INTRODUCTION

The major physiological regulators of aldosterone synthesis and secretion by adrenal glomerulosa cells are angiotensin II (Ang II)¹ and potassium (1). Although acting through different mechanisms, these two extracellular stimulatory factors trigger steroidogenesis by a process involving the calcium messenger system (2, 3, 4, 5, 6, 7, 8, 9, 10).

The crucial role of the Ca²⁺ messenger in the acute regulation of aldosterone production in adrenal glomerulosa cells is well recognized (4, 11, 12). However, the sites of action and the specific molecular targets of calcium along the complex steroidogenic cascade are poorly defined. A first direct indication of the involvement of the adrenal glomerulosa mitochondria as targets for the Ca²⁺ messenger has been obtained when our laboratory has shown using permeabilized bovine glomerulosa cells that changes of ambient Ca²⁺ within the range of the physiological cytosolic concentrations are able to activate aldosterone production and that this effect can be prevented by ruthenium red, a blocker of the mitochondrial Ca²⁺ uniport (13). Moreover, recent work has allowed us to narrow the potential target domain for Ca²⁺ to the very early steps of steroidogenesis (14), which occur inside the mitochondria.

Indeed, the acute response of steroidogenic cells to hormone stimulation involves the mobilization of cholesterol from intracellular lipid droplets to the mitochondrial inner membrane, where the first enzymatic step of steroidogenesis, namely the conversion of cholesterol to pregnenolone by the cytochrome P450_scc, occurs (15). The rate-limiting and hormonally regulated step in this process is the delivery of cholesterol from the outer to the inner mitochondrial membrane. This step is known to require de novo protein synthesis. Studies from several laboratories have shown that a family of hormone-induced and cycloheximide-sensitive 30-kDa mitochondrial proteins, described in different steroidogenic cell types (16, 17, 18, 19, 20), play a crucial role in the acute regulation of steroid synthesis. Recently, the steroidogenic acute regulatory (StAR) protein has been proposed as an essential mediator of the acute steroidogenic response. Cholesterol transfer is believed to be facilitated by contact sites that occur between the outer and the inner mitochondrial membranes during the import of the StAR protein precursor into the mitochondria (20). This hypothesis is strengthened by many observations showing that mitochondrial contact sites are involved in phospholipid and protein import into the mitochondria (21, 22).

Although the regulation of the cholesterol transfer steps has been almost exclusively investigated in response to elevated cAMP levels in adrenal fasciculata cells, many questions remain unanswered concerning the possible role of Ca²⁺ in the regulation of cholesterol mobilization in adrenal glomerulosa cells. The aims of the present study were firstly to determine whether the changes in intracellular calcium concentration triggered by activators of steroidogenesis in glomerulosa cells are accompanied by concomitant changes in cholesterol distribution in mitochondrial membranes and secondly to examine the effect of the inhibitor of protein translation, cycloheximide (CHX), on the calcium-mediated cholesterol transfer from the outer to the inner mitochondrial membrane.

EXPERIMENTAL PROCEDURES

Ionomycin was purchased from Calbiochem (Lucerne, Switzerland) and [Ile⁵]-Ang II from Bachem (Bubendorf, Switzerland). Cholesterol oxidase, peroxidase, p-hydroxyphenylacetic acid, aminogluthetimide, cycloheximide, and all other chemicals were obtained from Sigma.

Bovine adrenal glands were obtained from a local slaughterhouse. Zona glomerulosa cells were prepared by enzymatic dispersion with dispase and purified on Percoll density gradients as described in detail elsewhere (23). Purified glomerulosa cells were resuspended at a density of 10⁶ cells/ml in a modified Krebs-Ringer buffer (NaCl, 136 mM; NaHC0₃, 5 mM; KH₂PO₄, 1.2 mM; MgSO₄, 1.2 mM; KCl, 1.8 mM; CaCl₂, 1.2 mM; D-glucose, 5.5 mM; Hepes, 20 mM; pH 7.4) and preincubated at 37 °C for 1 h before being used in the subsequent experiments.

