INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IT IS WELL RECOGNIZED that the proliferation of vascular smooth muscle cells (VSMCs) is a key event in the pathogenesis of various vascular diseases, including atherosclerosis and postangioplasty restenosis (32). Among the factors that have been shown to be involved in the control of VSMC growth, thrombin appears to play an important role (24). The mechanisms whereby thrombin induced VSMC growth/proliferation have been documented (6, 24), and one of the most immediate signaling events has been demonstrated to be thrombin-evoked variations of intracellular free Ca²⁺ concentration ([Ca²⁺]_i) (1). [Ca²⁺]_i variations play a predominant role in VSMC proliferation (27, 36), and in this cell type the voltage-activated Ca²⁺ channels considerably participate in intracellular Ca²⁺ homeostasis (see Ref. 15 for review).

There is compelling evidence that Ca²⁺ channel blockers (CCBs) inhibit VSMC growth/proliferation (18), but the mechanisms underlying this inhibitory effect of CCBs remain to be determined. Recent data are consistent with the idea that CCBs interact with targets other than the L-type Ca²⁺ channel (4, 31). Of the various CCBs, the L-type Ca²⁺ channel antagonist amlodipine is of particular interest, because this dihydropyridine derivative endowed with antihypertensive and antiatherosclerotic properties exhibits a selectivity for the vasculature relative to the myocardium (7, 23, 29). Moreover, recent results that described the inhibitory effect of amlodipine on thrombin-induced proliferation of VSMCs from rat aortas (39) suggested that, in addition to its L-type Ca²⁺ channel inhibitory effect, amlodipine might inhibit other intracellular signaling pathways involved in VSMC proliferation.

This prompted us to investigate the influence of amlodipine on thrombin-elicited Ca²⁺ movements in rat aortic VSMCs compared with that of other CCBs such as isradipine, diltiazem, and verapamil.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reagents. Cell culture materials and media were obtained from Costar and Life Technologies, respectively; FCS was from Boehringer-Mannheim; thrombin and amlodipine were from Roche (Basel, Switzerland) and Pfizer (Orsay, France), respectively; and fura 2-AM was from Molecular Probes (Eugene, OR). All other chemicals, including the CCBs isradipine, diltiazem, and verapamil, were obtained from Sigma Chemical (St. Louis, MO).

Cells and culture conditions. VSMCs were isolated from the thoracic aortas of 10-wk-old Wistar-Kyoto rats by the explant technique of Ross (33), as previously described (13). They were grown in DMEM supplemented with 8 mM HEPES buffer, 100 U/ml penicillin, 100 µg/ml streptomycin, and 10% (vol/vol) FCS. Cells were subcultured weekly. Cells between passages 4 and 13 were used. VSMCs grew in a characteristic "hill-and-valley" pattern, and immunostaining procedures showed that cells were reactive to anti--actin antibodies. For each experiment, VSMCs were cultured for 3 days in 10% FCS and then made quiescent by serum deprivation for 2 days. Cell viability was assessed by the measurement of lactate dehydrogenase activity released from damaged cells by use of the cytotoxic detection kit (Boehringer-Mannheim); it was not affected by the various experimental conditions used in these investigations.

[Ca²+]_i measurement. [Ca²⁺]_i was determined by using the fluorescent indicator fura 2. VSMCs were allowed to attach and grow on glass coverslips in 10% FCS-containing medium. Then VSMCs were rendered quiescent, as described above. On the day of the experiment, the coverslips were washed twice with modified Hanks' buffered saline solution (in mM: 135 NaCl, 5.4 KCl, 44 NaHCO₃, 0.9 NaH₂PO₄, 10 HEPES, 0.8 MgSO₄, and 5 glucose), pH 7.4, and kept at 37°C before incubation for 40 min in a humidified incubator at 37°C with fura 2-AM (3 µM in DMSO containing 0.025% Pluronic F-127) for fluoroprobe loading. DMSO concentration was <0.1% and has no effect on [Ca²⁺]_i. Before each experiment, coverslips were washed with the modified Hanks' buffered saline solution for extracellular dye removal. Then they were inserted in a cuvette containing 3 ml of Ca²⁺-containing medium (i.e., the modified Hanks' buffered saline solution with 1 mM CaCl₂) or Ca²⁺-free medium [i.e., a modified EGTA-containing Hanks' buffered saline solution (17)] and the various agents to be tested. Fluorescence was monitored at 510 nm (excitations were at 340 and 380 nm) with a dual-excitation-wavelength spectrofluorometer (SPEX fluorolog, Jobin-Yvon, Longjumeau, France) equipped with a chamber thermostatically controlled at 37°C. After a stable basal value was monitored, cells were exposed to the agent(s) to be tested. Autofluorescence from unloaded cells and test agent(s) was subtracted from the measured value. [Ca²⁺]_i was calculated using the equation of Grynkiewicz et al. (11).

