EKODE, an epoxy-keto derivative of linoleic acid, was previously shown to stimulate aldosterone secretion in rat adrenal glomerulosa cells (15). In the present study, we investigated the effect of exogenous EKODE on cytosolic [Ca2+] increase and aimed to elucidate the mechanism involved in this process. Through the use of the fluorescent Ca2+-sensitive dye Fluo-4, EKODE was shown to rapidly increase intracellular [Ca2+] ([Ca2+]i) along a bell-shaped dose-response relationship with a maximum peak at 5 μM. Experiments performed in the presence or absence of Ca2+ revealed that this increase in [Ca2+]i originated exclusively from intracellular pools. EKODE-induced [Ca2+]i increase was blunted by prior application of angiotensin II, Xestospongin C, and cyclopiazonic acid, indicating that inositol trisphosphate (InsP3)-sensitive Ca2+ stores can be mobilized by EKODE despite the absence of InsP3 production. Accordingly, EKODE response was not sensitive to the phospholipase C inhibitor U-73122. EKODE mobilized a Ca2+ store included in the thapsigargin (TG)-sensitive stores, although the interaction between EKODE and TG appears complex, since EKODE added at the plateau response of TG induced a rapid drop in [Ca2+]i. 9-Oxo-octadecadienoic acid, another oxidized derivative of linoleic acid, also increases [Ca2+]i, with a dose-response curve similar to EKODE. However, arachidonic and linoleic acids at 10 μM failed to increase [Ca2+]i but did reduce the amplitude of the response to EKODE. It is concluded that EKODE mobilizes Ca2+ from an InsP3-sensitive store and that this [Ca2+]i increase is responsible for aldosterone secretion by glomerulosa cells. Similar bell-shaped dose-response curves for aldosterone and [Ca2+]i increases reinforce this hypothesis.
- intracellular calcium ion concentration
linoleic acid is the most prevalent polyunsaturated fatty acid in mammals, found mostly in adipocyte triglycerides. It is a potential source of metabolic energy through β-oxidation but is also vulnerable to addition of molecular oxygen at its double bonds. Reactions that add oxygen to linoleic acid may be catalyzed by lipoxygenases, cytochromes, or free metal ions. The initial products of this addition are hydroperoxides, which rapidly undergo rearrangement and modification to epoxides, hydroxides, and ketones, some of which are relatively stable in aqueous solution. These oxidation reactions probably play a role in the pathological state called “oxidative stress,” and, accordingly, some stable oxidation products of linoleic acid have toxic effects on mammalian cells and organelles (28).
In earlier reports, we tested some of the compounds formed from linoleic acid on rat adrenal cells and demonstrated that at least three of these reaction products stimulate secretion of aldosterone by glomerulosa cells and corticosterone by fasciculata cells (7, 14, 15). The experiments reported in the present study were aimed at investigating the signaling mechanism by which one of the products of linoleic acid oxidation, namely 12,13-epoxy-9-keto-10-(trans)-octadecenoic acid, abbreviated “EKODE,” affects steroidogenesis. Because increased intracellular calcium is a critical regulator of aldosterone secretion (39), focus was placed on this particular aspect of the signal transduction cascade.
MATERIALS AND METHODS
DNase was purchased from Sigma (St. Louis, MO); Fluo-4 AM was from Molecular Probes (Eugene, OR); collagenase, MEM-Eagle's medium, and OPTI-MEM were from Invitrogen (Burlington, ON, Canada); and Xestospongin C, cyclopiazonic acid (CPA), and thapsigargin (TG) were from Calbiochem-Novabiochem (San Diego, CA). EKODE was synthesized from linoleic acid in a two-step process catalyzed by lipoxygenase, cysteine, and iron, and the product was purified by silicic acid and reverse-phase chromatography (15).
Preparation of cells.
Rat adrenal glomerulosa cells were prepared as described earlier (13). All protocols were approved by the Animal Care and Ethics Committee of our faculty. In brief, female Long-Evans rats weighing 200–250 g were euthanized, adrenal glands were excised, and capsules were separated by an incision and gentle pressure to extrude the inner zones. Successive steps were performed in MEM-Eagle's medium. Capsules were incubated for 20 min at 37°C in collagenase (2 mg/ml) and DNase (25 μg/ml), and cells were dispersed by repeated aspiration with a pipette. After a filtering to remove debris, cells were pelleted at 100 g for 10 min and resuspended in OPTI-MEM medium containing 2% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were plated on poly-l-lysine-coated coverslips and maintained at 37°C in a humidified atmosphere of 95% O2-5% CO2. Studies were performed on 1- to 3-day-old cultures.
