- Split View
-
Views
-
Cite
Cite
Sheng-Nan Wu, Shiuh-Inn Liu, Mei-Han Huang, Cilostazol, an Inhibitor of Type 3 Phosphodiesterase, Stimulates Large-Conductance, Calcium-Activated Potassium Channels in Pituitary GH3 Cells and Pheochromocytoma PC12 Cells, Endocrinology, Volume 145, Issue 3, 1 March 2004, Pages 1175–1184, https://doi.org/10.1210/en.2003-1430
- Share Icon Share
Abstract
The effects of cilostazol, a dual inhibitor of type 3 phosphodiesterase and adenosine uptake, on ion currents were investigated in pituitary GH3 cells and pheochromocytoma PC12 cells. In whole-cell configuration, cilostazol (10 μm) reversibly increased the amplitude of Ca2+-activated K+ current [IK(Ca)]. Cilostazol-induced increase in IK(Ca) was suppressed by paxilline (1 μm) but not glibenclamide (10 μm), dequalinium dichloride (10 μm), or β-bungarotoxin (200 nm). Pretreatment of adenosine deaminase (1 U/ml) or α,β-methylene-ADP (100 μm) for 5 h did not alter the magnitude of cilostazol-stimulated IK(Ca). Cilostazol (30 μm) slightly suppressed voltage-dependent l-type Ca2+ current. In inside-out configuration, bath application of cilostazol (10 μm) into intracellular surface caused no change in single-channel conductance; however, it did increase the activity of large-conductance Ca2+-activated K+ (BKCa) channels. Cilostazol enhanced the channel activity in a concentration-dependent manner with an EC50 value of 3.5 μm. Cilostazol (10 μm) shifted the activation curve of BKCa channels to less positive membrane potentials. Changes in the kinetic behavior of BKCa channels caused by cilostazol were related to an increase in mean open time and a decrease in mean closed time. Under current-clamp configuration, cilostazol decreased the firing frequency of action potentials. In pheochromocytoma PC12 cells, cilostazol (10 μm) also increased BKCa channel activity. Cilostazol-mediated stimulation of IK(Ca) appeared to be not linked to its inhibition of adenosine uptake or phosphodiesterase. The channel-stimulating properties of cilostazol may, at least in part, contribute to the underlying mechanisms by which it affects neuroendocrine function.
CILOSTAZOL, A THROMBOLYTIC AND vasodilator drug, has been developed as a selective inhibitor of cyclic nucleotide phosphodiesterase type 3A (1, 2). It has been regarded as a useful tool to investigate the adenylyl cyclase-cAMP pathway (3–5). However, cilostazol appears to possess an additional effect that does not require the inhibition of phosphodiesterases (2, 5). It was reported to be an inhibitor of adenosine uptake in a variety of cells (5, 6). For example, this drug could enhance the adenosine-induced increase in cAMP in Chinese hamster ovary cells in which adenosine receptors were overexpressed (6). In Calu-3 cell monolayers, cilostazol has been shown to produce a short circuit current that was sensitive to inhibition by glibenclamide (5). It could potentiate the increase of this current caused by the stimulation of adenosine receptors with adenosine (5). In addition, it has been reported to suppress acetylcholine-induced catecholamine secretion in bovine adrenal chromaffin cells (7). The latter findings appeared not to be related to an increase in intracellular cAMP. On the other hand, this drug has been reported to improve blood flow, facilitate axonal regeneration, and prevent impairment of slow axonal transport in streptozotocin-induced diabetic rats (8–12). It has been found to have a neuroprotective effect against cerebral infarct (13, 14). However, the effect of cilostazol on ion currents in neurons or neuroendocrine cells has not been studied extensively, although a recent report (15) showed the ability to stimulate Ca2+-activated K+ currents [IK(Ca)] in human neuroblastoma SK-N-SH cells.
