Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

PTX isolated from a zoanthid of the genus Palythoa, is the most potent animal toxin known. Its lethality is profoundly associated with its action on vascular smooth muscles. Especially coronary artery is highly sensitive to PTX and its intense vasoconstriction leads to cardiac depression through global ischemia, resulting in acute death (Kaul et al., 1974; Ito et al., 1976; 1982). PTX depolarizes the plasma membrane and induces or enhances a contraction of striated and smooth muscles (Deguchi et al., 1976; Ito et al., 1976, 1979). ⁴⁵Ca⁺⁺ flux experiments showed that the contraction of vascular smooth muscle accompanied an increase in Ca⁺⁺ influx from extracellular space (Ito et al., 1977; Ozaki et al., 1983). The sensitivity of PTX-induced contraction to verapamil indicated that voltage-dependent Ca⁺⁺ channels were at least partly involved in the Ca⁺⁺ mobilization. However, the PTX-induced increase in [Ca⁺⁺]_i cannot be explained solely by Ca⁺⁺ influx through the Ca⁺⁺ channels because a component resistant to organic Ca⁺⁺ channel blockers exists in the PTX-induced Ca⁺⁺ mobilization (Ozaki et al., 1983; Robinson et al., 1992).

A mechanism for the PTX-induced rise in [Ca⁺⁺]_i which is independent of voltage-dependent Ca⁺⁺ channels has not yet been elucidated. It is known that in many types of cell PTX induces a cation-permeable channel (Ikeda et al., 1988; Van Rentherghen and Frelin, 1993), which is assumed to be located around Na⁺,K⁺-ATPase (Habermann, 1989). However, several studies showed that Ca⁺⁺ did not permeate the channel (Ikeda et al., 1988; Van Renterghem and Frelin, 1993; Ishii et al., 1997). Therefore, the Ca⁺⁺ entry through the channel is not responsible for the PTX-induced Ca⁺⁺ mobilization. Under physiological conditions, PTX permits the passage of Na⁺ through the channel, leading to a depolarization (Dubois and Cohen, 1977; Ito et al., 1985; Ikeda et al., 1988). Therefore, an increase in intracellular Na⁺ may be related to an increase in [Ca⁺⁺]_i. This issue remains to be examined in details.

Coronary artery is an important tissue in respect that it is related to ischemic heart diseases. The high sensitivity to PTX of this artery indicates that PTX could be a tool to make a model of coronary vasospasm or cardiac ischemia. The sensitivity also suggests that coronary artery may have a special feature in mechanisms for Ca⁺⁺ mobilization. In this study we aimed to clarify the detailed mechanism of PTX-induced Ca⁺⁺ mobilization. Using a Ca⁺⁺ indicator fura-PE3, we observed changes in [Ca⁺⁺]_i, tension and membrane potential caused by PTX in porcine coronary artery strips.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Pig hearts were obtained at a local abattoir and delivered to the laboratory in oxygenated PSS (table 1, <10°C) within 30 min. The left circumflex and the anterior descending branch of the coronary artery were dissected from surrounding tissues and cut into helical strips 3 to 4 mm in width and 12 to 15 mm in length. The endothelium was removed by gently rubbing the intimal surface with a cotton swab.

The tension and [Ca⁺⁺]_i were measured simultaneously as described by Sato et al. (1988) mainly in strips from the circumflex. We confirmed that the results from the descending branch were similar to those from the circumflex. The strips were loaded with 2.5 µM acetoxymethyl ester of fura-PE3 (fura-PE3/AM) dissolved in PSS containing 0.03% cremophor EL for 4 to 6 hr at 37°C. After loading, one end of the strip was fixed in an organ bath (37°C) constructed in a fluorimeter (CAF-100, JASCO, Tokyo, Japan) and the other end was connected to a force-displacement transducer (TB-611T, Nihon-Kohden, Tokyo, Japan). The basal tension was adjusted to 10 mN. We did not calculate the absolute [Ca⁺⁺]_i in coronary arteries because the dissociation constant of fura-PE3 for Ca⁺⁺ may be different in the cytosol from that obtained in vitro (Karaki, 1989) and the indicator slowly leaks out from cells. Instead, we used R340/380 (the ratio of fluorescence at emission of 500 nm after excitation at wavelength of 340 nm to that after excitation at 380 nm) as an index of [Ca⁺⁺]_i and expressed the magnitude of change in [Ca⁺⁺]_i as a relative value of R340/380 caused by high K⁺ solution (table 1), which had been observed before the application of PTX.

