- Email: [email protected]
The role of the sodium-calcium exchanger for calcium extrusion in coronary arteries
Life Sciences 67 (2000) 549Ð557
The role of the sodium-calcium exchanger for calcium extrusion in coronary arteries StŽphane Budel, Jean-Louis Beny* Department of Zoology and Animal Biology, University of Geneva, Sciences III, 30 quai Ernest Ansermet, 1211 Gen•ve 4, Switzerland
Abstract Calcium ionophores, such as the A23187, cause endothelium-dependent relaxation of arterial strips with intact endothelium, whereas the effect of the ionophore should result from the combination of a relaxation caused by the endothelium-dependent factors and of a contraction of the smooth muscles. In addition, the application of a calcium ionophore to a strip of pig coronary arteries without endothelium does not change cytosolic free calcium concentration and force developed by the smooth muscle cells. To explain these paradoxes, the hypothesis that active calcium extrusion would match the entry of extracellular calcium caused by the ionophore was tested. We see that the sodium-calcium exchanger extrudes calcium that enters the smooth muscle cells in the absence of the ionophore. This exchanger is efÞcient enough to expel the increased inßux of calcium created by the additional calcium carriers formed by the ionophore. This explains the inefÞciency of calcium ionophores to increase cytosolic free calcium of smooth muscle cells and consequently, the fact that the ionophore does not cause a contraction of a strip without endothelium. This makes evident that a calcium ionophore fully relaxes, in an endothelium-dependent manner, an intact strip of porcine coronary artery. © 2000 Elsevier Science Inc. All rights reserved. Keywords: A23187; Calcium; Sodium-calcium exchanger; Endothelium-derived relaxing factor; Pig coronary arteries
Introduction Paradoxically, a calcium ionophore fully relaxes an arterial strip with intact endothelium in an endothelium-dependent manner (1). The effect of the ionophore should be the result of the combination of an endothelium-dependent relaxing mechanism and of a contraction caused by the calcium entering the smooth muscle cells of the strip. In addition, to calibrate cytosolic free calcium concentration in ßuorometric techniques, * Corresponding author. Tel.: 141-22-702-67-66; fax: 141-22-781-17-47. E-mail address: [email protected] (J.-L. Beny) 0024-3205/00/$ Ð see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 0 2 4 - 3 2 0 5 ( 0 0 )0 0 6 4 6 -9
550
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
calcium ionophores are used. They allow the concentration of cytosolic free calcium to equilibrate with the extracellular concentration (2). Surprisingly, the application of an ionophore to porcine coronary artery strip does not change smooth muscle cell cytosolic free calcium concentration, as measured by the fura-2 technique (3). Furthermore, the strip without endothelium should fully contract when cytosolic free calcium increases in response to the ionophore, but no contractions are observed when the ionophore is applied to a porcine coronary strip (3). Our goal is to explain the inefÞciency of the calcium ionophore in increasing cytosolic free calcium and contracting arterial smooth muscles. A possible explanation would be that the ionophore does not permeabilize the arterial smooth muscle cell membrane to calcium. However, this hypothesis is unlikely, since the mechanism of action of a calcium ionophore is physicochemical. We therefore tested the hypothesis that even if the ionophore allows extracellular calcium to enter the smooth muscle cells, the cytosolic free calcium does not increase, because the plasmalemmal calcium extrusion pumps are active enough to maintain a low concentration of cytosolic free calcium. Methods Pharmacological experiments The anterior left descending branches of pig coronary arteries were taken from pigs freshly killed at the local slaughterhouse. The lumen was rinsed with ice cold oxygenated Krebs solution of the following composition (mM): NaCl 118.7, KCl 4.7, CaCl2 2.5, MgSO4 1.2, KH2PO4 1.2, NaHCO3 24.8, glucose 10.1 and gassed with 95% O2 and 5% CO2 at pH 7.4. The arteries were cleaned of all adherent fat and connective tissues. The endothelium was removed by rubbing the strip with a cotton tip. The endothelium was considered to be completely absent when the strip did not show any relaxation in response to substance P or bradykinin. The arterial segment was cut longitudinally and tension was measured in a 85 ml tissue bath by two silk threads tied to the extremities of the strip. One extremity was attached to the bottom of the bath and the other to a force displacement transducer (FTO3C). Changes in isometric tension were ampliÞed (Lectromed 3559) and recorded on a chart paper with polygraphs (W 1 W Electronics). Strips were continuously superfused with oxygenated Krebs solution (1.25 ml min21) maintained at 378C. The muscles were stretched up to 10 milliNewton (mN) and allowed to stabilize for about 1 hour. This tension was readjusted to 10 mN during stabilization and was taken as the baseline (4). The isometric forces developed by the strips are dispersed because the cross sections of the strips vary depending on the thickness of the vessels. Therefore, to normalize the forces, they are expressed as the percentage of the force developed during maximal phasic acetylcholine contraction. Cytosolic free Ca21 measurements Segments of about 8 mm long were cut longitudinally to form a strip. The endothelium was removed gently by rubbing the luminal face of the strip with a cotton tip. The adventitia was carefully removed, because it has a high autoßuorescence. Then the tissue was loaded with fura-2 AM (10 mM), the acetoxymethyl ester form of fura-2 for 2 hours in a HEPES
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
551