The large-conductance Ca2+-activated K+ (BKCa) channel, the gating of which is controlled by intracellular Ca2+ levels and/or membrane depolarization, plays a role in the regulation of neuronal excitability and the coupling of excitation-contraction and stimulus-secretion (16). Cloning and expression of α- and β-subunits, which these channels comprise, will enable us to elucidate the structural components that control their gating (16, 17). We have demonstrated that riluzole and ciglitazone, both of which were reported to prevent neuronal injuries, could enhance the activity of BKCa channels found in GH3 cells (18, 19). Importantly, previous studies have demonstrated that the opener of BKCa channels could counteract the deleterious effects of excitatory neurotransmitters after neurotoxic injuries (20).
Therefore, the goal of this study was to examine the effect of cilostazol on IK(Ca) in pituitary GH3 cells, address the issue whether cilostazol affects the activity and kinetic properties of BKCa channels, and determine whether cilostazol can affect the activity of BKCa channels in pheochromocytoma PC12 cells. Interestingly, these results indicate that cilostazol can enhance the activity of BKCa channels in a mechanism unlikely to be linked to the inhibition of phosphodiesterase or adenosine uptake.
Materials and Methods
Cell culture
The clonal strain GH3 cell line, originally derived from a rat anterior pituitary adenoma, was obtained from the Culture Collection and Research Center (CCRC-60015, Hsinchu, Taiwan). The detailed methods have been previously described (21). Briefly, cells were cultured in Ham’s F-12 medium (Life Technologies, Inc., Grand Island, NY) supplemented with 15% heat-inactivated horse serum (vol/vol), 2.5% fetal calf serum (vol/vol), and 2 mml-glutamine (Life Technologies) in a humidified environment of 5% CO2/95% air. The experiments were performed 5 or 6 d after cells were subcultured (60–80% confluence). In a separate series of experiments, GH3 cells were incubated with adenosine deaminase (1 U/ml) or α,β-methylene-ADP (AOPCP, 100 μm) at 37 C for 5 h.
Stock cultures of rat pheochromocytoma PC12 cells were also obtained from the Culture Collection and Research Center (CCRC-60048). PC12 cells were maintained in RPMI 1640 medium (Life Technologies) supplemented with 10% heat-inactivated horse serum (vol/vol) and 5% fetal bovine serum (vol/vol) in a 5% carbon dioxide atmosphere at 37 C.
Electrophysiological measurements
Immediately before each experiment, cells were dissociated and an aliquot of cell suspension was transferred to a recording chamber positioned on the stage of an inverted microscope (DM IL, Leica Microsystems, Wetzlar, Germany). Cells were bathed at room temperature (20–25 C) in normal Tyrode’s solution containing 1.8 mm CaCl2. The recording pipettes were pulled from thin-walled borosilicate glass capillaries (Kimax-51, Kimble Glass, Vineland, NJ) using a two-stage microelectrode puller (PP-830, Narishige, Tokyo, Japan) and the tips were fire-polished with a microforge (MF-83; Narishige). When filled with pipette solution, their resistance ranged between 3 and 5 mΩ. Ion currents were recorded in the cell-attached, inside-out, and whole-cell configurations of the patch-clamp technique, using an RK-400 patch-clamp amplifier (Bio-Logic, Claix, France) (21). All potentials were corrected for liquid junction potential, a value that would always develop at the tip of the pipette when the composition of the pipette solution was different from that in the bath.
Data recording and analysis
The signals were displayed on an analog/digital oscilloscope (HM 507, Hameg Inc., East Meadow, NY) and a liquid crystal projector (PJ550-2, ViewSonic Corp., Walnut, CA). The data were on-line stored in a Pentium III-grade laptop computer (Slimnote VX3, Lemel, Taipei, Taiwan) via a universal serial bus port at 10 kHz through a high-speed/low-noise analog/digital interface (Digidata 1322A, Axon Instruments, Union City, CA). This device was controlled by a commercially available software (pCLAMP 9.0, Axon Instruments). Currents were low-pass filtered at 1 or 3 kHz. Ion currents recorded during whole-cell experiments were stored without leakage correction and analyzed subsequently using the pCLAMP 9.0 software (Axon Instruments), the Origin 6.0 software (Microcal Software, Inc., Northampton, MA), SigmaPlot 7.0 software (SPSS, Inc., Apex, NC), or custom-made macros in Excel (Microsoft, Redmont, WA). The pCLAMP-generated voltage-step protocols were used to examine the current-voltage (I-V) relations for ion currents.