Intracellular calcium measurements.
To load cells with fluorescent calcium-sensitive dye, cells were transferred to a medium (“HBS”) containing the following: NaCl, 140 mM; KCl, 5.4 mM; CaCl2, 2 mM; MgCl2, 1 mM; HEPES, 10 mM, at pH 7.35; and Fluo-4 AM, 5 μM. After 30 min at 37°C, cells were washed with HBS containing 1% bovine serum albumin in the absence of dye and incubated in HBS for 30 min. Coverslips mounted with loaded cells were then placed in a chamber on an inverted Nikon Diaphot microscope at 25°C. The light source was generated by a 100-W mercury lamp. Band-pass filters (450–490 and 520–560 nm) were used for excitation and emission wavelengths, respectively, with emitted light recorded by a photon-counting unit. The amplitude of the signal is measured at the plateau level for sustained response or at the peak for transient response. Data are expressed as F/F0, where F0 is the base fluorescence signal level and F is the fluorescence level of the response.
Before each experiment, the medium of cultured cells was removed, and the cells were washed twice with cold HBS supplemented with 1 g/l glucose and 0.5% BSA. The cells were incubated in 1 ml of medium consisting of 0.9 ml of HBS-glucose (0.1 mg/ml) and 0.1 ml of stimulus. After a 2-h incubation at 37°C, the incubation medium was removed by aspiration and stored at −20°C until RIA determination of aldosterone in the medium, using specific anti-sera and tritiated steroid as tracer (13).
Inositol trisphosphate production.
The method used was described previously by Gallo-Payet et al. (12). Cells were incubated in growth medium with 2 μCi/ml [3H]myoinositol for 3 days, after which the radioactive medium was discarded, and cells were washed and incubated in isotope-free medium. After 1 h, the cells were washed again and incubated for 5 min at 37°C in PBS supplemented with ACTH or various concentrations of EKODE. The inositol phosphates were separated as described previously (4) by chromatography on Dowex 1 × 8 columns.
Intracellular cAMP production was determined by measuring conversion of [3H]ATP to [3H]cAMP, as described previously (13). Briefly, cultured cells were incubated at 37°C in OPTI-MEM culture medium containing 2 μCi/ml [3H]adenine. After 1 h, the cultures were washed and incubated with ACTH or various concentrations of EKODE in HBSS buffer containing 1 mM isobutylmethylxanthine. After 15 min, cells were collected, solubilized, and chromatographed on Dowex and alumina columns according to the method of Salomon et al. (34).
Data are presented as means ± SE. Comparisons were tested for significance by the Student's t-test, with P values >0.05 considered as statistically significant.
Dose-response and temporal relationships.
Application of EKODE to cultured rat adrenal glomerulosa cells preloaded with the Ca2+-sensitive dye Fluo-4 rapidly increased the fluorescent signal (Fig. 1A). The bulk of the increase was achieved within 25 s and was maintained over the course of the experiment, a span of 5–10 min. The amplitude of the intracellular calcium concentration ([Ca2+]i) response, in a Ca2+-containing medium, increased with stepwise concentrations of EKODE from 0.1 to 5 μM and then decreased at EKODE concentrations >5 μM (Fig. 1B, solid squares). This biphasic response mirrors the dose-response relationship between EKODE and aldosterone secretion (15).
In a calcium-free medium, the amplitude of the [Ca2+]i response to EKODE demonstrated the same bell-shaped curve (Fig. 1B, open circles) as it did in a Ca2+-containing medium. However, the [Ca2+]i increase was lower, but the difference was not statistically significant, with P values of 0.299, 0.344, and 0.07 for EKODE concentrations of 3, 5, and 10 μM, respectively. To further assess whether Ca2+ influx is activated by EKODE, experiments were initially performed in a Ca2+-free medium that was then substituted with 2 mM Ca2+ medium once cells had responded to EKODE. Figure 1C shows that no additional [Ca2+]i increase was triggered by the introduction of Ca2+-containing medium (n = 4). Moreover, the response to EKODE in a Ca2+ medium was not abrogated by EGTA (data not shown).
Calcium signaling reagents.