After having been washed in Krebs-Ringer buffer, glomerulosa cells were Ca²⁺-clamped as reported elsewhere (14) in the presence of 2 µM ionomycin, 1 mM total extracellular Ca²⁺, and 0.2 mM EGTA, in order to achieve an intracellular [Ca²⁺]_c of 600 nM. 500 µM of aminogluthetimide (AMG) was included in the incubation medium to avoid cholesterol side chain cleavage by the cytochrome P450_scc located in the inner mitochondrial membrane. At the end of a 2-h incubation period at 37 °C, the cells were sedimented at 200 × g for 15 min. All subsequent operations were conducted at 4 °C, in buffers containing 500 µM AMG.

Glomerulosa cells were homogenized with a Potter-Elvehjem homogenizer (1200 rpm, 35 strokes) in a 5 mM Tris-HCl buffer, pH 7.4, containing 275 mM sucrose. The homogenate was centrifuged at 200 × g for 15 min to remove large debris and nuclei. Further centrifugation of the supernatant at 10,000 × g for 10 min yielded the mitochondria. The mitochondrial pellet was washed twice at 8,000 × g with the same buffer.

Submitochondrial particles were prepared as described elsewhere (24). The washed mitochondrial pellets were exposed to a swelling procedure by incubation in 10 mM sodium phosphate buffer, pH 7.4, for 20 min (final protein concentration, ~1 mg/ml), followed by the addition of 61.5% sucrose in order to obtain a 0.45 M sucrose medium. After a 20-min incubation period, 10-ml portions were mildly sonicated (3 × 30 s), using a probe sonicator (Branson Sonifier 250). The suspension was centrifuged at 8,000 × g to remove unbroken mitochondria. The supernatant was collected and centrifuged again at 150,000 × g for 90 min. The pellet containing the submitochondrial membrane fraction was resuspended in 10 mM sodium phosphate buffer, pH 7.4, containing 0.45 M sucrose (final protein concentration, 5 mg/ml) using a Teflon homogenizer. The membrane suspension (1-2 mg of protein) was loaded onto a linear (15-50%) sucrose density gradient (10 ml) and centrifuged for 20 h at 100,000 × g. Subsequently, the gradients were divided into 500-µl fractions that were assayed for marker enzyme activities. Protein was quantified using the Bio-Rad protein micro assay and bovine serum albumin as a standard.

Cytochrome c oxidase (EC 1.9.3.1) and monoamine oxidase (EC 1.4.3.4) were assayed according to Appelmans et al. (25) and Otsuka and Kobayashi (26), respectively. Nucleoside-diphosphate kinase (EC 2.7.4.6) was assayed as described elsewhere (24).

The cholesterol content of each submitochondrial fraction of the gradient was determined by a coupled cholesterol oxidase-peroxidase assay with cholesterol as a standard (27). Aliquots of the fractions (200 µl) were transferred to glass tubes. To each sample, 20 µl of 20 mM cholate and 1% Triton X-100 in 100 mM potassium phosphate buffer, pH 7.4, were added, followed by the addition of 25 µl of 95% ethanol. The reaction mixture containing potassium phosphate buffer (100 mM), pH 7.4, cholesterol oxidase, peroxidase, and p-hydroxyphenylacetic acid was then added to each fraction in a final volume of 1 ml. Assay tubes were incubated for 1 h at 37 °C. Cholesterol oxidase generates H₂O₂, and peroxidase catalyzes the reaction of H₂O₂ with p-hydroxyphenylacetic acid to yield a stable fluorescent product. The fluorescence was measured in a Jasco CAF-110 fluorimeter (excitation, 325 nm; emission, 405 nm).

The aldosterone content of the incubation medium was measured by radioimmunoassay using a commercially available kit (Diagnostic Systems Laboratories, Webster, TX).

Results are expressed as the means ± S.E. The mean values were compared by analysis of variance using Fisher's test. A value of p < 0.05 was considered as statistically significant.