The amplitudes of single BKCa channel currents were determined by fitting Gaussian distributions to the amplitude histograms of the closed and the open state, respectively. Channel dwell times were determined by applying a standard half-amplitude crossing protocol using the pCLAMP 9.0 software (Axon Instruments). The activity of the channel in a patch was expressed as N Po, which can be estimated using the following equation: N Po = (A1 + 2A2 + 3A3 + … + nAn)/(A0 + A1 +A2 +A3 + … + An), where N is the number of active channels in the patch, A0 is the area under the curve of an all-points histogram corresponding to the closed state, and A1…. An represent the histogram areas reflecting the levels of distinct open state for 1 to n channels in the patch. The single-channel conductance was calculated by linear regression using mean current amplitudes measured at different voltages.
To calculate the concentration-response relationship for cilostazol-stimulated BKCa channel activity, the potential was held at the level of +60 mV and the bath medium contained 0.1 μm Ca2+. The probability of channel openings measured during the exposure to cilostazol (100 μm) was taken as 100%. The concentration-dependent relationship of cilostazol on the channel open probability was fitted to a Hill equation: y = (Emax × [C]nh)/(EC50nh +[C]nh), where [C] is the concentration of cilostazol; EC50 and nh are the concentration required for a 50% inhibition and the Hill coefficient, respectively; and Emax is the cilostazol-induced maximal increase in channel activity. The Solver subroutine in Microsoft Excel was used to fit data by a least-squares minimization procedure.
To determine the effect of cilostazol on the activation curve of BKCa channels, the ramp pulses from 0 to +80 mV with a duration of 1 sec were delivered from pCLAMP 9.0 software. The activation curve was calculated by averaging current traces in response to 20 voltage ramps and dividing each point of the mean current by the unitary amplitude of each potential after the leakage component was corrected. The activation curves obtained before and after the addition of cilostazol were fitted with Boltzmann function of the form: relative N Po = nP/{1 + exp[-(V − a)/b]}, where nP is the maximal relative N Po, b is the slope factor of the voltage-dependent activation (i.e. the change in the potential required to result in an e-fold increase in the activation), and a is the voltage at which there is half-maximal activation.
The averaged results are presented as the mean values ± sem. The paired or unpaired t test and one-way ANOVA with the least significance difference method for multiple comparisons were used for the statistical evaluation of differences among the mean values. Differences between values were considered significant when P < 0.05.
Drugs and solutions
Cilostazol [Pletal, 6-(4-[1-cyclohexyl-1H-tetrazol-5-yl]butoxy)-3,4-dihydro-2-(1H)-quinolinone] and dequalinium dichloride were obtained from Tocris Cookson Ltd. (Bristol, UK), and sp-adenosine-3′,5′-cyclic monophosphorothioate (sp-cAMPS) was from Biomol (Plymouth Meeting, PA). Adenosine, OPCP, glibenclamide, and tetraethylammonium chloride were purchased from Sigma (St. Louis, MO). Tetrodotoxin was obtained from Alomone Labs (Jerusalem, Israel), adenosine deaminase was from Roche Molecular Biochemicals (Indianapolis, IN), and tetrandrine was from Aldrich Chemical (Milwaukee, WI). β-Bungarotoxin was a kind gift of Dr. Long-Sen Chang (Institute of Biomedical Sciences, National Sun Yat-sen University, Kaohsiung City, Taiwan). Magnolol was kindly provided by Dr. Chien-Chich Chen (National Institute of Chinese Medicine, Taipei City, Taiwan). All other chemicals were commercially available and of reagent grade. The twice-distilled water that had been deionized through a Millipore-Q system (Millipore, Bedford, MA) was used in all experiments.