Statistical analysis. Values are means ± SE of n experiments. Each experiment involved one distinct culture. Multiple comparisons and dose-response and time-dependence effects were examined by one-way ANOVA and post hoc Fisher's test. Unless otherwise stated, tests of significance were performed using the unpaired Student's t-test.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Assessment of thrombin-induced [Ca²+]_i responses. Preliminary experiments indicated that, irrespective of the presence of Ca²⁺ in the solution, [Ca²⁺]_i responses to various doses of thrombin reached a maximum for thrombin concentration of 0.5-1 U/ml (results not shown). Therefore, further experiments were carried out with 1 U/ml thrombin. Figure 1A shows that thrombin-induced [Ca²⁺]_i variations considerably depended on the presence of Ca²⁺ in the external medium. In Ca²⁺-containing medium, thrombin elicited an initial transient peak followed by a sustained phase, whereas in Ca²⁺-free medium, the [Ca²⁺]_i response consisted only in the transient peak (Fig. 1A, traces a and b). A quantitative analysis of data revealed that, irrespective of the presence of Ca²⁺ in the external medium, [Ca²⁺]_i values at basal (before stimulation) and at peak levels did not differ significantly (Table 1).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Thrombin-induced intracellular free Ca2+ concentration ([Ca2+]i) responses in cultured aortic vascular smooth muscle cells (VSMCs) under various conditions. A: VSMCs were stimulated with 1 U/ml thrombin (arrow) in a Ca2+-containing medium (trace a) or in Ca2+-free medium (trace b). B: in Ca2+-containing medium, VSMCs were pretreated for 20 min with the diluent (trace a) or with 1 µM thapsigargin (Tg, trace b) before the addition of 1 U/ml thrombin (arrow). C: VSMCs were placed in a Ca2+-free medium and treated with 1 µM thapsigargin (left arrow) and then with 1 U/ml thrombin (right arrow). Each trace is representative of 4 experiments.

Effects of amlodipine on voltage-operated Ca²+ channels. In a first series of experiments, we ascertained that cultured VSMCs used in our investigations expressed functional dihydropyridine-sensitive voltage-operated Ca²⁺ channels. To do so, cells were pretreated with amlodipine or its diluent (for control) and then depolarized by the addition of a high (80 mM) KCl concentration to the medium. Under these experimental conditions, addition of KCl to control cells produced a transient elevation in [Ca²⁺]_i ([Ca²⁺]_i ~300 nM) followed by a sustained increase (Fig. 2). When VSMCs were pretreated with 10 nM amlodipine, the Ca²⁺ response to KCl addition decreased in a time-dependent manner. After 10 and 60 min of pretreatment with the drug, the KCl-induced increase in [Ca²⁺]_i was 95 ± 3 and 40 ± 6% of controls (n = 3-6; not shown); preincubation of VSMCs for 2 h with amlodipine abolished the Ca²⁺ response to KCl (Fig. 2), in agreement with preceding reports (23, 38). Thus, unless otherwise specified, in the following experiments, VSMCs were pretreated with amlodipine for 2 h.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Inhibition by amlodipine of voltage-dependent Ca2+ influx in cultured VSMCs. In Ca2+-containing medium, quiescent VSMCs were pretreated for 2 h with 10 nM amlodipine (trace b) or its diluent (control, trace a) before stimulation by 80 mM KCl to induce depolarization.

