L- and T-type calcium channels control aldosterone production from human adrenals

Aldosterone, which plays a key role in the regulation of blood pressure, is produced by zona glomerulosa (ZG) cells of the adrenal cortex. Exaggerated overproduction of aldosterone from ZG cells causes primary hyperaldosteronism. In ZG cells, calcium entry through voltage-gated calcium channels plays a central role in the regulation of aldosterone secretion. Previous studies in animal adrenals and human adrenal adrenocortical cell lines suggest that the T-type but not the L-type calcium channel activity drives aldosterone production. However, recent clinical studies show that somatic mutations in L-type calcium channels are the second most prevalent cause of aldosterone-producing adenoma. Our objective was to define the roles of T and L-type calcium channels in regulating aldosterone secretion from human adrenals. We find that human adrenal ZG cells mainly express T-type CaV3.2/3.3 and L-type CaV1.2/1.3 calcium channels. TTA-P2, a specific inhibitor of T-type calcium channel subtypes, reduced basal aldosterone secretion from acutely prepared slices of human adrenals. Surprisingly, nifedipine, the prototypic inhibitor of L-type calcium channels, also decreased basal aldosterone secretion, suggesting that L-type calcium channels are active under basal conditions. In addition, TTA-P2 or nifedipine also inhibited aldosterone secretion stimulated by angiotensin IIor elevations in extracellular K+. Remarkably, blockade of either Lor T-type calcium channels inhibits basal and stimulated aldosterone production to a similar extent. Low concentrations of TTA-P2 and nifedipine showed additive inhibitory effect on aldosterone secretion. We conclude that Tand L-type calcium channels play equally important roles in controlling aldosterone production from human adrenals.


Introduction
Aldosterone, a steroid hormone produced by zona glomerulosa (ZG) cells of the adrenal cortex, plays a key role in maintaining electrolyte and water balance, and thereby the control of blood pressure. Primary aldosteronism (PA), Rossi 2011, Hannemann & Wallaschofski 2012, compared to 20% of individuals with resistant hypertension (Calhoun et al. 2002, Douma et al. 2008, Rossi 2011, Hannemann & Wallaschofski 2012. Idiopathic hyperaldosteronism and aldosterone-producing adenoma are the two most common subtypes of PA (Funder et al. 2008, Monticone et al. 2017. Notably, somatic mutations of potassium and calcium channels have been found in more than half of aldosterone-producing adenomas (Yang et al. 2018).
Intracellular Ca 2+ regulates key steps in aldosterone biosynthesis from ZG cells . Calcium controls the early rate-limiting step, the conversion of cholesterol to pregnenolone, by stimulating the transfer of cholesterol to CYP11A1 on the inner mitochondrial membrane (Cherradi et al. 1996). Calcium controls the late rate-limiting step, the conversion of deoxycorticosterone to aldosterone, by stimulating the production of NADPH, a required cofactor for CYP11B2 activity (Lifton et al. 1992, Rossier et al. 1996a, Lalevee et al. 2003, Wiederkehr et al. 2011. In ZG cells, a rise of intracellular Ca 2+ can be generated either by releasing Ca 2+ from intracellular stores or by increasing extracellular Ca 2+ entry through voltagegated calcium channels (Lotshaw 2001).
Extracellular Ca 2+ entry through ZG voltage-gated calcium channels is necessary to maintain stimulated secretion induced by angiotensin II (Ang II) or K + , which are the two primary physiological aldosterone secretagogues in vivo , Yang et al. 2018. ZG voltage-gated calcium channels are differentially expressed among mammalian species. Mainly two types of calcium currents have been identified in human, rat, mouse, and bovine ZG cells: high voltage-activated (L-type, including Ca V 1.1-1.4 subtypes) and low voltageactivated (T-type, including Ca V 3.1-3.3 subtypes) calcium channel currents (Cohen et al. 1988, Payet et al. 1994, Rossier et al. 1996b, Schrier et al. 2001, Hu et al. 2012. However, previous studies indicate that the activity of the T-type, but not that of the L-type channel, correlates with aldosterone secretion in dispersed preparations of animal ZG cells and human adrenocortical cell lines (Barrett et al. 1995, Rossier et al. 1996b, Imagawa et al. 2006, Akizuki et al. 2008. Thus, surprisingly recent clinical studies have found more than 25 different somatic Ca V 1.3 L-type calcium channels mutations in patients with PA (Yang et al. 2018), suggesting that in humans L-type calcium channels may also play an equally important role in controlling aldosterone production.
In this study, we assessed the contribution of T-and L-type calcium channels to the control of basal and stimulated aldosterone production from human adrenal slices. We find that both channel types participate in the control of aldosterone production from human ZG cells that remained within adrenal tissue slices.

