺ levels compared to vector control cells (b). Replacement of Na⁺ by NMDG⁺ stopped the Ca²⁺ oscillations and reduced the mean Ca²⁺ level in mutant cells, while leading to a Ca²⁺ increase in wildtype and control cells (d). Numbers in brackets represent individual dishes measured in independent experiments. Control cells in (d) represent untransfected cells (without CD8-beads) from WT and MUT dishes. * indicates P < 0.05 comparing wildtype (WT) with mutant (MUT) CACNA1H–expressing cells, # indicates P < 0.05 comparing MUT cells with vector control cells (tested by 1-way ANOVA and Tukey post-test).

The Ca²⁺ response of transfected NCI-H295R cells to physiological stimulation of aldosterone production was tested by increasing the extracellular K⁺ concentration in a stepwise manner from 3.6 up to 15 mM (Fig. 5B). In control cells carrying the empty vector, a slight increase in Ca²⁺ activity was triggered by a K⁺ concentration of 7.5 mM. This was markedly intensifying at K⁺ concentrations of 10 mM to 15 mM.

Expression of the wildtype CACNA1H caused a shift in K⁺ sensitivity towards lower K⁺ concentrations. Correspondingly, an increase in Ca²⁺ was observed in these cells as early as 6 mM K⁺ and the maximum of Ca²⁺ activity was already reached at 12.5 mM K⁺. In cells with mutant CACNA1H, the influence of K⁺ elevation on Ca²⁺ activity was different depending on whether the cells featured basal Ca²⁺ oscillations or not. In cells that showed oscillations before, the frequency of oscillations and the baseline levels were increased by applying 4.5 to 6 mM K⁺ (Fig. 5C, upper trace). On the other hand, a further increase in K⁺ then led to a disappearance (or fusion) of the Ca²⁺ peaks, with the mean Ca²⁺ activity still remaining much higher than that in wildtype or control cells. In mutant cells with no basal Ca²⁺ peaks, elevated extracellular K⁺ in the range of 4.5 to 6 mM led to the formation of Ca²⁺ oscillations, which disappeared in line with further K⁺ increase (Fig. 5C, lower trace). Overall, the expression of the mutant CACNA1H severely affected cellular Ca²⁺ homeostasis, causing oscillatory Ca²⁺ changes and an even more severe shift in K⁺ sensitivity towards lower K⁺ concentrations than the wildtype CACNA1H.

Na⁺ dependence of Ca²⁺ oscillation in cells expressing mutant CACNA1H

Next, the influence of extracellular Na⁺ on cytosolic Ca²⁺ activity was investigated (Fig. 5D). Replacement of extracellular Na⁺ by NMDG⁺ resulted in a reduction of the mean cytosolic Ca²⁺ activity in mutant CACNA1H–expressing cells. In parallel, the oscillations disappeared in the Na⁺-free solution in cells showing basal oscillatory Ca²⁺ changes. In contrast, wildtype CACNA1H or untransfected control cells showed an increase in Ca²⁺ activity in the Na⁺-free solution, reaching a similar Ca²⁺ level as found in mutant cells under these conditions.

Discussion

Understanding of the pathophysiological mechanisms underlying primary aldosteronism has made significant progress in the past few years by identifying and functionally characterizing recurrent somatic and germline mutations of ion channels and transporters found in patients with primary aldosteronism. It is believed that most of these mutations lead, directly or indirectly, to a stimulus-independent increase in cytosolic Ca²⁺ activity in adrenocortical glomerulosa cells, which drives autonomous aldosterone production and possibly also proliferation (60). Mutations of the voltage-activated T-type Ca²⁺ channel CACNA1H (Cav3.2) have been suggested to induce aldosterone production also by an increased Ca²⁺ signaling (51, 52). Here, we further investigated the pathophysiological mechanism using adrenal NCI H295R cells expressing wildtype and mutant CACNA1H by considering the following points: i) characterization of the electrophysiological properties of CACNA1H in an adrenal cell context, ii) examination of the effects of mutant CACNA1H on the control of the membrane potential, and iii) direct measurements of cytosolic Ca²⁺ activity under control and stimulated conditions.