The composition of normal Tyrode’s solution was 136.5 mm NaCl. 5.4 mm KCl, 1.8 mm CaCl2, 0.53 mm MgCl2, 5.5 mm glucose, and 5.5 mm HEPES-NaOH buffer, pH 7.4. To record K+ currents or membrane potential, the recording pipette was backfilled with a solution consisting of 140 mm KCl, 1 mm MgCl2, 3 mm Na2ATP, 0.1 mm Na2GTP, 0.1 mm EGTA, and 5 mm HEPES-KOH buffer (pH 7.2). To measure Ca2+ current, KCl inside the pipette solution was replaced with equimolar CsCl, and pH was adjusted to 7.2 with CsOH.
For single-channel current recordings, the high-K+ bathing solution contained 145 mm KCl, 0.53 MgCl2, and 5 mm HEPES-KOH buffer (pH 7.4), and the pipette solution contained 145 mm KCl, 2 mm MgCl2, and 5 mm HEPES-KOH buffer (pH 7.2). The value of free Ca2+ concentration was calculated assuming a dissociation constant for EGTA and Ca2+ (at pH 7.2) of 0.1 μm. The pipette solution was filtered on the day of use with a 0.22-μm pore size syringe filter (Millipore).
Results
Effect of cilostazol on IK(Ca) in GH3 cells
In the first series of experiments, the whole-cell configuration of the patch-clamp technique was used to investigate the effect of cilostazol on ion currents in these cells. Cells were bathed in normal Tyrode’s solution containing 1.8 mm CaCl2, and the pipette solution contained a low concentration (0.1 mm) of EGTA and 3 mm ATP. ATP (3 mm) contained in the pipette solution could be effective in suppressing ATP-sensitive K+ channels (22). To inactivate other types of voltage-dependent K+ currents, each cell was held at the level of 0 mV. As illustrated in Fig. 1, when the cell was held at 0 mV and different potentials ranging from +10 to +70 mV with 20-mV increments were applied, a family of large, noisy, outward currents were elicited. These outward currents have been previously identified as Ca2+-activated K+ currents [IK(Ca)]. Interestingly, within 1 min of exposing the cells to cilostazol (10 μm), the amplitude of outward currents was greatly increased throughout the entire range of voltage-clamp step. This stimulatory effect was readily reversed after the removal of cilostazol (Fig. 1). Figure 1B illustrates the averaged I-V relations for IK(Ca) in the absence and presence of cilostazol (10 μm).
Glibenclamide (10 μm), dequalinium dichloride (10 μm), and β-bungarotoxin (200 nm) were found to have no effect on the amplitude of IK(Ca) increased by cilostazol (10 μm), whereas paxilline (1 μm) and tetrandrine (10 μm) suppressed it significantly (Fig. 1C). Dequalinium dichloride is a blocker of small-conductance Ca2+-activated K+ channels (23), whereas paxilline and tetrandrine can inhibit the activity of BKCa channels (24, 25). Glibenclamide can reduce the activity of ATP-sensitive K+ channels in GH3 cells (22), whereas β-bungarostoxin is a blocker of voltage-dependent K+ current (26).
Effect of cilostazol on IK(Ca) in cells preincubated with AOPCP or adenosine deaminase
It has been shown that the effect of cilostazol was associated with its inhibition of adenosine uptake (1, 6, 27, 28). The binding of adenosine receptor by adenosine analogs was also found to inhibit prolactin release from GH3 cells (29). The effect of cilostazol on IK(Ca) was thus assessed in cells treated with AOPCP (100 μm) or adenosine deaminase (1 U/ml) for 5 h. AOPCP is an inhibitor of ecto-5′-nucleotidase, whereas adenosine deaminase can degrade adenosine to inosine. However, unexpectedly, in these GH3 cells preincubated with adenosine deaminase or AOPCP for 5 h, the stimulatory effect of cilostazol on the I-V relationship of IK(Ca) was little altered (Fig. 2). For example, in cells pretreated with adenosine deaminase (1 U/ml), cilostazol (10 μm) could increase the amplitude of IK(Ca) at the level of +50 mV from 281 ± 37 to 560 ± 71 pA (n = 8). There was no significant difference in the magnitude of cilostazol-stimulated effect on IK(Ca) between control cells and cells treated with AOPCP or adenosine deaminase. Therefore, these results suggest that the stimulatory effect of cilostazol on IK(Ca) is not attributed to its inhibition of adenosine uptake.
