Evidence for the Presence of a Proteinase-Activated Receptor Distinct From the Thrombin Receptor in Vascular Endothelial Cells
Abstract The thrombin receptor was the first cloned G protein–coupled receptor reported to be activated by proteolytic cleavage of its extracellular amino terminus. A second proteinase-activated receptor (PAR-2) was cloned recently and expressed in Xenopus laevis oocytes. PAR-2 was activated by trypsin and by a peptide (SLIGRL) derived from the new amino terminus. Since PAR-2 mRNA was detected in highly vascularized organs, we compared the physiological functions of the thrombin receptor and PAR-2 in vascular endothelium. Thrombin and trypsin both elicited endothelium-dependent relaxations in prostaglandin F2α (PGF2α)–contracted strips of porcine coronary artery. Whereas high doses of both thrombin or trypsin (10 U/mL) caused homologous desensitization, trypsin caused further relaxation of thrombin-desensitized tissues. Thrombin and PAR-2–derived peptides (SFLLRN and SLIGRL) both induced endothelium-dependent relaxations in PGF2α-contracted porcine coronary arteries. SFLLRN or SLIGRL (30 μmol/L) also showed homologous desensitization but not cross desensitization. In the presence of the NO synthase inhibitor NG-monomethyl-l-arginine (1 mmol/L), both SFLLRN- and SLIGRL-induced relaxations were partially inhibited. SFLLRN elicited weak contraction in coronary arteries without endothelium, whereas SLIGRL had no effect. Intravenous injection of SFLLRN (1 mg/kg, bolus) into anesthetized rats elicited a transient depressor response followed by pronounced pressor response. In contrast, intravenous administration of SLIGRL (1 mg/kg, bolus) produced only a marked depressor response. Consistent with the in vivo data, SFLLRN contracted the endothelium-rubbed rat aortic rings and aggregated human platelets in vitro, whereas SLIGRL had no effect. The finding that both trypsin and SLIGRL induced endothelium-dependent relaxations indicates the presence of PAR-2 on endothelial cells. In addition, both trypsin and SLIGRL elicited relaxations in thrombin- or SFLLRN-desensitized tissue, suggesting that PAR-2 is distinct from thrombin receptor in vascular endothelium. The lack of PAR-2–mediated platelet aggregation or smooth muscle contraction suggested it might not share the pathogenic properties associated with the thrombin receptor in the vasculature.
Endothelial cells represent a major interface with plasma proteins and play a key role in modulating platelet adhesion, aggregation, and fibrinolysis.1 Endothelial cells also regulate vascular tone by releasing contractile agents (eg, endothelin) and EDRFs.2 A variety of neurohumoral mediators trigger the release of EDRFs by activating specific receptors. For example, serotonin,3 ADP,4 and thrombin5 released from aggregating platelets respectively activate 5-HT1, P2 purinergic, and thrombin receptors on endothelial cells and subsequently trigger the release of EDRFs. Interestingly, the signal transduction pathways of receptors mediating the release of EDRFs are diverse. Whereas 5-HT1 and α2-adrenergic receptors in porcine and canine coronary arteries are coupled to pertussis toxin–sensitive G proteins, P2 purinergic and bradykinin receptors are insensitive to pertussis toxin.6
Cell-surface receptors are typically activated by ligand binding in a reversible fashion, since the ligand can dissociate from its receptor. The thrombin receptor was the first cloned G protein–coupled receptor reported to be activated by proteolytic cleavage of its extracellular amino terminus.7 This irreversible proteolytic event unmasks a new amino terminus that serves as a tethered peptide ligand, binding intramolecularly to other receptor domains to activate the receptor.8 Recently, Nystedt et al9 cloned a PAR that was 30% identical to the thrombin receptor, with a similar proteolytic activation mechanism. Because the physiological function and endogenous activator for the receptor were unknown, it was provisionally named PAR-2.9
PAR-2, expressed in Xenopus laevis oocytes, is activated by low concentrations of trypsin and the newly formed amino terminal hexapeptide, SLIGRL, but not by thrombin.9 Although the human thrombin receptor is activated by trypsin,8 alignment of the PAR-2/thrombin receptor sequence revealed that the PAR-2 sequence lacks the thrombin/hirudin anion–binding exosite present in the thrombin receptor.9 Therefore, PAR-2 appears to be a novel receptor on the basis of amino acid sequence and ligand specificity.