Altered electrophysiological properties of mutant CACNA1H in adrenal cells

So far, 5 different germline mutations of CACNA1H have been identified in patients with primary aldosteronism (S196L, M1549V, M1549I, V1951E, P2083L) (51, 52). Electrophysiological characterization of these mutants using nonadrenal cell models revealed changes in activation and inactivation of the mutant channels. The most prominent effect induced by the germline mutant CACNA1H^M1549V was seen in a delayed inactivation during sustained depolarization (51). In addition, the voltage-dependent activation of the mutant CACNA1H was shifted to less depolarized potentials (51). As this effect was only minor, a relevant activation of the mutant at the hyperpolarized resting membrane potential normally found in native glomerulosa cells appeared unlikely. Accordingly, an increase in Ca²⁺ activity in cells expressing mutant CACNA1H was supposed to be primarily triggered during stimulation or due to oscillations in the membrane potential.

Here, we found both wildtype and mutant CACNA1H^M1549V to be activated at more hyperpolarized potentials in adrenal NCI-H295R cells compared to the study in nonadrenal cells (Fig. 1). Thereby, the mutant channel was already strongly active in the range of the hyperpolarized resting membrane potential of native glomerulosa cells. Accordingly, the greater shift in activation of mutant CACNA1H could result in autonomous aldosterone production in patients carrying the mutation without the need for additional stimulation, eg, by fluctuations in plasma K⁺ concentration.

Which factors could underlie the different findings about CACNA1H electrophysiology in adrenal NCI-H295R cells? The electrophysiological properties of CACNA1H and other voltage-dependent Ca²⁺ channels were reported to be different depending on the cell type and the extracellular ion composition (61, 62). In the study using nonadrenal cells (51), patch-clamp measurements were performed with a solution containing a nonphysiologically high extracellular Ca²⁺ concentration (5 mM), while Na⁺ and K⁺ currents were inhibited. Such conditions are usually being applied to improve detection of voltage-dependent Ca²⁺ currents. However, 2 studies reported a shift of voltage dependence of T-type Ca²⁺ channel currents to more depolarized membrane potentials with increasing extracellular Ca²⁺ concentrations (63, 64). Moreover, in contrast to L-type Ca²⁺ channels, T-type Ca²⁺ channels also conduct a relevant amount of Na⁺ ions (59, 65, 66). These effects could contribute to the difference in activation of CACNA1H obtained from our data, as we included Na⁺ and K⁺ ions and used a physiological extracellular Ca²⁺ concentration (1.3 mM). On the other hand, we found CACNA1H current to be activated at a less hyperpolarized potential when the channel was expressed in nonadrenal cells (CHO cells) compared with adrenal cells (Fig. 1E). Thus, ion composition–dependent effects as well as adrenal cell–specific factors contribute to the shift in activation of CACNA1H. These findings should be considered when evaluating functional data on CACNA1H obtained in nonadrenal cell models.

Induction of membrane potential oscillations in adrenal NCI-H295R cells with mutant CACNA1H

Membrane potential oscillations in adrenal cells are supposed to be important for the control of aldosterone production in vivo (25). Unfortunately, their functional characterization is hampered by the fact that primary cultured adrenal cells and adrenal cell lines do not normally exhibit oscillations, while they were found in freshly prepared adrenal slices (25). Here, we were able to elicit strong oscillatory changes of the membrane potential in NCI-H295R cells expressing mutant CACNA1H by hyperpolarizing the baseline potential (Fig. 2 and 3). In contrast, oscillations were rarely inducible in cells with wildtype CACNA1H. Similarly, the membrane potential oscillations in glomerulosa cells of patients are possibly intensified by mutant CACNA1H, thereby driving autonomous aldosterone production.