Effect of cilostazol on the activity of BKCa channels in pituitary GH3 cells
The results from our whole-cell experiments suggest that IK(Ca) may be K+ flux through the BKCa channel because the cilostazol-induced increase in IK(Ca) could be suppressed by paxilline. Therefore, to elucidate how cilostazol can act to affect IK(Ca), the effect of cilostazol on BKCa channels was further investigated. The single-channel recordings with inside-out configuration were performed in symmetrical K+ concentration (145 mm). The bath solution contained 0.1 μm Ca2+ and the potential was held at +60 mV. As shown in Fig. 3, the activity of BKCa channels can be readily observed in excised patches. An increase in channel activity could also be obtained in cell-attached patches when cells were exposed to ionomycin (10 μm) or squamocin (10 μm). These two agents have been previously reported to be Ca2+ ionophores (30). When cilostazol (10 μm) was applied to the intracellular surface of detached patches, the channel open probability was increased (Fig. 3). The open probability at +60 mV obtained in the absence of cilostazol was 0.014 ± 0.002 (n = 8). The application of cilostazol (10 μm) significantly increased the activity to 0.031 ± 0.025 (n = 7). When cilostazol was washed out, the channel activity returned to the control level. However, the single-channel amplitude remained unaltered in the presence of cilostazol (10 μm) (Fig. 3B).
The response to other known regulators of BKCa channels was also examined in GH3 cells (Fig. 3C). Magnolol (10 μm) applied to the intracellular surface of inside-out patch significantly increased the channel activity, whereas paxilline (1 μm) reduced it. Magnolol has been shown to be an opener of BKCa channels in smooth muscle cells (31). However, sp-cAMPS, an analog of cAMP, was found to have no effect on the channel open probability. In addition, in continued presence of paxilline (1 μm), a subsequent application of cilostazole (10 μm) was found to have no effect on BKCa channel activity. It thus appears that the cilostazol-stimulated increase in the channel activity shown here does not require the presence of cytoplasmic factors and is primarily due to the result of a direct binding to the inner surface of the channel.
The relationship between the concentration of cilostazol and the channel open probability is shown in Fig. 4A. This drug increased the channel open probability in a concentration-dependent manner with an EC50 value of 3.5 μm. At a concentration of 100 μm cilostazol fully increased the channel activity. The Hill coefficient was found to be 2.3, suggesting that there was a positive cooperativity for the stimulation of BKCa channels.
Effect of cilostazol on the activation curve of BKCa channels
Figure 4B shows the activation curve of BKCa channels in the absence and presence of cilostazol (10 μm). In these experiments, the activation curves were obtained with the aid of the voltage ramp protocols. The ramp pulses were delivered from +20 to +120 mV with a duration of 1 sec. The plots of open probability of BKCa channels as a function of membrane potential were fit with Boltzmann function as described in Materials and Methods. In control, nP = 0.34 ± 0.03, a = 85.1 ± 1.5 mV, and b = 10.9 ± 0.4 mV (n = 6), whereas in the presence of cilostazol (10 μm), nP = 0.88 ± 0.06, a = 76.5 ± 2.1 mV, and b = 11.1 ± 0.5 mV (n = 6). Thus, cilostazol not only caused a 2.6-fold increase in the maximal open probability of these channels, but it also shifted the activation curve to a lower membrane potential by approximately 10 mV. However, there was no significant effect on the slope (i.e. b value) of the activation curve in the presence of cilostazol.