The membrane potential was measured in the left descending branch of the coronary artery using a microelectrode technique. The deendothelialized artery strip was placed with the adventitial side down in a perfusion bath and was superfused with PSS at a rate of 3 ml/min. A microelectrode having a tip resistance of 40 to 60 M when filled with 3 M KCl was impaled into a cell from the intimal side. The voltage was fed to a microelectrode amplifier (MEZ-7200, Nihon-Kohden) and recorded on a pen-writing recorder (Recti-Horiz, San-ei, Tokyo Japan).

The composition of solutions used in this study is listed in table 1. All solutions were adjusted to pH 7.3 to 7.4 and gassed with 95% O₂ and 5% CO₂. The following drugs were used: PTX, isolated from Palythoa tuberculosa, fura-PE3/AM (TEFLAB, Austin, TX), verapamil (Eisai, Tokyo, Japan), nicardipine (Sigma Chemical Co., St. Louis, MO), prazosin (Sigma), ryanodine (S.B. Penick, Lyndhurst, NJ) and caffeine (Nacarai Tesque, Kyoto, Japan). PTX was dissolved in distilled water with 0.1% bovine serum albumin and kept frozen as a 100 µM stock solution. The final dilution was made immediately before use.

Summarized data are expressed as mean ± S.E.M. Statistical difference was determined by Student's t test at the level of P < .05 for nonpaired samples.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Dose-dependent effects of PTX on [Ca⁺⁺]_i and tension. PTX (1 pM-10 nM) induced a rise in [Ca⁺⁺]_i and a contraction in a concentration-dependent manner in porcine coronary arteries as summarized in figure 1. The threshold concentration required to increase [Ca⁺⁺]_i was 100 fM, although that to induce a contraction was 1 pM. With lower concentrations of PTX (1-100 pM), [Ca⁺⁺]_i and the tension rose slowly and remained constant after 15 min. With the highest concentration (10 nM), [Ca⁺⁺]_i and the tension gradually declined after peaking at 3 and 10 min, respectively. The [Ca⁺⁺]_i level reached by PTX at more than 100 pM was higher than that due to 65.4 mM KCl, although the tension development was higher at only 10 nM than 65.4 mM KCl-induced one. Subsequent addition of 10 µM verapamil or 1 µM nicardipine during the sustained increase in [Ca⁺⁺]_i due to PTX reduced [Ca⁺⁺]_i and the contraction significantly but slightly. When the medium was replaced with Ca⁺⁺-free, EGTA (0.3 mM) PSS after the verapamil effect reached a steady state, [Ca⁺⁺]_i and the tension decreased very close to the resting levels in 8 to 10 min (fig. 1).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   The increase in [Ca++]i (R340/380, a) and the contraction (b) induced by PTX in porcine coronary arteries. A single dose of PTX at 1 pM (; n = 5), 10 pM (, n = 6), 100 pM (, n = 6), 1 nM (, n = 6) or 10 nM (, n = 8) was applied to each muscle. After responses to PTX were observed for 20 min, 10 µM verapamil (Verap) were added and 10 min later the medium was switched to Ca++-free, EGTA (0.3 mM) PSS (0-Ca++). R340/380 and contraction are expressed as % responses to high K+-solution that had been observed before PTX. Each point represents mean ± S.E.M.