Effect of amlodipine on thrombin-induced Ca²+ movements. To determine the influence of amlodipine on thrombin-triggered Ca²⁺ responses, VSMCs were pretreated in Ca²⁺-containing medium with dihydropyridine for 2 h before thrombin stimulation. Then [Ca²⁺]_i was determined by placement of the cells in the fluorometer in Ca²⁺-containing medium or the Ca²⁺-free medium and the various agents to be tested. Under these conditions, CCBs such as amlodipine, isradipine, diltiazem, or verapamil (up to 100 nM) did not significantly modify the basal [Ca²⁺]_i (results not shown). After thrombin stimulation of VSMCs in Ca²⁺-containing medium, only the transient [Ca²⁺]_i elevation was diminished by amlodipine in a concentration-dependent manner (Fig. 3A, Table 2). Amlodipine (10-1,000 nM) did not significantly affect thrombin-elicited Ca²⁺ influx, since after subtraction of basal [Ca²⁺]_i, [Ca²⁺]_i values obtained at plateau (measured 3 min after thrombin stimulation) were 140 ± 12, 146 ± 36, 180 ± 33, and 112 ± 23 (SE) nM (n = 5-9, not significant) for amlodipine additions of 0, 10, 100, and 1,000 nM, respectively. Under these experimental conditions, 100 nM isradipine did not significantly influence thrombin-induced Ca²⁺ responses (Table 2); diltiazem and verapamil behaved as isradipine (not shown). Experiments carried out in Ca²⁺-free medium (Fig. 3B, Table 2) confirmed that although 100 nM isradipine was without effect, amlodipine inhibited thrombin-induced Ca²⁺ mobilization from internal stores in a concentration-dependent manner; they also revealed that as low as 10 nM amlodipine significantly reduced Ca²⁺ mobilization by ~40-45% without modifying basal and plateau [Ca²⁺]_i values.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effects of various concentrations of amlodipine (amlo) on thrombin-induced Ca2+ responses. A: in Ca2+-containing medium, VSMCs were pretreated for 2 h with amlodipine or the solvent (control) before stimulation by 1 U/ml thrombin (arrow). B: VSMCs were pretreated with amlodipine or the solvent (control) for 2 h in a Ca2+-containing medium and then placed in a Ca2+-free medium before stimulation by 1 U/ml thrombin (arrow). In A and B, traces a, b, and c correspond to pretreatment of cells with the solvent, 100 nM amlodipine, and 1,000 nM amlodipine, respectively, and are representative of 4 experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Effect of amlodipine on the thapsigargin-sensitive Ca2+ stores. VSMCs were pretreated with the diluent or 100 nM amlodipine (traces a and b, respectively) for 2 h in Ca2+-containing medium. Then cells were placed in a Ca2+-free medium before stimulation by 1 U/ml thrombin at time 0 and by 1 µM thapsigargin. Traces are representative of 4 experiments.

Effect of amlodipine on thapsigargin-induced [Ca²+]_i increase. Because amlodipine was capable of inhibiting thrombin-induced Ca²⁺ mobilization from internal stores and because such a mobilization appeared to be thapsigargin sensitive, we investigated whether amlodipine could directly affect thapsigargin-induced Ca²⁺ movements. When VSMCs were treated with 1 µM thapsigargin in Ca²⁺-containing medium (Fig. 5A), we observed a slow rise in [Ca²⁺]_i that reached a plateau ([Ca²⁺]i = 540 ± 35 nM, n = 12) within 2-3 min. This sustained Ca²⁺ response depended on the presence of extracellular Ca²⁺, since when experiments were performed in Ca²⁺-free medium, only a small and transient [Ca²⁺]_i increase could be observed (Fig. 5B). Pretreatment of VSMCs with amlodipine for 2 h blunted the thapsigargin-evoked [Ca²⁺]_i increase in a concentration-dependent manner, irrespective of the presence of Ca²⁺ in the extracellular medium (Fig. 5, Table 3). Thus as low as 1 nM amlodipine significantly reduced (by ~10-15%) the thapsigargin-induced Ca²⁺ response (Table 3). In the absence of Ca²⁺ in the extracellular medium, experiments with amlodipine at <100 nM could not be interpreted because of the weakness of the signal.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Effect of amlodipine on thapsigargin-induced Ca2+ responses. VSMCs were pretreated with the diluent or 1, 10, 100, and 1,000 nM amlodipine for 2 h in Ca2+-containing medium before stimulation by 1 µM thapsigargin in the same medium (A) or in Ca2+-free medium (B). Traces are representative of 4 experiments.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effect of amlodipine on store-operated Ca2+ entry. Store-operated Ca2+ entry was induced by depletion of intracellular Ca2+ stores with thapsigargin in Ca2+-free medium and then addition of 1 mM CaCl2. VSMCs were pretreated with the diluent or 100 nM amlodipine (traces a and b, respectively) for 20 min in Ca2+-containing medium. Cells were then placed in Ca2+-free medium before addition of 1 µM thapsigargin. The transmembrane Ca2+ gradient was reestablished 3 min later by the addition of 1 mM CaCl2. For trace c, VSMCs received neither amlodipine nor thapsigargin but did receive their solvents. Traces are representative of 3 experiments.