(N-[2-hydroxyethyl]piperazine-N9-[2-ethanosulphonic acid])- buffered solution containing (mM): NaCl 145, KCl 5, CaCl2 1, MgSO4 0.5, NaH2PO4 1, HEPES 20 and glucose 10.1 at pH 7.4, 378C in the presence of pluronic F-127 (0.25%). The tissue was then rinsed for 1 hour in the solution used for pharmacological experiments (described above) but with a modiÞed gas mixture of 75% N2; 20% O2 and 5% CO2, since high O2 concentrations greatly reduce the intensity of fura-2 ßuorescence (5). The artery strip was then Þxed in an observation chamber to reduce movement. The intimal face was put facing down onto the glass coverslip used for ßuorescence measurements. The chamber was continuously perfused (1.5 ml min21), which ensured that drug concentrations remained stable and removed the dye from the extracellular ßuid. The ßuorescence emitted by the cells corresponds to the Þeld of the lens (magniÞcation 103) and was measured through a glass coverslip with a Nikon inverted microscope (Diaphot) equipped with a P1 Photometer. We used dual excitation wavelengths of 340 and 380 nm. These wavelengths can be changed by manually sliding the double Þlter holder about every minute and more rapidly when the changes in [Ca21]i occur. During the advancement of this research, the excitation wavelengths were changed by rotating the Þlters mounted on an automatic Þlter wheel, and the P1 Photometer was replaced by an Extended Isis IntensiÞed CCD camera (Photonic Science Ltd, East Sussex, UK). Changes in ßuorescence were digitized with an analog-digital interface (Mac Lab; World Precision Instrument) and later by a video frame grabber (Neotech) connected to a computer (Macintosh II fx). The [Ca21]i levels were estimated by calculating the ratio of the emissions at 510 nm. These emissions were caused by the excitation of the probe at 340 and 380 nm (2). The ratio was calculated each time that the Þlters were alternated. Short transient changes in calcium could not be detected with our low frequency sampling. However, continuous recording at one excitation wavelength during contraction-relaxation cycles demonstrated that no such events happen in our experiment. The [Ca21]i was calculated using the equation of Grynkiewicz et al. (2). The dissociation constant for the fura-2-Ca21 complex was considered to be 225 nM; the minimal and maximal ßuorescence ratios were measured in a medium where sodium chloride was replaced by choline chloride in order to inhibit the Na1-Ca21 exchanger and the cells were permeabilized to calcium with 1.7 mM 4-bromo A23187. In this medium, the maximal ßuorescence (Rmax) was determined in the presence of 2.5 mM Ca21 and the minimal ßuorescence (Rmin) in a solution without calcium and with 2 mM EGTA. Drugs used in this study Thapsigargin, A23187, 4-bromo-A23187, ruthenium red, eosin and acetylcholine were obtained from Sigma Chemical (St. Louis, MO). Pluronic F-127 was obtained from Calbiochem (San Diego, CA). 29,49-dichlorobenzamil hydrochloride was obtained from Molecular Probe (Eugene, OR). Stock solutions of acetylcholine, 29,49-dichlorobenzamil and ruthenium red were prepared in Krebs solution. Stock solutions of thapsigargin, eosin, A23187 and 4bromo A23187 were prepared in dimethyl sulfoxide. Data analysis Data were calculated as the mean 6 standard error of the mean (SEM). ÒnÓ represents the number of strips; each strip comes from a different pig. The StudentÕs t test was used to compare results; p , 0.05 deemed to be signiÞcant.
552
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
Fig. 1. Effect of the calcium ionophore A23187 (A23) (190 nM) without and with inhibitors of calcium pumps on isometric force developed by a transversal strip of pig coronary artery without endothelium. The forces are expressed as the percentage of the maximal phasic acetylcholine (ACh) (10 mM) contraction. ruth.: ruthenium red 100 mM; 0 Na: medium without sodium.
Fig. 2. Isometric force developed by a strip of porcine coronary artery without endothelium. Effect of different plasmalemmal calcium pump inhibitors on force development due to calcium inßux caused by A23187. Eosin is an inhibitor of the plasmalemmal Ca21-ATPase, and a medium without sodium (0 Na) is used to inhibit the Na1-Ca21 exchanger. The effect of a supramaximal concentration of acetylcholine (ACh)(10 mM) is shown for comparison.
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
553
Results Effect of the calcium ionophore A23187 The calcium ionophore A23187 superfused at a concentration of 190 nM during 20 minutes did not signiÞcantly change the isometric force produced by a coronary artery strip without endothelium (n520 observations) (Figs. 1 and 2). The smooth muscle cell cytosolic free calcium concentration was 125632 nM (n59). This concentration did not signiÞcantly change during the application of 190 nM 4-bromo A23187 where it reached 124636 nM, (n59) (Fig. 3). 4-bromo A23187, which is not ßuorescent, is used in calcium measurement experiments instead of A23187. We veriÞed that these two forms of the ionophore have the same effect on contraction (n54). To explain this lack of effect, the hypothesis that active calcium extrusion would match the entry of extracellular calcium caused by the ionophore was tested using different inhibitors of calcium pumps in the presence of the calcium ionophore. Effect of inhibition of calcium pumps Plasmalemmal Ca21-ATPase It was reported that 100 mM ruthenium red inhibits the plasmalemmal Ca21-ATPase (6). However, the addition of this inhibitor to a strip incubated in the presence of A23187 had no
Fig. 3. Effect of the calcium ionophore A23187 (A23) (190 nM) without and with inhibitors of calcium pumps on smooth muscle cell cytosolic free calcium. The effect of a supramaximal concentration of acetylcholine (ACh)(10 mM) is shown for comparison. The calcium concentration is expressed as the percentage of the resting level (124636 nM, n59). 0 Na: medium without sodium.