Verapamil-resistant component in high K⁺- and PTX-induced contraction. In the following experiments, we analyzed the verapamil-resistant Ca⁺⁺ mobilization caused by PTX. First, we verified the effectiveness of verapamil on high K⁺-induced Ca⁺⁺ mobilization. Pretreatment with 10 µM verapamil for 3 min inhibited the responses of [Ca⁺⁺]_i and the tension to high K⁺-solution to 42.4 ± 5.7 and 32.7 ± 3.2% (n = 8), respectively (fig. 2a). When verapamil was pretreated for 15 min, the inhibition was the same ([Ca⁺⁺]_i and the contraction were inhibited to 46.6 ± 6.5 and 28.5 ± 8.9%, respectively, n = 4). Effect of verapamil at 10 µM was maximal because inhibition by 30 µM verapamil was similar to 10 µM verapamil (with 30 µM verapamil [Ca⁺⁺]_i and contraction were decreased to 41.3 ± 5.6 and 29.3 ± 2.2%, respectively, n = 6). Therefore, we thought that 3 min pretreatment with 10 µM verapamil exerted a consistent effect. Prazosin (1 µM) did not affect the verapamil-resistant increase in [Ca⁺⁺]_i and contraction, excluding a possibility that endogenously released catecholamines were involved in the contraction. To see the nature of a component independent of L-type Ca⁺⁺ channels in the high K⁺-induced Ca⁺⁺ mobilization, the external medium was replaced with Li⁺ solution (Na⁺-free). Li⁺ solution slightly elevated [Ca⁺⁺]_i but did not change the tension. Ten min later, 60 mM Li⁺ in the solution was substituted with 60 mM K⁺, which caused an increase in [Ca⁺⁺]_i and tension if verapamil was absent. Pretreatment with verapamil abolished the high K⁺-induced increases in [Ca⁺⁺]_i and tension in Li⁺-solution (fig. 2b). These results suggest that verapamil at 10 µM was sufficient to suppress L-type Ca⁺⁺ channels, and that K⁺-induced Ca⁺⁺ mobilization involved a component independent of voltage-dependent Ca⁺⁺ channel but related to Na⁺ in the medium.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Inhibitory effect of verapamil on high K+-induced rise in [Ca++]i and contraction. In each panel, the upper trace is [Ca++]i (R340/380) and the lower trace is tension. Left traces are control responses and right traces are responses in the presence of 10 µM verapamil. a, Responses in normal PSS. b, Responses in Li+ solution that had been replaced for normal PSS 10 min before the addition of high K+. In a and b, 65.4 mM K+ was applied as high K+ solution and K+/Li+ solution (table 1), respectively. Calibration bar on the right of R340/380 trace represents the control response to high K+ solution.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effects of replacement of external medium with high K+ solution on the PTX-induced increase in [Ca++]i and contraction in the presence of 10 µM verapamil. The medium was replaced with high K+ solution after the responses to 10 pM (a) or 10 nM (b) PTX reached a steady state. A broken line represents the resting level. In the upper trace of each panel, [Ca++]i is expressed as R340/380, which is % change induced by high K+-solution that was observed before application of verapamil and PTX.

Effects of PTX in Ca⁺⁺-free solution. In Ca⁺⁺-free (nominally Ca⁺⁺-free) PSS containing 10 µM verapamil, PTX (10 nM) induced a rise in [Ca⁺⁺]_i and a contraction, which reached the respective maximum at about 2 min and then decreased gradually (fig. 4a). The peak values of [Ca⁺⁺]_i and the contraction in the Ca⁺⁺-free PSS were 38 ± 6 and 56 ± 7% (n = 4), respectively, of those observed in normal PSS. On restoration of 2.5 mM Ca⁺⁺ in the medium, [Ca⁺⁺]_i and the tension increased. Figure 4b shows the effect of Ca⁺⁺ store depletion by ryanodine and caffeine (Ito et al., 1991) on the PTX-induced responses. When ryanodine (10 µM) and caffeine (10 mM) were simultaneously applied in the Ca⁺⁺-free PSS, [Ca⁺⁺]_i and tension increased transiently, then decreased to below the basal level. After ryanodine and caffeine were removed, application of PTX (10 nM) did not change [Ca⁺⁺]_i but slightly increased the tension. In Ca⁺⁺-free, EGTA (1 mM) PSS, PTX (10 nM) did not affect [Ca⁺⁺]_i, but induced a small, transient contraction (fig. 4c). This small contraction without an increase in [Ca⁺⁺]_i was not observed with PTX at less than 1 nM. However, 10 mM caffeine induced a transient rise in [Ca⁺⁺]_i and tension in this solution (data not shown), indicating that Ca⁺⁺ was present in the store.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Dependence of the PTX- (10 nM) induced increase in [Ca++]i and contraction on external Ca++. All experiments were performed in the presence of 10 µM verapamil. a, PTX-induced responses in nominally Ca++-free PSS (no EGTA). b, The effect of Ca++ store depletion on the PTX-induced responses in nominally Ca++-free PSS. Ryanodine (10 µM) and caffeine (10 mM) were simultaneously applied in nominally Ca++-free PSS to discharge Ca++ in the stores. c, Responses to PTX in Ca++-free, EGTA- (1 mM) PSS. A broken line represents the resting level. In the upper trace of each panel, R340/380 is expressed as % change induced by high K+ solution. PTX was applied at the upward arrow and washed at the downward arrow.

