European Journal of Pharmacology, 216 (1992) 135-137
135
© 1992 Elsevier Science Publishers B.V. All rights reserved 0014-2999/92/$05.00
EJP 0349R
Rapid communication
Venous and arterial endothelial cells respond differently to thrombin and its endogenous receptor agonist Serge Simonet, Edith Bonhommc, Michel Laubie, Christophe Thurieau, Jean-Luc Fauch~re a n d T o n y J. V e r b e u r e n lnstitut de Recherches Serl'ier, Dit,isions of Angiology and Peptide Chemistry, 11 Rue des Moulineaux, 92150 Suresnes, France
Received 15 April 1992, accepted 21 April 1992 The effects of thrombin and a peptide mimicking the amino terminus of its receptor, Res (42-55), on vascular reactivity were compared in isolated canine blood vessels. In saphenous veins contracted with endothelin-1, both thrombin and Res (42-55) caused relaxation in rings with endothelium and contraction in rings without endothelium. In coronary arteries, thrombin caused similar responses while Res (42-55) only caused contraction. These data suggest that different thrombin receptors are present on venous and arterial endothelial cells. Endothelium-dependent relaxations; Thrombin; Thrombin receptor agonist (Res 42-55) The serine protease thrombin catalyzes the cleavage of fibrinogen to fibrin and potently activates blood platelets, endothelial cells and vascular smooth muscle cells; the enzyme plays a critical role in hemostasis and in thrombosis. Recently, a functional thrombin receptor was cloned and a proteolytic mechanism for the activation of this receptor was described. Thrombin binds to its receptor via its anionic binding exosite and creates a new amino terminus by cleavage of a single peptide bond located 41 amino acids carboxyl to the receptor's start methionine (Vu et al., 1991a,b). A 14-amino acid peptide Ser-Phe-Leu-Leu-Arg-Asn-ProAsn-Asp-Lys-Tyr-Glu-Pro-Phe which mimics the unmasked new amino terminal is a full agonist for thrombin receptor activation in human platelets. This peptide can be considered as the endogenous thrombin agonist (Vu et al., 1991a,b). The present study was designed to compare the activities of thrombin and its endogenous agonist, which we have called Res (42-55), on venous and arterial thrombin receptors present on either endothelial or vascular smooth muscle ceils. Segments of saphenous veins and coronary arteries were isolated from mongrel dogs (10-20 kg) anesthetized with sodium pentobarbital (30 m g / k g i.v.). Rings of these blood vessels were mounted for isometric tension recording in organ chambers filled with
Correspondence to: T.J. Verbeuren, Division of Angiology, Institut de Recherches Servier, 11 Rue des Moulineaux, 92150 Suresnes, France.
oxygenated physiological saline solution at 37°C in the presence of indomethacin (10/xM) as described earlier (Verbeuren et al., 1990). The endothelium was mechanically removed from some rings by rubbing the intimal surface. After stabilization (60 rain) the segments were briefly contracted with KCI (60 mM). After washout and recovery, a stable, long-lasting contraction was produced in each preparation with human endothelin-1 (ET-1, 0.5 or 1 nM). During these contractions, increasing concentrations of thrombin or the endogenous trombin-receptor agonist, Res (42-55), were added to the blood vessels and the resulting responses were recorded. At the end of the experiments, the presence or absence of the endothelium was verified with 0.1 /xM of acetylcholine (coronary arteries) or 1 IzM of ionophore A 23187 (calcimycin; saphenous veins). A final response to nitroglycerin (10 /xM) or sodium nitroprusside (10/xM) was also recorded in all tissues. Thrombin and Res (42-55) were evaluated in parallel on separate vessel segments. Student's t-test for paired or unpaired observations was used for statistical analyses. H u m a n a-thrombin was purchased from Sigma. Res (42-55) was synthesized by automated solid phase synthesis on a Milligen 9050 apparatus, using the Fmoc. strategy, and t-butyl groups for side-chain protection. In two groups of 12 saphenous vein rings, one group with and one group without endothelium, ET-1 (0.5 nM) evoked stable contractions that averaged respectively 4.61 + 0.51 and 4.40 _+ 0.30 g (values not significantly different). In the veins with endothelium, A 23187 (1 /xM) caused relaxations by 73.1 _ 4.8% of the
136 SAPHENOUS VEIN
10o.
60,
20. 9 8
88
,
7.8
THROMBIN,
-log
M
, 7
Res(42-55),
, 8
, 5 -log
M
C O R O N A R Y ARTERY
_
~2001
200j
0:1
00
9'.'