PAR-2 mRNA is present in highly vascularized organs, including the kidney, small intestine, and stomach, and has been speculated to play a role in the regulation of blood vessel tone.9 In the present study, we compared the physiological functions of thrombin and PAR-2 receptors using in vivo hemodynamic studies and isometric tension recordings of isolated blood vessels. Both trypsin and SLIGRL elicited endothelium-dependent relaxations in PGF2α-contracted porcine coronary artery, suggesting the presence of PAR-2 on vascular endothelium. Trypsin-induced relaxation was blocked by soybean trypsin inhibitors, suggesting that proteolytic activity was essential for activating PAR-2. Furthermore, both trypsin and SLIGRL caused additional relaxations in thrombin- or SFLLRN-desensitized tissue, indicating that PAR-2 is a distinct proteolytically activated receptor.
Materials and Methods
Porcine Coronary Artery
Hearts from adult male swine were obtained from a local abattoir (Dealaman Inc, Warren, NJ) and transported in ice-cold Krebs’ physiological salt solution containing (mmol/L) NaCl 118, KCl 4.7, CaCl2 2.5, MgSO4 1.2, KH2PO4 1.2, NaHCO3 25, disodium EDTA 0.03, and glucose 11. The right coronary artery was dissected, and the adhering perivascular tissue was carefully removed. Typically four or five strips (3 mm) were obtained from each animal.
Rat Thoracic Aorta
Male Sprague-Dawley rats (200 to 300 g, Charles River, Kingston, NY) were euthanized by CO2 inhalation. The aorta was removed quickly from each animal, placed in Krebs’ solution, and cleaned of adhering perivascular tissue. Five or six aortic rings (3 mm) were attained from each animal.
In Vitro Experimental Protocol
Vessels were suspended between a fixed base and a strain gauge for measurement of isometric circumferential tension. Tension changes were measured with a Grass force-displacement transducer (model FT03) and recorded on a Grass polygraph (model 7DA). The tissue bath was filled with Krebs’ physiological salt solution that was kept at 37°C and continuously gassed with 95% O2/5% CO2 to maintain the pH at 7.4. The length of the smooth muscle was increased in a stepwise manner over a 90-minute period to adjust basal tension to 6 g for coronary strips and to 2 g for aortic rings. This tension was found to be optimal for contractions in control experiments determining the length-tension relationship with KCl challenge. In some of the preparations, the endothelium was removed by gentle scraping of the luminal surface.
After equilibration for 40 minutes at optimal tension, the vessels were challenged with 40 mmol/L KCl to confirm viability. Vessels were washed several times with Krebs’ solution and equilibrated for an additional 40 minutes. All experiments were performed in the presence of indomethacin (10 μmol/L) to inhibit PGF2α-induced synthesis of vasoactive products of the cyclooxygenase pathway. Porcine coronary arteries were contracted with 4 μmol/L of PGF2α, and rat aortic rings were contracted with 0.3 μmol/L of phenylephrine before eliciting endothelium-dependent relaxation responses to a single dose of thrombin (1 to 10 U/mL), trypsin (1 to 10 U/mL), SFLLRN (0.3 to 10 μmol/L), SLIGRL (0.3 to 10 μmol/L), or other related peptides. Results were expressed as percent relaxation of the PGF2α- or phenylephrine-induced contraction. To study endothelium-independent responses, a single dose of thrombin (1 to 10 U/mL), trypsin (1 to 10 U/mL), SFLLRN (0.3 to 10 μmol/L), SLIGRL (0.3 to 10 μmol/L), or other related peptides was tested directly on strips without endothelium. Endothelium-independent contractions were expressed as percentage of KCl (120 mmol/L)–induced contraction. Because thrombin and PAR-2 produced a state of homologous desensitization,10 each strip was exposed to a single dose of thrombin or PAR-2 agonist. The IC50 values were calculated as the concentration of thrombin/PAR-2 analogues required to produce 50% relaxation of PGF2α-contracted coronary arteries or phenylephrine-contracted rat aortas.
