INTRODUCTION

Repetitive Ca²⁺ transients occur in a variety of cell types following stimulation with low concentrations of Ca²⁺-mobilizing agonists. This type of Ca²⁺ signaling consists of a series of Ca²⁺ spikes that occur at intervals of 0.5-5 min, depending on the agonist dose and the cell type. Such oscillatory signals, which often persist for 1 h or more, have been observed in cell types ranging from electrically nonexcitable cells such as hepatocytes through partially excitable endocrine and neuroendocrine cells (e.g. pancreatic  cells, pituitary cells, and adrenal glomerulosa cells) to classical excitable cells (muscle cells, neurons) (1). In many cases the spiking frequency, rather than the amplitude of the oscillations, is regulated by the agonist concentration, providing the basis of a frequency-encoded Ca²⁺ signaling mechanism. Although Ca²⁺ release from internal stores is a major source of the cytoplasmic Ca²⁺ transients, [Ca²⁺]_i levels are also influenced by Ca²⁺ entry mechanisms operating through plasma membrane Ca²⁺ channels. These include a ubiquitous form of Ca²⁺ influx in which the filling state of the Ca²⁺ stores regulates Ca²⁺ influx (``capacitative Ca²⁺ entry''; see Ref. 2) as well as various types of voltage-sensitive calcium channels (VSCCs)¹ that determine Ca²⁺ influx as a function of the membrane potential in excitable cells.

Recent models to explain the generation of Ca²⁺ oscillations have focused on the dual regulation of the InsP₃ receptor channel by its ambient cytosolic Ca²⁺ concentrations, with positive and negative feedback by low and high [Ca²⁺]_i, respectively. This mechanism has been proposed to initiate and terminate short bursts of Ca²⁺ release even at constant InsP₃ levels (3, 4, 5). The restoration of basal [Ca²⁺]_i between spikes depends on diverse Ca²⁺ homeostatic mechanisms, including Ca²⁺ pumps located in the endoplasmic reticulum (ER) and plasma membrane (6, 7). These processes, and possibly uptake by mitochondria (7, 8, 9, 10), reestablish the resting [Ca²⁺]_i levels and also refill the Ca²⁺ pools so that a new cycle can be initiated. Due to the high Ca²⁺ sensitivity of the InsP₃ receptor channel, disturbances in the Ca²⁺ homeostatic mechanism(s) of the cell frequently affect the rate and character of Ca²⁺ oscillations. For example, the Ca²⁺ wave frequency in Xenopus oocytes can be accelerated by enhancement of Ca²⁺ influx (11) as well as by overexpression of a SERCA type Ca²⁺ pump to promote the rapid resetting of basal [Ca²⁺]_i levels (12). These findings indicate the involvement of both Ca²⁺ entry and reuptake mechanisms in the regulation of periodic calcium signaling in these large cells. Recent microfluorimetric studies revealed that small microdomains, or so-called hot spots, with little impact on the average [Ca²⁺]_i of the cell may have a crucial role in the regulation of intracellular Ca²⁺ signaling events (9, 13). This implies that if the Ca²⁺ stores with their InsP₃ receptors and the plasma membrane Ca²⁺ entry channels are close to one another (14), even small amounts of Ca²⁺ entering through these channels could be an important factor in the regulation of Ca²⁺ signals by the local effects of Ca²⁺ influx.