Lack of effect of cilostazol on single-channel conductance of BKCa channels
Whether cilostazol affects single-channel conductance was also examined. To construct the plots of current amplitude as a function of membrane potential, the voltage ramp pulses from 0 to +50 mV with a duration of 1 sec were applied at a rate of 0.05 Hz. Figure 4C illustrates the I-V relationships of BKCa channels in the absence and presence of cilostazol (3 and 10 μm). The single-channel conductance calculated from the linear I-V relationship in control was 185 ± 7 pS (n = 8) with a reversal potential of 0 ± 2 mV (n = 8). The value for these channels in the control did not differ significantly from that obtained in the presence of 10 μm cilostazol (187 ± 7 pS, n = 8).
Effect of cilostazol on kinetic behavior of BKCa channels
The effect of cilostazol on the gating of these channels was analyzed because of its inability to modify single-channel conductance. In an excised patch of control cell, open and closed time histograms at the level of +30 mV can be fitted by a one or two exponential curve (Fig. 5). The time constant for the open-time histogram was 2.3 ± 0.5 msec, whereas those for the fast and slow components of the closed time histogram were 6.9 ± 1.2 and 85 ± 4 msec, respectively (n = 7). In inside-out patches, the application of cilostazol (10 μm) to the bath increased the time constant of the open state to 4.8 ± 0.6 msec and decreased the mean closed times to 3.2 ± 0.2 and 24 ± 2 msec (n = 7). Thus, the results showed that this drug produced an increase in channel open time and a decease in channel closed time. These changes may primarily explain its stimulatory effect on the channel activity.
Effect of cilostazol on voltage-dependent l-type Ca2+ current (ICa,L) in pituitary GH3 cells
IK(Ca) can be coupled with Ca2+ influx through plasmalemmal voltage-dependent Ca2+ channels (32). The increase in intracellular cAMP was found to enhance Ca2+ influx through l-type Ca2+ channels in pituitary tumor cells (33). A recent report also demonstrated that inhibitory action of cilostazol on catecholamine secretion could be due to its blockade of Ca2+ movement in adrenal chromaffin cells (7). For these reasons, we further investigated whether cilostazol exerts any effect on ICa,L present in these cells. These experiments were conducted with a Cs+-containing solution. As shown in Fig. 6, cilostazol (10 μm) was found to be have no effect on the amplitude of ICa,L. However, cilostazol at a concentration of 30 μm slightly suppressed ICa,L, but it did not modify the I-V relationship of ICa,L. Cilostazol (30 μm) significantly decreased the amplitude of ICa,L to 168 ± 15 pA from a control value of 198 ± 18 pA (n = 7). Therefore, cilostazol increased IK(Ca) in a manner unlikely to be linked to the increase in the amplitude of ICa,L.
Effect of cilostazol on spontaneous action potentials in pituitary GH3 cells
In another series of experiments, the effect of cilostazol on repetitive firing of action potentials was investigated. Cells were bathed in normal Tyrode’s solution containing 1.8 mm CaCl2. Current-clamp configuration was performed with a K+-containing pipette solution. When cells were exposed to cilostazol (10 μm), the membrane became hyperpolarized and the repetitive firing of action potentials was decreased (Fig. 7). Cilostazol (10 μm) significantly reduced the firing frequency from 1.19 ± 0.09 to 0.46 ± 0.04 Hz (n = 5). However, adenosine (100 μm) had little or no effect on spontaneous action potentials. Magnolol (10 μm), an opener of BKCa channels (31), reduced the frequency to 0.42 ± 0.03 Hz (n = 7), and paxilline (1 μm) increased it to 1.34 ± 0.11 Hz (n = 6). Thus, cilostazol can regulate the firing of action potentials in these cells. Such an effect appeared to be not associated with its inhibition of adenosine uptake and could be explained by its stimulation of BKCa channels.
