Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Aspirin (acetylsalicylic acid) has antiplatelet, anti-inflammatory, and neuroprotective actions, but its molecular mechanisms of action are complex. Aspirin was reported two decades ago to inhibit platelet cyclooxygenase activity and the formation of thromboxanes (Patrono, 1994). A growing number of studies suggest, however, that anti-inflammatory actions of aspirin may not be related to the inhibition of prostaglandin synthesis. For instance, salicylates activate p38 mitogen-activated protein kinase and inhibit tumor necrosis factor-induced IB phosphorylation (Schwenger et al., 1997, 1998).

A major clinical use of aspirin therapy is to improve the survival of patients with myocardial infarction and unstable angina (ISIS-2, 1988). This action is usually attributed to the antiplatelet action of aspirin, but participation of other mechanisms of action cannot be excluded. For instance, aspirin has recently been shown to protect endothelial cells against oxidative stresses, possibly by promoting the synthesis of ferritin (Oberle et al., 1998).

Endothelin-1 (ET-1) is a potent vasoconstrictor peptide (Yanagisawa et al., 1988) that plays an important role in several diseases thought to be associated with vasoconstrictions. These are coronary vasospasm (Toyo-oka et al., 1991), unstable angina (Weiczorek et al., 1994), myocardial infarction (Omland et al., 1994), cardiac insufficiency (Mulder et al., 1997), and cerebral vasospasm associated with subarachnoid hemorrhage (Clozel et al., 1993). The fact that protective actions of aspirin are observed in conditions that are associated to increased circulating levels of ET-1 suggested to us that aspirin may have anti-ET-1 properties. This study was designed to test this hypothesis. It shows that salicylates are allosteric inhibitors of ETA receptors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. ET-1, indo-1/AM, indomethacin, neuromedin B, UTP, aspirin, and sodium salicylate (SA) were obtained from Sigma Chemical Co. (St. Louis, MO). ¹²⁵I-ET-1 (2200 Ci/mmol) was prepared as described previously (Desmarets et al., 1996). Bosentan was provided by Dr. M. Clozel (Actelion, Basel, Switzerland). All effectors were dissolved in an Earle's salt solution. Composition of the solution was: 140 mM NaCl, 5 mM KCl, 0.8 mM MgSO₄, 1.8 mM CaCl₂, 25 mM HEPES, pH 7.4.

Cell Cultures. Rat brain capillary endothelial cells of the B7 clone were grown as described previously (Vigne et al., 1995). Stable transfectant CCl39 cells expressing functional rat ETA or rat ETB receptors were prepared as described previously (Gresser et al., 1996). Cell homogenates were prepared as described previously (Desmarets et al., 1996) and stored at 20°C. Proteins were determined according to Bradford (1976) using BSA as standard. Receptor densities were 0.4 ± 0.1 pmol/mg of proteins in membranes prepared from B7 cells (Vigne et al., 1990). They were 26 ± 1 and 4 ± 0.5 pmol/mg in membranes prepared from ETA and ETB receptor-expressing fibroblasts, respectively.

Contraction Experiments. Thoracic aorta from 200-g female Wistar rats were cleaned of adherent fat, cut into rings, and the endothelium was removed. Human internal mammary arteries from seven patients (six men and one woman, age 56-83, mean 63.5 ± 4.3 years) undergoing bypass surgery were collected. All the procedures followed were in accordance with institutional guidelines. In the operating room, fragments were immediately placed in a Krebs' solution buffered at pH 7.4 with 25 mM HEPES/NaOH. Fragments were cleaned of adherent connective tissues. The endothelium was removed by gently rubbing the intimal surface with the tips of small forceps and 3- to 4-mm wide rings were prepared.

Binding Experiments. All experiments were performed at 20°C. Cell homogenates (10 µg of protein/ml) were incubated in the presence of 4 to 20 pM ¹²⁵I-ET-1 and effectors in an Earle's salt solution supplemented with a cocktail of protease inhibitors (0.1 mM bacitracin, 0.1 mM phenylmethylsulfonyl fluoride, 1 µM leupeptin). After selected times of incubation, aliquots of the incubation solution were filtered under reduced pressure onto Sartorius 0.2-µm filters and washed three times with 4 ml of 0.1 M MgCl₂. Filters were then counted. Nonspecific binding was measured in parallel experiments using 100 nM unlabeled ET-1. Equilibrium binding experiments and dissociation experiments were performed after 3 h of equilibration of receptors with ¹²⁵I-ET-1. Triplicate experiments were performed.