Why did cells have to be hyperpolarized by current injection in order to induce oscillations? Membrane potential oscillations in native glomerulosa cells were only observed at membrane potentials between −87 and −77 mV (25). In most cells overexpressing wildtype or mutant CACNA1H, however, the membrane potential was depolarized and outside this range (Fig. 1). Several factors probably contributed to the depolarization and prevented the generation of membrane potential oscillations at the control condition: i) alterations induced by the pipette solution, such as exchange of ions and loss of K⁺ channel activity, ii) overexpression of CACNA1H promoting a persistent depolarizing window current of Ca²⁺, and iii) a lack of further factors, such as electrical coupling to adjacent cells.

On the other hand, oscillatory Ca²⁺ changes detected by Fura-2 measurements occurred without current injection. Possibly, the electrical properties of the cells have been better preserved during the Ca²⁺ measurements. Compensation of depolarization during patch-clamp experiments by injection of a hyperpolarizing current, however, induced strong oscillations in cells expressing mutant CACNA1H. This was most likely caused by the abnormally strong activation of the mutant CACNA1H at more hyperpolarized membrane potentials (Fig. 1.). Thereby, the current generated by the mutant could serve as a depolarizing pace-making current and increase the probability and frequency of the oscillations. On the other hand, T-type Ca²⁺ channels indirectly contribute to the repolarization by supporting the activation of Ca²⁺-dependent K⁺ channels. However, the repolarization in mutant cells took longer and correspondingly broader depolarizing peaks were detected. Possibly, repolarization by Ca²⁺-dependent K⁺ channels was outweighed by the delayed inactivation of mutant CACNA1H and the higher influx of positively charged Ca²⁺ ions. The effects described may be less pronounced in patients because only 1 CACNA1H allele is mutated, while the experiments were carried out on NCI-H295R cells overexpressing CACNA1H. On the other hand, baseline depolarization is likely to be smaller in glomerulosa cells. This would favor the activation of mutant CACNA1H, amplifying membrane potential oscillations and inducing autonomous aldosterone production.

Role of partial Na⁺ conductance of CACNA1H for the adrenal phenotype

Our data suggest that part of the pace-making current in adrenal cells expressing mutant CACNA1H depends on the partial Na⁺ conductivity of CACNA1H. A previous study found about 25% of the CACNA1H current to be carried by Na⁺ ions (59). Here, we found a similar Na⁺-dependent inward current in both wildtype and mutant CACNA1H–expressing adrenal cells (Fig. 4). However, the mutant CACNA1H, and consequently the Na⁺-dependent current, was activated at more hyperpolarized membrane potentials. Therefore, the Na⁺ permeability of mutant CACNA1H could particularly contribute to the initial depolarization during oscillations near the hyperpolarized resting membrane potential of adrenal cells. Accordingly, the membrane voltage oscillations in mutant CACNA1H cells were reduced in the Na⁺-free solution, and the baseline membrane potential was hyperpolarized (Fig. 4). The role of Na⁺ current in the mutant CACNA1H is supported by the finding that Na⁺ removal also led to a reduction of the abnormally high cytosolic Ca²⁺ activity as well as the silencing of Ca²⁺ oscillations in NCI-H295R cells with mutant CACNA1H (Fig. 5, further discussed later on). Additional experiments are needed to examine the physiological impact of Na⁺ conductance of wildtype CACNA1H regarding the control of membrane potential in glomerulosa cells. Na⁺-dependent CACNA1H activity may also affect the cellular pathophysiology of other diseases that are associated with CACNA1H mutations (eg, epilepsy (67)).

Correlation between abnormal CACNA1H channel activity and disturbed Ca²⁺ homeostasis