554
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
signiÞcant effect on isometric force developed by the coronary artery strips (Fig. 1). Neither had this inhibitor any effect on force developed by a strip incubated in the absence of the calcium ionophore. Since ruthenium red could not be speciÞc enough, eosin, a more speciÞc inhibitor of the plasmalemmal Ca21-ATPase, was also tested (7). Eosin at concentrations of 0.3 and 3 mM did not contract pig coronary arteries in the presence or absence of 190 nM of A23187 (for both situations: n54 observations) (Fig. 2). The effect of these inhibitors on cytosolic free calcium could not be tested using our approach since these molecules only exist in a ßuorescent form. These results indicate that the plasmalemmal Ca21-ATPase is not crucially implicated in the homeostasis of cytosolic free calcium in porcine coronary arteries. An alternative explanation would be that when the Ca21-ATPase is inhibited, the Na1-Ca21 exchanger is accelerated, thus maintaining the level of calcium low. Na1-Ca21 exchanger A way to inhibit this plasmalemmal pump which actively extrudes calcium from the cell is to remove the extracellular sodium, since it is the trans-plasmalemmal concentration gradient of this ion which is the driving force for calcium extrusion (for review see 8). A strip incubated in a medium where sodium chloride was replaced by choline chloride developed an isometric force which reached 8864% of a maximal ACh contraction (n54) (Figs. 1 and 2). The smooth muscle cell cytosolic calcium concentration was increased by 666 6 78% (n54) (Fig. 3). When the strip, already in the presence of the ionophore A23187, was incubated in a medium where sodium chloride was replaced by choline chloride, isometric force reached 87.563% of the maximal ACh contraction (n54). In this situation, the smooth muscle cell cytosolic calcium concentration was increased by 19316528% (n54). This increase is not signiÞcantly different from the one observed during supra maximal acetylcholine contraction (19526655%, n59). So the addition of the ionophore increases the cytosolic free calcium from 34% to 99% when the exchanger is inhibited (expressed in % of a supramaximal ACh concentration), because the inßux of calcium is strongly increased by the added carriers. However the resultant contractions were the same with and without the ionophore showing that a maximal calcium increase is not necessary to induce a maximal contraction (9,10). Depending upon the electrochemical driving forces, the Na1-Ca21 exchanger can invert the direction of calcium pumping. Particularly in the experiments without extracellular sodium, this would inßuence calcium entry and hence the contraction. Therefore, we tested a pharmacological inhibitor of the Na1-Ca21 exchanger, 29,49-dichlorobenzamil, although this inhibitor is less speciÞc than the removal of sodium. The contraction caused by 100 mM 29,49-dichlorobenzamil was 1263% in the absence of the ionophore (n54) and 53614% in its presence (n57) (these percentages refer to the contraction caused by a medium without sodium). This shows that the contraction is not only caused by an entry of calcium through the inverted exchanger in the case of extracellular sodium removal, but also by an inhibition of the exchanger. This extrusion of calcium seems very efÞcient. The cells can be depolarized by incubation in a medium containing 120 mM potassium instead of 4.7 mM, thus opening the voltagedependent calcium channels. The consequent fast inßux of calcium causes a transient contraction of 8167% of a maximal ACh contraction (n54) (Fig. 4). After this, the force is sta-
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
555
Fig. 4. Typical recording of the effect of high potassium and a medium without sodium (0 Na) on isometric force developed by a strip of porcine coronary arteries without endothelium. The opening of voltage dependant calcium channels caused by depolarization (with high potassium) is partially compensated by the activity of the Na1-Ca21 exchanger as demonstrated by the inhibition of this pump using a medium without sodium (0 Na). The effect of a supramaximal concentration of acetylcholine (ACh) (10 mM) is shown for comparison.
bilized to a plateau of 2166% of a maximal ACh contraction (n54). In this situation, when sodium is removed, a new steady contraction reaching 3669% of a maximal ACh contraction is observed (n54). Our interpretation is that the fast entry of extracellular calcium into the smooth muscle can not be immediately compensated by calcium extrusion. However, this inßux diminishes progressively. It is likely to be due to the activation of the Na1-Ca21 exchanger as shown by the increased force developed by the strip resulting from the removal of extracellular sodium. Indeed, using the equation Eexchanger 5 3ENa Ð 2ECa we calculated that this depolarization is not important enough to invert the direction of the Na1-Ca21 exchanger, as the membrane potential only reaches Ð5 mV (4) (Eexchanger5 equilibrium potential of the pump; ENa 5 equilibrium potential for the sodium ion ECa 5 equilibrium potential for the calcium ion). We veriÞed that in the absence of extracellular sodium, the endothelium-dependent relaxation induced by 190 nM A23187 was indeed reduced. The control relaxation was 8967% (n54) and in the absence of sodium, it was only 462% (n54). Indeed, when the main way of calcium extrusion from the smooth muscle cells is inhibited, being no more able to extrude calcium, they tend to contract. This contraction is opposed to the concomitant endothelium-
556
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