Kinetics of the inhibition by amlodipine of thrombin- and thapsigargin-induced [Ca²+]_i increase. In the experiments presented above, amlodipine was always added 2 h before VSMC stimulation by thrombin or thapsigargin. We therefore studied the time dependence of amlodipine pretreatment on thrombin- or thapsigargin-induced [Ca²⁺]_i increases. The kinetics of action of amlodipine on thrombin-induced intracellular Ca²⁺ mobilization were observed in Ca²⁺-free medium (Fig. 7A). The kinetics of action of amlodipine on the thapsigargin-induced increase in Ca²⁺ were studied in Ca²⁺-containing medium (Fig. 7B), because the amplitude of thapsigargin-induced [Ca²⁺]_i increase in Ca²⁺-free medium was too low (Fig. 5, Table 3). Nevertheless, Fig. 7 shows that the kinetics of action of amlodipine (100 nM) were similar, irrespective of stimulus. Thrombin- and thapsigargin-induced [Ca²⁺]_i increases were not significantly modified by the addition of amlodipine 10 min before VSMC stimulation but were almost maximally inhibited by pretreatment of VSMCs for 20 min with amlodipine.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Kinetics of the effect of amlodipine on thrombin- and thapsigargin-induced Ca2+ responses. For the effects on thrombin (A) and thapsigargin (B), the experimental protocols were as described in the legends of Figs. 3B and 5A, respectively, except VSMCs were pretreated with amlodipine for the time indicated. Results (means ± SE of 4 experiments) are expressed as percent inhibition of Ca2+ increase that was determined at peak (A) and at plateau (B).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The results reported here show that thrombin-induced Ca²⁺ responses clearly consisted of two phases (Fig. 1): 1) a transient Ca²⁺ increase that corresponded to the mobilization of thapsigargin-sensitive intracellular Ca²⁺ stores, i.e., those Ca²⁺ ions that were sequestered in the sarcoplasmic reticulum and could be released by inositol trisphosphate (IP₃), and 2) a sustained Ca²⁺ increase that corresponded to an influx that was not mediated by the voltage-operated channels (Figs. 1 and 3; see below). This is in agreement with previous reports (2, 26, 30, 41). Figure 2 also demonstrates that, under our experimental conditions, VSMCs expressed functional L-type Ca²⁺ channels, as expected (10), and that amlodipine behaves as a CCB.

Treatment of VSMCs with amlodipine, but not with isradipine, before their stimulation with thrombin, affected in a concentration-dependent manner only that Ca²⁺ response corresponding to mobilization from internal stores (Fig. 3, Table 2). Treatment of VSMCs with amlodipine, isradipine, diltiazem, or verapamil did not affect thrombin-induced Ca²⁺ influx from external medium (Fig. 3A; results not shown). Amlodipine behaved similarly in VSMCs isolated from human internal mammary artery and stimulated by thrombin (results not shown). However, amlodipine has no effect in vasopressin-stimulated A7r5 cells (16). The intracellular Ca²⁺ pool mobilized by thrombin and sensitive to amlodipine in a concentration-dependent manner has been identified as a thapsigargin-sensitive pool (Fig. 1, B and C, and Figs. 4-6; see below). This is to our knowledge the first time that an L-type CCB has been shown to be capable of altering the Ca²⁺ responses of VSMCs to thrombin. Nevertheless, neither amlodipine (Fig. 3A), isradipine, diltiazem, nor verapamil could inhibit thrombin-induced Ca²⁺ influx. This indicated that, although VSMCs did express functional L-type Ca²⁺ channels, thrombin-induced Ca²⁺ influx was not mediated by such voltage-operated channels, consistent with results reported 1) in neonatal rat VSMCs, where the Ca²⁺ responses to thrombin were not affected by nicardipine or (+)-BAY K 8644 inhibitor (26), and 2) in the dog coronary artery, where the contractile response to thrombin was not affected by CCBs (12).

In Ca²⁺-free medium, the addition of thapsigargin to VSMCs that had previously been stimulated by a maximal dose of thrombin induced a Ca²⁺ response unless the cells had been pretreated with amlodipine (Fig. 4). The additional Ca²⁺ fraction mobilized by thapsigargin may be an IP₃-insensitive Ca²⁺ pool or a portion of the IP₃-sensitive pool that is not mobilized by a maximum concentration of thrombin, as previously observed with endothelin (42). Our finding is also consistent with the recent observation that IP₃-induced Ca²⁺ release reflects partial emptying of intracellular stores in A7r5 VSMCs (25). That the pretreatment of VSMCs with amlodipine could completely blunt the effect of thapsigargin (Fig. 4) 1) demonstrated that the effects of amlodipine on thrombin-induced Ca²⁺ responses could not be accounted for by an inhibition of thrombin binding to its membrane receptors and 2) indicated that amlodipine interfered with thapsigargin in a manner that prevents the Ca²⁺ release.

