L-type Ca2+ Channels and K+ Channels Specifically Modulate the Frequency and Amplitude of Spontaneous Ca2+ Oscillations and Have Distinct Roles in Prolactin Release in GH3 Cells*

GH3 cells showed spontaneous rhythmic oscillations in intracellular calcium concentration ([Ca2+] i ) and spontaneous prolactin release. The L-type Ca2+ channel inhibitor nimodipine reduced the frequency of Ca2+ oscillations at lower concentrations (100nm–1 μm), whereas at higher concentrations (10 μm), it completely abolished them. Ca2+ oscillations persisted following exposure to thapsigargin, indicating that inositol 1,4,5-trisphosphate-sensitive intracellular Ca2+ stores were not required for spontaneous activity. The K+ channel inhibitors Ba2+, Cs+, and tetraethylammonium (TEA) had distinct effects on different K+ currents, as well as on Ca2+oscillations and prolactin release. Cs+ inhibited the inward rectifier K+ current (KIR) and increased the frequency of Ca2+ oscillations. TEA inhibited outward K+ currents activated at voltages above –40 mV (grouped within the category of Ca2+ and voltage-activated currents, KCa,V) and increased the amplitude of Ca2+oscillations. Ba2+ inhibited both KIR and KCa,V and increased both the amplitude and the frequency of Ca2+ oscillations. Prolactin release was increased by Ba2+ and Cs+ but not by TEA. These results indicate that L-type Ca2+ channels and KIRchannels modulate the frequency of Ca2+ oscillations and prolactin release, whereas TEA-sensitive KCa,V channels modulate the amplitude of Ca2+ oscillations without altering prolactin release. Differential regulation of these channels can produce frequency or amplitude modulation of calcium signaling that stimulates specific pituitary cell functions.

There is increasing evidence for distinct roles of different spatial and temporal patterns of intracellular free Ca 2ϩ concentration ([Ca 2ϩ ] i ) in the regulation of cellular processes (1). Many cell types exhibit oscillations of [Ca 2ϩ ] i that may be differentially modulated to produce highly specific intracellular signals. For example, the frequency of Ca 2ϩ oscillations has been shown to regulate secretion, whereas the amplitude of Ca 2ϩ oscillations has been shown to regulate gene expression in different cell systems (1,2). The multifunctional enzyme calmodulin kinase II has been shown to be capable of decoding different patterns of Ca 2ϩ signaling into different functional responses. (3).
Endocrine cells have the intrinsic capacity for extensive spontaneous activity that is independent of stimulation by external factors. In pituitary cells, this activity is characterized by membrane potential oscillations, action potentials, and Ca 2ϩ oscillations (4 -7). It is likely that this spontaneous, intrinsic signaling plays a role in basal hormone release by pituitary cells and other endocrine cells, although this role has yet to be clearly defined (8 -10). In addition, the individual components of this intrinsic signaling may be targets of modulation through which diverse signals can induce specific changes in cellular activity and hormone release.
The rat pituitary growth hormone-and prolactin-secreting GH 3 cell line is a useful and well studied model system for the study of pituitary cell signaling. GH 3 cells express L-type Ca 2ϩ channels as well as inwardly rectifying and Ca 2ϩ -and voltageactivated K ϩ channels. The biophysical and pharmacological properties of these channels have been extensively characterized in previous studies (11)(12)(13)(14)(15)(16)(17)(18)(19). GH 3 cells also show spontaneous activity that is generated by the coordinated action of ion channels, Ca 2ϩ influx, Ca 2ϩ release from intracellular stores, and other second messengers including IP 3 1 and cAMP. In these and other pituitary cells, the relative contributions of each of these cellular signaling components to overall cellular activity may vary from cell to cell and under stimulated versus unstimulated conditions (20 -22). It is well established that increases in [Ca 2ϩ ] i directly mediate hormone release in GH 3 cells and other endocrine cell types (10,23,24). Oscillatory patterns of Ca 2ϩ signaling provide the opportunity for a cell to respond to individual components of a Ca 2ϩ signal (e.g. baseline [Ca 2ϩ ] i , oscillation frequency, oscillation duration, or oscillation amplitude). The functional response of the cell may be different if it depends on a "frequency-modulated" signal versus an "amplitude-modulated" signal (1). Individual ion channels and second messengers may play specific roles in generating specific patterns of spontaneous Ca 2ϩ signaling and in turn may generate different patterns of hormone release.