Cell culture and chemicals
Human adrenocortical cell line, H295R (ATCC#: CRL-2128), was purchased from American Type Culture Collection. Cells were cultured in Dulbecco's modified Eagle's medium: Nutrient Mixture F-12 medium (GIBCO) supplemented with 10% fetal bovine serum (GIBCO), 1% antibiotic solution, and 0.1% insulin, transferrin and selenium premix (BD Biosciences). Human embryonic kidney cell line, HEK293, was purchased from The Cell Bank of Chinese Academy of Science. Cells were cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum and 1% antibiotic antimycotic solution. TTA-P2 was purchased from Alomone Laboratories; Angiotensin II and nifedipine were purchased from Sigma-Aldrich.

Aldosterone secretion of human adrenal gland
Adrenal gland samples were obtained from 28 patients (16 men and 12 women, age 36-70 years old) undergoing radical nephrectomy to remove kidney cancer and adjacent ipsilateral adrenal. The protocol for obtaining and using human adrenal tissues in this work was approved by the Ethics Committee of Shanghai Cancer Center (050434-4-1212B) and School of Life Sciences (2017-601), Fudan University. Written informed consent from all patients was obtained. All protocols were in accordance with institutional guidelines.
Human adrenal samples were prepared as previously described (Yang et al. 2016). Briefly, human adrenal samples were transported in cold PBS, and then kept in ice-cold bicarbonate-buffered saline containing: 140 mmol/L NaCl, 2 mmol/L KCl, 0.1 mmol/L CaCl 2 , 5 mmol/L MgCl 2 , 26 mmol/L NaHCO 3 , and 10 mmol/L glucose, bubbled with 95% O 2 and 5% CO 2 . Fat tissues around adrenal glands were removed. Tissues were embedded in 2.5% low-melting temperature agarose (2-hydroxyethylagarose, Sigma-Aldrich) and sectioned into 300 µm slices using DSK microslicer (DOSAKA). The slices were placed on Millicell-CM membranes (pore size 0.4 µm, Millipore) and incubated in Hanks MEM (Gibco) containing 1 mmol/L l-glutamine, 1 mmol/L CaCl 2 , 2 mmol/L MgCl 2 , 30 mmol/L HEPES, 12.8 mmol/L glucose, 5.24 mmol/L NaHCO 3 , 12 mg/L bovine insulin, 0.12% ascorbic acid, and 20% horse serum (Gibco) at 37°C with 5% CO 2 . Because the steroid synthesis profile of ZG cells may change during primary culture (Williams et al. 2014), in this study, we conducted relatively short time course secretion experiments. Slices were incubated for 2 h to determine basal production (control), and then treated with the indicated agents for 6 h to evaluate stimulated production. Slice culture supernatants were collected before and after treatment. Supernatant aldosterone concentrations were measured by an Aldosterone ELISA Kit (ENZO Life Science) as previously reported (Yang et al. 2016). For 62.5 pg/mL and 125 pg/mL aldosterone, the intra-assay co-efficients of variation were 5.2 and 6.2, and the inter-assay co-efficients of variation were 11.7 and 7.9. A standard curve (ranging from 3.9 pg/mL to 500 pg/mL) was conducted along with samples for every measurement. 10 µL supernatant (1:10 dilution in assay buffer) was used for aldosterone ELISA assay. In this condition, the mean percent recovery for 150 pg/mL and 50 pg/mL aldosterone were 93 and 98% respectively. Data were presented as fold over baseline to avoid ZG mass differences among the adrenal slices.