Human Platelet Aggregation
Washed human platelets were prepared according to the methods described by Radomski and Moncada.11 Briefly, blood (60 mL) was obtained from healthy human volunteers (who had not ingested any platelet-altering drugs for 2 weeks before donation) by venous puncture into Vacutainers containing 4.0% acid citrate dextrose (1 mL for 10 mL whole blood). Prostaglandin (PGE1, 2 μg/mL whole blood) was added, and the blood was centrifuged at 250g for 15 minutes at 15°C. Platelet-rich plasma was removed, supplemented with 0.3 μg/mL PGE1, and centrifuged at 900g for 7 minutes at 15°C. The supernatant was decanted, and platelets were resuspended in 10 mL of wash buffer containing (mmol/L) NaCl 130, KCl 4.74, glucose 11.5, bovine serum albumin (0.2%), HEPES 10, and EGTA 0.02, along with 0.3 μg/mL PGE1. After centrifugation (900g), platelets were resuspended in suspension buffer containing (mmol/L) NaCl 130, KCl 4.74, KH2PO4 1.2, NaHCO3 4, glucose 11.5, bovine serum albumin (0.2%), HEPES 10, MgCl2 1.2, and CaCl2 1.8 without PGE1 at a final concentration of 3×108 platelets per milliliter. The platelet suspension was supplemented with human fibrinogen (400 μg/mL) and stored at 4°C. Platelet aggregation was performed in a dual-channel aggregometer (model 440, Chrono-Log Corp) 1 hour after the final resuspension.
In Vivo Blood Pressure Responses in Anesthetized Rats
Animal experimentation in the present study was carried out in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the Animal Welfare Act in a program accredited by the American Association for Accreditation of Laboratory Animal Care. Male Charles River CD rats (250 to 300 g) were anesthetized with Inactin (100 mg/kg IP). The trachea was catheterized (PE-240) to facilitate spontaneous respiration. The left carotid artery was cannulated (PE-50), and blood pressure was recorded with a Statham pressure transducer connected to a grass polygraph. The right jugular vein was cannulated (PE-50) for injection of drugs and/or vehicle, and body temperature was maintained at 37°C with a Yellow Springs thermostat-controlled heating pad sensitive to rectal temperature changes. After a 30-minute equilibration period, vehicle (saline) or test agents were administered by intravenous bolus, and the changes in blood pressure were monitored for 30 minutes.
The PAR-2–derived hexapeptide (SLIGRL), its retro analogue (LRGILS), and the scrambled analogue (LSRLGI) were assembled from a Rink Amide MBHA resin (Novabiochem) on an ABI model 431A peptide synthesizer using FastMoc chemistry. The side-chain protecting groups for Ser and Arg were, respectively, tert-butyl and 2,2,5,7,8-pentamethylchroman-6-sulfonyl. The peptides were cleaved off the resin and deprotected by trifluoroacetic acid with scavengers (82.5% trifluoroacetic acid/5% H2O/5% phenol/5% thioanisole/2.5% ethanedithiol). The cleaved and deprotected peptides were separated from the resin by filtration, washed, and precipitated by anhydrous ethyl ether. The precipitated peptides were dissolved in H2O, rotary-evaporated to remove the ether, and lyophilized. Crude peptides were purified to 99% purity by reverse-phase HPLC, and the molecular weights were confirmed by mass spectroscopy.
L-NMMA, PGF2α, trypsin (type IX from porcine pancreas, 15 900 U/mg), trypsin inhibitor (type I-S from soybean), hirudin (leech, recombinant), phenylephrine, and indomethacin were obtained from Sigma Chemical Co. The thrombin receptor–derived peptide (SFLLRN) and an inactive analogue (FLLRN) were obtained from Bachem Bioscience Inc. Human thrombin (3080 U/mg) was obtained from Enzyme Research Laboratories, Inc.
Data are expressed as mean±SEM. ANOVA and Student’s t test were performed on an Apple IICi using Microsoft Excel 4.0 package. Values of P<.05 were regarded as statistically significant.
Thrombin- Versus Trypsin-Induced Relaxations
Thrombin (10 U/mL) caused a transient relaxation of PGF2α-contracted porcine coronary arteries with intact endothelium (Fig 1A⇓). Thrombin-induced relaxations were endothelium-dependent, with an IC50 value of 0.45±0.17 U/mL (n=5) (Fig 2A⇓).