Angiotensin II is a potent stimulus of aldosterone secretion, increasing intracellular calcium by mobilizing calcium from intracellular stores and opening calcium channels that communicate with the extracellular milieu (39). This angiotensin action on intracellular pools is mediated by inositol trisphosphate (InsP3). As seen in Fig. 2A, a maximal dose of angiotensin II applied repeatedly to rat adrenal cells in a calcium-free medium induced a transient increase in [Ca2+]i that decreased with each of the two successive doses (Fig. 2A). When EKODE was applied to cells after the second dose of angiotensin II, there was a sustained response in [Ca2+]i, although it was smaller than that observed when EKODE was applied before angiotensin II (Fig. 2B; 13.1 ± 2%, n = 4, P = 0.022).
The above results with angiotensin II pretreatment thus suggested that EKODE releases calcium from an InsP3-sensitive intracellular pool that is also mobilized by angiotensin II. To test this hypothesis, EKODE was applied to cells pretreated with Xestospongin C, a potent blocker of InsP3 receptors (24). Pretreatment of cells with Xestospongin C at a concentration of 10 μM for 30 min markedly blunted the subsequent response to EKODE. Although the time course of the increased [Ca2+]i response was typical, the magnitude was only 7.1 ± 1.8% (n = 6), significantly less than control levels (P = 0.004; Fig. 3).
Two additional signal transduction pathways were also tested to investigate the putative mechanism of action of EKODE. Pretreatment of rat adrenal cells with U-73122, a specific inhibitor of phospholipase Cβ (33), at a concentration of 1 μM for 18 h reduced the [Ca2+]i response to EKODE to 27.8 ± 8.3% (n = 6); however, the difference with controls was not statistically significant (P = 0.35; Fig. 3). Caffeine is known to activate ryanodine-sensitive calcium channels and to induce an increase in [Ca2+]i in numerous cell types (19). In glomerulosa cells, caffeine was previously shown not to induce Ca2+ release but did interfere with the action of angiotensin II (40). In the present experiments, caffeine (20 mM) had no effect on its own and no significant effect on the response to EKODE (mean increase 41.2 ± 4.4%, n = 3, P = 0.613; Fig. 3). Furthermore, a first application of 30 μM CPA, a known inhibitor of the Ca2+-ATPase sarcoplasmic reticulum (25, 36), in a Ca2+-free medium induced a substantial increase in [Ca2+]i, as shown in Fig. 4A. To completely empty the Ca2+ store, a second treatment with 30 μM CPA was applied. Under these conditions, further application of EKODE (5 μM) was unable to induce an increase in [Ca2+]i (Fig. 4A; n = 5). TG is also an inhibitor of the sarcoplasmic/endoplasmic reticulum Ca2+-ATPase pumps (23). In adrenal cells, TG at a concentration of 4 μM in a calcium-containing medium induced a large increase in [Ca2+]i (83.8 ± 13.6%, n = 5). When EKODE (5 μM) was applied before the addition of TG, the TG-induced [Ca2+]i increase was lowered to 59.9 ± 9.1% (n = 6, P = 0.22; Fig. 4, B and C). Finally, when 5 μM EKODE was applied during the plateau phase of the TG response, a rapid drop in [Ca2+]i occurred (−27.4 ± 5.3%, n = 10, P = 0.001; Fig. 4B).
Aldosterone production, InsP3, and cAMP levels.
Aldosterone secretion was measured as previously described (13) on glomerulosa cells in culture in two different experiments in triplicate. Various EKODE concentrations were tested in a Ca2+-containing medium, and Fig. 5A shows that aldosterone production increased until 5 μM EKODE and then decreased for higher concentrations. Angiotensin II (10−9 M) was used as a positive control; note that the maximum effect of EKODE (5 μM) was similar in potency to that of angiotensin II (10−9 M).
InsP production was measured in EKODE-stimulated cells, using angiotensin II as a positive control. The data from three separate experiments are summarized in Fig. 5B, which shows that EKODE at concentrations ranging from 1 to 10 μM had no effect on InsP production. However, under the same conditions, angiotensin II at 10−9 M significantly increased InsP.
cAMP is a well-known second messenger involved in aldosterone production under ACTH stimulation. Figure 5C demonstrates that EKODE concentrations ranging from 1 to 10 μM were ineffective on cAMP production compared with ACTH at 10−8 M, taken as positive control.
Other fatty acids.
EKODE is an oxidized metabolite of linoleic acid. Linoleic acid (10 μM) itself did not affect [Ca2+]i (Fig. 6A) nor did arachidonic acid (10 μM; data not shown) in a free-Ca2+ medium. In fact, pretreatment of rat glomerulosa cells with linoleic or arachidonic acid blunted the response to subsequent doses of EKODE (Fig. 6B). The increase in [Ca2+]i in response to EKODE was only 11.8 ± 2.1% (n = 7, P = 0.038) in the presence of linoleic acid and 14.4 ± 5% (n = 4, P = 0.045) in the presence of arachidonic acid.