Dependence of PTX effects on external Na⁺. Next, the dependence of PTX-induced rise in [Ca⁺⁺]_i on extracellular Na⁺ was tested. In Na⁺-deficient solution (136.8 mM Li⁺ and 11.9 mM Na⁺ with 10 µM verapamil), PTX (10 nM) induced sustained increases in [Ca⁺⁺]_i and tension (66 ± 4 and 31 ± 2%, n = 3, respectively, of those observed in normal PSS, fig. 5a). On returning to normal PSS, [Ca⁺⁺]_i transiently dropped a little then increased. In the other experiment, when PSS was switched to Li⁺ solution, where Na⁺ was completely omitted, the baseline [Ca⁺⁺]_i tended to increase gradually in the presence of 10 µM verapamil. Subsequent application of 10 nM PTX did not change the course of gradual rise in [Ca⁺⁺]_i, thus PTX did not affect [Ca⁺⁺]_i or the tension (fig. 5b). On returning to PSS, [Ca⁺⁺]_i and the tension increased, but the contraction after exposure to Li⁺ solution was significantly smaller than that observed in PSS (46 ± 19% after exposure to Li⁺ solution and 122 ± 6% in PSS without preexposure to Li⁺ solution; 100% is referred to the high K⁺-induced contraction in normal PSS, n = 6).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effects of PTX on [Ca++]i and the tension in Na+-deficient and Na+-free solutions with normal K+. All experiments were performed in the presence of 10 µM verapamil. Eight min after switching to Na+-deficient or Na+-free solution, PTX (10 nM) was applied. a, Responses in Na+-deficient solution (136.8 mM Li+ and 11.9 mM Na+). After the PTX-effect reached a steady state, the medium was replaced with normal PSS. b, Responses in Li+-solution, where Na+ was completely omitted. Eight min after PTX, Li+ solution was switched to normal PSS. A broken line represents the resting level in normal PSS. In the upper trace of each panel, R340/380 is expressed as % change induced by high K+ solution that had been observed before application of verapamil and PTX.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Effects of PTX on [Ca++]i and the tension in Na+-free solutions with high K+. All experiments were performed in the presence of 10 µM verapamil. Eight min after switching to Na+-free solutions, PTX (10 nM) was applied. a, Responses in K+/Li+ solution (65.4 mM K+ and 76.8 mM Li+). b, responses in K+/NMDG+ solution (65.4 mM K+ and 76.8 mM NMDG+). After the responses to Na+-free solutions subsided, PTX (10 nM) was applied. Seven min after the application of PTX the medium was switched to Na+/Li+ solution (a) or Na+/NMDG+-solution (b). A broken line represents the resting level. In the upper trace of each panel, R340/380 is expressed as % change induced by high K+ solution that had been observed before application of verapamil and PTX.

View larger version (13K): [in this window] [in a new window]   Fig. 7.   Effects of ouabain on the PTX- (10 nM) induced changes in [Ca++]i and tension. a, Control responses to 10 nM PTX in normal PSS. b, Effects of 15 min pretreatment with 100 µM ouabain. In both cases verapamil (10 µM) and prazosin (1 µM) were present throughout the experiments. In the upper trace of each panel, R340/380 is expressed as % change induced by high K+ solution which had been observed before application of verapamil and PTX.

Effects of PTX on membrane potential. The resting membrane potential of porcine coronary arteries was -50.3 ± 0.9 mV (n = 11). PTX at 10 and 100 pM caused a sustained depolarization (fig. 8). The maximum depolarization was 4.5 ± 0.8 mV (n = 4) and 18.6 ± 1.7 mV (n = 7) with 10 and 100 pM PTX, respectively. PTX-induced depolarization was not reversible upon washout of PTX.

View larger version (14K): [in this window] [in a new window]   Fig. 8.   Time course of change in membrane potential after treatment with PTX. Closed circle, 10 pM PTX; open circle, 100 pM PTX. Each point represents the mean ± S.E.M. of four (10 pM) or seven (100 pM) experiments.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

PTX induced a rise in [Ca⁺⁺]_i and a contraction at above 1 pM in porcine coronary arteries. These changes were accompanied with membrane depolarization. PTX-induced Ca⁺⁺ mobilization and contraction depended on both external Ca⁺⁺ and Na⁺.