In the adrenal glomerulosa, the presence of both the capacitative Ca²⁺ entry pathway (15, 16) and various types of VSCC (17, 18, 19, 20, 21) have been reported. Both of these Ca²⁺ influx mechanisms are active during Ang II stimulation of bovine glomerulosa cells (15, 22) providing two complementary means to regulate intracellular Ca²⁺ signaling through Ca²⁺ entry. Thus, dihydropyridines have been shown to exert minor but consistent effects on the [Ca²⁺]_i plateau following stimulation with high doses of Ang II in suspensions of adrenal zona glomerulosa cells (22, 23, 24). Such results have suggested that voltage-dependent Ca²⁺ channels make relatively little contribution to the Ca²⁺ signals generated by Ang II in adrenal cells, which contrasts with the sensitivity of the secretory response of these cells to dihydropyridines, especially at low Ang II concentrations (16). This apparent controversy could be resolved, if the role of VSCC activation is not to increase the overall cytosolic Ca²⁺ levels but rather to modulate Ca²⁺ oscillations that occur primarily at lower, more physiological doses (25, 26). To explore the relationship between Ca²⁺ entry through VSCC and Ca²⁺ oscillations, we examined the relative contributions of the two Ca²⁺ entry pathways (VSCC and capacitative Ca²⁺ influx) to the regulation of periodic Ca²⁺ release from internal Ca²⁺ stores in Ang II-stimulated bovine adrenal glomerulosa cells.

EXPERIMENTAL PROCEDURES

Fura-2/AM, Indo-2/AM, and pluronic acid were purchased from Molecular Probes, Inc. (Eugene, OR). Angiotensin II was purchased from Peninsula Laboratories, Inc. (San Carlos, CA). BAY K 8644, nifedipine, and thapsigargin were obtained from Calbiochem. All other chemicals were from Sigma. Cell culture media were prepared by the NIH Media Unit (Bethesda, MD) or were supplied by Biofluids (Rockville, MD).

Bovine adrenal zona glomerulosa cells were prepared and cultured as described previously (27). Cells were plated on poly-L-lysine-coated circular coverslips placed in the individual wells of 6-well culture dishes (Corning, NY) at a density of 10⁵ cells/ml (3 ml of cell suspension/well). Cells were cultured for 2-5 days in Dulbecco's modified Eagle's medium supplemented with donor horse serum (10%) and fetal bovine serum (2%) before Ca²⁺ measurements.

Cells attached to coverslips were loaded with 0.5-5 µM fura-2/AM or indo-1/AM for 2-5 h at room temperature in a modified Krebs-Ringer buffer solution (KRBS) containing 118 mM NaCl, 2.42 mM KCl, 1.8 mM Ca²⁺, 1.18 mM KH₂PO₄, 0.8 mM MgSO₄, 20 mM Na-Hepes (pH 7.4), 5 mM NaHCO₃, 10 mM glucose, and 1 mg/ml bovine serum albumin, supplemented with 0.25% pluronic acid. For [Ca²⁺]_i measurements, the loading buffer was replaced by fresh KRBS or Ca²⁺-free KRBS. The coverslips were then placed in a 35-mm diameter Teflon culture dish (Medical Systems Corporation, Greenvale, NY), and the [Ca²⁺]_i was determined by microfluorometry using either a single excitation, two emission wavelength procedure with indo-1 (28), or a dual excitation single emission wavelength protocol with fura-2 (29). All stimulations and pharmacological treatments of the cells were performed at room temperature by the addition of the substance diluted in 1 ml of buffer.

Inositol phosphates were measured in glomerulosa cells, prelabeled with myo-[³H]inositol (20 µCi/ml) for 24 h, after extraction and separation by high pressure liquid chromatography as described previously (24). Because of the small inositol phosphate response of the cells to 100 pM Ang II, cells were preincubated with LiCl (10 mM) for 30 min and were stimulated with Ang II for 30 min to capture the metabolites of inositol 1,4,5-trisphosphate. Experiments were also performed without lithium, which showed similar results but with smaller changes.

Values are given as mean ± S.E. Analyses of statistical significance were performed using paired or unpaired t tests. For multiple comparisons we used analysis of variance combined with the Duncan test.