Stimulatory effect of cilostazol on BKCa channels in pheochromocytoma PC12 cells
BKCa channels observed in GH3 cells may be different from those in other types of neuroendocrine cells. Cilostazol was reported to decrease acetylcholine-induced catecholamines secretion in adrenal chromaffin cells (7). In this regard, the effect of cilostazol in pheochromocytoma PC12 cells was further examined. PC12 cells were also bathed in symmetrical K+ concentration (145 mm). The holding potential was +60 mV and bath medium contained 0.1 μm Ca2+. Similar to that found in GH3 cells, the activity of BKCa channels was observed in these cells (Fig. 8). When paxilline (1 μm) was applied to the bath, the channel open probability was significantly reduced (data not shown); however, when the intracellular surface of excised patch was exposed to cilostazol (10 μm), the probability of channel openings was elevated. After addition of 10 μm cilostazol into the bath, the channel activity was significantly increased from 0.011 ± 0.001 to 0.026 ± 0.001 (n = 6). However, as shown in Fig. 8, the single-channel conductance between the absence and presence of cilostazol did not differ significantly (174 ± 4 pS vs. 176 ± 4 pS, n = 6). In addition, paxilline (1 μm) could suppress the increased channel activity caused by cilostazol or magnolol (Fig. 8C). Consistent with previous observations in GH3 cells, these data indicate that cilostazol can stimulate the activity of BKCa channels in pheochromocytoma PC12 cells.
Discussion
This study shows that cilostazol: 1) effectively increases IK(Ca) in GH3 cells; 2) increases the activity of BKCa channels; 3) causes a left shift in the midpoint for the activation curve of BKCa channels; 4) depresses the firing of action potentials; and 5) stimulates BKCa channel activity in pheochromocytoma PC12 cells. These findings suggest that the effects on ion channels may be one of the mechanisms underlying cilostazol-induced actions if similar results occur in neuroendocrine cells in vivo.
Several lines of evidence have been presented to indicate that cilostazol may block the uptake of adenosine into the cells (1, 6, 27, 28). The activation of adenosine receptors caused by adenosine agonists has been previously reported to stimulate prolactin secretion in GH3 cells (29). Therefore, the question arises as of whether the stimulatory effect of cilostazol on IK(Ca) observed in GH3 cells could be due to an increase in the level of endogenous adenosine when adenosine uptake was blocked by cilostazol. However, this idea is difficult to reconcile with the present results showing that in GH3 cells preincubated with adenosine deaminase or AOPCP, the stimulatory effect of cilostazol on IK(Ca) remained unaltered. Adenosine (100 μm) was found to have no effect on the firing of action potentials. Therefore, under our experimental conditions, the stimulation of IK(Ca) by cilostazol in GH3 cells was not associated with changes in the level of endogenous adenosine. It is unlikely that cilostazol stimulated IK(Ca) in a manner independent on its inhibition of adenosine uptake. Indeed, single-channel experiments with inside out configuration showing that cilostazol applied to the intracellular surface was found to enhance BKCa channel activity also exclude this possibility.
The present results also provide the evidence that the cilostazol-induced increase in IK(Ca) does not depend on the increased availability of intracellular Ca2+ resulting from the enhanced Ca2+ influx through voltage-dependent Ca2+ channels because cilostazol was not found to increase the amplitude of ICa,L. These observations are compatible with a previous report showing the ability of this drug to suppress the tonic contraction induced by high K+ solution in vascular smooth muscle (34). In addition, because the increased level of intracellular cAMP was found to increase Ca2+ influx through l-type Ca2+ channels in pituitary tumor (GH4C1) cells (33), the inhibition of ICa,L caused by cilostazol at a concentration greater than 30 μm appears to be direct and unlikely to be linked to its inhibition of phosphodiesterase.
The results from single-channel experiments with excised patches led to the interpretation that the stimulatory effect of cilostazol on BKCa channel activity does not require cytosolic diffusible substances (e.g. cAMP) inside the cell and is primarily due to the result of direct binding to the inner surface of the channel. Inability of sp-cAMPS, an analog of cAMP, to stimulate BKCa channels also supports this notion. However, the binding site for cilostazol could not be located intracellular to the central pore cavity because of no modification of single-channel conductance in the presence of cilostazol. This drug may thus bind to the BKCa channel at the sites that are important for channel gating, given that there was an increase in mean open time and a decrease in mean closed time in the presence of cilostazol.