91a'
"rHROMBIN,
7'a -log
; M
;
Re$(42-55),
'8
; -log
M
Fig. 1. Concentration-response curves for thrombin (]eft) and Res (42-55) (right) in canine saphenous veins (upper) and coronary arteries (lower), previously contracted with ET-1 (0.5 and 1 n M ) respectively. Data from tissues with ( o ) and without (e) endothelium are compared, n = 6 for each group o f saphenous veins; n = 4 for each group o f coronary arteries.
ET-1 response; A 23187 did not cause a significant relaxation (15.2 _+ 12.9%) in the veins without endothelium. In both groups of veins, sodium nitroprusside (10 /~M) caused complete relaxation (98.0 _+ 1.3 and 98.0 _+ 1.1% respectively). In the veins with endothelium, both thrombin (0.13-13 nM) and Res (42-55) (0.04-10/~M) evoked concentration-dependent relaxations of the ET-l-induced contractions (fig. 1). The maximal relaxation caused with thrombin was obtained with the concentration of 1.3 nM. No relaxation was obtained in the veins without endothelium; higher concentrations of thrombin and Res (42-55) evoked additional contractions in these tissues (fig. 1). in two groups of coronary arteries, one group with and one group without endothelium, ET-1 (1 nM) caused stable contractions that averaged 5.85 _+ 0.36 and 4.31 + 0.36 g respectively. The contractions were slightly but significantly lower in the arteries without endothelium than in those with endothelium. A similar decrease was noted for contractions caused by 60 mM of KCh 6.20 +_ 0.42 and 4.00 _+ 0.41 g in arteries with and without endothelium respectively. In the coronary artery rings with endothelium, acetylcholine (0.1 /~M) caused a relaxation of 90.6_+ 3.2% of the ET-1 contraction; no relaxation was obtained in the arteries without endothelium (1.5 _+2.1%). Nitroglycerin (10
/~M) caused complete relaxation in both groups of arteries (by 98.5 _+ 2.3 and 96.3 + 2.6% respectively). In arteries with endothelium, thrombin (0.13-13 nM) evoked concentration-dependent relaxations while Res (42-55) (0.01-10/~M) evoked concentration-dependent contractions (fig. 1). In the arteries without endothelium, both thrombin and Res (42-55) evoked concentration-dependent contractions; with Res (42-55), these contractions were not significantly different from those obtained in arteries with endothelium (fig. 1). For thrombin, the contractions were only significant with the highest concentration (13 riM) used. Our data are the first demonstration of contraction and relaxation caused by the endogenous thrombin receptor agonist, Ser-Phe-Leu-Leu-Arg-Asn-Pro-AsnAsp-Lys-Tyr-Glu-Pro-Phe, which we named Res (4255), in isolated arteries and veins. The data presented above confirm the well known vascular actions of thrombin: endothelium-dependent relaxation in both coronary arteries and saphenous veins and endothelium-independent contraction with higher concentrations; the contractions were more pronounced in the veins than in the arteries (see also Janssens and Verhaeghe, 1982; Eber et al., 1990). The thrombin receptor agonist, Res (42-55), mimicked the relaxations and the contractions induced by thrombin in the saphenous veins. In the coronary arteries, however, Res (42-55) only produced contractions, in tissues both with and without endothelium. The endothelium-dependent relaxations caused by thrombin in the dog coronary arteries could not be demonstrated with Res (42-55). These results illustrate that Res (42-55) can be an important tool for the study of the vascular actions of thrombin. The peptide was able to mimic the contractile effects of thrombin in both arteries and veins as it did in platelets (Vu et al., 1991a,b). A surprising finding was the differential effect of Res (42-55) on the venous and arterial endothelial cells: in the veins the peptide mimicked the endothelium-dependent relaxations induced by thrombin but did not do so in the arteries. A likely explanation is that different thrombin receptors are implicated in the venous and arterial endothelium-dependent relaxations.
Acknowledgement The authors thank Mrs Karine Baudelocq for her excellent secretarial assistance.
References Eber, B., W. Bowie, B. Rowbotham and P.M. Vanhoune, 1990, Recombinant hirudin inhibits endothelium-mediated relaxations evoked by thrombin in isolated canine coronary arteries, J. Vasc. Med. Biol. 2, 62.
137 Janssens, W.J. and R.H. Verhaeghe, 1992, Effect of thrombin on isolated canine blood vessels, Blood Vessels. 19, 126. Verbeuren, T.J., F.H. Jordaens, C.E. Van Hove, A.-E. Van Hoydonck and A.G. Herman, 1990, Release and vascular activity of endothelium-derived relaxing factor in atherosclerotic rabbit aorta, Eur. J. Pharmacol. 191, 173.
Vu, T.-K.H., D.T. Hung, V.I. Wheaton and S.R. Coughlin, 1991a, Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation, Cell 64, 1057. Vu, T.-K.H., V.I. Wheaton, Hung D.T., 1. Charo and S.R. Coughlin, 1991b, Domains specifying thrombin-receptor interaction, Nature 353, 674.