RESULTS

In bovine adrenal zona glomerulosa cells, stimulation with 100 pM Ang II induced repetitive [Ca²⁺]_i transients with an average frequency of 0.38 ± 0.03 min¹ (n = 40 cells, Fig. 1A). The cytosolic Ca²⁺ concentration returned to basal levels between individual spikes, which were usually maintained for >1 h with only a slight decrease in amplitude. A 3-fold increase in Ang II concentration caused a significant increase in oscillatory frequency from 0.38 ± 0.05 to 0.63 ± 0.09 min¹ (p < 0.005, n = 5 cells, Fig. 1A) with no major change in spike amplitude. The subsequent addition of 1 nM Ang II elicited a large peak followed by a sustained plateau (Fig. 1B). Direct application of higher concentrations of Ang II (10 nM) usually induced similar biphasic Ca²⁺ responses with an initial spike followed by a plateau (n = 12, Fig. 1C), but in a subset of cells (35%) it elicited rapid Ca²⁺ oscillations. This oscillatory signal was preceded by an initial rise in [Ca²⁺]_i with rapid superimposed oscillations of irregular amplitude and an average frequency of 2.07 ± 0.40 min¹ (n = 4, Fig. 1D). These high frequency spikes often merged to resemble the early phase of a sustained plateau response (n = 3, Fig. 1E). After a few such peaks the signal switched to regular base-line spiking with a frequency of 1.36 ± 0.25 min¹ (n = 7, Fig. 1, D and E). The responsiveness of individual cells showed some variation, and plateau elevations in [Ca²⁺]_i were occasionally observed even with 100 pM Ang II. Regular oscillations of very low frequency could be elicited with Ang II concentrations as low as 25 pM, which is within the physiological range of the octapeptide present in the circulation (data not shown).

To study the role of extracellular Ca²⁺ in the regulation of Ca²⁺ oscillations, glomerulosa cells were stimulated with 100 pM Ang II in Ca²⁺-free medium (no added Ca²⁺). Under these conditions, sustained Ca²⁺ oscillations were again evident for >1 h in most of the cells. The frequency of these sustained oscillations (but not their amplitude) was significantly lower than that observed in the presence of extracellular Ca²⁺ (0.23 ± 0.01 min¹, n = 45, p < 0.001).

The addition of 0.6 mM Ca²⁺ to such Ang II-stimulated cells immediately increased the oscillatory frequency from 0.27 ± 0.02 to 0.52 ± 0.06 min¹ (n = 16 cells, p < 0.001, Fig. 2A). Subsequent increases of [Ca²⁺]_o to 1.2 mM or higher did not cause any further increase in frequency (data not shown). In a subset of cells, addition of extracellular Ca²⁺ caused an immediate spike followed by a plateau Ca²⁺ elevation (n = 16, Fig. 2B). Interestingly, such cells showed a pronounced decrease in the amplitude of their Ca²⁺ transients during stimulation in Ca²⁺-free medium (Fig. 2B). The average decrease in spike amplitude (19 ± 3% per spike) was significantly higher in this group of cells than in cells that continued to oscillate after the addition of Ca²⁺ (6 ± 2%, n = 16, p < 0.001, Fig. 3A). The loss of spiking amplitude observed in Ca²⁺-free medium was positively correlated with the magnitude of the [Ca²⁺]_i rise after the addition of Ca²⁺ (Fig. 3B, linear fit, p < 0.05). The amplitude of the biphasic [Ca²⁺]_i increase upon addition of Ca²⁺ also correlated with the average width of the Ca²⁺ spikes observed in Ca²⁺-free medium (Fig. 3C, linear fit, p < 0.01), as well as with the interspike interval length (Fig. 3D, linear fit, p < 0.05). The average spike duration was usually significantly higher (101 ± 13 s, n = 16) in cells that responded with a biphasic Ca²⁺ increase when the [Ca²⁺]_o was increased fom 0 to 0.6 mM than in those that continued to oscillate after the readdition of Ca²⁺ (34 ± 4 s, n = 16, p < 0.001, Fig. 3A). There was a significant correlation between the average spike duration and the average interspike interval, based on the data from all cells stimulated with 100 pM Ang II in Ca²⁺-free medium (Fig. 3E, linear fit, p < 0.01). That is, the shorter the Ca²⁺ spike duration, the more rapid were the oscillations.