dependent relaxation induced by the ionophore. The change in force is therefore the result of a combination of these two opposite effects. Sarcoplasmic reticulum Ca21-ATPase The superÞcial buffer barrier concept states that the superÞcial sarcoplasmic reticulum accumulates a portion of the calcium which enters the smooth muscle cells through the cell membrane and that calcium is then continually extruded from the sarcoplasmic reticulum to the extracellular space (11). This implies that calcium is Þrst pumped into the sarcoplasmic reticulum before being extruded. Indeed, removal of external sodium inhibits the calcium extrusion from the rabbit vena cava only when the sarcoplasmic reticulum calcium pump is active. We tested whether the sarcoplasmic reticulum is a necessary pathway for calcium extrusion in the porcine coronary artery by inhibiting the sarcoplasmic reticulum Ca21-ATPase using thapsigargin. Thapsigargin alone contracts the coronary strip by 2.361.8% of a maximal ACh contraction (n54). Thapsigargin applied before sodium removal did not signiÞcantly affect the effect of the Na1-Ca21 exchanger inhibition. In that case the contraction reached 89612% (n58), and was not signiÞcantly different from that observed in the absence of thapsigargin (88%) (for review see 11). Thus, our results show that the sarcoplasmic reticulum is not the only way of calcium extrusion in the porcine coronary artery as already demonstrated using a rabbit vena cava (12). Discussion It is difÞcult to speciÞcally inhibit an ionic pump. Particularly, the removal of extracellular sodium can cause side effects. First of all, it was proposed that the inhibition of the Na1-H1 exchanger acidiÞes the cytosol and thus modiÞes the cellular response (13). However, the acidiÞcation is slow (14). Therefore this phenomenon cannot interfere with the contraction we observed which is immediate and quickly reaches a plateau. Secondly, the removal of extracellular sodium could reverse the direction of the Na1-Ca21 exchanger thus facilitating calcium diffusion from the outside to the inside of the cell (15). This would participate to calcium entry. However we observed that the inhibition of the Na1-Ca21 exchanger by dichlorobenzamil, which does not facilitate calcium entry, contracts the strip in the presence and in the absence of the ionophore. Furthermore, the contraction, which is weak in the absence of the ionophore, becomes stronger in its presence. This allows us to conclude that the Na1Ca21 exchanger is important when large amounts of calcium have to be extruded from the cells. The smooth muscle Na1-Ca21 exchanger is a low afÞnity high capacity exchanger; therefore, it could strongly increase its activity when cytosolic free calcium increases (16). This extrusion mechanism would be efÞcient enough to match the progressive entry of calcium caused by the ionophore, but not to immediately match the explosive increase in calcium caused by either the ACh or the opening of voltage gated calcium channels. Three distinct ways for the extracellular calcium to diffuse into the cytosol are considered: the leak calcium channel, the carrier formed by the ionophore and the voltage dependent calcium channel opened by the depolarization caused by a high extracellular potassium concentration (8). For these three ways of calcium entry, our results indicate that it is mainly the Na1-Ca21 exchanger which is responsible for the extrusion of cytosolic free calcium from smooth muscle cells of porcine coronary arteries.
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
557
The importance of the Na1-Ca21 exchanger differs from one smooth muscle type to another. So, in the airway smooth muscle, the Na1-Ca21 exchanger makes a minor contribution to calcium homeostasis (17). However, the importance of this exchanger for the cell calcium homeostasis was already demonstrated for arterial smooth muscle cells. In the swine common carotid arteries and in the rat aorta, the Na1-Ca21 exchanger has been shown to be the major extrusion mechanism (13Ð15,18). On the contrary, in the porcine coronary endothelial cells, it is likely that the calcium extrusion mechanisms are not so important. As a matter of fact, a calcium ionophore increases cytosolic free calcium and hyperpolarizes the endothelial cells without the necessity to inhibit any calcium extrusion mechanism (1,19). Altogether these results suggest that in porcine coronary arteries, the homeostasis of calcium results from the potential important change of activity of the Na1-Ca21 exchanger, which controles the extrusion of calcium and thus the contraction of the artery. This model resolves the paradoxes evoked in the introduction: Þrst of all, the inefÞciency of a calcium ionophore to increase cytosolic free calcium for the calibration of this parameter using the ßuorescence technique; secondly: the fact that a calcium ionophore fully relaxes in an endothelium-dependent manner an intact strip of porcine coronary artery; thirdly: the fact that the ionophore does not cause a contraction of a strip without endothelium. Acknowledgments We thank Fran•oise Gribi for her excellent technical assistance and Alexander Schuster for reviewing the manuscript. This work was supported by the Swiss National Science Foundation (grant 3100-49163.96). References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
P-Y VON DER WEID, J-L BƒNY, Am J Physiol 262 H1823ÐH1831 (1992). G GRYNKIEWICZ, M POENIE, RY TSIEN, J Biol Chem 260 3440Ð3450 (1985). F CHABAUD, M DANNA, J-L BƒNY, Life Sci 54 1449Ð1458 (1994). V VAN DER BENT, J-L BƒNY, Br J Pharmacol 103 1877Ð1882 (1991). PL BECKER & FS FAY Am. J. Physiol. 253 C613ÐC618 (1987). L MISSIAEN, H DE SMEDT, G DROOGMANS, F WUYTACK, L RAEYMAEKERS, R CASTEELS, Biochim Biophys Acta 1023 449Ð454 (1990). C GATTO, CC HALE, W WU, MA MILANICK, Biochemistry 34 965Ð972 (1995). ME OÕDONNELL, NE OWEN, Physiological reviews 74 683Ð721 (1994). AP SOMLYO & AV SOMLYO, Nature 372 231Ð236 (1994). RA MURPHY, FASEB Journal 8 311Ð318 (1994). C VAN BREEMEN, Q CHEN, M LAHER, Trends Pharmacol Sci 16 98Ð105 (1995). M NAZER, C VAN BREEMEN, Am J Physiol 274 H123Ð131 (1998). S BOVA, WF GOLDMAN, XJ YAUAN, MP BLAUSTEIN, Am J Physiol 259 H409Ð423 (1990). DC BATLLE, M GODINICH, MS LAPOINTE, E MUNOZ, F CARONE, N MEHRING, Am J Physiol 261 C845ÐC856 (1991). T ASHIDA, MP BLAUSTEIN, J Physiol 392 617Ð635 (1987). MJ TAGGART, S WRAY, Cell calcium 22 333Ð341 (1997). LJ JANSSEN, DK WALTRES and J WATTIC, Am J Physiol 273 L322ÐL330 (1997). CM REMBOLD, H RICHARD, XL CHEN Hypertension 19 308Ð313 (1992). A DOMENIGHETTI, JL BENY, F CHABAUD AND M FRIEDEN, J Physiol 513 103Ð116 (1998).