135
© 1992 Elsevier Science Publishers B.V. All rights reserved 0014-2999/92/$05.00
EJP 0349R
Rapid communication
Venous and arterial endothelial cells respond differently to thrombin and its endogenous receptor agonist Serge Simonet, Edith Bonhommc, Michel Laubie, Christophe Thurieau, Jean-Luc Fauch~re a n d T o n y J. V e r b e u r e n lnstitut de Recherches Serl'ier, Dit,isions of Angiology and Peptide Chemistry, 11 Rue des Moulineaux, 92150 Suresnes, France
Received 15 April 1992, accepted 21 April 1992 The effects of thrombin and a peptide mimicking the amino terminus of its receptor, Res (42-55), on vascular reactivity were compared in isolated canine blood vessels. In saphenous veins contracted with endothelin-1, both thrombin and Res (42-55) caused relaxation in rings with endothelium and contraction in rings without endothelium. In coronary arteries, thrombin caused similar responses while Res (42-55) only caused contraction. These data suggest that different thrombin receptors are present on venous and arterial endothelial cells. Endothelium-dependent relaxations; Thrombin; Thrombin receptor agonist (Res 42-55) The serine protease thrombin catalyzes the cleavage of fibrinogen to fibrin and potently activates blood platelets, endothelial cells and vascular smooth muscle cells; the enzyme plays a critical role in hemostasis and in thrombosis. Recently, a functional thrombin receptor was cloned and a proteolytic mechanism for the activation of this receptor was described. Thrombin binds to its receptor via its anionic binding exosite and creates a new amino terminus by cleavage of a single peptide bond located 41 amino acids carboxyl to the receptor's start methionine (Vu et al., 1991a,b). A 14-amino acid peptide Ser-Phe-Leu-Leu-Arg-Asn-ProAsn-Asp-Lys-Tyr-Glu-Pro-Phe which mimics the unmasked new amino terminal is a full agonist for thrombin receptor activation in human platelets. This peptide can be considered as the endogenous thrombin agonist (Vu et al., 1991a,b). The present study was designed to compare the activities of thrombin and its endogenous agonist, which we have called Res (42-55), on venous and arterial thrombin receptors present on either endothelial or vascular smooth muscle ceils. Segments of saphenous veins and coronary arteries were isolated from mongrel dogs (10-20 kg) anesthetized with sodium pentobarbital (30 m g / k g i.v.). Rings of these blood vessels were mounted for isometric tension recording in organ chambers filled with
Correspondence to: T.J. Verbeuren, Division of Angiology, Institut de Recherches Servier, 11 Rue des Moulineaux, 92150 Suresnes, France.
oxygenated physiological saline solution at 37°C in the presence of indomethacin (10/xM) as described earlier (Verbeuren et al., 1990). The endothelium was mechanically removed from some rings by rubbing the intimal surface. After stabilization (60 rain) the segments were briefly contracted with KCI (60 mM). After washout and recovery, a stable, long-lasting contraction was produced in each preparation with human endothelin-1 (ET-1, 0.5 or 1 nM). During these contractions, increasing concentrations of thrombin or the endogenous trombin-receptor agonist, Res (42-55), were added to the blood vessels and the resulting responses were recorded. At the end of the experiments, the presence or absence of the endothelium was verified with 0.1 /xM of acetylcholine (coronary arteries) or 1 IzM of ionophore A 23187 (calcimycin; saphenous veins). A final response to nitroglycerin (10 /xM) or sodium nitroprusside (10/xM) was also recorded in all tissues. Thrombin and Res (42-55) were evaluated in parallel on separate vessel segments. Student's t-test for paired or unpaired observations was used for statistical analyses. H u m a n a-thrombin was purchased from Sigma. Res (42-55) was synthesized by automated solid phase synthesis on a Milligen 9050 apparatus, using the Fmoc. strategy, and t-butyl groups for side-chain protection. In two groups of 12 saphenous vein rings, one group with and one group without endothelium, ET-1 (0.5 nM) evoked stable contractions that averaged respectively 4.61 + 0.51 and 4.40 _+ 0.30 g (values not significantly different). In the veins with endothelium, A 23187 (1 /xM) caused relaxations by 73.1 _ 4.8% of the
136 SAPHENOUS VEIN
10o.
60,
20. 9 8
88
,
7.8
THROMBIN,
-log
M
, 7
Res(42-55),
, 8
, 5 -log
M
C O R O N A R Y ARTERY
_
~2001
200j
0:1
00
9'.'
91a'
"rHROMBIN,
7'a -log
; M
;
Re$(42-55),
'8
; -log
M
Fig. 1. Concentration-response curves for thrombin (]eft) and Res (42-55) (right) in canine saphenous veins (upper) and coronary arteries (lower), previously contracted with ET-1 (0.5 and 1 n M ) respectively. Data from tissues with ( o ) and without (e) endothelium are compared, n = 6 for each group o f saphenous veins; n = 4 for each group o f coronary arteries.