Another oxidized derivative of linoleic acid, 9-oxo-octadecadienoic acid (9-oxo-ODE), increased intracellular calcium. Figure 6C, top, shows that application of 5 μM 9-oxo-ODE in a Ca2+ medium triggered a sustained [Ca2+]i increase; in some cells, the 9-oxo-ODE-induced [Ca2+]i increase was transient (Fig. 6C, bottom). On 20 cells tested, 55% displayed a sustained [Ca2+]i increase, whereas it was transient in the 45% remaining cells. The relation between the 9-oxo-ODE concentration and the [Ca2+]i increase is illustrated in Fig. 6D. This relation is bell shaped like the relationship between EKODE and [Ca2+]i increase. However, the maximal effect was found for 10 μM 9-oxo-ODE compared with 5 μM for EKODE.
Our experiments show that EKODE, a naturally occurring oxidized derivative of linoleic acid, increases intracellular levels of calcium in rat adrenal glomerulosa cells, and the calcium increase comes primarily from intracellular pools. This is probably the mechanism by which EKODE stimulates aldosterone production by these cells. That conclusion is supported by the shape of the dose-response curves. Both [Ca2+]i and aldosterone production are increased maximally by 5 μM EKODE and decreased by concentrations greater than that (Fig. 5A and Refs. 14, 15).
Aldosterone secretion is under the control of several factors, including angiotensin II, ACTH, extracellular K+, and several paracrine factors (38, 39). Angiotensin II, by binding to the AT1 receptor subtype, induces the production of InsP3, which triggers Ca2+ release from InsP3-sensitive intracellular stores (1). After the intracellular release, increased Ca2+ influx explains the sustained [Ca2+]i observed with angiotensin II stimulation. A dihydropyridine-insensitive Ca2+ influx was first reported by Spat et al. (38), similar to the capacitative influx described by Putney et al. (29). Later, it was shown that angiotensin II induces Ca2+ influx through voltage-dependent T-type Ca2+ channels, either by depolarizing the membrane (20) or by changing the voltage dependence of the channels (8). An increase in [Ca2+]i is also observed with ACTH stimulation. That response is caused by activation of voltage-dependent L-type Ca2+ channels (11, 42). Cyclic AMP, the main second messenger of ACTH action, activates protein kinase A and increases the transport of cholesterol in adrenal mitochodria (41). A positive-feedback mechanism also exists by which increased Ca2+ increases cAMP production through Ca2+-sensitive adenylyl cyclases (9). In glomerulosa cells, aldosterone production is also stimulated by increased extracellular K+ concentration (6, 43). Elevated extracellular [K+] depolarizes the membrane and results in a Ca2+ influx (2, 31). T-type and L-type voltage-dependent Ca2+ channels are activated by low and high K+ concentrations, respectively (3).
The mechanism of EKODE is apparently different from those described above. Unlike the response to other agonists, cytosolic Ca2+ increases at EKODE concentrations up to 5 μM and decreases at higher EKODE concentrations. We investigated the nature of the intracellular Ca2+ pools mobilized by EKODE and the possibility of an EKODE-activated calcium influx. EKODE did not induce production of InsP3, which could be taken as evidence to exclude InsP3-sensitive Ca2+ stores as a source of [Ca2+]i increase. Furthermore, there is no inhibition by U-73132, an inhibitor of phospholipase Cβ (33). However, application of Xestospongin C, which blocks InsP3 receptors (24), significantly blunted the Ca2+ response to EKODE. Taken together, these results support the idea that EKODE releases Ca2+ from InsP3-sensitive stores, but without directly stimulating production of InsP3.
Native fatty acids have been shown to increase [Ca2+]i in various cell types. Arachidonic, linoleic, and palmitic acids liberate Ca2+ from InsP3-insensitive stores in Dictyostelium discoideum (35), and arachidonic acid triggers Ca2+ release from both InsP3-sensitive and mitochondrial Ca2+ pools of rat lactotrophs, rat cerebellum, and intestinal smooth muscle (16, 27, 32). Several reports indicate that arachidonic acid induces a transient [Ca2+]i increase (32, 35); a sustained [Ca2+]i increase was found in permeabilized cells (35). There is no consensus concerning arachidonic acid and Ca2+ influx (37, 22). Our results indicate that neither arachidonic acid nor linoleic acid induced a [Ca2+]i increase in glomerulosa cells. However, another oxidized derivative of linoleic acid, 9-oxo-ODE, is able to trigger a Ca2+ response. Compared with EKODE, the dose-response relationship is also bell shaped, but the maximum is found at 10 μM. Moreover, EKODE is 39.9% more potent than 9-oxo-ODE.