The role of the sodium-calcium exchanger for calcium extrusion in coronary arteries StŽphane Budel, Jean-Louis Beny* Department of Zoology and Animal Biology, University of Geneva, Sciences III, 30 quai Ernest Ansermet, 1211 Gen•ve 4, Switzerland
Abstract Calcium ionophores, such as the A23187, cause endothelium-dependent relaxation of arterial strips with intact endothelium, whereas the effect of the ionophore should result from the combination of a relaxation caused by the endothelium-dependent factors and of a contraction of the smooth muscles. In addition, the application of a calcium ionophore to a strip of pig coronary arteries without endothelium does not change cytosolic free calcium concentration and force developed by the smooth muscle cells. To explain these paradoxes, the hypothesis that active calcium extrusion would match the entry of extracellular calcium caused by the ionophore was tested. We see that the sodium-calcium exchanger extrudes calcium that enters the smooth muscle cells in the absence of the ionophore. This exchanger is efÞcient enough to expel the increased inßux of calcium created by the additional calcium carriers formed by the ionophore. This explains the inefÞciency of calcium ionophores to increase cytosolic free calcium of smooth muscle cells and consequently, the fact that the ionophore does not cause a contraction of a strip without endothelium. This makes evident that a calcium ionophore fully relaxes, in an endothelium-dependent manner, an intact strip of porcine coronary artery. © 2000 Elsevier Science Inc. All rights reserved. Keywords: A23187; Calcium; Sodium-calcium exchanger; Endothelium-derived relaxing factor; Pig coronary arteries
Introduction Paradoxically, a calcium ionophore fully relaxes an arterial strip with intact endothelium in an endothelium-dependent manner (1). The effect of the ionophore should be the result of the combination of an endothelium-dependent relaxing mechanism and of a contraction caused by the calcium entering the smooth muscle cells of the strip. In addition, to calibrate cytosolic free calcium concentration in ßuorometric techniques, * Corresponding author. Tel.: 141-22-702-67-66; fax: 141-22-781-17-47. E-mail address: [email protected] (J.-L. Beny) 0024-3205/00/$ Ð see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 0 2 4 - 3 2 0 5 ( 0 0 )0 0 6 4 6 -9
550
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
calcium ionophores are used. They allow the concentration of cytosolic free calcium to equilibrate with the extracellular concentration (2). Surprisingly, the application of an ionophore to porcine coronary artery strip does not change smooth muscle cell cytosolic free calcium concentration, as measured by the fura-2 technique (3). Furthermore, the strip without endothelium should fully contract when cytosolic free calcium increases in response to the ionophore, but no contractions are observed when the ionophore is applied to a porcine coronary strip (3). Our goal is to explain the inefÞciency of the calcium ionophore in increasing cytosolic free calcium and contracting arterial smooth muscles. A possible explanation would be that the ionophore does not permeabilize the arterial smooth muscle cell membrane to calcium. However, this hypothesis is unlikely, since the mechanism of action of a calcium ionophore is physicochemical. We therefore tested the hypothesis that even if the ionophore allows extracellular calcium to enter the smooth muscle cells, the cytosolic free calcium does not increase, because the plasmalemmal calcium extrusion pumps are active enough to maintain a low concentration of cytosolic free calcium. Methods Pharmacological experiments The anterior left descending branches of pig coronary arteries were taken from pigs freshly killed at the local slaughterhouse. The lumen was rinsed with ice cold oxygenated Krebs solution of the following composition (mM): NaCl 118.7, KCl 4.7, CaCl2 2.5, MgSO4 1.2, KH2PO4 1.2, NaHCO3 24.8, glucose 10.1 and gassed with 95% O2 and 5% CO2 at pH 7.4. The arteries were cleaned of all adherent fat and connective tissues. The endothelium was removed by rubbing the strip with a cotton tip. The endothelium was considered to be completely absent when the strip did not show any relaxation in response to substance P or bradykinin. The arterial segment was cut longitudinally and tension was measured in a 85 ml tissue bath by two silk threads tied to the extremities of the strip. One extremity was attached to the bottom of the bath and the other to a force displacement transducer (FTO3C). Changes in isometric tension were ampliÞed (Lectromed 3559) and recorded on a chart paper with polygraphs (W 1 W Electronics). Strips were continuously superfused with oxygenated Krebs solution (1.25 ml min21) maintained at 378C. The muscles were stretched up to 10 milliNewton (mN) and allowed to stabilize for about 1 hour. This tension was readjusted to 10 mN during stabilization and was taken as the baseline (4). The isometric forces developed by the strips are dispersed because the cross sections of the strips vary depending on the thickness of the vessels. Therefore, to normalize the forces, they are expressed as the percentage of the force developed during maximal phasic acetylcholine contraction. Cytosolic free Ca21 measurements Segments of about 8 mm long were cut longitudinally to form a strip. The endothelium was removed gently by rubbing the luminal face of the strip with a cotton tip. The adventitia was carefully removed, because it has a high autoßuorescence. Then the tissue was loaded with fura-2 AM (10 mM), the acetoxymethyl ester form of fura-2 for 2 hours in a HEPES
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
551