PTX has been reported to increase both [Na⁺]_i and [Ca⁺⁺]_i under physiological conditions (Ito et al., 1977; Ozaki et al., 1984; Frelin et al., 1990a; Monroe and Tashjian, 1995). Frelin et al. (1990a) suggested that the primary action of PTX in chick ventricular cells is an elevation of [Ca⁺⁺]_i and a rise in [Na⁺]_i is a consequence of Na⁺-Ca⁺⁺ exchange. From a similar standpoint, Monroe and Tashjian (1995) postulated that PTX directly stimulates Ca⁺⁺ entry through a channel created around the Na⁺,K⁺-ATPase. In contrast, the present study showed that PTX-induced rise in [Ca⁺⁺]_i was dependent on extracellular Na⁺ as the rise was abolished in Na⁺-free media (Li⁺ solution, K⁺/Li⁺ solution and K⁺/NMDG⁺ solution). The dependence of increase in [Ca⁺⁺]_i or contraction caused by PTX on extracellular Na⁺ has also been shown in other papers (Ito et al., 1979; Ozaki et al., 1983; Yoshizumi et al., 1991). In Monroe and Tashjian's work (1995), PTX-induced rise in [Ca⁺⁺]_i in Saos-2 cells also depended on extracellular Na⁺ and they interpreted the data as that the formation of PTX-induced channel required the presence of extracellular Na⁺. However, we observed that PTX induced a cation channel in Na⁺-free solutions (K⁺ or Li⁺ solution) in smooth muscle cells from rabbit portal vein (Ishii et al., 1997), consistent with findings from ventricular myocytes (Ikeda et al., 1988; Kinoshita et al., 1991) or aortic myocytes (Van Renterghem and Frelin, 1993). Thus, at least in muscle cells, the induction of channel by PTX does not require Na⁺ in an external medium. It has been reported that the PTX-induced channel permitted the passage of monovalent cations such as Na⁺, K⁺ or Li⁺ but not Ca⁺⁺ (Ikeda et al., 1988; Van Renterghem and Frelin, 1993; Ishii et al., 1997). Taken together, the dependence of PTX-induced Ca⁺⁺ mobilization on extracellular Na⁺ suggests that Na⁺ entry is a prerequisite for PTX to increase [Ca⁺⁺]_i.

An event supposed to be a consequence of increased permeability to Na⁺ is membrane depolarization, which did happen in porcine coronary arteries in this study. Depolarization would lead to an opening of voltage-dependent Ca⁺⁺ channels. PTX-induced rise in [Ca⁺⁺]_i in coronary arteries was partially inhibited by verapamil, so it is evident that voltage-dependent Ca⁺⁺ channels are at least partly involved in the PTX-induced Ca⁺⁺ mobilization, consistent with findings from rabbit aorta (Ito et al., 1977) or guinea-pig aorta (Ozaki et al., 1983). However, PTX-induced increase in [Ca⁺⁺]_i obviously involved an L-type Ca⁺⁺ channel blockers-resistant component.

A Ca⁺⁺ transport mechanism related to Na⁺ movement is Na⁺-Ca⁺⁺ exchange. A charge movement through Na⁺-Ca⁺⁺ exchange in smooth muscle cell is electrogenic in nature, since the stoichiometry of the exchanger is considered to be 3Na⁺: 1Ca⁺⁺ (McCarron et al., 1994). If we assume [Na⁺]_i and [Ca⁺⁺]_i before application of PTX as 10 mM and 150 nM (Jelicks and Gupta, 1990), respectively, the reversal potential of Na⁺-Ca⁺⁺ exchange (E_r) in normal PSS at 37°C is calculated to be -49 mV according to the following equation (Mullins, 1981); where ENa and ECa represent the equilibrium potential for Na⁺ and Ca⁺⁺, respectively. E_r is close to the resting membrane potential of porcine coronary arteries (-50 to -52 mV) before application of PTX. PTX (1 pM - 10 nM) has been reported to increase cellular Na⁺ content or [Na⁺]_i two- to several-fold from the resting level (Ozaki et al., 1984; Wattenberg et al., 1989; Frelin et al., 1990a). If PTX at a given concentration increases [Na⁺]_i to more than 20 mM, E_r would be negative to -115 mV. In this study, PTX at 10 pM or 100 pM depolarized the membrane to -45 mV or -32 mV, respectively. The negative shift of the reversal potential and the concomitant depolarization increase the driving force for Ca⁺⁺ influx through Na⁺-Ca⁺⁺ exchange in a reverse mode (Aaronson and Benham, 1989). Even so, since the affinity of Na⁺ to the internal Na⁺ binding site is low (K_D is 28 mM, Smith et al., 1991), Na⁺-Ca⁺⁺ exchanger could be latent unless [Na⁺]_i increases above the physiological level (Smith et al., 1989). Increase in [Na⁺]_i due to PTX can assist the exchanger to permit Ca⁺⁺ influx (reverse mode). In Na⁺-deficient solution (11.9 mM Na⁺) PTX increased [Ca⁺⁺]_i. When NaCl in PSS was replaced with LiCl, E_r can be calculated as negative to -250 mV. If PTX increases [Na⁺]_i, the exchanger may easily start to promote Ca⁺⁺ influx in this solution. Based on the above discussion, we conclude that Ca⁺⁺ influx via the Na⁺-Ca⁺⁺ exchange is another factor increasing [Ca⁺⁺]_i in response to PTX.