Trypsin (10 U/mL) also caused a transient relaxation of PGF2α-contracted porcine coronary arteries with intact endothelium (Fig 1B⇑). The trypsin-induced relaxations were endothelium dependent, with an IC50 value of 0.52±0.20 U/mL (n=5) (Fig 2B⇑). Unlike the thrombin concentration-response curve, the trypsin concentration-response profile was biphasic (Fig 2B⇑). These results suggest that trypsin may evoke endothelium-dependent relaxations by multiple mechanisms.
Tissues exposed to thrombin (10 U/mL) for 20 minutes were desensitized to a second thrombin challenge (Fig 1A⇑). In contrast, relaxation was observed when trypsin (10 U/mL) was added to thrombin-desensitized tissue, suggesting that trypsin can activate other proteinase-activated receptor(s) (Fig 1A⇑). Analogous to the effects of thrombin, a high dose of trypsin (10 U/mL) for 20 minutes also caused refractory desensitization to subsequent stimulation by trypsin (10 U/mL) (Fig 1B⇑). Thrombin was not able to elicit any response in the trypsin-desensitized tissue (Fig 2B⇑). These results are consistent with the possibility that trypsin activates both the thrombin receptor and other proteinase-activated receptor(s).
The thrombin-induced relaxation was completely inhibited by hirudin (Fig 1C⇑), a thrombin anion–binding exosite inhibitor. In contrast, the trypsin-induced relaxation was not blocked by hirudin. In the presence of soybean trypsin inhibitor, trypsin-induced relaxation was inhibited, whereas thrombin-induced relaxation was not (Fig 1D⇑).
Effects of Thrombin Receptor–Derived Peptides
In the presence of intact endothelium, SFLLRN evoked a transient relaxation of PGF2α-contracted porcine coronary arteries. The mean IC50 value of SFLLRN-induced relaxations was 2.0±0.98 μmol/L (n=5) (Fig 3A⇓). The relaxations were endothelium dependent (Fig 3A⇓) and were partially sensitive to the NO synthase inhibitor L-NMMA (1 mmol/L) (Fig 4A⇓).
Eliminating serine from SFLLRN produces an inactive analogue for the platelet thrombin receptor.12 FLLRN (100 μmol/L) had no effect on PGF2α-contracted porcine coronary arteries with intact endothelium (Fig 5⇓), suggesting that the thrombin receptor on vascular endothelium has a selectivity to peptide agonists similar to that of the human platelet thrombin receptor.
Effects of PAR-2–Derived Peptides
SLIGRL elicited a transient relaxation in PGF2α-contracted porcine coronary arteries, with an IC50 value of 2.1±1.07 μmol/L (Fig 3B⇑). No response was observed in endothelium-rubbed arteries (Fig 3B⇑). The relaxation induced by SLIGRL was partially blocked by L-NMMA (Fig 4B⇑). In addition, peptides with the reversed sequence (LRGILS) or scrambled sequence (LSRLGI) had no effect on PGF2α-contracted porcine coronary arteries (Fig 5⇑), indicating that the endothelial response was sequence-selective.
Desensitization of PAR-2 or Thrombin Receptors
Porcine coronary arteries exposed to 30 μmol/L SFLLRN for 20 minutes were unresponsive to a subsequent challenge with SFLLRN (Fig 6A⇓). However, SLIGRL was able to elicit relaxation in the SFLLRN-desensitized tissue (Fig 6A⇓). Correspondingly, exposure of SLIGRL (30 μmol/L) to coronary arterial strips for 20 minutes caused desensitization to subsequent stimulation by SLIGRL (Fig 6B⇓). The SLIGRL-desensitized coronary strips were sensitive to SFLLRN (30 μmol/L) (Fig 6B⇓). These experiments further support the notion that SFLLRN and SLIGRL are acting on distinct receptors.
We used cross-desensitization experiments to determine the selectivity of SFLLRN/thrombin and SLIGRL/trypsin for their respective receptors. Porcine coronary arteries desensitized with SFLLRN were unresponsive to thrombin (Fig 6C⇑). In contrast, trypsin induced relaxations in the SFLLRN-desensitized tissues (Fig 6D⇑). Trypsin-induced relaxations were completely suppressed in SLIGRL-densensitized tissue (Fig 6E⇑), whereas the thrombin-induced responses were not affected (Fig 6F⇑).