(N-[2-hydroxyethyl]piperazine-N9-[2-ethanosulphonic acid])- buffered solution containing (mM): NaCl 145, KCl 5, CaCl2 1, MgSO4 0.5, NaH2PO4 1, HEPES 20 and glucose 10.1 at pH 7.4, 378C in the presence of pluronic F-127 (0.25%). The tissue was then rinsed for 1 hour in the solution used for pharmacological experiments (described above) but with a modiÞed gas mixture of 75% N2; 20% O2 and 5% CO2, since high O2 concentrations greatly reduce the intensity of fura-2 ßuorescence (5). The artery strip was then Þxed in an observation chamber to reduce movement. The intimal face was put facing down onto the glass coverslip used for ßuorescence measurements. The chamber was continuously perfused (1.5 ml min21), which ensured that drug concentrations remained stable and removed the dye from the extracellular ßuid. The ßuorescence emitted by the cells corresponds to the Þeld of the lens (magniÞcation 103) and was measured through a glass coverslip with a Nikon inverted microscope (Diaphot) equipped with a P1 Photometer. We used dual excitation wavelengths of 340 and 380 nm. These wavelengths can be changed by manually sliding the double Þlter holder about every minute and more rapidly when the changes in [Ca21]i occur. During the advancement of this research, the excitation wavelengths were changed by rotating the Þlters mounted on an automatic Þlter wheel, and the P1 Photometer was replaced by an Extended Isis IntensiÞed CCD camera (Photonic Science Ltd, East Sussex, UK). Changes in ßuorescence were digitized with an analog-digital interface (Mac Lab; World Precision Instrument) and later by a video frame grabber (Neotech) connected to a computer (Macintosh II fx). The [Ca21]i levels were estimated by calculating the ratio of the emissions at 510 nm. These emissions were caused by the excitation of the probe at 340 and 380 nm (2). The ratio was calculated each time that the Þlters were alternated. Short transient changes in calcium could not be detected with our low frequency sampling. However, continuous recording at one excitation wavelength during contraction-relaxation cycles demonstrated that no such events happen in our experiment. The [Ca21]i was calculated using the equation of Grynkiewicz et al. (2). The dissociation constant for the fura-2-Ca21 complex was considered to be 225 nM; the minimal and maximal ßuorescence ratios were measured in a medium where sodium chloride was replaced by choline chloride in order to inhibit the Na1-Ca21 exchanger and the cells were permeabilized to calcium with 1.7 mM 4-bromo A23187. In this medium, the maximal ßuorescence (Rmax) was determined in the presence of 2.5 mM Ca21 and the minimal ßuorescence (Rmin) in a solution without calcium and with 2 mM EGTA. Drugs used in this study Thapsigargin, A23187, 4-bromo-A23187, ruthenium red, eosin and acetylcholine were obtained from Sigma Chemical (St. Louis, MO). Pluronic F-127 was obtained from Calbiochem (San Diego, CA). 29,49-dichlorobenzamil hydrochloride was obtained from Molecular Probe (Eugene, OR). Stock solutions of acetylcholine, 29,49-dichlorobenzamil and ruthenium red were prepared in Krebs solution. Stock solutions of thapsigargin, eosin, A23187 and 4bromo A23187 were prepared in dimethyl sulfoxide. Data analysis Data were calculated as the mean 6 standard error of the mean (SEM). ÒnÓ represents the number of strips; each strip comes from a different pig. The StudentÕs t test was used to compare results; p , 0.05 deemed to be signiÞcant.
552
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
Fig. 1. Effect of the calcium ionophore A23187 (A23) (190 nM) without and with inhibitors of calcium pumps on isometric force developed by a transversal strip of pig coronary artery without endothelium. The forces are expressed as the percentage of the maximal phasic acetylcholine (ACh) (10 mM) contraction. ruth.: ruthenium red 100 mM; 0 Na: medium without sodium.
Fig. 2. Isometric force developed by a strip of porcine coronary artery without endothelium. Effect of different plasmalemmal calcium pump inhibitors on force development due to calcium inßux caused by A23187. Eosin is an inhibitor of the plasmalemmal Ca21-ATPase, and a medium without sodium (0 Na) is used to inhibit the Na1-Ca21 exchanger. The effect of a supramaximal concentration of acetylcholine (ACh)(10 mM) is shown for comparison.
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
553
Results Effect of the calcium ionophore A23187 The calcium ionophore A23187 superfused at a concentration of 190 nM during 20 minutes did not signiÞcantly change the isometric force produced by a coronary artery strip without endothelium (n520 observations) (Figs. 1 and 2). The smooth muscle cell cytosolic free calcium concentration was 125632 nM (n59). This concentration did not signiÞcantly change during the application of 190 nM 4-bromo A23187 where it reached 124636 nM, (n59) (Fig. 3). 4-bromo A23187, which is not ßuorescent, is used in calcium measurement experiments instead of A23187. We veriÞed that these two forms of the ionophore have the same effect on contraction (n54). To explain this lack of effect, the hypothesis that active calcium extrusion would match the entry of extracellular calcium caused by the ionophore was tested using different inhibitors of calcium pumps in the presence of the calcium ionophore. Effect of inhibition of calcium pumps Plasmalemmal Ca21-ATPase It was reported that 100 mM ruthenium red inhibits the plasmalemmal Ca21-ATPase (6). However, the addition of this inhibitor to a strip incubated in the presence of A23187 had no
Fig. 3. Effect of the calcium ionophore A23187 (A23) (190 nM) without and with inhibitors of calcium pumps on smooth muscle cell cytosolic free calcium. The effect of a supramaximal concentration of acetylcholine (ACh)(10 mM) is shown for comparison. The calcium concentration is expressed as the percentage of the resting level (124636 nM, n59). 0 Na: medium without sodium.