Intracellular Ca²⁺ Measurements. For intracellular Ca²⁺ measurements, suspended endothelial cells were incubated for 30 min in the presence of 5 µM indo-1/AM, centrifuged at 1000g, and resuspended into an Earle's salt solution at a density of 10⁶ cells/ml. ET-1, neuromedin B, or UTP was added to the cell suspension. After mild vortexing, tubes were inserted into a FacStar Plus cytometer (Becton Dickinson) (Vigne et al., 1994). Mean fluorescence ratios were determined for 1000 cells sampled between 8 and 10 s after the addition of agonists. This time corresponded to the peak of the intracellular Ca²⁺ transients. Fluorescence ratios were calculated in arbitrary units set to a value of 100 for unstimulated cells.

Nonreactivity of ET-1 and SA. It was important to define whether SA interacted and modified ET-1. ET-1 (4 nmol) was added to an Earle's salt solution supplemented with 20 mM SA. After 75 min of equilibration at room temperature, the mixture was injected on a C18 Lichrocart 250-4 HPLC column (Merck, Darmstadt, Germany). The column was eluted for 60 min at a rate of 1 ml/min with a linear 10 to 50% acetonitrile gradient in the presence of 0.1% trifluoroacetic acid. No evidence for a change in the retention time of ET-1 (49 min) or for a degradation of ET-1 could be obtained.

Data Presentation and Statistical Analysis. Data are given as mean ± S.E. and the number of experiments performed. Statistical analysis was performed with the paired t test. Values of P < .05 were considered statistically significant. Binding data were analyzed using the Ligand software (Jandel Scientifics, Corte Madera, CA). When no error bar is present on a figure, it was smaller than the size of the points.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Anti-ET-1 Actions of Salicylates on Rat Aortic Rings. ET-1 induced large and irreversible contractions of isolated rat aortic rings. These contractions are mediated by ETA receptors (Marsault et al., 1991). Figure 1 presents typical recordings that show that addition of aspirin (Fig. 1A) or SA (Fig. 1B) to arteries that had been precontracted with ET-1 induced rapid and partial relaxations. Data obtained are summarized in Fig. 1, C and D. Figure 1C shows that the half-time for salicylate induced relaxations was independent of the concentration used. The overall half-time of relaxations induced by SA was 2.9 ± 0.6 min (n = 22). The overall half-time of relaxations induced by aspirin was 3.4 ± 0.4 min (n = 21). Figure 1D shows a dose-dependent increase in the maximum extent of the relaxations induced both by aspirin and SA.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Anti-ET-1 actions of salicylates in isolated rat aorta. A and B, representative traces showing the contractile action of 100 nM ET-1 and the relaxing actions of the indicated concentrations of aspirin (A) and SA (B). Horizontal bar: 4 min, vertical bar: 0.5 g. C, half-times for salicylate-induced relaxations. Rings were precontracted with 100 nM ET-1 and relaxations were induced by 3, 10, or 30 mM aspirin (open columns) or SA (filled columns). Half-times of the relaxations were determined graphically. Mean ± S.E. (n = 6-9) are shown. D, maximum relaxations induced by salicylates. Rings were precontracted with 100 nM Et-1 and relaxations were induced by 3, 10, or 30 mM aspirin (open columns) or SA (filled columns). Mean ± S.E. (n = 6-9) are shown.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Cumulative ET-1 dose-response curves in the absence or the presence of 10 mM SA (A) or 30 mM SA (B). Two aortic rings from the same rat were mounted in two parallel organ chambers. One ring served as a control. The other was treated with 10 or 30 mM SA. The two rings were then exposed to increasing doses of ET-1 (1, 3, 10, 30, and 100 nM) and changes in tension were recorded. Mean ± S.E. (n = 12 in A, n = 4 in B) are indicated. **P < .01, *P < .05 using a paired t test. Tensions were normalized to that produced by the last KCl contraction.