In this study, we have used the combination of patch clamp measurements, fluorescence imaging of intracellular Ca 2ϩ concentration, and a sensitive enzyme-linked immunosorbent assay (ELISA) for prolactin to study the ionic mechanisms controlling hormone release from GH 3 cells. We have investigated the role of specific patterns of spontaneous Ca 2ϩ signaling in prolactin release from GH 3 cells by using Ca 2ϩ and K ϩ channel antagonists to modulate the patterns of Ca 2ϩ signaling.

EXPERIMENTAL PROCEDURES
Cell Culture-GH 3 cells, obtained from American Type Culture Collection, Manassas, VA (CCL 82.1), were maintained in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal calf serum, penicillin (0.05 IU/ml), and streptomycin (50 g/ml) and incubated in a humid atmosphere of 5% CO 2 , 95% O 2 at 37°C. Cells were harvested once a week by treatment with a phosphate-buffered saline containing EDTA (1 mM) and reseeded at 20% original density, either into 6-well plates for prolactin release assays, 35-mm diameter culture dishes for electrophysiological studies, or poly-D-lysine-coated coverslips for Ca 2ϩimaging studies. The incubation medium was changed every 2-3 days.
Electrophysiological Recordings-Single cells were voltage-clamped, and voltage-activated K ϩ channel activity was recorded from whole GH 3 cells using a List EPC-7 patch-clamp amplifier. For the recording of K Ca,V channel activity, cells were superfused with a solution containing 140 mM NaCl, 2.8 mM KCl, 2 mM MgCl 2 , 1 mM CaCl 2 , 10 mM HEPES, 6 mM glucose, 5 ϫ 10 Ϫ4 mM tetrodotoxin (pH 7.2 with NaOH). The recording electrode contained 120 mM KCl, 1 mM EGTA, 1 mM MgCl 2 , 3 mM Mg-ATP, 10 mM HEPES (pH 7.2 with KOH) (all from Sigma). Currents were activated by step depolarizations of membrane potential from a holding potential of Ϫ80 mV for 100 ms every 10 s. Capacitance compensations were achieved using the patch-clamp amplifier. Residual artifacts and leakage currents were nulled using a P/4 subtraction.

Measurement of [Ca 2ϩ ] i -[Ca 2ϩ
] i was measured using a fluorescence imaging system that has previously been described in detail (25). Briefly, cells on poly(D-lysine)-coated glass coverslips were loaded with fura2 by incubation in 5 M fura2-AM for 40 min. Cells were then washed and maintained in normal medium for 30 min before experimentation. Coverslips were excited with a mercury lamp through 340and 380-nm band-pass filters, and fluorescence at 510 nm was recorded through a 10ϫ or 20ϫ objective with a silicon intensified target camera to an optical memory disc recorder. Images were then digitized and subjected to background subtraction and shading correction, after which [Ca 2ϩ ] i was calculated on a pixel-by-pixel basis, as described previously, by a frame grabber and image analysis board (Data Translation). Data acquisition and analysis software were written by Dr. Michael Sanderson. Tracings in all figures are based upon fluorescence of a 4 ϫ 4 pixel area located within each cell body.
Experiments were carried out in Hanks' balanced salt solution with 10 mM HEPES buffer, pH 7.4 (HBSS/HEPES) at 22°C. Agents were applied by perfusion of coverslips with at least 2 ml of HBSS/HEPES containing the particular agent.