Our electrophysiological measurements on adrenal NCI-H295R cells using a solution with physiological ion concentrations revealed that mutant CACNA1H was already activated in the range of the resting potential of adrenal glomerulosa cells. Consequently, NCI-H295R cells expressing mutant CACNA1H showed strongly elevated cytosolic Ca²⁺ levels at basal conditions. (Fig. 5A). In addition, oscillatory Ca²⁺ changes were detected in mutant cells, which disappeared at higher extracellular K⁺ concentrations (Fig. 5B) as was the case for the membrane potential oscillations in patch-clamp measurements (Fig. 3 and 5C). In contrast, basal Ca²⁺ levels were not significantly increased in cells with wildtype CACNA1H, and Ca²⁺ oscillations were rarely detected. This corresponds to the low activity of the wildtype channel at the resting potential of NCI-H295R cells. Nevertheless, overexpression of wildtype CACNA1H increased the sensitivity of cytosolic Ca²⁺ to lower extracellular K⁺ concentrations, indicating that wildtype CACNA1H was activated by the K⁺ induced depolarization (Fig. 5B). At least 2 factors contributed to the increased basal Ca²⁺ level in mutant cells: i) the prolonged Ca²⁺ influx through the mutant CACNA1H itself, which was activated at a more hyperpolarized membrane potential and inactivated more slowly, and ii) the stronger basal depolarization and the induction of membrane voltage oscillations, which probably increased the activation of mutant CACNA1H and of high voltage-gated L-type Ca²⁺ channels. L-type channels are possibly also activated in depolarized wildtype cells, but an increased Ca²⁺ efflux may compensate for the higher Ca²⁺ influx, at least at basal conditions. In addition, there are only a few oscillations of V_m and Ca²⁺ in cells with wildtype CACNA1H, which, however, may be decisive for the repeated opening of the L-type Ca²⁺ channels.

Moreover, our data indicate that the partial Na⁺ conductance of CACNA1H contributed to the depolarization of the membrane potential promoting the Ca²⁺ influx in mutant cells. Accordingly, the elevated basal Ca²⁺ levels and Ca²⁺ oscillations in cells with mutant CACNA1H were reduced by removing extracellular Na⁺ (Fig. 5D). At the same time, Ca²⁺ export by the Na⁺/Ca²⁺ exchanger was either inhibited or its direction of transport changed leading to abnormal Ca²⁺ import upon removal of extracellular Na⁺. Thereby, the reduction of the Ca²⁺ influx through the mutant CACNA1H in Na⁺-free solution was possibly masked to some extent. Inhibition of the Na⁺/Ca²⁺ exchanger may also explain the increase of Ca²⁺ in wildtype CACNA1H–expressing and vector control cells upon removal of extracellular Na⁺.

Conclusion

This study provides new insights into the pathophysiological mechanisms induced by mutations of the voltage-dependent T-type Ca²⁺ channel CACNA1H in patients with primary aldosteronism. The mutant CACNA1H is already activated at the resting membrane potential of adrenal cells in the presence of physiological ion concentrations. This leads to a severely disturbed control of membrane potential and Ca²⁺ homeostasis, including oscillatory changes in both parameters. These data provide an explanation for the stimulus-independent autonomous aldosterone production in patients carrying CACNA1H mutations. Moreover, the partial Na⁺ conductivity of CACNA1H contributes to the phenotype of adrenal cells. Overall, our findings underline the importance of CACNA1H for the physiological control of aldosterone production and support the concept of the glomerulosa cell as an electrical oscillator. Finally, our study illustrates the importance of the cellular context for the functional characterization of adrenal membrane proteins.

Abbreviations

Abbreviations
 
• ang II

  angiotensin II

 
• APA

  adrenal adenoma

 
• CaV

  voltage-activated calcium channels

 
• CHO

  Chinese hamster ovary

 
• ER

  endoplasmic reticulum

 
• I

  inward current

 
• L-type

  long-lasting type (calcium channel)

 
• NMDG⁺

  N-methyl-D-glucamine

 
• T-type

  transient type (calcium channel)

 
• V_m

  membrane potential

 
• V_c

  clamped potential in patch-clamp measurements

 
• V_mc

  membrane potential at constant current injection

Acknowledgments

Financial Support: The study was supported by the Deutsche Forschungsgemeinschaft (project number 403208210 to R.W. and project number 392121970 to S.B.)

Additional Information

Disclosure Summary: The authors have nothing to disclose.

Data Availability: All data generated or analyzed during this study are included in this published article or in the data repositories listed in References.

Author notes

These authors contributed equally.