The above correlation between the average duration of [Ca²⁺]_i spikes and the oscillatory frequency was relevant to recent observations in Xenopus laevis oocytes, in which overexpression of a SERCA Ca-ATPase caused an increase in the frequency of InsP₃-induced [Ca²⁺]_i waves and a decrease in the average wave duration (12). The effect of changes in SERCA Ca²⁺-ATPase activity on Ca²⁺ oscillations in glomerulosa cells during stimulation with 100 pM Ang II in Ca²⁺-free medium was evaluated by treatment with low concentrations of the microsomal Ca²⁺-ATPase inhibitor, thapsigargin. The impairment of Ca²⁺ sequestering ability caused by the addition of 20 nM thapsigargin to cells exhibiting narrow Ca²⁺ spikes caused an increase in spike duration and a more rapid loss of amplitude (Fig. 2C). This effect resembled the oscillatory pattern of cells that responded with biphasic Ca²⁺ increases after the addition of Ca²⁺ (Fig. 3A and Table I). Such thapsigargin-treated cells always displayed a biphasic rise in [Ca²⁺]_i upon addition of 0.6 mM [Ca²⁺]_o and never continued to oscillate (n = 13, Fig. 2C). However, it should be noted that the mean amplitude of the rise in [Ca²⁺]_i in response to extracellular Ca²⁺ was significantly smaller than in cells in which a biphasic [Ca²⁺]_i rise in response to extracellular Ca²⁺ occurred in the absence of thapsigargin (Table I). Thapsigargin also caused a transient increase of spiking frequency before slowing and eventually terminating the oscillatory Ca²⁺ signal (Fig. 2C and Table I).

Table I. Ca2+ oscillations elicited by low (100 pM) and high (10 nM) concentrations of Ang II in Ca2+-free medium Single glomerulosa cells were stimulated in Ca2+-free medium with the indicated concentrations of Ang II. Once an oscillatory [Ca2+]i signal had been established, 0.6 mM [Ca2+]o was added to the medium. At both Ang II concentrations, two types of responses were observed upon addition of extracellular Ca2+ ions. Cells stimulated with 100 pM Ang II displayed either an acceleration in their oscillatory frequency or a biphasic [Ca2+]i increase. Treatment of cells that belonged to the former group with thapsigargin (20 nM) caused a change to an oscillatory pattern that was characteristic of the latter group and also made these cells to respond with biphasic [Ca2+]i increase after Ca2+ addition. Ca2+ oscillations driven by 10 nM Ang II either were independent of extracellular Ca2+ or showed a biphasic rise in [Ca2+]i. The latter group was similar in its parameters to the cells that responded with a biphasic Ca2+ increase upon addition of Ca2+ after stimulation with 100 pM Ang II. The former group, however, resembled in its spike parameters the one that was stimulated by 100 pM Ang II and was accelerated by external Ca2+]. (*, p < 0.01; **, p < 0.001; ***, p < 0.001 compared with the amplitude observed in the absence of thapsigargin). 100 pM Ang II 10 nM Ang II Acceleration of spiking frequency (n = 16) Biphasic increase (n = 16) Thapsigargin (20 nM) Oscillations independent of Ca2+ (n = 5) Biphasic increase (n = 8) Before After Average frequency during Ca2+-free period (min1) 0.27  ± 0.02 0.21  ± 0.02 0.22  ± 0.02 0.34  ± 0.03** 1.00  ± 0.14 0.28  ± 0.05 Average spike duration before Ca2+ addition (s) 34  ± 4 101  ± 13 50  ± 5 86  ± 13* 35  ± 5 85  ± 11 Loss of amplitude per spike (% of the preceding spike) 6  ± 2 19  ± 3 10  ± 1 36  ± 5** 5  ± 1 23  ± 3 [Ca2+]i rise upon Ca2+ addition ( ratio) NA 0.52  ± 0.06 0.23  ± 0.02*** NAa 0.61  ± 0.10 Acceleration in spiking frequency (%) +85 NA NA  6 NA a  NA, not applicable.