ELISA-A competitive ELISA has been developed for measuring prolactin secreted by GH 3 cells. The assay utilizes an antibody (raised in rabbit) against rat prolactin. Both antisera (prolactinS-9) and standards (prolactin RP-3) were kindly provided by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). For the collection of samples, GH 3 cells were seeded into 6-well tissue culture plates 2 days before experiments (0.5-0.7 million cells/well). Release experiments were conducted at 37°C in a humidified incubator with 5% CO 2 . Immediately before each experiment, cells were washed gently with media (Dulbecco's modified Eagle's medium with 20 mM HEPES and 0.1% bovine serum albumin, pH 7.4, with NaOH). After washing, aliquots of media (1 ml) were added to each well for 0.5-h time points such that release could be monitored before, during, and after exposure to drugs. The amount of prolactin (ng/ml/10 6 cells) released in 0.5 h in the presence of drugs was expressed as a percentage of release from the same cells during 0.5 h under control conditions before exposure to drugs. After incubation with the cells, each media aliquot was centrifuged at 1800 rpm at 4°C then stored at Ϫ20°C or assayed directly by ELISA for prolactin. For the competitive prolactin ELISA, 96-well Nunc-Immuno Maxisorp Plates from Life Sciences, Denver, CO were used. Each well was coated with prolactin by incubation of 100 l of 0.1 M NaHCO 3 , pH 9.5, containing 1 ng of prolactin for 20 -24 h at 4°C. Before assay, prolactin-coated plates were washed with assay buffer containing 0.5 M NaCl, 20 mM NaH 2 PO 4 , 0.05% Tween 20, 0.5% bovine serum albumin, pH to 7.4, then incubated with assay buffer for 30 min at room temperature to remove prolactin binding weakly to the plate. After further washing with assay buffer, undiluted samples (100 l) or standards (0.02-40 ng) dissolved in 100 l of media were added to the wells, followed by the addition of 50 l of prolactin antibody at a dilution of 1:40,000. After incubation for 2 h at room temperature, bound antibody was detected using peroxidase-conjugated anti-rabbit antibody (Vector, Burlingame, Ca) with tetramethylbenzidine (Life Technologies, Inc.) as substrate. The peroxidase reaction was terminated by 1 N H 2 SO 4 , and absorbance was measured by a microplate reader (Molecular Devices) at 450 nm. To determine the amount of prolactin present in the samples, a standard curve was generated. The percent of maximum absorbance (corresponding to no prolactin added) was plotted against known amounts of prolactin (0.02-40 ng). All samples were assayed in quadruplicates from three separate determinations.

RESULTS
The majority of GH 3 cells (approximately 70%, n Ͼ 1500 cells in 50 experiments) showed spontaneous oscillations in [Ca 2ϩ ] i . The pattern of these oscillations in [Ca 2ϩ ] i varied considerably from cell to cell. Ca 2ϩ oscillations had a periodicity ranging from 3-30 s and a peak amplitude ranging from 40 -300 nM in different cells. Some cells showed Ca 2ϩ oscillations with a relatively consistent frequency, amplitude, and shape (e.g. Fig.  1, Cell #2), whereas other cells showed a more random pattern of oscillations (e.g. Fig. 1, Cell #28). Increasing the temperature of the bath solution from room temperature to 37°C increased the frequency of spontaneous Ca 2ϩ oscillations but did not result in any qualitative changes in spontaneous Ca 2ϩ oscillations or in their response to the conditions described below (data not shown).
Ca 2ϩ oscillations were completely abolished in all cells during perfusion with medium containing no added Ca 2ϩ , and they were restored immediately upon replacement of Ca 2ϩ (n ϭ 120 cells in three experiments, data not shown). We have previously shown that GH 3 cells show predominantly L-type Ca 2ϩ currents that are inhibited by nimodipine (19). Nimodipine inhibited spontaneous Ca 2ϩ oscillations in a concentration-dependent fashion (Fig. 1). Nimodipine (1 M) reduced the frequency of Ca 2ϩ oscillations in some cells and abolished them in other cells. Interestingly, although 1 M nimodipine reduced the frequency of Ca 2ϩ oscillations in some cells, in most of these cells it did not significantly affect their amplitude. Nimodipine (10 M) abolished Ca 2ϩ oscillations in all cells (n ϭ 100 cells in three experiments). Nimodipine (10 M) also reduced spontaneous prolactin release by 32.9 Ϯ 9.1% (see Fig. 9). The persistence of prolactin release despite the abolition of Ca 2ϩ oscillations shows that a significant portion of the spontaneous hormone release does not require spontaneous Ca 2ϩ signaling.