Anti-ET-1 Actions of Salicylates on Human Mammary Arteries. Similar experiments were performed using rings of human internal mammary arteries. This preparation has already been reported to respond to ET-1 by large and sustained contractions that are mainly due to ETA receptors (Seo et al., 1994; Maguire and Davenport, 1995). Experiments were performed using endothelium-denuded arteries and in the presence of 10 µM indomethacin to prevent a possible action of aspirin on cyclooxygenase. Figure 3 shows that SA and aspirin relaxed arteries that had been precontracted with ET-1. The time courses of the relaxations were very similar to those observed with rat aortic rings. Figure 3C also shows cumulative dose-response curves for the actions of ET-1 in the absence or presence of SA. SA inhibited actions of all concentrations of ET-1 >1 nM.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Anti-ET-1 actions of salicylates in isolated human mammary arteries. A, representative traces showing relaxing actions of 30 mM SA and 30 mM aspirin on endothelium-denuded rings of mammary arteries. Horizontal bar: 2 min, vertical bar: 0.6 g. C, cumulative ET-1 dose-response curves in the absence () or the presence () of 10 mM SA. Experiments were performed as described in the legend of Fig. 2. Mean ± S.E. (n = 5) are indicated. **P < .01, *P < .05 using a paired t test. Tensions were expressed relative to that observed in response to 100 nM ET-1 in the control ring.

Salicylates Prevented ETA Receptor-Mediated Intracellular Ca²⁺ Mobilization. We next looked for anti-ET-1 actions of salicylates at a cellular level. Endothelial cells from brain capillaries were chosen for different reasons: 1) unlike endothelial cells from large peripheral vessels, they express ETA receptors (Vigne et al., 1990); 2) they express two other phospholipase C-coupled heptahelical receptors: neuromedin B-preferring bombesin receptors (Vigne et al., 1995) and P2Y₂ purinergic receptors (Frelin et al., 1993); and 3) pharmacological properties of receptors can be conveniently monitored by flow cytometric analyses of indo-1-loaded cells (Vigne et al., 1990). Figure 4A shows that SA and aspirin decreased the ET-1 induced [Ca²⁺]i rise in indo-1-loaded cells. Half maximum inhibition was observed at 2 mM SA and 3 mM aspirin.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   SA specifically inhibited ET-1 induced intracellular Ca2+ mobilization in brain capillary endothelial cells. Left: dose-response curve for the inhibition by SA () and aspirin () of ET-1 (100 nM)-induced intracellular Ca2+ mobilization. Right: dose-response curves for the Ca2+ mobilizing action of ET-1. Experiments were performed in the absence () or the presence () of 20 mM SA. Mean ± S.E. (n = 3) are indicated.

Salicylates Inhibited ET-1 Binding to Recombinant ETA Receptors. To further analyze the mechanism of action of salicylates, we performed binding experiments using ¹²⁵I-ET-1 and recombinant rat ETA receptors. Figure 5 shows that salicylates prevented ¹²⁵I-ET-1 binding. Half maximum inhibition was observed at 10 mM aspirin or SA. Bosentan was much more potent.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Inhibition by salicylates of 125I-ET-1 binding to recombinant ETA receptors.125I-ET-1 binding to ETA receptors was measured in the presence of increasing doses of aspirin (), SA (), and bosentan (). The concentration of 125I-ET-1 was 10 pM.