Our results with CPA indicate that EKODE liberates Ca2+ from an ATPase-dependent store, probably the InsP3-sensitive one. Moreover, emptying the InsP3-sensitive Ca2+ store with a repeated dose of angiotensin II severely blunted the Ca2+ response to EKODE. Incidentally, this result suggests that EKODE would add little steroidogenesis to a maximal dose of angiotensin II, and that is exactly what is found when aldosterone production is measured (14).
It could be postulated that EKODE first liberates a small amount of Ca2+ from an intracellular pool by a nonspecific mechanism not involving InsP3 receptors. This small increase in [Ca2+]i in the presence of a low basal InsP3 concentration (see Fig. 5B) could activate the type 1 InsP3 receptor, the most abundant receptor isoform in rat glomerulosa cells (10), by a calcium-induced calcium release mechanism (18); this would explain the Xestospongin C sensitivity of the Ca2+ response to EKODE. A second possibility could be that EKODE induces a sensitization of the InsP3 receptors that, considering the basal level of InsP3 in the cell, would be sufficient to trigger a Ca2+ release from the InsP3-sensitive stores. Thimerosal, a thiol-reactive agent, worked in this way. Indeed, thimerosal has both sensitizing (<2 μM) and inhibitory (>2 μM) effects on InsP3-induced Ca2+ release from cerebellar microsomes (26).
Although the intracellular targets of angiotensin II and EKODE action seem to be similar, it should be noted that the kinetics of the Ca2+ increase are quiet different. In Ca2+-free media, angiotensin II triggers a transient response, whereas EKODE gives rise to a sustained response. Another unusual feature of EKODE action is the biphasic dose-response curve. Cytosolic [Ca2+] increases at EKODE concentrations up to 5 μM and then decreases at higher concentrations. Aldosterone production in cultured glomerulosa cells from rat displays exactly the same pattern (Fig. 5A and Refs. 14, 15). This is reminiscent of the actions of 2,5-di-(tert-butyl)-1,4-hydroquinone (tBHQ). This compound is unrelated chemically to TG, but like TG, it blocks Ca2+-ATPase. In a Ca2+-containing medium, tBHQ displayed a biphasic [Ca2+]i-concentration relationship. This was interpreted as a release of Ca2+ from Ca2+ stores together with a Ca2+ influx for low concentrations, but a release with a blockage of a Ca2+ influx at high tBHQ concentrations; similar results were reported for CPA (25). When tBHQ was added at the plateau response of TG, it induced a rapid drop in [Ca2+]i. Our results show that 5 μM EKODE added at the plateau response of TG also induced a [Ca2+]i drop, which can be interpreted as the blockage by EKODE of a Ca2+ influx. However, similar results were observed in a Ca2+-free medium (data not shown), so the interaction between EKODE and TG remains unexplained.
Our results indicate that EKODE did not induce the production of InsP3 or cAMP as did angiotensin II and ACTH, respectively. Intracellular calcium alone seems to be the second messenger of EKODE, so the question arises whether a [Ca2+]i increase alone is sufficient to trigger aldosterone production in glomerulosa cells. It has been demonstrated by others that nonspecific Ca2+ entry induced by Ca2+ ionophores (ionomycin, A23187) is sufficient to induce aldosterone production (21, 30). Our results indicate that EKODE does not induce the production of classical second messengers of the steroidogenic pathway and does not activate PKA or PKC, but does activate the steroidogenic pathway only by increasing [Ca2+]i.
In conclusion, we have demonstrated in female rat glomerulosa cells that EKODE increases [Ca2+]i by mobilization of an InsP3-sensitive Ca2+ pool without a noticeable effect on InsP3 production or Ca2+ influx. The biphasic relationship between [Ca2+]i increase and the concentration of EKODE, and the minimal effect of EKODE added to angiotensin II, mirror the relationships of EKODE and aldosterone production observed by Goodfriend et al. (14, 15). In summary the data provide a strong argument in the favor of the postulate that EKODE activates aldosterone production only by increasing [Ca2+]i.
N. Gallo-Payet is a recipient of the Canada Research Chair in Endocrinology of the Adrenal Gland. This work was supported by grants from the Canadian Institute for Health Research to M. D. Payet (MOP-6813).
We thank Pierre Pothier for the critical reading of the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 by American Physiological Society