554
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
signiÞcant effect on isometric force developed by the coronary artery strips (Fig. 1). Neither had this inhibitor any effect on force developed by a strip incubated in the absence of the calcium ionophore. Since ruthenium red could not be speciÞc enough, eosin, a more speciÞc inhibitor of the plasmalemmal Ca21-ATPase, was also tested (7). Eosin at concentrations of 0.3 and 3 mM did not contract pig coronary arteries in the presence or absence of 190 nM of A23187 (for both situations: n54 observations) (Fig. 2). The effect of these inhibitors on cytosolic free calcium could not be tested using our approach since these molecules only exist in a ßuorescent form. These results indicate that the plasmalemmal Ca21-ATPase is not crucially implicated in the homeostasis of cytosolic free calcium in porcine coronary arteries. An alternative explanation would be that when the Ca21-ATPase is inhibited, the Na1-Ca21 exchanger is accelerated, thus maintaining the level of calcium low. Na1-Ca21 exchanger A way to inhibit this plasmalemmal pump which actively extrudes calcium from the cell is to remove the extracellular sodium, since it is the trans-plasmalemmal concentration gradient of this ion which is the driving force for calcium extrusion (for review see 8). A strip incubated in a medium where sodium chloride was replaced by choline chloride developed an isometric force which reached 8864% of a maximal ACh contraction (n54) (Figs. 1 and 2). The smooth muscle cell cytosolic calcium concentration was increased by 666 6 78% (n54) (Fig. 3). When the strip, already in the presence of the ionophore A23187, was incubated in a medium where sodium chloride was replaced by choline chloride, isometric force reached 87.563% of the maximal ACh contraction (n54). In this situation, the smooth muscle cell cytosolic calcium concentration was increased by 19316528% (n54). This increase is not signiÞcantly different from the one observed during supra maximal acetylcholine contraction (19526655%, n59). So the addition of the ionophore increases the cytosolic free calcium from 34% to 99% when the exchanger is inhibited (expressed in % of a supramaximal ACh concentration), because the inßux of calcium is strongly increased by the added carriers. However the resultant contractions were the same with and without the ionophore showing that a maximal calcium increase is not necessary to induce a maximal contraction (9,10). Depending upon the electrochemical driving forces, the Na1-Ca21 exchanger can invert the direction of calcium pumping. Particularly in the experiments without extracellular sodium, this would inßuence calcium entry and hence the contraction. Therefore, we tested a pharmacological inhibitor of the Na1-Ca21 exchanger, 29,49-dichlorobenzamil, although this inhibitor is less speciÞc than the removal of sodium. The contraction caused by 100 mM 29,49-dichlorobenzamil was 1263% in the absence of the ionophore (n54) and 53614% in its presence (n57) (these percentages refer to the contraction caused by a medium without sodium). This shows that the contraction is not only caused by an entry of calcium through the inverted exchanger in the case of extracellular sodium removal, but also by an inhibition of the exchanger. This extrusion of calcium seems very efÞcient. The cells can be depolarized by incubation in a medium containing 120 mM potassium instead of 4.7 mM, thus opening the voltagedependent calcium channels. The consequent fast inßux of calcium causes a transient contraction of 8167% of a maximal ACh contraction (n54) (Fig. 4). After this, the force is sta-
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
555
Fig. 4. Typical recording of the effect of high potassium and a medium without sodium (0 Na) on isometric force developed by a strip of porcine coronary arteries without endothelium. The opening of voltage dependant calcium channels caused by depolarization (with high potassium) is partially compensated by the activity of the Na1-Ca21 exchanger as demonstrated by the inhibition of this pump using a medium without sodium (0 Na). The effect of a supramaximal concentration of acetylcholine (ACh) (10 mM) is shown for comparison.
bilized to a plateau of 2166% of a maximal ACh contraction (n54). In this situation, when sodium is removed, a new steady contraction reaching 3669% of a maximal ACh contraction is observed (n54). Our interpretation is that the fast entry of extracellular calcium into the smooth muscle can not be immediately compensated by calcium extrusion. However, this inßux diminishes progressively. It is likely to be due to the activation of the Na1-Ca21 exchanger as shown by the increased force developed by the strip resulting from the removal of extracellular sodium. Indeed, using the equation Eexchanger 5 3ENa Ð 2ECa we calculated that this depolarization is not important enough to invert the direction of the Na1-Ca21 exchanger, as the membrane potential only reaches Ð5 mV (4) (Eexchanger5 equilibrium potential of the pump; ENa 5 equilibrium potential for the sodium ion ECa 5 equilibrium potential for the calcium ion). We veriÞed that in the absence of extracellular sodium, the endothelium-dependent relaxation induced by 190 nM A23187 was indeed reduced. The control relaxation was 8967% (n54) and in the absence of sodium, it was only 462% (n54). Indeed, when the main way of calcium extrusion from the smooth muscle cells is inhibited, being no more able to extrude calcium, they tend to contract. This contraction is opposed to the concomitant endothelium-
556
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