Salicylates Modified the Properties of Interaction of ET-1 with ETA Receptors. We next analyzed the kinetics of association of ¹²⁵I-ET-1 to ETA receptors and the kinetics of dissociation of the complexes in the presence of different concentrations of SA. Results are shown in Fig. 6. Association of ¹²⁵I-ET-1 to ETA receptors in the absence of SA did not follow pseudo first order kinetics for reasons that have been described previously (Desmarets et al., 1996). We observed, however, that association kinetics in the presence of SA could be linearized according to pseudo first order processes. Table 1 shows that increasing doses of SA did not modify the second order rate constant of association of ¹²⁵I-ET-1 to its receptors. Dissociation kinetics in the absence or the presence of SA were monoexponential. The half-life of ¹²⁵I-ET-1 ETA receptor complexes was 12.4 ± 0.5 h (n = 11) in the absence of SA, meaning an almost irreversible binding. SA increased up to 10-fold the rate constant of dissociation of ¹²⁵I-ET-1 receptor complexes (Table 1). The apparent equilibrium dissociation constant (K_d) of ET-1 ETA receptor complexes was estimated to be 1.4 pM in the absence of SA. It increased to 12 pM in the presence of 50 mM SA. A Schild plot of the data indicated an apparent K_i value for SA of 8 mM. Thus SA decreased the apparent affinity of ET-1 for ETA receptors by favoring dissociation of ET-1 receptor complexes.

View larger version (25K): [in this window] [in a new window]   Fig. 6.   Association and dissociation kinetics of 125I-ET-1 to ETA receptors. A, association of 125I-ET-1 (4 pM) to ETA receptors was followed in the absence () or the presence of 20 (), 30 (), or 50 mM () SA. B, after association of 125I-ET-1 to ETA receptors in the absence () or the presence of 20 () or 30 mM SA (), dissociation kinetics were initiated by the addition of 100 nM unlabeled ET-1 to the tubes.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   SA induced a dissociation of 125I-ET-1 ETA receptor complexes. 125I-ET-1 (10 pM) and ETA receptor complexes were first allowed to form. SA was then added and dissociation of the complexes was followed. Concentrations of SA used were: 5 (), 20 (), and 50 mM (). Note that no unlabeled ET-1 was used in these experiments.

View larger version (11K): [in this window] [in a new window]   Fig. 8.   Salicylates did not induce dissociation of 125I-ET-1 ETB receptor complexes. 125I-ET-1 and ETB receptor complexes were first allowed to form. Dissociation kinetics were then initiated by the addition of 100 nM unlabeled ET-1 in the absence () or the presence of 50 mM aspirin () or of 50 mM SA ().

Salicylates Potentiated Actions of Receptor Antagonists. The major interest of an allosteric inhibitor of ETA receptors is that it transforms an almost irreversible ET-1 binding into a reversible process. One consequence is that salicylates should favor replacement of bound ET-1 by receptor antagonist and thus potentiate actions of receptor antagonists. This hypothesis was tested using binding experiments and contraction experiments.

View larger version (12K): [in this window] [in a new window]   Fig. 9.   Salicylates promoted replacement of bound 125I-ET-1 by unlabeled ET-1 and bosentan. A, 125I-ET-1 and ETA receptor complexes were first allowed to form. Complexes were then treated with 100 nM unlabeled ET-1 in the absence () or the presence of 20 mM SA () or aspirin (). B, same as in (A) except that complexes were treated with 10 µM bosentan in the absence () or the presence () of 20 mM SA.

View larger version (11K): [in this window] [in a new window]   Fig. 10.   SA potentiated relaxing actions of bosentan. Aortic rings were exposed to 100 nM ET-1 and then to 10 mM SA and 10 µM bosentan as indicated and changes in tension were recorded. Horizontal bar: 10 min, vertical bar: 0.2 g.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Why Is It Important to Relieve Irreversible ET-1 Binding? Quasi-irreversible ET-1 binding was recognized very early but its consequences are not yet fully understood. Waggoner et al. (1992) first described that the slow rate of dissociation of ET-1 receptor complexes implies K_d values of ET-1 receptor complexes in the pM range. We also showed that equilibrium binding studies can greatly underestimate affinity constants for both ET-1 and competing ligands unless special care is not taken to avoid time-limited second order kinetic conditions (Desmarets et al., 1996). Irreversible binding is responsible for atypical pharmacological responses such as the lack of action of guanine nucleotides (Nambi et al., 1996) and actions of receptor antagonists (Leite et al., 1994). It is responsible for long-lasting actions of ET-1 and the development of long-term refractoriness. Recovery from refractoriness is variable from tissue to tissue. It is very slow in the cardiac muscle (Leite et al., 1994; Hilal-Dandan et al., 1997) but faster in the rat aorta (Marsault et al., 1993) and probably limited by the rate of externalization of newly formed or recycling receptors (Marsault et al., 1993). Quasi-irreversible binding may account for the fact that functional ET receptors serve as clearance receptors (Frelin and Guédin, 1994). It imposes conditions in which autocrine actions (that is, vasodilating) of ET-1 should be favored over paracrine actions (that is, vasoconstricting) as observed in ET-1 +/ mouse and provides a mechanism for the genesis of ET-1 induced vasospasms (Frelin and Guédin, 1994). Finally, irreversible binding implies that ET-1 is more likely to contribute to long-term physiological or pathophysiological regulations than to short-term regulations (Hilal-Dandan et al., 1997). Relieving the irreversibility of ET-1 binding is therefore important for a better understanding of the role of ET-1. It can be achieved using allosteric inhibitors of the receptors. This study shows that aspirin and SA are allosteric inhibitors of ETA receptors.