Cilostazol could increase the activity of BKCa channels in a voltage-dependent fashion possibly by acting at a site that is accessible from the intracellular side of the channel. Therefore, it seems reasonable to assume that the sensitivity of neurons or neuroendocrine cells to this drug would depend on the level of resting membrane potential, the firing of action potentials, and the concentration of cilostazol used, if the cilostazol action in vivo is the same as those in GH3 cells shown herein. Indeed, we also found that, like magnolol (31), it was effective in suppressing the repetitive firing of action potentials in GH3 cells.
The EC50 value of cilostazol required for the stimulation of BKCa channels was 3.5 μm, a value that is similar to that required for the inhibition of adenosine uptake into the cells (6, 27) but is lower than that used to inhibit platelet aggregation (35). Plasma concentrations of cilostazol in humans are 1–5 μm (36). Because its effect shown here may occur at a concentration achievable in humans (36), the present findings suggest that the BKCa channel may be a target for the action of cilostazol. Regardless of the mechanism of its actions, the present results should also be noted with caution in relation to its increasing use as an inhibitor of type 3 phosphodiesterase or adenosine uptake (3, 6, 27, 28, 37–39).
The concentration-response curve of channel activation by cilostazol had a Hill coefficient of 2.3, suggesting that more than one cilostazol molecule bind to the channel, although the stoichiometry of cilostazol binding to the channel is currently unknown. It is possible that cilostazol has a unique structure that interacts with the BKCa channel to increase the amplitude of IK(Ca) in GH3 cells. Interestingly, cilostazol is noted to have a structure that bears resemblance to chlorzoxazone, riluzole, and the benzimidazolones, the structurally related compounds of which have been known to be BKCa channel openers (40). A recent report showed that hemin could bind to the heme-binding amino acid sequence motif found in Slo BK and native BKCa channels with a high affinity, thus leading to a decrease in channel activity (41). It will be of interest to explore whether cilostazol can interact with this similar motif to affect the activity of BKCa channels.
The results of this study show a significant increase of BKCa channel activity by cilostazol in GH3 and PC12 cells. This effect is presumably not mediated by its inhibitory effects on the uptake of adenosine or on the activity of phosphodiesterases. It is possible that cilostazol-induced effects on neurons or neuroendocrine cells are partly, if not entirely, attributed to a direct simulation of the BKCa channels. Furthermore, these results suggest that in addition to the increase in intracellular cAMP, the direct effect of cilostazol on the stimulation of BKCa channels may have important implications in understanding its relaxation of vascular smooth muscle (42, 43). In fact, this drug has been used for the management of chronic peripheral arterial occlusive disease (14, 27, 42–45).
Recent observations showed that the BKCa channel activity might play a role in time- and cell-dependent modulation of physiological outputs in anterior pituitary cells (46). The increase of these channel caused by cilostazol might reduce hormonal secretion. The activity of these channels was found to be activated in the course of steroidogenesis induced by human chorionic gonadotropin (47). Therefore, it will be of interest to determine to what extent cilostazol affects the steroidogenesis in the ovary.
Abbreviations:
- AOPCP,
α,β-Methylene-ADP;
- BKCa,
large-conductance Ca2+-activated K+;
- ICa, L,
voltage-dependent l-type Ca2+ current;
- IK(Ca),
Ca2+-activated K+ current;
- I-V,
current-voltage;
- sp-cAMPS,
sp-adenosine-3′,5′-cyclic monophosphorothioate.
Acknowledgements
The authors would like to thank Su-Rong Yang for technical assistance in the preparation of cultured cells.
This work was supported by grants from the National Science Council (NSC-91-2320B-006-106 and NSC-92-2320B-006-041), Taiwan.