To study the role of extracellular Ca²⁺ during stimulation with high Ang II concentrations, glomerulosa cells were treated with 10 nM Ang II in Ca²⁺-free medium (no added Ca²⁺, without EGTA). This hormone concentration usually caused a biphasic [Ca²⁺]_i increase when added in the presence of Ca²⁺ (Fig. 1C). However, under Ca²⁺-free conditions 10 nM Ang II evoked Ca²⁺ oscillations, although of a different character than those elicited by 100 pM Ang II (Fig. 4 and Table I). Such responses were typically large and rapid Ca²⁺ elevations in which the individual spikes were superimposed and showed a progressive decrease in amplitude. This initial phase was usually followed by base-line oscillations of higher frequency and of variable amplitude and spike duration, depending on the individual cell. The responses of these cells to the addition of Ca²⁺ could be classified as follows. In about 30% of the cells (5 of 15) in which the Ang II-evoked signal stabilized as regular oscillations of high amplitude, narrow spike duration, and high frequency, the response was unaffected by extracellular Ca²⁺ (Fig. 4A), being 1.00 ± 0.14 and 0.94 ± 0.14 min¹ in the absence and presence of 0.6 mM [Ca²⁺]_o, respectively. In a second group of cells (8 of 15) that showed wide, low frequency spikes of greatly diminished amplitude, the addition of Ca²⁺ caused a biphasic [Ca²⁺]_i increase (Fig. 4C). The remaining two cells showed an ``intermediate'' response in that the addition of 0.6 mM Ca²⁺ caused a transient run of high frequency oscillations with an elevated interspike [Ca²⁺]-level, followed by a return to high frequency base-line oscillations (Fig. 4B).

The contribution of VSCCs to the regulation of Ca²⁺ oscillations was examined in Ang II-stimulated cells treated with dihydropyridine agonist (BAY K 8644) and antagonist (nifedipine) derivatives. As shown in Fig. 5A, BAY K 8644 (10 nM) significantly increased the frequency of Ang II (100 pM)-induced Ca²⁺ oscillations (in 7 of 13 cells) by 86% from 0.43 ± 0.08 to 0.80 ± 0.18 min¹ (p < 0.02) or changed the oscillatory Ca²⁺ signal to a plateau type response (in 6 of 13 cells, Fig. 5B). Conversely, nifedipine (2 µM) significantly decreased the frequency of Ca²⁺ oscillations by 49% from 0.76 ± 0.08 to 0.39 ± 0.05 min¹ (p < 0.02, n = 6 out of 6 cells, Fig. 5C). In an additional subset of cells that exhibited Ca²⁺ plateaus even at low Ang II concentrations, these were converted by nifedipine to regular repetitive Ca²⁺ transients (n = 7 cells, Fig. 5D). In contrast, the plateau components of biphasic Ca²⁺ responses elicited by high Ang II concentrations (10 nM) were not significantly affected by either nifedipine (Fig. 1C) or BAY K 8644 (not shown). In the few cells that displayed spontaneous [Ca²⁺]_i signals in the absence of any stimulus (in the form of elevated, noisy base lines), these were terminated by treatment with nifedipine (Fig. 5E), suggesting their dependence on Ca²⁺ entry through VSCCs. The reciprocal effects of BAY K 8644 and nifedipine on [Ca²⁺]_i signaling following stimulation with a high [K⁺]_o (30 mM) to activate VSCCs in glomerulosa cells are shown in Fig. 5F.

To estimate the extent to which Ca²⁺ influx through VSCC contributes to the rise in frequency caused by increases in extracellular Ca²⁺ concentration, nifedipine (2 µM) was added after the restoration of [Ca²⁺]_o in Ang II (100 pM)-stimulated cells (Fig. 2A). Comparison of the oscillatory frequencies observed before and after Ca²⁺ addition and after administration of nifedipine in the presence of external Ca²⁺, gives an index of the VSCC component of extracellular Ca²⁺-dependent frequency modulation. Such comparison was achieved by expressing the oscillation frequency as a percentage of that observed in the presence of Ca²⁺ after Ca²⁺ restoration. This analysis showed that while the frequency was decreased by 30% after administration of nifedipine in the Ca²⁺-restored state it was still significantly higher than that recorded before Ca²⁺ addition (50%, n = 6 cells; Fig. 2A and Table II). Nifedipine (2 µM) had no effect on the plateau level of [Ca²⁺]_i when the same experimental protocol was used with a high dose (10 nM) of Ang II (not shown).