Data presentation and statistical analysis
Data analysis was performed with Clampfit 10.2 (Axon Instruments) and Origin 8.0 (OriginLab). The normal distribution of the data was checked by the Shapiro-Wilk test. Statistical analysis was performed using paired Student t tests unless indicated otherwise. The nonparametric Kruskal-Wallis test was used to compare groups that did not meet the assumption of normality. Values are given as means ± s.e.m., n indicated the number of tested cells, patient samples or independent tests. P < 0.05 was considered statistically significant.

L-and T-type calcium channels regulate basal aldosterone production from human adrenal slices
To determine which L-type calcium channel subtypes were expressed in human ZG cells, we isolated ZG cells by laser capture microdissection and tested for mRNA expression. As shown in Fig. 1A, mRNA for Ca V 1.2 and Ca V 1.3 but not Ca V 1.1 and Ca V 1.4 was detected in ZG cells. All PCR products were confirmed by sequencing. To corroborate protein expression of Ca V 1.2 and Ca V 1.3 channels we used immunofluorescence. Immunodetection showed high expression of both Ca V 1.2 and Ca V 1.3 channel protein in the ZG layer of the human adrenal (Fig. 1B). By contrast, although the mRNA for all three T-type calcium channel subtypes was detected in ZG cells ( Fig. 2A), the expression of T-type calcium channel protein, as reported by immunofluorescence, showed only strong expression for Ca V 3.2 and Ca V 3.3 channel protein. Signals for Ca V 3.1 channels were not detected (Fig. 2B, C and D) in the ZG layer of the human adrenal cortex, indicating a discrepancy between mRNA and protein expression. To assess the selectivity of these antibodies, and thus verify our putative expression findings, HEK293 cells were transfected with a vector encoding one of five channel subtypes to provide both positive and negative immunodetection controls. HEK293 cells expressing Ca V 3.1, Ca V 3.2, Ca V 3.3, Ca V 1.2 and Ca V 1.3 channels showed strong immunofluorescence signals only with antibodies against their own antigenic target; no signals were detected in either un-transfected HEK293 cells or in cells transfected with other channel family members ( Supplementary Figs 1 and 2, see section on supplementary materials given at the end of this article). Thus, the normal human adrenal cortex, indeed expresses proteins for both Ca V 1.x and Ca V 3.x channels.
Next, we used the L-type calcium channel-specific inhibitor, nifedipine and the T-type calcium channelspecific inhibitor, TTA-P2, to test the contribution of each channel class to basal aldosterone secretion from human adrenal slices. 500 nmol/L nifedipine significantly and consistently reduced aldosterone secretion from adrenal tissue slices prepared from ten individual patient adrenals (percent inhibition: 46 ± 6%, n = 10, P < 0.01 compared with control by Student t tests, Fig. 3A). 2 µmol/L TTA-P2 also significantly inhibited aldosterone secretion from adrenal tissue slices prepared from 11 individual patient adrenals (percent inhibition: 44 ± 5%, n = 11, P < 0.01, Fig. 3B).
Although nifedipine is considered to be a highly selective L-type calcium channels blocker (Stengel et al. 1998), we tested if nifedipine could also partially inhibit T-type calcium channels at high nanomolar concentration. Because human adrenal ZG cells express both Ca V 3.2 and Ca V 3.3 channels, we evaluated the inhibition of Ca V 3.2 and Ca V 3.3 currents in H295R cells by 500 nmol/L nifedipine. Calcium currents were elicited by a 50-ms depolarizing pulse to −20 mV from a holding potential of −90 mV at 10 s intervals. 500 nmol/L nifedipine did not inhibit either Ca V 3.2 or Ca V 3.3 channel currents, whereas 2 µmol/L TTA-P2 fully inhibited the currents ( Fig. 4A and B, n = 5-6). To confirm the efficacy of nitrendipine, we tested its effects on Ca V 1.2 and Ca V 1.3 channel currents at 500 nmol/L. As shown in Fig. 4C and D, 500 nmol/L nifedipine almost fully inhibited Ca V 1.2 and Ca V 1.3 channel currents in H295R cells, in contrast to 2 µmol/L TTA-P2 which had little effect (n = 5). To rule out a nonselective inhibitory effect of these drugs at voltages other than −20 mV, we recorded the current-voltage relationship for each channel (Ca V 1.2, 1.3, 3.2, 3.3) and tested for drug-induced changes. At all voltages, 500 nmol/L nifedipine or 2 µmol/L TTA-P2 remained selective for their channel targets ( Supplementary Fig. 3). The data above indicate that the inhibition of L-or T-type calcium channel activity should account for the induced effects of 500 nmol/L nifedipine or 2 µmol/L TTA-P2 reported here.