Exposure to 1 M thapsigargin resulted in an increase in base-line [Ca 2ϩ ] i and a slight increase in the frequency of Ca 2ϩ oscillations in most cells (n ϭ 120 cells in four experiments). Base-line [Ca 2ϩ ] i returned to previous levels after 3-5 min in thapsigargin. After exposure to thapsigargin, Ca 2ϩ oscillations continued to occur in 69% of cells showing spontaneous oscillations (59/87 cells in three experiments), although in some cells their amplitude and frequency were reduced (Fig. 2). The phospholipase C inhibitor U73122 had no significant effect on Ca 2ϩ oscillations, suggesting that the active formation of IP 3 is not required for spontaneous Ca 2ϩ oscillations (n ϭ 100 cells in three experiments, data not shown). The activity of both thapsigargin and U73122 were verified by the observation that each agent completely inhibited the rapid peak increase in [Ca 2ϩ ] i induced by thyrotropin-releasing hormone (data not shown). These results show that the activity of phospholipase C and the release of thapsigargin-sensitive intracellular Ca 2ϩ are not required for spontaneous Ca 2ϩ oscillations in most cells but may contribute to modulation of their frequency and amplitude.
The effect of nimodipine suggests that Ca 2ϩ oscillations are generated by spontaneous depolarizations with resultant influx of Ca 2ϩ through voltage-gated channels. Consistent with this hypothesis, spontaneous Ca 2ϩ oscillations were highly sensitive to depolarization induced by increasing extracellular [K ϩ ]. Increasing extracellular [K ϩ ] by as little as 3 mM resulted in a increase in base-line [Ca 2ϩ ] i as well as an increase in the frequency of Ca 2ϩ oscillations (n ϭ 100 cells in three experiments, not shown). GH 3 cells display multiple types of K ϩ currents, including an inward rectifying (K IR ) current and outward currents mediated by Ca 2ϩ and voltage-activated channels (15)(16)(17)26). For the purposes of this study, the different outward K ϩ currents recorded were grouped into a single category of Ca 2ϩ and voltageactivated (K Ca,V ) currents. K Ca,V channel activity was recorded using the whole-cell patch-clamp configuration by depolarizing GH 3 cells held at Ϫ80 mV to between Ϫ50 and 60 mV in 10-mV increments (Fig. 3). Using the specified internal and external solutions (see "Experimental Procedures"), outward K ϩ cur-rents were observed with a threshold of activation of approximately Ϫ40 mV. Different recording solutions (see "Experimental Procedures") and hyperpolarizing steps from a Ϫ40 mV holding potential to between Ϫ50 and Ϫ120 mV (10 mV decrements) were used to specifically activate currents mediated by K IR channels. K IR channel activity recorded at more depolarized potentials than Ϫ40 mV was overwhelmed by currents through K Ca,V channels. K IR and K Ca,V currents were inhibited selectively by different agents (Fig. 4). TEA (1 mM) inhibited outward K Ca,V currents but not K IR currents. The outward K ϩ currents had both transient and sustained components; the inhibition of outward K ϩ current by TEA was most marked at the end of the voltage step, indicating a relatively selective block of the sustained current component. Ba 2ϩ (1 mM) inhibited both K IR and K Ca,V (with no obvious specificity for the transient or sustained components), whereas Cs ϩ inhibited only K IR currents (Fig. 4).