RESULTS

Membranes of osmotically lysed mitochondria isolated from glomerulosa cells in which [Ca²⁺]_c had been previously clamped for 2 h at either low levels (<100 nM, control cells) or at high levels (600 nM) were separated into various fractions by continuous sucrose density gradient centrifugation as described under ``Experimental Procedures.'' The protein distribution in the gradient under each condition is shown in Fig. 1A. Fig. 1B illustrates the separation of mitochondrial membranes from control (low [Ca²⁺]_c) glomerulosa cells. The activity profile of specific mitochondrial marker enzymes in the gradient led to the characterization of three distinct membrane populations. A first population with the lowest density (fractions 5-8) showed the highest monoamine oxidase activity, which is specific of the outer mitochondrial membranes. The population of membranes with the highest density (fractions 13-15) exhibits the highest content of cytochrome c oxidase activity, as expected for the inner mitochondrial membranes. In addition, a third membrane population of intermediate density (fractions 9-12) possessing both monoamine oxidase and cytochrome c oxidase activities displayed the highest nucleoside-diphosphate kinase activity, which is characteristic of mitochondrial intermembrane contact sites. A similar profile of mitochondrial membrane marker enzymes was obtained after fractionation of mitochondria isolated from high Ca²⁺-clamped cells; fractions 5, 11, and 14 contained the bulk of monoamine oxidase, NDP kinase, and cytochrome c oxidase activities, respectively (data not shown).

We have previously shown in bovine adrenal zona glomerulosa cells that the calcium-ionophore, ionomycin, can be effectively used at low concentration to clamp the cytosolic free Ca²⁺ concentration, [Ca²⁺]_c, at various physiological levels (50-1000 nM) (14). This submicromolar [Ca²⁺]_c stimulates the early mitochondrial steps of steroidogenesis, namely pregnenolone formation from cholesterol side chain cleavage, as well as aldosterone synthesis (14).

Cholesterol content was determined in the various fractions of the sucrose gradient. In Fig. 2, the total cholesterol content in each submitochondrial fraction from high Ca²⁺-clamped cells has been expressed as a percentage of the cholesterol content measured in the corresponding submitochondrial fractions from control cells. The stimulation of intact glomerulosa cells with Ca²⁺ led to a marked decrease of cholesterol content in the outer mitochondrial membranes (fraction 5, corresponding to the peak of monoamine oxidase activity: 54 ± 24% of controls, n = 5) with a concomitant increase in contact sites (fraction 11, corresponding to the peak of NDP kinase activity: 145 ± 14% of controls) and a less pronounced augmentation in the inner mitochondrial membranes (fraction 14, corresponding to the peak of cytochrome c oxidase activity: 125 ± 5% of controls).

These results demonstrate that Ca²⁺ activates mitochondrial endogenous cholesterol transfer from the outer membrane to the contact sites and inner membrane, a process that is accompanied by an increased aldosterone production. Indeed, the aldosterone output measured in Ca²⁺-clamped cells incubated in the absence of AMG amounted to 345 ± 42% of controls (n = 8).

Earlier reports have indicated that ACTH-activated cholesterol transport to the mitochondrial inner membrane of steroidogenic cells is blocked by inhibitors of protein synthesis such as CHX, with a resulting ACTH-induced accumulation of cholesterol in the outer mitochondrial membrane (15, 28). Moreover, it has been reported that the increase in pregnenolone and aldosterone synthesis triggered by Ang II is inhibited by the same compound (29), suggesting that at the mitochondrial level, both hormones operate through similar mechanisms. We have therefore tested whether CHX may also inhibit Ca²⁺-stimulated intramitochondrial cholesterol transfer in glomerulosa cells. Fig. 3 shows that when CHX was added concomitantly with Ca²⁺, the outer membrane fractions 5 and 6 contained 82 ± 4.4 and 82 ± 4.9% of the cholesterol content of their respective control fractions, as compared with only 68.3 ± 4.8% and 67 ± 3.6%, respectively, when the Ca²⁺ clamp was performed in the absence of CHX (p < 0.05, n = 4). By contrast, CHX significantly reduced Ca²⁺-activated cholesterol transfer to contact sites (fractions 11-12) and inner membranes (fractions 13-14) (Fig. 3). Cycloheximide similarly prevented Ang II-induced cholesterol transfer to contact sites and inner membranes (data not shown).