Effect of nifedipine and TTA-P2 on Ang II-and K + -induced aldosterone secretion from human adrenal slices
To assess the overall importance of L-type and T-type calcium channels in the production of aldosterone in human adrenals, we tested for nifedipine or TTA-P2 evoked modulation of Ang II-or K + -induced aldosterone secretion. As shown in Fig. 5A, 500 nmol/L nifedipine abrogated the stimulation of aldosterone production by 10 nmol/L Ang II (fold increase: Ang II, 2.4 ± 0.6; Ang II + nifedipine, 0.9 ± 0.1, n = 6, P < 0.05 compared with

Journal of Endocrinology
Ang II alone, Fig. 5B). 500 nmol/L nifedipine also abolished aldosterone production induced by High K + (15 mmol/L) (fold increase: High K + , 2.4 ± 0.4; High K + + Nifedipine, 1.0 ± 0.2, n = 6, P < 0.05 compared with High K + alone, Fig.  5C and D). Interestingly, 2 µmol/L TTA-P2 also abrogated the stimulation of aldosterone production by Ang II or High K + from acutely prepared slices of human adrenals (P < 0.05, Fig. 6). Together, these data indicate that in human adrenals both T-and L-type calcium channel activity is required for stimulated aldosterone synthesis.

Additivity of effects of nifedipine and TTA-P2 on aldosterone production from human adrenal slices
Finally, we tested whether the combined application of nifedipine and TTA-P2 could reduce aldosterone secretion from human adrenal slices further than the level of inhibition induced by either drug alone. As shown in Fig. 7A, 2 µmol/L TTA-P2 did not augment the degree of inhibition induced by nifedipine (P > 0.05 compared with nifedipine alone). Similarly, 500 nmol/L nifedipine did not increase the inhibition of basal aldosterone production from human adrenal slices produced by TTA-P2 (n = 5, P > 0.05 compared with TTA-P2 alone, Fig. 7B). To further test for drug additivity, we used lower concentrations of each antagonist, near the IC 50 for T or L channel current inhibition (Fig. 7C). As shown in Fig. 7D, the inhibition of basal aldosterone secretion from human adrenal slices by 200 nmol/L TTA-P2 was additive to that of 100 nm/L nifedipine (n = 3). Drug additivity was also demonstrable against agonist-stimulated secretion. The inhibition of Ang II (10 nmol/L) or high K + (15 mM)-stimulated aldosterone secretion from human adrenal slices induced by either TTA-P2 (200 nmol/L) or nifedipine (100 nm/L) was less than that effected by these channel antagonists in combination (Fig. 8). These data indicate that both Tand L-type calcium channels maybe necessary to drive aldosterone production. With aging, the cellular capacity to synthesize aldosterone declines (Hegstad et al. 1983, Nanba et al. 2018. Less aldosterone synthase (CYP11B2) is expressed in the human ZG and more is expressed in aldosteroneproducing clusters (APCCs) found beneath the adrenal capsule (Nishimoto et al. 2010(Nishimoto et al. , 2016. APCCs, formed by non-neoplastic cells, increase with age and harbor mutations in the CACNA1D gene that encodes Ca V 1.3 channels (Nishimoto et al. 2015, Nanba et al. 2017). Therefore, we calculated the size effect (percent inhibition by nifedipine) as a function of patient age to determine if the nifedipine-induced inhibition of aldosterone production reported here increased with patient age. In the relatively small sample size of our study, we found no significant correlations with aging suggesting a role for normal human L-type calcium channels in the control of aldosterone synthesis (basal production: R = −0.21; P = 0.56 ( Supplementary Fig. 4); Ang II stimulated: R = 0.23; P = 0.66 (data not shown); K-stimulated: R = 0.24; P = 0.64 (data not shown)).