Table II. DHP-sensitive component of extracellular Ca2+-mediated regulation of Ang II-induced Ca2+ oscillations Comparison of the intertransient intervals (mean ± S.E.) in six individual bovine adrenal glomerulosa cells following stimulation with 100 pM Ang II in Ca2+-free medium, before, and after the addition of 0.6 mM [Ca2+]o and following the subsequent addition of nifedipine (2 µM). The frequencies are expressed as percentage of that observed in 0.6 mM [Ca2+]o-containing medium (control, 100%). Statistical significance was calculated by analysis of variance with the Duncan test (*, p < 0.005 compared with 0.6 mM [Ca2+]o; **, p < 0.005 compared with 0.6 mM [Ca2+]o and p < 0.05 compared with Ca2+-free, F, nifedipine-sensitive part of the Ca2+-dependent frequency enhancement as percentage of the entire frequency increase; n, number of intervals counted in each trace. Cell Ca2+-free 0.6 mM [Ca2+]o Nifedipine (2 µM) n min1 % s n min1 % s n min1 % s F 1 5 0.46 31 131  ± 6 10 1.5 100 40  ± 1 10 0.54 36 112  ± 2 92 2 3 0.18 38 341  ± 5 6 0.48 100 126  ± 4 5 0.41 85 148  ± 11 23 3 5 0.29 67 208  ± 45 6 0.43 100 138  ± 5 7 0.36 84 169  ± 19 50 4 4 0.23 72 256  ± 17 4 0.32 100 188  ± 6 4 0.24 75 250  ± 6 89 5 5 0.28 43 217  ± 11 8 0.65 100 93  ± 6 7 0.48 74 126  ± 6 46 6 4 0.23 45 262  ± 24 7 0.51 100 117  ± 5 6 0.34 67 175  ± 8 61 Mean ± S.E. of six cells 0.28  ± 0.04 49  ± 7 236  ± 28 0.65  ± 0.18 100  117  ± 20 0.40  ± 0.04 70  ± 7    163  ± 20** 60  ± 11

Since Ca²⁺ influx has been shown to affect InsP₃ formation in agonist-stimulated cells (24, 30, 31), we determined whether the effects of dihydropyridines on Ca²⁺ signaling were due to the increased formation of this second messenger. Glomerulosa cells prelabeled with myo-[³H]inositol were stimulated with low concentrations (100 or 300 pM) of Ang II in the presence of 10 mM Li⁺ to quantitate the small amount of InsP₃ (mostly in the form of its metabolites) that is produced under these conditions. The effects of the dihydropyridines on inositol phosphate responses to 100 pM Ang II were negligible compared with the prominent response elicited by raising the Ang II concentration from 100 pM to 300 pM (Fig. 6). Since the addition of BAY K 8644 and elevation of Ang II concentration from 100 to 300 pM caused a comparable increase in Ca²⁺ oscillatory frequency (86 and 75% increases, respectively), these results indicated that the effects of dihydropyridines on Ca²⁺ oscillations are not attributable to changes in InsP₃ production.

DISCUSSION

The characteristics of the oscillatory Ca²⁺ responses of adrenal glomerulosa cells to low doses of Ang II were similar to those of the ``base line-spiking'' Ca<sup>2+</sup> oscillations first described in single hepatocytes (32). These oscillations consisted of Ca<sup>2+</sup> transients separated by long interspike base-line intervals and showed frequency modulation at low agonist concentrations. Higher concentrations of Ang II usually evoked nonoscillatory signals that consisted of a rapid rise to a peak [Ca<sup>2+</sup>]<sub>i</sub> followed by a fall to a lower plateau elevation (``biphasic response''). The rapid initial [Ca²⁺]_i increase appeared to consist of superimposed Ca²⁺ transients that sometimes could be resolved into partially separated spikes. Individual Ca²⁺ responses showed a degree of heterogeneity, and in many cases low agonist doses initiated regular oscillations that subsequently changed to nonoscillatory plateau increases. Conversely, a small number of cells responded to high dose Ang II stimulation with high frequency oscillations that were maintained for a prolonged period of time.