Effects of SFLLRN and SLIGRL on In Vivo Blood Pressure Responses in Anesthetized Rats
Intravenous injection of SFLLRN (1 mg/kg, bolus) resulted in a biphasic blood pressure response characterized by an initial depressor response (−25±3 mm Hg, ≈10 to 30 seconds) followed by a pronounced pressor response (50±7 mm Hg, ≈1 to 2 minutes) (Fig 7A⇓). In contrast, intravenous administration of SLIGRL (1 mg/kg, bolus) produced a marked depressor response (−60±4 mm Hg, n=5), which lasted for 2 to 3 minutes (Fig 7B⇓). Because SFLLRN fails to cause activation and/or aggregation of rat platelets in vitro or in vivo in the rat pulmonary microcirculation (2 mg/kg IV),13 it is unlikely that the biphasic response to SFLLRN can be attributed to platelet activation.
Effects of SFLLRN and SLIGRL in Rat Aorta
In the presence of endothelium, both SFLLRN and SLIGRL elicited transient relaxations in phenylephrine (0.3 μmol/L)–contracted rat aortic rings. The maximum relaxation induced by SFLLRN was 74.8±8.0% of phenylephrine-induced contraction, with an IC50 value of 2.05±0.11 μmol/L (n=4) (Fig 8A⇓). In comparison, the maximum relaxation induced by SLIGRL was 97.7±2.4% of phenylephrine-induced contraction, with an IC50 value of 0.70±0.19 μmol/L (n=4) (Fig 8A⇓).
In the absence of endothelium, SFLLRN caused concentration-dependent contractions in rat aortic rings. Maximum SFLLRN (100 μmol/L)–induced contractions were ≈60.3±12.0% of high KCl (120 mmol/L)–induced contractions (Fig 8B⇑). In contrast, SLIGRL did not elicit significant contraction in rat aortic rings (Fig 8B⇑).
Effects of SFLLRN and SLIGRL on Platelet Aggregation
The thrombin receptor–derived peptide (SFLLRN) caused aggregation of washed human platelets, with an EC50 of 6 μmol/L (Fig 9⇓). In contrast, the PAR-2 peptide (SLIGRL) failed to aggregate washed human platelets at concentrations up to 300 μmol/L (Fig 9⇓).
A second, proteolytically activated, seven-transmembrane G protein–coupled receptor was recently cloned from a mouse genomic library.9 This receptor is activated by trypsin and a peptide fragment (SLIGRL) having the sequence of the new amino terminal after proteolytic cleavage but is not activated by thrombin. In the present study, we demonstrate the presence of a PAR in vascular endothelial cells distinct from the thrombin receptor. Although trypsin can activate the human thrombin receptor,7 trypsin-induced relaxation responses in PGF2α-contracted porcine coronary artery strips were not solely mediated by the thrombin receptor. The evidence is that (1) both thrombin- and trypsin-induced relaxations exhibited homologous desensitization but not cross desensitization; (2) trypsin-induced relaxations were completely suppressed in SLIGRL-desensitized tissues, whereas thrombin-induced responses were not affected; and (3) trypsin induced a biphasic concentration-relaxation profile in porcine coronary artery, suggesting that trypsin has a distinct site of action.
Our data with thrombin and PAR-2–activating peptides (SFLLRN and SLIGRL) further suggested that these hexapeptides activated distinct receptors on vascular endothelium. Both SFLLRN and SLIGRL induced endothelium-dependent relaxations in PGF2α-contracted porcine coronary artery strips. The IC50 value for SFLLRN (2.0±0.98 μmol/L) was very similar to that of SLIGRL (2.1±1.07 μmol/L) in the porcine coronary artery. However, SLIGRL (IC50, 0.70±0.19 μmol/L) was more potent than SFLLRN (IC50, 2.05±0.11 μmol/L) in eliciting relaxations of phenylephrine-contracted rat aortic rings (P<.01). FLLRN, an inactive analogue for the human platelet thrombin receptor,12 did not induce relaxation in the PGF2α-contracted porcine coronary artery strips. These results indicated that serine was essential for activating the vascular endothelial thrombin receptor, similar to the observation for the human platelet thrombin receptor. In order to verify the selectivity of PAR-2–activating peptide (SLIGRL), peptides with a reversed sequence (LRGILS) and a scrambled sequence (LSRLGI) were prepared. Both LRGILS and LSRLGI were inactive in inducing endothelium-dependent relaxation, suggesting that the PAR-2–activating peptide had a unique sequence for activating its receptor.