Discussion
Here, we report the novel findings that in human adrenal glomerulosa cells both basal and stimulated rates of aldosterone production depend on voltage-gated

Figure 5
Effects of nifedipine on Angiotensin II (Ang II) -or high K + -stimulated aldosterone secretion from human adrenal slices. (A) Nifedipine abrogates Ang II-stimulated aldosterone secretion from human adrenal slices. (B) Statistical analysis of effect of nifedipine-induced inhibition of Ang II-stimulated aldosterone secretion. Data are presented as fold over control. *P < 0.05 compared with control (n = 6). # P < 0.05 compared with Ang II alone (n = 6). (C) Nifedipine abrogated high K + -stimulated aldosterone secretion from human adrenal slices. (D) Statistical analysis of nifedipine-induced inhibition of high K + -stimulated aldosterone secretion. Data are presented as fold over control. *P < 0.05 compared with control (n = 6). # P < 0.05 compared with high K + alone (n = 6).

Figure 6
Effect of TTA-P2 on Angiotensin II (Ang II) -or high K + -stimulated aldosterone secretion from human adrenal slices. (A) TTA-P2 abrogates Ang II-stimulated aldosterone secretion from human adrenal slices. (B) Statistical analysis of TTA-P2 induced inhibition of Ang II-stimulated aldosterone secretion. Data are presented as fold over control. *P < 0.05 compared with control. # P < 0.05 compared with Ang II alone (n = 4). (C) TTA-P2 abrogates high K + -stimulated aldosterone secretion from human adrenal slices. (D) Statistical analysis of TTA-P2 induced inhibition of high K + -stimulated aldosterone secretion. *P < 0.05 compared with control. # P < 0.05 compared with high K + alone (n = 5).
calcium currents through T-type as well as L-type calcium channels. These findings are surprising because patchclamp studies consistently report the basal membrane potential of ZG cells to be very hyperpolarized (around −80 mV), a potential at which high threshold L-type calcium channels are not expected to be active. In contrast, low threshold T-type calcium channels are activated at more hyperpolarized potentials and have a permissive window of steady-state activity at more hyperpolarized potentials (Perez-Reyes 2003). When stimulated by low physiological concentrations of Ang II or small increases in extracellular potassium, the membrane potential of ZG cells can rapidly reach the permissive voltage window for persistent T-type, but not L-type, calcium channel activity, that allows steady-state calcium influx (Rossier 2016). Therefore, in primary cultured ZG cells (Python et al. 1993, Rossier et al. 1993, 1996b, 1998, Lotshaw 2001 or H295R cells (Akizuki et al. 2008), it is not surprising that Ang II-or K + -stimulated aldosterone secretion consistently has been reported to be efficiently inhibited by various T-type calcium channel blockers, and not inhibited by highly selective L-type calcium channel inhibitors such as nifedipine. Nevertheless, previously we reported that mouse ZG cells when retained within an adrenal slice spontaneously generate Vm oscillations (Hu et al. 2012), of high amplitude, which can reach depolarized membrane voltages close to but always below 0 mV, voltages at which L-type calcium channels are active. Therefore, the contribution of L-type calcium channels to the regulation  of aldosterone production previously may have been underestimated in studies of dispersed ZG cells that have lost the intrinsic ability to oscillate persistently.
Indeed, here we find that within human adrenal slices the L-type-specific blocker nifedipine significantly inhibited secretagogue-stimulated aldosterone secretion from ZG cells. Although some dihydropyridines that are considered to be specific L-type channel inhibitors, such as nicardipine and niguldipine also inhibit T-type calcium channels, nifedipine discriminates between T-and L-type calcium channel families (Stengel et al. 1998). Nifedipine selectively inhibits L-type calcium channels in human adrenocortical H295R cells (Rossier et al. 1996b, Imagawa et al. 2006, Akizuki et al. 2008, Yang et al. 2016) and in primary