In order to test whether Ang II-mediated cholesterol mobilization in mitochondria is similar to that triggered by the cytosolic Ca²⁺ clamp, glomerulosa cells were incubated for 2 h in the presence of 10 nM Ang II and 500 µM AMG. Submitochondrial membranes were prepared and analyzed as above. No change in protein profile of sucrose gradients could be observed upon Ang II stimulation (data not shown). Fig. 4 illustrates the distribution of cholesterol content in the submitochondrial fractions of glomerulosa cells exposed to Ang II. Firstly, the hormone induced a pronounced increase of cholesterol content in the contact site-enriched fractions (fractions 11 and 12, 172 ± 16 and 169 ± 6% of controls, respectively, n = 5). In separate experiments, we have observed a significant (28 ± 5%, n = 5) increase in total mitochondrial cholesterol of Ang II-stimulated-cells. Secondly, no significant changes were detected in the outer mitochondrial membranes (fractions 5 and 6, 83 ± 6 and 80 ± 11% of controls, respectively, n = 5). Thirdly, no increase in cholesterol content was observed in the inner membrane fractions (fractions 13 and 14, 109 ± 13 and 76 ± 22% of controls, respectively, n = 5). Interestingly, a subpopulation of the inner membranes revealed a significant decrease in cholesterol content, as compared with the respective control fractions (fractions 15, 16, and 17, 64 ± 4, 52 ± 6, and 57 ± 13% of controls, respectively, n = 5).

DISCUSSION

In the present study, we took advantage of the cytosolic Ca²⁺ clamp technique to investigate the Ca²⁺ sensitivity of intramitochondrial cholesterol transfer in bovine glomerulosa cells. Recently, using the ionomycin-mediated Ca²⁺ clamp, our laboratory has provided a first direct demonstration that [Ca²⁺]_c in the submicromolar range stimulates aldosterone synthesis in intact glomerulosa cells (14). Because Ca²⁺ affects the formation of pregnenolone, an early mitochondrial step of aldosterone production (14), one could expect that one or several target(s) of the Ca²⁺ messenger are located within the mitochondria.

The overall rate-limiting step of the steroidogenic cascade is the transport of cholesterol from a presteroidogenic pool in the outer membrane to a steroidogenic pool in the inner membrane (15, 28, 30). The regulation by ACTH of cholesterol distribution in adrenal mitochondria has been studied by several groups using mitochondrial disruption to yield outer and inner membranes (28, 31, 32). The hormone has been shown to activate cholesterol supply to the P450_scc enzyme in the inner membrane. However, restricting the separation to two major fractions may fail to uncover additional, functionally relevant membranous structures. We therefore fractionated bovine adrenal glomerulosa mitochondria into outer membranes, inner membranes, and intermembrane contact sites. The latter fraction contains marker enzymes for both membranes, in addition to NDP kinase activity (Fig. 1), which has been shown to be specific of mitochondrial contact sites in several tissues (33).

The data presented here show that endogenous cholesterol transfer from the outer to the inner mitochondrial membrane and contact sites is substantially stimulated by physiological levels of cytosolic Ca²⁺. The supply of cholesterol to the P450_scc appears thus to be a Ca²⁺-sensitive step in the early steroidogenic pathway. This observation is in agreement with the data recently reported by Kowluru and colleagues (34), showing that Ca²⁺ stimulates the metabolism of endogenous cholesterol to pregnenolone in rat adrenal mitochondria. Our previous studies have demonstrated that the P450_scc enzyme is located in the contact sites, in addition to being in the inner membrane of bovine adrenocortical mitochondria (24). Interestingly, the present data indicate that the Ca²⁺-induced increase in cholesterol content is greater in the contact sites than in the inner membrane (Fig. 2). One interpretation could be that the P450_scc located in the contact sites is more active than the P450_scc of the inner membrane. Indeed, contact sites are known to be enriched in cardiolipin, a phospholipid that enhances the affinity of the P450_scc for cholesterol (35), although other explanations may be envisaged. Moreover, our results are in agreement with the data reported by Stevens et al. (36), suggesting that in addition to the outer and inner membrane pools, a third pool of steroidogenic cholesterol may be found in contact sites. On the other hand, Ca²⁺ induces direct contacts between the outer and the inner mitochondrial membranes, presumably by promoting a nonbilayer configuration leading to membrane fusion (37, 38, 39). Such an increase in the number of contact sites could result in an increase in the rate of cholesterol transfer.