dependent relaxation induced by the ionophore. The change in force is therefore the result of a combination of these two opposite effects. Sarcoplasmic reticulum Ca21-ATPase The superÞcial buffer barrier concept states that the superÞcial sarcoplasmic reticulum accumulates a portion of the calcium which enters the smooth muscle cells through the cell membrane and that calcium is then continually extruded from the sarcoplasmic reticulum to the extracellular space (11). This implies that calcium is Þrst pumped into the sarcoplasmic reticulum before being extruded. Indeed, removal of external sodium inhibits the calcium extrusion from the rabbit vena cava only when the sarcoplasmic reticulum calcium pump is active. We tested whether the sarcoplasmic reticulum is a necessary pathway for calcium extrusion in the porcine coronary artery by inhibiting the sarcoplasmic reticulum Ca21-ATPase using thapsigargin. Thapsigargin alone contracts the coronary strip by 2.361.8% of a maximal ACh contraction (n54). Thapsigargin applied before sodium removal did not signiÞcantly affect the effect of the Na1-Ca21 exchanger inhibition. In that case the contraction reached 89612% (n58), and was not signiÞcantly different from that observed in the absence of thapsigargin (88%) (for review see 11). Thus, our results show that the sarcoplasmic reticulum is not the only way of calcium extrusion in the porcine coronary artery as already demonstrated using a rabbit vena cava (12). Discussion It is difÞcult to speciÞcally inhibit an ionic pump. Particularly, the removal of extracellular sodium can cause side effects. First of all, it was proposed that the inhibition of the Na1-H1 exchanger acidiÞes the cytosol and thus modiÞes the cellular response (13). However, the acidiÞcation is slow (14). Therefore this phenomenon cannot interfere with the contraction we observed which is immediate and quickly reaches a plateau. Secondly, the removal of extracellular sodium could reverse the direction of the Na1-Ca21 exchanger thus facilitating calcium diffusion from the outside to the inside of the cell (15). This would participate to calcium entry. However we observed that the inhibition of the Na1-Ca21 exchanger by dichlorobenzamil, which does not facilitate calcium entry, contracts the strip in the presence and in the absence of the ionophore. Furthermore, the contraction, which is weak in the absence of the ionophore, becomes stronger in its presence. This allows us to conclude that the Na1Ca21 exchanger is important when large amounts of calcium have to be extruded from the cells. The smooth muscle Na1-Ca21 exchanger is a low afÞnity high capacity exchanger; therefore, it could strongly increase its activity when cytosolic free calcium increases (16). This extrusion mechanism would be efÞcient enough to match the progressive entry of calcium caused by the ionophore, but not to immediately match the explosive increase in calcium caused by either the ACh or the opening of voltage gated calcium channels. Three distinct ways for the extracellular calcium to diffuse into the cytosol are considered: the leak calcium channel, the carrier formed by the ionophore and the voltage dependent calcium channel opened by the depolarization caused by a high extracellular potassium concentration (8). For these three ways of calcium entry, our results indicate that it is mainly the Na1-Ca21 exchanger which is responsible for the extrusion of cytosolic free calcium from smooth muscle cells of porcine coronary arteries.
S. Budel, J-L. Beny / Life Sciences 67 (2000) 549Ð557
557
The importance of the Na1-Ca21 exchanger differs from one smooth muscle type to another. So, in the airway smooth muscle, the Na1-Ca21 exchanger makes a minor contribution to calcium homeostasis (17). However, the importance of this exchanger for the cell calcium homeostasis was already demonstrated for arterial smooth muscle cells. In the swine common carotid arteries and in the rat aorta, the Na1-Ca21 exchanger has been shown to be the major extrusion mechanism (13Ð15,18). On the contrary, in the porcine coronary endothelial cells, it is likely that the calcium extrusion mechanisms are not so important. As a matter of fact, a calcium ionophore increases cytosolic free calcium and hyperpolarizes the endothelial cells without the necessity to inhibit any calcium extrusion mechanism (1,19). Altogether these results suggest that in porcine coronary arteries, the homeostasis of calcium results from the potential important change of activity of the Na1-Ca21 exchanger, which controles the extrusion of calcium and thus the contraction of the artery. This model resolves the paradoxes evoked in the introduction: Þrst of all, the inefÞciency of a calcium ionophore to increase cytosolic free calcium for the calibration of this parameter using the ßuorescence technique; secondly: the fact that a calcium ionophore fully relaxes in an endothelium-dependent manner an intact strip of porcine coronary artery; thirdly: the fact that the ionophore does not cause a contraction of a strip without endothelium. Acknowledgments We thank Fran•oise Gribi for her excellent technical assistance and Alexander Schuster for reviewing the manuscript. This work was supported by the Swiss National Science Foundation (grant 3100-49163.96). References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
P-Y VON DER WEID, J-L BƒNY, Am J Physiol 262 H1823ÐH1831 (1992). G GRYNKIEWICZ, M POENIE, RY TSIEN, J Biol Chem 260 3440Ð3450 (1985). F CHABAUD, M DANNA, J-L BƒNY, Life Sci 54 1449Ð1458 (1994). V VAN DER BENT, J-L BƒNY, Br J Pharmacol 103 1877Ð1882 (1991). PL BECKER & FS FAY Am. J. Physiol. 253 C613ÐC618 (1987). L MISSIAEN, H DE SMEDT, G DROOGMANS, F WUYTACK, L RAEYMAEKERS, R CASTEELS, Biochim Biophys Acta 1023 449Ð454 (1990). C GATTO, CC HALE, W WU, MA MILANICK, Biochemistry 34 965Ð972 (1995). ME OÕDONNELL, NE OWEN, Physiological reviews 74 683Ð721 (1994). AP SOMLYO & AV SOMLYO, Nature 372 231Ð236 (1994). RA MURPHY, FASEB Journal 8 311Ð318 (1994). C VAN BREEMEN, Q CHEN, M LAHER, Trends Pharmacol Sci 16 98Ð105 (1995). M NAZER, C VAN BREEMEN, Am J Physiol 274 H123Ð131 (1998). S BOVA, WF GOLDMAN, XJ YAUAN, MP BLAUSTEIN, Am J Physiol 259 H409Ð423 (1990). DC BATLLE, M GODINICH, MS LAPOINTE, E MUNOZ, F CARONE, N MEHRING, Am J Physiol 261 C845ÐC856 (1991). T ASHIDA, MP BLAUSTEIN, J Physiol 392 617Ð635 (1987). MJ TAGGART, S WRAY, Cell calcium 22 333Ð341 (1997). LJ JANSSEN, DK WALTRES and J WATTIC, Am J Physiol 273 L322ÐL330 (1997). CM REMBOLD, H RICHARD, XL CHEN Hypertension 19 308Ð313 (1992). A DOMENIGHETTI, JL BENY, F CHABAUD AND M FRIEDEN, J Physiol 513 103Ð116 (1998).