One of the characteristics of cellular responses to thrombin was that receptor activation produced a state of homologous desensitization in which the readdition of thrombin failed to evoke a second response.10 In porcine coronary artery, a high dose of thrombin or trypsin (10 U/mL) caused homologous desensitization. Furthermore, SFLLRN or SLIGRL (30 μmol/L) also elicited homologous desensitization. The observation that thrombin-mediated relaxations were blocked in SFLLRN-desensitized porcine coronary artery confirmed that thrombin and SFLLRN both activated and desensitized the porcine coronary endothelial thrombin receptor. Since trypsin- or SLIGRL-induced relaxations were not affected in the thrombin- or SFLLRN-desensitized tissues, these responses were not likely to be mediated by the thrombin receptor. The finding that trypsin and SLIGRL demonstrated cross desensitization further indicated that they both activated and desensitized PAR-2.
The mechanisms for thrombin- and trypsin-induced vasodilation have not been delineated clearly. Thrombin receptor–derived peptide and thrombin have been shown to directly activate human endothelial thrombin receptor and cause generation of inositol 1,4,5-tris-phosphate and diacylglycerol, with elevation of [Ca2+]i, and increased prostacyclin production.14 Since our experiments were performed in the presence of indomethacin, the contribution of endothelium-derived prostanoids to the relaxation responses of the peptides is unlikely. The thrombin receptor–mediated endothelium-dependent relaxation in porcine coronary artery was partially inhibited by L-NMMA (Fig 4A⇑), an NO synthase inhibitor, suggesting that NO may play a role in the relaxation. Nagao and Vanhoutte15 reported that thrombin also induced hyperpolarization of smooth muscle cells in the porcine coronary artery. This hyperpolarization-induced relaxation was resistant to nitro-l-arginine but was inhibited by high K+ or tetrabutylammonium. Therefore, both NO and an endothelium-dependent hyperpolarization factor may mediate the thrombin-induced relaxation in porcine coronary arteries.
Little is known about the physiological function of PAR-2. When PAR-2 mRNA was expressed in Xenopus laevis oocytes, trypsin or SLIGRL stimulated 45Ca2+ efflux of the cells.9 We observed that trypsin and SLIGRL both elicited endothelium-dependent relaxations in porcine coronary artery. This relaxation was partially blocked by L-NMMA, suggesting that NO may be released upon PAR-2 activation in vascular endothelium. This finding is further supported by earlier data in rat aorta, showing that trypsin-induced relaxation was endothelium dependent and correlated with increased levels of cGMP.16 This would be consistent with a role for PAR-2 in the vascular endothelium, which regulates blood vessel tone and permeability through NO-dependent and NO-independent pathways. Recently, PAR-2 has been identified in human umbilical vein endothelial cells17 (K. Emilsson, unpublished data, 1995). These data correlate well with our findings that PAR-2 activation plays a role in circulatory control.
The endogenous activator of PAR-2 has not been elucidated. The PAR-2 sequence does not contain the anionic binding site for thrombin, and Chinese hamster ovary cells transfected with PAR-2 cDNA did not respond to thrombin,18 suggesting that PAR-2 is not another receptor for thrombin.9 Since PAR-2 was activated by trypsin at concentrations as low as 300 pmol/L,9 the endothelial cell PAR-2 activator may be trypsin or trypsin-like enzymes. Although the trypsin concentration of normal serum is negligible,19 trypsin-like enzymes are released by mast cells or by T-cell activation.20 Additionally, porcine aortic smooth muscle cells have been reported to secrete serine proteases, which significantly modify insulin-like growth factor function in smooth muscle cells.21 Moreover, a number of cancer cell lines produce novel trypsinogen isoenzymes, eg, tumor-associated trypsinogen 2, which is involved in tumor cell invasion and degradation of extracellular matrix.22 Whether these mechanisms are responsible for physiological or pathophysiological activation of PAR-2 in endothelial cells remains to be determined.