Cs ϩ , Ba 2ϩ , and TEA also had distinct effects on Ca 2ϩ oscillations. TEA (1 mM) induced a marked increase in the ampli- These K ϩ channel antagonists also had distinct effects on prolactin release (Fig. 9). TEA had no significant effect on prolactin release. By contrast, Ba 2ϩ evoked a large increase in prolactin release, whereas Cs ϩ evoked a small but significant increase in prolactin release. DISCUSSION GH 3 cells show spontaneous activity characterized by rhythmic Ca 2ϩ oscillations and prolactin release. Ca 2ϩ channels, K ϩ channels, Ca 2ϩ influx, and release of intracellular Ca 2ϩ play distinct roles in this spontaneous activity. A number of previous studies have demonstrated a primary role for release of Ca 2ϩ from intracellular stores in the spontaneous and stimulated activity of pituitary cells (20,23,(27)(28)(29)(30). Our results show that spontaneous Ca 2ϩ oscillations in GH 3 cells are completely dependent on influx of Ca 2ϩ and in the majority of cells, do not require release of intracellular Ca 2ϩ from thapsigargin-sensitive stores. The efficacy of thapsigargin in depleting Ca 2ϩ stores was verified by the complete inhibition of the peak increase in [Ca 2ϩ ] i in response to thyrotropin-releasing hormone, as reported previously by Nelson and Hinkle (31). The frequency and amplitude of Ca 2ϩ oscillations were diminished in some cells in response to thapsigargin, suggesting that intracellular Ca 2ϩ stores may play a modulatory rather than a primary role in Ca 2ϩ oscillations. This modulatory role may involve Ca 2ϩ -induced Ca 2ϩ release through interaction of Ca 2ϩ with the IP 3 receptor (28,30,32). Thapsigargin did abolish spontaneous Ca 2ϩ oscillations in a significant proportion of cells (approximately 30%), suggesting that there is a subset of cells whose spontaneous Ca 2ϩ signaling does require intracellular Ca 2ϩ release from thapsigargin-sensitive stores. A similar dependence on IP 3 -sensitive intracellular Ca 2ϩ stores in a subset of GH 3 cells is suggested by Varney et al. (33), who found that chronic treatment with Li 2ϩ , which reduces basal levels of IP 3 , decreased the number of cells showing spontaneous Ca 2ϩ oscillations.
The effects of nimodipine indicate that spontaneous Ca 2ϩ oscillations are generated by influx of Ca 2ϩ through L-type channels that have previously been shown to be inhibited by nimodipine in GH 3 cells (19). At lower concentrations, nimodipine often reduced the frequency of Ca 2ϩ oscillations without significantly altering their amplitude. This observation suggests that the frequency of Ca 2ϩ oscillations is dependent upon the activation state of Ca 2ϩ channels, whereas the base-to-peak amplitude of spontaneous Ca 2ϩ oscillations is an all-or-none phenomenon that is not regulated by the state of Ca 2ϩ channels under unstimulated conditions. The central role of Ca 2ϩ channels in the overall activity of the cell makes them a logical target for inhibitory or excitatory modulation by external ligands. The observations that somatostatin receptors and expressed opioid receptors in GH 3 cells modulate Ca 2ϩ channels is consistent with their role as a target for receptor-mediated modulation (19,34).
Nimodipine inhibited basal prolactin release by approximately 30%, indicating that a significant proportion of basal prolactin release is stimulated by spontaneous Ca 2ϩ signaling. However, the persistence of prolactin release in the absence of any detectable Ca 2ϩ signaling shows that a high proportion of basal prolactin release is independent of spontaneous Ca 2ϩ oscillations. This finding is consistent with the report of Masumoto et al. (10), who found that basal exocytosis continued FIG. 3. Whole-cell K ؉ currents recorded from GH 3 cells. A, outward K ϩ currents through K Ca,V channels were recorded using the whole-cell patch-clamp technique. Currents were evoked by depolarizing from a Ϫ80 mV holding potential to voltages between Ϫ50 to 60 mV (10 mV increments) for 100 ms. Traces represent averaged currents recorded from 6 cells. B, relationship between the peak K ϩ current amplitude and test pulse are shown graphically. Data were obtained from the same cells as in A. Vertical error bars, when bigger than symbols, represent ϮS.E. C, K IR currents were activated by hyperpolarizing pulses (duration 1.5 s) to voltages between Ϫ50 to Ϫ120 mV (10 mV decrements) from a Ϫ40 mV holding potential. Equimolar KCl (120 mM) solutions were used to record K IR channel activity (see "Experimental Procedures"). Tracings represent currents averaged from 5 cells. D, the graph shows the current-voltage relationship of K IR channel activity recorded from the same cells as in C.
Both peak (f) and sustained (Ⅺ) current amplitude is illustrated. Vertical error bars represent Ϯ S.E.
in the absence of increases in [Ca 2ϩ ] i in pituitary gonadotrophs. It is likely that this represents an unregulated pathway of secretion that has been reported in GH 3 cells (35) and other cell types.