Salicylates Are Allosteric Modifiers of ETA Receptors. Salicylates reverse ET-1-induced contractions in isolated rat aortic rings and human mammary artery rings (Figs. 1 and 3). These actions were observed in the absence of endothelium and for human tissues, in the presence of indomethacin. This, together with the observation (in the different assays that have been performed) that SA is slightly more potent than aspirin suggests that actions of salicylates are unlikely to be mediated by an inhibition of cyclooxygenase.

Specificity of the Action of Salicylates. Actions of salicylates reported here were observed at large millimolar concentrations of the drugs. At such concentrations, weak organic acids and salicylates are known to alter membrane fluidity (Watala and Gwozdzinski, 1993; Balasubramanian et al., 1997) and to act as mild chaotropic agents. The following evidence suggests that actions of salicylates described here are not nonspecific: 1) SA selectively modifies the rate of dissociation of ET-1 ETA receptor complexes; 2) SA does not alter ETB receptors, neuromedin B-preferring bombesin receptors, and P2Y2 purinergic receptors; 3) Na benzoate did not induce dissociation of ET-1 receptor complexes; and 4) SA promotes dissociation of solubilized ET-1 receptor complexes (data not shown). The possibility that SA induced a degradation of free or receptor-bound ET-1 was also ruled out.

Relaxing Actions of Salicylates and Bosentan. The proposed mechanism of action of SA and aspirin can account (at least qualitatively) for the results obtained in functional assays. These are: 1) that fact that salicylates produced faster relaxations than bosentan, and 2) apparent competitive antagonism observed using intact vessels (Figs. 2 and 3) and endothelial cells (Fig. 4).

Pharmacological Implications. The most important observation of this study is that SA potentiates relaxing actions of bosentan (Fig. 10). Several potent antagonists of ET receptors have been devised (Cheng et al., 1994). They are usually potent in protocols in which animals are treated at the same time with ET-1 and with receptor antagonists, i.e., under conditions in which antagonists and ET-1 compete for the same free receptors. Antagonists are usually less potent when administered to prevent actions of endogenous, receptor-bound, ET-1. One alleged reason is that ET-1 plays a minor role in the physiopathological conditions that have been assessed. The observation that SA potentiates relaxing actions of bosentan (Fig. 10) suggests that actions of receptor antagonists may be limited by the slow rate of dissociation of ET-1 receptor complexes. Indeed, a competitive antagonist can only act if receptors are free from endogenous ET-1. Our results thus suggest that an association of salicylates and ETA receptor antagonists could be useful to evaluate the physiopathological role of ET-1 and may be of therapeutic interest is the treatment of ischemic heart disease.

Clinical Implications. Aspirin has clear beneficial actions in coronary heart disease. Recommended daily doses of aspirin for the prevention of myocardial infarction and vascular death in patients with coronary heart disease are about 200 mg. Larger doses may be needed in patients with cerebrovascular diseases (Patrono and Roth, 1996). This study shows that millimolar concentrations of SA are needed to inhibit contractile actions of ET-1. Similar plasma concentrations of SA are only obtained with high anti-inflammatory doses of aspirin. This indicates that the mechanisms described in this article are unlikely to contribute to the beneficial actions of antiplatelet doses of aspirin in coronary heart disease.