Recently, Clark and colleagues (40) have observed that the elevations of cytosolic calcium triggered by Ang II, K⁺, and the Ca²⁺ channel agonist, BayK8644, are accompanied by increases in the level of the 30-kDa StAR protein in the human H295R adrenocarcinoma cell line. StAR protein import into the mitochondria via contact sites is thought to be a crucial event promoting cholesterol transfer to the inner membrane (20). We have also observed an increase in StAR protein content of mitochondria isolated from Ca²⁺-clamped bovine glomerulosa cells.² Our results therefore lead us to conclude that the StAR protein participates in the Ca²⁺-induced cholesterol transfer in bovine glomerulosa mitochondria.

The above hypothesis is strengthened by the sensitivity of Ca²⁺-induced cholesterol transfer to cycloheximide, an inhibitor of protein translation that depletes a set of 30-kDa proteins involved in the activation of steroidogenesis (16). From our study, it appears that CHX partially inhibits Ca²⁺-stimulated cholesterol transfer from the outer membrane to the contact sites and inner membranes, suggesting that at least one part of the calcium-activated cholesterol transfer process requires protein synthesis. This finding is in agreement with the results reported by Kowluru et al. (34), suggesting that two intramitochondrial cholesterol transfer processes are mediated by Ca²⁺. One of these mechanisms involves mitochondrial membrane sites accessible to activation by Ca²⁺ and GTP (presumably the sites of StAR-mediated cholesterol transfer). In the presence of CHX, this process is blocked, leading to redistribution of cholesterol to other sites involving mitochondrial membrane contact sites resulting from Ca²⁺-induced matrix swelling.

Several conclusions can be drawn from the experiments with Ang II stimulation. The lack of a decrease of cholesterol content in the outer membranes can be explained by a stimulation of cholesterol supply to the mitochondria in response to Ang II. Furthermore, a striking feature of the present study was the marked increase of cholesterol content in the contact sites, as compared with inner membranes, when glomerulosa cells were challenged with Ang II. This finding suggests that the hormone markedly enhances cholesterol availability in regions where the intermembrane space barrier is abolished. Interestingly, a subpopulation of the inner membranes even showed a decrease of cholesterol content, suggesting either a possible diffusion of cholesterol to specific sites in the inner membrane or a more unlikely selective loss of cholesterol during the fractionation procedure of mitochondria. Although we have no clear explanation at the present time for this result, it is worth mentioning that contact sites may be preferential sites for cholesterol transfer and metabolism, because they are enriched in the phospholipid cardiolipin, which promotes nonbilayer structures and enhances the affinity of P450_scc for cholesterol (35). One could therefore speculate that in addition to activating cholesterol transfer from the outer to the inner membrane, Ang II may also affect cholesterol movement within the inner membrane.

In conclusion, using the ionomycin-mediated cytosolic Ca²⁺ clamp, we have shown that Ca²⁺ itself is able to activate cholesterol transfer from the outer to the inner mitochondrial membrane and to intermembrane contact sites in bovine glomerulosa cells, a process that occurs even in the absence of cholesterol metabolism to pregnenolone. In fact, cholesterol flux could be even more important if steroidogenesis were allowed to proceed. We also demonstrate that Ang II, a Ca²⁺-mobilizing hormone, markedly increases cholesterol content in contact sites. Our results strongly suggest that one of the main functions of the Ca²⁺ messenger under hormonal stimulation is to increase cholesterol supply to the P450_scc enzyme by enhancing intermembrane cholesterol transfer, thus promoting the activation of the subsequent steroidogenic cascade.