Activation of thrombin receptor in a variety of cells, such as platelets and endothelial and vascular smooth muscle cells, has been reported to evoke important biological responses. In the case of platelets, these responses include granule secretion, fibrinogen receptor expression, and the formation of multicellular aggregates. Our data show that SFLLRN, but not SLIGRL, caused aggregation of washed human platelets. Therefore, PAR-2 may not mediate the hemostatic functions associated with thrombin receptor activation in platelets. In the absence of intact endothelium, SFLLRN caused vascular smooth muscle contraction in the rat thoracic aorta. This observation corresponds well with our in vivo data, suggesting that the thrombin receptor is present in vascular smooth muscle cells. In contrast, PAR-2–activating peptide did not have any effect on the endothelium-removed rat thoracic aorta. These results suggest that SLIGRL cannot activate vascular thrombin receptor and that PAR-2 is not present on vascular smooth muscle cells.
In conclusion, the present studies show that PAR-2–derived peptide and trypsin both elicited endothelium-dependent relaxations in porcine coronary artery. These relaxations were not blocked by exposure to thrombin, which leads to homologous receptor desensitization, indicating that the thrombin receptor was not involved in the relaxation. Since tissues exposed to PAR-2–derived peptide were cross-desensitized to trypsin, trypsin-induced relaxation may be mediated through the vascular endothelium PAR-2. The observation that trypsin-induced relaxation was blocked by soybean trypsin inhibitor further suggested that the proteolytic activity was essential for activating PAR-2. PAR-2 activation does not lead to platelet aggregation or smooth muscle contraction; therefore, in contrast to thrombin, it is unlikely to play a pathogenic role in the vasculature.
Selected Abbreviations and Acronyms
|EDRF||=||endothelium-derived relaxing factor|
|PGE1, PGF2α||=||prostaglandins E1 and F2α|
- Received August 18, 1995.
- Accepted January 16, 1996.
- © 1996 American Heart Association, Inc.
De Mey J, Vanhoutte PM. Role of the intima in cholinergic and purinergic relaxation of isolated canine femoral arteries. J Physiol (Lond). 1981;316:437-455.
Ku DD. Coronary vascular reactivity after acute myocardial infarction. Science. 1982;218:576-578.
Nystedt S, Emilsson K, Wahlestedt C, Sundelin J. Molecular cloning of a potential proteinase activated receptor. Proc Natl Acad Sci U S A.. 1994;91:9208-9212.
Hoxie JA, Ahuja M, Belmonte E, Pizarro S, Parton, Brass LF. Internalization and recycling of activated thrombin receptors. J Biol Chem.. 1993;268:13756-13763.
Vassallo RR Jr, Kieber-Emmons T, Cichowsk K, Brass LF. Structure-function relationships in the activation of platelet thrombin receptors by receptor-derived peptides. J Biol Chem.. 1992;267:6081-6085.
Chintala M, Chiu PJS, Bernardino V, Tetzloff G, Tedesco R, Watkins R. Disparate vascular and platelet effects of thrombin receptor activating peptide (TRAP) in rats. FASEB J. 1995;9:A562. Abstract.
Rapoport RM, Draznin MB, Murad F. Mechanisms of adenosine triphosphate, thrombin-, and trypsin-induced relaxation of rat thoracic aorta. Circ Res.. 1984;55:468-479.
Nystedt S, Larsson A-K, Aberg H, Sundelin J. The mouse proteinase-activated receptor-2 cDNA and gene. J Biol Chem.. 1995;270:5950-5955.
Caughey GH. Serine proteinases of mast cell and leucocyte granules: a league of their own. Am J Respir Crit Care Med. 1994;150(pt 2):s138-s142.
Gockerman A, Clemmons DR. Porcine aortic smooth muscle cells secrete a serine protease for insulin-like growth factor binding protein-2. Circ Res.. 1995;76:514-521.
Koivunen E, Ristimaki A, Itkonen O, Osman S, Vuento M, Stenman UH. Tumor-associated trypsin participates in cancer cell-mediated degradation of extracellular matrix. Cancer Res.. 1991;51:2107-2112.