Our results also showed unambiguously that pretreatment of cells with amlodipine, but not with isradipine, diltiazem, or verapamil, inhibited in a concentration-dependent manner thapsigargin-induced Ca²⁺ responses, irrespective of the presence of Ca²⁺ in the external medium (Fig. 5, Table 3; data not shown). This confirmed the preceding observation and, in addition, demonstrated that amlodipine could not exert its effect through the inhibition of IP₃ binding to its receptors. Moreover, the kinetics of inhibition by amlodipine of thrombin- and thapsigargin-induced Ca²⁺ responses were similar (Fig. 7). Furthermore, the comparison of kinetics of inhibition by amlodipine of voltage-dependent Ca²⁺ influx from the external medium and of thrombin/thapsigargin-induced Ca²⁺ mobilization from internal stores clearly indicated that these effects of amlodipine are dissociated from each other.

Taken together, our results therefore suggest that the mechanism of action of amlodipine in thrombin- and thapsigargin-induced Ca²⁺ responses in VSMCs is similar and does not involve L-type Ca²⁺ channel blockade. They led us to hypothesize that to interfere with thrombin- and thapsigargin-induced Ca²⁺ mobilization from internal stores, amlodipine interacts directly or indirectly with the enzyme activities that are closely involved in the control of such a mobilization, namely, the SERCAs. Further observations are reported that support this hypothesis. 1) Amlodipine also inhibited Ca²⁺ mobilization induced by BHQ, a compound that, like thapsigargin, is known to mobilize intracellular Ca²⁺ stored in the sarcoplasmic reticulum, irrespective of the production of IP₃, by inhibiting the SERCAs (19, 40, 42). 2) The so-called store-operated Ca²⁺ entry that participates in the sustained Ca²⁺ response observed in thapsigargin-stimulated VSMCs and that is known to be closely linked to the filling state of internal Ca²⁺ stores (14) was markedly reduced in amlodipine-pretreated VSMCs (Figs. 5A and 6). The store-operated Ca²⁺ entry in VSMCs has been reported to be insensitive to nifedipine and nicardipine (42, 43), consistent with our observations with isradipine, diltiazem, and verapamil. At variance, observations in guinea pig intestinal smooth muscle showed that the thapsigargin-evoked Ca²⁺ increase and contraction were blocked by CCBs such as nimodipine and D-600 (5). On the other hand, an interaction between thapsigargin and the voltage-dependent Ca²⁺ channel has been reported in the A7r5 cell line and in adrenal glomerulosa cells (3, 34). Some specific structural characteristics of amlodipine (vs. other CCBs) have been envisaged to account for its antioxidant property (22). One may hence speculate that such a structure might also confer on amlodipine a particular physicochemical interaction with the sarcoplasmic reticulum. Nevertheless, the mechanism of action of amlodipine and the modulation of SERCA activity by amlodipine remain to be proved.

In conclusion, the agent amlodipine was demonstrated to inhibit thrombin- and thapsigargin-induced Ca²⁺ responses from intracellular Ca²⁺ stores and thapsigargin-induced Ca²⁺ influx from the external medium. The effects of amlodipine on thrombin- and thapsigargin-evoked Ca²⁺ responses appeared to be specific, since they could not be reproduced with isradipine, diltiazem, or verapamil. In rat aortic VSMCs, this is the first time that a CCB has been reported to alter Ca²⁺ responses of VSMCs to thapsigargin and thrombin. Such findings might be related to the inhibitory potency of amlodipine on thrombin-induced VSMC growth (39), since SERCA inhibitors, including thapsigargin, cyclopiazonic acid, and BHQ, have been reported to inhibit serum-induced VSMC and DDT₁MF-2 cell proliferation (8, 20, 36, 37), in agreement with the concept that the intracellular Ca²⁺ pool content is linked to control of cell growth (9, 21, 36). In this respect, we have observed that amlodipine and thapsigargin similarly inhibited thrombin-induced DNA synthesis in cultured rat aortic VSMCs (results not shown). Furthermore, our findings might be of clinical relevance, since 10 nM amlodipine has been reported to correspond to the tissue concentration achieved after antihypertensive treatment of rats (28, 35). One may therefore envisage that amlodipine effects on Ca²⁺ movements participate in the antiatherogenic properties of this drug.