The basic mechanism of the oscillations observed in cells stimulated by Ca²⁺-mobilizing hormones is dependent on the InsP₃-gated Ca²⁺ release from intracellular stores through the InsP₃ receptor-Ca²⁺ channel, and its subsequent reuptake into the same stores. The periodicity of Ca²⁺ release at any given InsP₃ concentration is believed to result from the rapid stimulatory and slower inhibitory effects of small and large [Ca²⁺]_i increases, respectively, on the activity of the InsP₃ receptor channel (3, 33). Accordingly, in addition to agonist-induced increases in intracellular InsP₃ concentration, changes in [Ca²⁺]_i have a major impact on the character of Ca²⁺ oscillations. Theoretically, InsP₃-induced oscillations can be maintained in a closed system in which there is no net movement of Ca²⁺ through the plasma membrane and only the periodic emptying and refilling of the intracellular Ca²⁺ stores contributes to the cytosolic Ca²⁺ changes (34). However, in real cells the plasma membrane Ca²⁺ pumps and the various Ca²⁺ influx mechanisms make major contributions to the changes in [Ca²⁺]_i. It was previously assumed that the activation of voltage-sensitive Ca²⁺ influx mechanism(s) by periodic depolarization (action potentials), as well as capacitative Ca²⁺ entry pathways activated by the depletion of intracellular Ca²⁺ stores, cause significant increases in [Ca²⁺]_i. However, recent data suggest that under physiological conditions such Ca²⁺ influx can modulate Ca²⁺ release from internal stores without necessarily changing the average cytosolic Ca²⁺ concentration (11). This concept posits that the limited amount of Ca²⁺ that enters the cell during agonist stimulation does not cause a measurable increase in [Ca²⁺]_i but rather a local elevation that influences the Ca²⁺ release process by acting specifically at the Ca²⁺-regulatory site of the InsP₃ receptor.

The results of the present study clearly demonstrate that Ca²⁺ influx regulates the frequency of Ca²⁺ oscillations in adrenal glomerulosa cells stimulated by low agonist concentrations. Similar findings have been recently reported in the exocrine avian nasal gland and in HeLa cells (35, 36). However, in glomerulosa cells, nifedipine and BAY K 8644 had a pronounced effect on Ca²⁺ oscillations, suggesting that the extracellular Ca²⁺-dependent frequency modulation involves dihydropyridine-sensitive VSCCs. Furthermore, the increased oscillatory frequency caused by the addition of Ca²⁺ was not usually accompanied by a significant increase in base-line [Ca²⁺]_i, indicating that the amount of Ca²⁺ entering the cells did not cause an overall increase in the cytosolic Ca²⁺ concentration. Rather, such Ca²⁺ influx changed the sensitivity of the InsP₃-dependent triggering process that determines the oscillation frequency. Such a change could result from either an increase in the InsP₃ concentration caused by Ca²⁺ influx (24, 31) or local increases of Ca²⁺ at the microdomains adjacent to the Ca²⁺-regulatory site of the InsP₃-receptor/Ca²⁺ channel (3).