When PSS was changed to high K⁺-solution after treatment with verapamil and PTX, [Ca⁺⁺]_i and the contraction were further increased (fig. 3). This could be because Ca⁺⁺ influx through Na⁺-Ca⁺⁺ exchange was enhanced by further depolarizing the membrane and reducing the Na⁺ gradient. Therefore, this enhancement is an indirect sign that Na⁺-Ca⁺⁺ exchange in a reverse mode was functioning during the PTX-induced contraction. When Na⁺-deficient solution or Na⁺-free solution was replaced by normal PSS in the presence of PTX, [Ca⁺⁺]_i transiently dropped then increased. Restoration of Na⁺ in the medium increased the Na⁺ gradient, and thereby promoted Ca⁺⁺ extrusion through the exchanger in a forward mode (Ca⁺⁺ extrusion mode), but the subsequent accumulation of [Na⁺]_i might have increased [Ca⁺⁺]_i through the reversed Na⁺-Ca⁺⁺ exchange.

A part of the response of [Ca⁺⁺]_i to high K⁺-solution which contained 88.7 mM Na⁺ was also resistant to verapamil. The situation in which reduction of transmembrane Na⁺ gradient and depolarization occurred was similar to that caused by PTX and favorable for Na⁺-Ca⁺⁺ exchange to reverse into Ca⁺⁺ influx mode (Ashida and Blaustein, 1987; Smith et al., 1989; Battle et al., 1991). High K⁺ could have another benefit for the exchanger; it decreases the potency of Mg⁺⁺ as an inhibitor of Na⁺-Ca⁺⁺ exchange (Smith et al., 1989). Therefore, the most probable explanation for the verapamil-resistant Ca⁺⁺ mobilization induced by high K⁺ solution is that Ca⁺⁺ entry through Na⁺-Ca⁺⁺ exchange in a reverse mode contributed to the increase in [Ca⁺⁺]_i. During incubation in Li⁺-solution, however, [Na⁺]_i might have lowered and thereby was not sufficient to activate Na⁺-Ca⁺⁺ exchange (Smith et al., 1991), so that only voltage-dependent Ca⁺⁺ channels worked to stimulate the Ca⁺⁺ entry during exposure to K⁺/Li⁺-solution. When normal PSS was switched to K⁺/Li⁺-solution or K⁺/NMDG⁺-solution [Ca⁺⁺]_i was greatly but transiently increased. The rise could be due to Ca⁺⁺ entry through Na⁺-Ca⁺⁺ exchange and the fall due to a decline of Na⁺, which was excluded by Na⁺ pump and Na⁺-Ca⁺⁺ exchange. In contrast, when Li⁺-solution with normal K⁺ was replaced for PSS, [Ca⁺⁺]_i only slightly increased. This is also the case that a reduction of Na⁺ gradient alone is not sufficient for the exchanger to increase [Ca⁺⁺]_i (Smith et al., 1989; Ganitkevich and Isenberg, 1993). From the present results, it is likely that the Na⁺-Ca⁺⁺ exchange system functions to regulate [Ca⁺⁺]_i in porcine coronary arteries.