ET-1 response; A 23187 did not cause a significant relaxation (15.2 _+ 12.9%) in the veins without endothelium. In both groups of veins, sodium nitroprusside (10 /~M) caused complete relaxation (98.0 _+ 1.3 and 98.0 _+ 1.1% respectively). In the veins with endothelium, both thrombin (0.13-13 nM) and Res (42-55) (0.04-10/~M) evoked concentration-dependent relaxations of the ET-l-induced contractions (fig. 1). The maximal relaxation caused with thrombin was obtained with the concentration of 1.3 nM. No relaxation was obtained in the veins without endothelium; higher concentrations of thrombin and Res (42-55) evoked additional contractions in these tissues (fig. 1). in two groups of coronary arteries, one group with and one group without endothelium, ET-1 (1 nM) caused stable contractions that averaged 5.85 _+ 0.36 and 4.31 + 0.36 g respectively. The contractions were slightly but significantly lower in the arteries without endothelium than in those with endothelium. A similar decrease was noted for contractions caused by 60 mM of KCh 6.20 +_ 0.42 and 4.00 _+ 0.41 g in arteries with and without endothelium respectively. In the coronary artery rings with endothelium, acetylcholine (0.1 /~M) caused a relaxation of 90.6_+ 3.2% of the ET-1 contraction; no relaxation was obtained in the arteries without endothelium (1.5 _+2.1%). Nitroglycerin (10
/~M) caused complete relaxation in both groups of arteries (by 98.5 _+ 2.3 and 96.3 + 2.6% respectively). In arteries with endothelium, thrombin (0.13-13 nM) evoked concentration-dependent relaxations while Res (42-55) (0.01-10/~M) evoked concentration-dependent contractions (fig. 1). In the arteries without endothelium, both thrombin and Res (42-55) evoked concentration-dependent contractions; with Res (42-55), these contractions were not significantly different from those obtained in arteries with endothelium (fig. 1). For thrombin, the contractions were only significant with the highest concentration (13 riM) used. Our data are the first demonstration of contraction and relaxation caused by the endogenous thrombin receptor agonist, Ser-Phe-Leu-Leu-Arg-Asn-Pro-AsnAsp-Lys-Tyr-Glu-Pro-Phe, which we named Res (4255), in isolated arteries and veins. The data presented above confirm the well known vascular actions of thrombin: endothelium-dependent relaxation in both coronary arteries and saphenous veins and endothelium-independent contraction with higher concentrations; the contractions were more pronounced in the veins than in the arteries (see also Janssens and Verhaeghe, 1982; Eber et al., 1990). The thrombin receptor agonist, Res (42-55), mimicked the relaxations and the contractions induced by thrombin in the saphenous veins. In the coronary arteries, however, Res (42-55) only produced contractions, in tissues both with and without endothelium. The endothelium-dependent relaxations caused by thrombin in the dog coronary arteries could not be demonstrated with Res (42-55). These results illustrate that Res (42-55) can be an important tool for the study of the vascular actions of thrombin. The peptide was able to mimic the contractile effects of thrombin in both arteries and veins as it did in platelets (Vu et al., 1991a,b). A surprising finding was the differential effect of Res (42-55) on the venous and arterial endothelial cells: in the veins the peptide mimicked the endothelium-dependent relaxations induced by thrombin but did not do so in the arteries. A likely explanation is that different thrombin receptors are implicated in the venous and arterial endothelium-dependent relaxations.
Acknowledgement The authors thank Mrs Karine Baudelocq for her excellent secretarial assistance.
References Eber, B., W. Bowie, B. Rowbotham and P.M. Vanhoune, 1990, Recombinant hirudin inhibits endothelium-mediated relaxations evoked by thrombin in isolated canine coronary arteries, J. Vasc. Med. Biol. 2, 62.
137 Janssens, W.J. and R.H. Verhaeghe, 1992, Effect of thrombin on isolated canine blood vessels, Blood Vessels. 19, 126. Verbeuren, T.J., F.H. Jordaens, C.E. Van Hove, A.-E. Van Hoydonck and A.G. Herman, 1990, Release and vascular activity of endothelium-derived relaxing factor in atherosclerotic rabbit aorta, Eur. J. Pharmacol. 191, 173.
Vu, T.-K.H., D.T. Hung, V.I. Wheaton and S.R. Coughlin, 1991a, Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation, Cell 64, 1057. Vu, T.-K.H., V.I. Wheaton, Hung D.T., 1. Charo and S.R. Coughlin, 1991b, Domains specifying thrombin-receptor interaction, Nature 353, 674.