cultured human adrenal glomerulosa cells (Payet et al. 1994), and at 500 nM in our study had no effect on recombinant Ca V 3.2 and Ca V 3.3 channel currents expressed in human adrenal cells. Therefore, inhibition of L-type calcium channel activity should account for the nifedipine-induced reduction of aldosterone production reported here. Similar to previous studies (Dreyfus et al. 2010, Enyeart & Enyeart 2015, 2 µm TTAP2 fully inhibits T-type calcium channels and has no effect on L-type calcium channels. That nifedipine also inhibited aldosterone secretion from unstimulated human adrenal slices implies that L-type calcium channels are active even under basal condition. Again, we hypothesize that if human ZG cells within adrenal slices spontaneously generate Vm oscillations, as mouse ZG cells, this would enable L-type calcium channels to participate in regulating basal production. The direct proof of this prediction requires voltage recordings of human ZG cells by patch clamp. However, unfortunately we found that obtaining a gigaseal on adult human ZG cells (data not shown) was not possible. Rather, we used the very limited human adrenal samples for aldosterone secretion experiments. It is surprising that the degree of inhibition of aldosterone secretion evoked by nifedipine was equivalent to that of TTA-P2. At concentrations that produced maximal inhibition of L or T channel currents, no additivity of drug effects or synergy was observed, whereas at near half-maximal concentrations, nifedipine and TTA-P2 showed additive inhibitory effects on both basal and agonist-stimulated aldosterone secretion. Together, these observations suggest that in the human ZG layer, L-and T-type calcium channel flux may not be redundant but rather may each subserve a unique function that is required for the control of aldosterone production. Therefore, we propose a model for the regulation of aldosterone secretion from human ZG cells by T-and L-type channels that can explain our findings. We predict that like mouse ZG cells, human ZG cells may also spontaneously oscillate, allowing the unique biophysical properties of each channel class to have distinct roles in controlling the oscillatory cycle. We hypothesize that by opening at more hyperpolarized potentials T-type channels may be principally responsible for recruiting L-type channels that, in turn, may provide the major calcium source of calcium required for the production of aldosterone; AngII or high K + may increase oscillation frequency and thus control the magnitude of calcium entry and the production of aldosterone.
The samples used in this study were obtained from patients spanning a broad age range and would be expected to contain an age-dependent increase in APCCs. Indeed, with tissue remaining from some patient samples, we detected a very large CYP11B2 expressing APCCs in one of three samples ( Supplementary Fig. 5). Approximately onethird of APCCs (Nishimoto et al. 2015) harbor mutations in Ca V 1.3 channels that lower the voltage dependence for channel activation or reduce channel inactivation (Azizan et al. 2013, Scholl et al. 2013, biophysical alterations that would make mutant channels more likely to participate in regulating calcium at hyperpolarized potentials. If mutant channels were the sole target of nifedipine action reported here, we would have expected an age-dependent increase in the efficacy of nifedipine, which we did not observe ( Supplementary Fig. 4). Thus, we conclude that normal Ca V 1.3 channels as well as mutant Ca V 1.3 channels participated in the response to nifedipine. The relative contribution of each channel type to the inhibition of aldosterone synthesis induced by nifedipine in normal adrenal tissue is worthy of a future expanded study with a larger sample size.
In conclusion, our data demonstrate that L-type calcium channels play an important role in the regulation of aldosterone production from human ZG cells within a tissue context and that both T-and L-type calcium channels critically contribute to aldosterone production from human adrenals.