Ouabain can increase [Ca⁺⁺]_i as a result of inhibition of Na⁺,K⁺-ATPase (Bova et al., 1990; Iwamoto et al., 1992; Stewart et al., 1993). PTX and ouabain are common in increasing [Ca⁺⁺]_i through Na⁺-Ca⁺⁺ exchange. However, the interaction of ouabain and PTX was not additive, but contrarily ouabain antagonized the action of PTX to increase [Ca⁺⁺]_i, consistent with the data in other cells (Habermann, 1989; Kinoshita et al., 1991; Van Renterghem and Frelin, 1993; Ishii et al., 1997). We showed that ouabain inhibited the current through PTX-induced channel in rabbit portal vein cells (Ishii et al., 1997). This inhibition is probably the cause for the inhibition of PTX-induced increase in [Ca⁺⁺]_i. The effect of ouabain itself on [Ca⁺⁺]_i was slow and small as compared with that of PTX. This may be because ouabain needs a longer time to accumulate Na⁺ inside cells (Stewart et al., 1993).

Interestingly, PTX increased [Ca⁺⁺]_i in nominally Ca⁺⁺-free PSS (fig. 4a). When Ca⁺⁺ in the sarcoplasmic reticulum (SR) had been depleted by caffeine and ryanodine, PTX did not increase [Ca⁺⁺]_i (fig. 4b). These indicate that [Ca⁺⁺]_i increased by PTX in nominally Ca⁺⁺-free PSS was due to Ca⁺⁺ release from the SR. PTX did not increase [Ca⁺⁺]_i in Ca⁺⁺-free, EGTA-PSS, where the SR still retained Ca⁺⁺ as shown by the response to caffeine, so that it is evident that PTX does not directly release Ca⁺⁺ from the SR. Since µM order of Ca⁺⁺ present in nominally Ca⁺⁺-free PSS and membrane-bound Ca⁺⁺ could be a source for Ca⁺⁺ influx, the data suggest that the Ca⁺⁺ release was triggered by transmembrane Ca⁺⁺ influx. The fact that we could not detect an increase in [Ca⁺⁺]_i on application of PTX in nominally Ca⁺⁺-free PSS after depletion of the SR suggests that although Ca⁺⁺ influx caused by PTX in the solution was very small or an increase in [Ca⁺⁺]_i was local, it could trigger Ca⁺⁺ release. It was shown that Ca⁺⁺ influx in a small amount can trigger Ca⁺⁺ release in a larger amount from the SR of vascular smooth muscle (Ito et al., 1991). In cardiac muscle cells, it has been proposed that Ca⁺⁺ influx through Na⁺-Ca⁺⁺ exchange, which itself is too small to be detected by indo-1, works as a trigger to induce Ca⁺⁺ release from the SR (Leblanc and Hume, 1990). It was shown that when [Na⁺]_i was high, Ca⁺⁺ influx through Na⁺-Ca⁺⁺ exchange triggers Ca⁺⁺ release from the SR in guinea-pig coronary artery cells (Ganitkevich and Isenberg, 1993).

PTX at 10 nM slightly increased the tension when [Ca⁺⁺]_i was not elevated in Ca⁺⁺-free PSS (fig. 4), whereas such an effect was not observed with lower concentrations of PTX. The contraction independent of [Ca⁺⁺]_i was not observed when PTX was applied in Na⁺-free solutions, where [Ca⁺⁺]_i was also not elevated (fig. 6). It was shown that PTX caused cytosolic acidification in chick ventricular myocytes (Frelin et al., 1990b), Saos-2 cells (Monroe and Tashjian, 1996) and adrenal chromaffin cells (Yoshizumi et al., 1991), probably as a result of increased [Na⁺]_i. Decrease in intracellular pH increases tension at a given [Ca⁺⁺]_i (Gardner and Diecke, 1988). Although we did not investigate the mechanism of the contraction which did not accompany an increase in [Ca⁺⁺]_i, one cause may be related to a change in cytosolic pH.

In contrast to the results in porcine coronary arteries, the contraction of rabbit aorta to PTX was totally sensitive to verapamil (Ito et al., 1977). Similarly, in our preliminary study the PTX-induced contraction of rabbit carotid artery was abolished in the presence of verapamil. These suggest that the significance of Na⁺-Ca⁺⁺ exchange to regulate [Ca⁺⁺]_i is different among vascular tissues. The high sensitivity to PTX of coronary artery may be related to the functional significance of the exchange in this muscle. The present data indicate that PTX can be a tool to investigate the functional role of Na⁺-Ca⁺⁺ exchange.