Different K ϩ channel antagonists have distinct effects on Ca 2ϩ signaling in GH 3 cells. TEA evoked a dramatic increase in the amplitude of Ca 2ϩ oscillations; this effect is likely because of prolongation of the action potential and subsequent increase FIG. 4. The effect of K ؉ channel antagonists on K Ca,V and K IR channels in GH 3 cells. A, control and Ba 2ϩ (1 mM)-inhibited K Ca,V currents. Ba 2ϩ -inhibited peak and sustained current components indicating blockade of both a transient A current and longer lasting currents, consistent with delayed rectifying and Ca 2ϩ -activated K ϩ currents. B, TEA (10 mM) predominantly inhibited the sustained current component. In both A and B, superimposed traces are the averages of two currents activated by depolarizing from -80 to 0 mV in the presence and absence of the inhibitors indicated. C, the graph shows the effects of TEA, Ba 2ϩ , and Cs ϩ (all 1 mM) on K Ca,V current amplitude. Current amplitudes were averaged over 5 ms at the end of each depolarizing (-80 to 0 mV) voltage step. Bars represent current amplitudes in the presence of the inhibitors expressed as a percentage of the control current amplitudes. Currents through K Ca,V channels were inhibited by TEA (n ϭ 5, p Ͻ 0.005) and Ba 2ϩ (n ϭ 4, p Ͻ 0.005) but were not modulated by Cs ϩ (n ϭ 4, p Ͻ 0.005). D, Ba 2ϩ (1 mM)-and (E) Cs ϩ (1 mM)-inhibited K IR currents activated by hyperpolarizing cells from -40 to -90 mV with equal K ϩ concentrations in the electrode and bath solutions (see "Experimental Procedures"). Superimposed traces are the averages of two currents recorded in the presence and absence of the inhibitors indicated. F, inward rectifying K IR currents were not sensitive to TEA (n ϭ 6 cells) but were inhibited by the application of Ba 2ϩ (n ϭ 6, p Ͻ 0.05) and Cs ϩ (n ϭ 4, p Ͻ 0.005). p values were determined by a paired sample t test. Current measurements were made at the end of the hyperpolarizing step (as described in C). in the action potential-induced influx of Ca 2ϩ . This is consistent with the relatively selective effect TEA has on sustained outward K ϩ currents recorded from GH 3 cells, with little effect on the A current (15). Interestingly, the large increase in the amplitude of Ca 2ϩ oscillations was not associated with a large increase in base-line [Ca 2ϩ ] i , showing that base-line [Ca 2ϩ ] i is regulated by a mechanism other than TEA-sensitive outward K ϩ currents. Both Ba 2ϩ and Cs ϩ increased the frequency of Ca 2ϩ oscillations, whereas TEA did not. These results indicate that the inward rectifier K ϩ channel plays a primary role in setting the frequency of spontaneous Ca 2ϩ oscillations. They are consistent with the results of Barros et al. (36,37), who report a primary role for the inward rectifier in the regulation of GH 3 cell excitability. Base-line [Ca 2ϩ ] i was increased dramatically by Ba 2ϩ , and to a lesser extent, by Cs ϩ and TEA. These results suggest that Ba 2ϩ modulates base-line [Ca 2ϩ ] i by a mechanism that is distinct from the inward rectifier, and as discussed above, distinct from TEA-sensitive outward K ϩ currents.
Prolactin release was also differentially modulated by different K ϩ channel antagonists. Despite the fact that TEA caused a marked increase in the amplitude of Ca 2ϩ oscillations and a significant increase in average [Ca 2ϩ ] i , it did not increase prolactin release. By contrast, Cs ϩ , which increased the frequency of Ca 2ϩ oscillations but caused a much smaller increase in average [Ca 2ϩ ] i than TEA, did cause a significant increase in prolactin release. These findings are consistent with the stimulation of hormone release by an increase in the frequency of spontaneous Ca 2ϩ oscillations but not in their amplitude. The marked effect of Ba 2ϩ suggests that the combination of increased base-line [Ca 2ϩ ] i and increased Ca 2ϩ oscillation frequency can cause a much greater increase in hormone release than increased Ca 2ϩ oscillation frequency alone.