Studies in [³H]inositol-labeled glomerulosa cells indicated that the frequency modulation caused by dihydropyridines was not accompanied by corresponding changes in InsP₃ production. In cells stimulated with 100 pM Ang II, the addition of BAY K 8644 caused a similar frequency increase as elevation of the Ang II concentration to 300 pM but had no significant effect on inositol phosphate production. These findings suggest that changes in Ca²⁺ concentration due to increased Ca²⁺ influx have an important regulatory function at the level of the InsP₃ receptor. In this regard, it is interesting to note that a significant negative correlation exists between the duration of Ca²⁺ spikes and the interspike intervals in cells stimulated with low Ang II concentrations in Ca²⁺-free medium, so that relatively narrow spikes were accompanied by higher spiking frequency. Since cells that show these narrow spikes do not appear to lose significant amounts of Ca²⁺ during oscillations, this observation indicates that the activity of the SERCA Ca²⁺ pump is an important determinant of the refractory period. This finding is in agreement with recent data obtained in X. laevis oocytes, in which overexpression of a SERCA Ca²⁺ pump caused synchronization and increased frequency of oscillations (12).

While regular oscillations were characteristic of most glomerulosa cells stimulated with low agonist concentrations, these usually changed to sustained [Ca²⁺]_i elevations at higher agonist doses, and in some cases even at low Ang II concentrations. Such nonoscillatory Ca²⁺ responses appeared to be a characteristic of cells in which the intracellular Ca²⁺ pools are unable to completely refill, with consequently more active capacitative Ca²⁺ influx pathways. This is clearly the case when cells are stimulated with high Ang II concentrations in the presence of Ca²⁺, where the high levels of InsP₃ prevent pool refilling. Cells exposed to high concentrations of Ang II are still able to oscillate in Ca²⁺-deficient medium, partly due to a smaller InsP₃ increase (24) and a lesser Ca²⁺ load in the cytosol due to the lack of Ca²⁺ influx. A similar extracellular Ca²⁺-dependent switch from oscillatory to nonoscillatory response was seen in some cells stimulated at low Ang II concentration after the addition of dihydropyridines. Thus, the addition of BAY K 8644 sometimes converted the oscillatory response to a sustained plateau, while nifedipine reestablished regular oscillations in cells that showed sustained [Ca²⁺]_i increases.

The heterogeneity of the [Ca²⁺]_i responses to the addition of external Ca²⁺ in Ang II-stimulated cells also shed some light on the conditions under which Ca²⁺ oscillations are not maintained in the presence of Ca²⁺. Thus, the biphasic Ca²⁺ increase with no oscillations was a characteristic response to Ca²⁺ addition of cells that showed broad spikes and/or rapidly decreasing spike amplitudes when stimulated with Ang II in Ca²⁺-free medium. On the other hand, cells with narrow and regular spikes responded to the addition of Ca²⁺ with an increase in oscillatory frequency but could behave like the former cells after minor inhibition of their SERCA Ca²⁺ pump by thapsigargin. These observations also implicate the Ca²⁺ reuptake mechanism(s), and its ability to restore basal Ca²⁺ levels and refill the Ca²⁺ pools, as a major prerequisite for the maintenance of regular oscillations in the presence of Ca²⁺. Under conditions of pool depletion, the major form of Ca²⁺ entry is the capacitative Ca²⁺ influx pathway, as evidenced by the large capacitative refilling event observed in such cells after the readdition of Ca²⁺. The minor effects of dihydropyridines on the sustained [Ca²⁺]_i elevations in these cells are consistent with the notion that the amount of Ca²⁺ entering through VSCCs during Ang II stimulation is not sufficient to produce a major increase of the average cytosolic Ca²⁺ level.

In summary, the present study demonstrates that in adrenal glomerulosa cells stimulated by low Ang II concentrations, Ca²⁺ entry through VSCCs acts as a major regulator of the cytosolic Ca²⁺ response, primarily by modifying the characteristics of the Ca²⁺ oscillations. At higher agonist concentrations, or when intracellular Ca²⁺ pools are subject to depletion, the capacitative entry pathway, which is a major contributor to the plateau increase of [Ca²⁺]_i, is activated with no further oscillations. Local changes in the Ca²⁺ concentration at the Ca²⁺-regulatory site of the InsP₃ receptor Ca²⁺ channel probably represent the mechanism through which extracellular Ca²⁺ affects Ca²⁺ oscillations.