Simultaneous measurements of membrane capacitance and [Ca 2ϩ ] i by multiple investigators have provided strong evidence that each oscillatory increase in [Ca 2ϩ ] i is capable of evoking secretion of hormone (24, 38 -42). Our results suggest that increasing the peak [Ca 2ϩ ] i associated with each sponta- neous Ca 2ϩ oscillation does not increase hormone release; this may be because there is a maximum number of vesicles released per Ca 2ϩ oscillation once the peak [Ca 2ϩ ] i has reached a certain threshold or because buffering of Ca 2ϩ prevents this increased peak [Ca 2ϩ ] i from being sensed by the secretory apparatus. By contrast, the increase in hormone release associated with an increase in Ca 2ϩ oscillation frequency suggests that modulation of Ca 2ϩ oscillation frequency does represent a mechanism for stimulation of hormone release. Increasing base-line [Ca 2ϩ ] i , in addition to increasing Ca 2ϩ oscillation frequency, results in a further increase in prolactin release. Possible explanations for the specific effects of base-line [Ca 2ϩ ] i and oscillation frequency on secretion include direct effects on the exocytotic trigger mechanism (43) or the recruitment of additional pools of vesicles with different Ca 2ϩ sensitivities (44,45).
Ca 2ϩ channels, K ϩ channels, Ca 2ϩ influx, and intracellular Ca 2ϩ release play specific roles in the generation of spontaneous activity in GH 3 cells. In turn, specific characteristics of spontaneous Ca 2ϩ signaling, namely base-line [Ca 2ϩ ] i and the frequency of Ca 2ϩ oscillations, are correlated with changes in hormone release. Our studies identify L-type Ca 2ϩ channels and the inward rectifier K ϩ channel as key components of the cellular signaling machinery whose modulation results in the specific changes in the patterns of spontaneous cellular activity that regulate changes in basal prolactin release. TEA-sensitive outward K ϩ currents may affect other cellular processes through amplitude-modulated Ca 2ϩ signaling. ] i values for TEA, Ba 2ϩ , and Cs ϩ were 63 Ϯ 28, 63 Ϯ 32, and 67 Ϯ 33 nM Ϯ S.D., respectively). TEA evoked a large increase in Ca 2ϩ oscillation amplitude (p Ͻ 0.001), Ba 2ϩ evoked a slight but significant (p Ͻ 0.05) increase in Ca 2ϩ oscillation amplitude, and Cs ϩ did not have a significant effect on Ca 2ϩ oscillation amplitude (middle, average control Ca 2ϩ oscillation amplitude values for TEA, Ba 2ϩ , and Cs ϩ were 121 Ϯ 48, 131 Ϯ 52, and 158 Ϯ 50 nM, respectively). Both Ba 2ϩ and Cs ϩ evoked a significant increase in Ca 2ϩ oscillation frequency (p Ͻ 0.005), whereas TEA did not have a significant effect (right, average control Ca 2ϩ oscillation frequency values for TEA, Ba 2ϩ , and Cs ϩ were 0.17 Ϯ .06, 0.16 Ϯ .04, and 0.14 Ϯ .05 Hz, respectively). p values were determined by a paired-sample t test.
FIG. 9. The effects of K ؉ and Ca 2؉ channel antagonists on prolactin release. Histogram bars represent percent change in prolactin release compared with control release measured immediately before antagonist application (100%). Each value is the average of at least four experiments. Control values for prolactin secretion in 0.5 h from cells before their exposure to TEA, Ba 2ϩ , and Cs ϩ were 2.6 Ϯ 1.2, 1.7 Ϯ 0.3, and 2.2 Ϯ 0.1 ng/ml/10 6 cells, respectively. Both Ba 2ϩ (1 mM) and Cs ϩ (1 mM) caused a significant increase in spontaneous prolactin release (p Ͻ 0.05), whereas nimodipine (10 M) significantly inhibited release (p Ͻ 0.05). TEA did not have a significant effect on prolactin release. Error bars represent S.E..