Donate Help Contact The AHA Sign In Home
American Heart Association
Arteriosclerosis, Thrombosis, and Vascular Biology
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents | Next Article »
Arteriosclerosis, Thrombosis, and Vascular Biology. 1997;17:3519-3526

This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Henriksen, R. A.
Right arrow Articles by Tracy, P. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Henriksen, R. A.
Right arrow Articles by Tracy, P. B.
(Arteriosclerosis, Thrombosis, and Vascular Biology. 1997;17:3519-3526.)
© 1997 American Heart Association, Inc.

Articles

Thrombin-Induced Thromboxane Synthesis by Human Platelets

Properties of an Anion Binding Exosite I– Independent Receptor

Ruth Ann Henriksen; Gennady P. Samokhin; ; Paula B. Tracy

From the Section of Allergy, Asthma, and Immunology, Department of Medicine, East Carolina University, Greenville, NC (R.A.H., G.P.S.); and the Department of Biochemistry and Cell and Molecular Biology Program, University of Vermont, Burlington (P.B.T.).

Correspondence to Ruth Ann Henriksen, Section of Allergy, Asthma, and Immunology, Department of Medicine, East Carolina University, Greenville, NC 27858-4354. E-mail rhenriksen{at}brody.med.ecu.edu


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Abstract These studies have examined the effects of thrombin-related agonists in stimulating thromboxane production by human platelets. The results presented show that (1) the maximal response elicited by thrombin receptor agonist peptide (TRAP) stimulation was 40% to 50% of that seen with thrombin or the thrombin mutant Thrombin Quick I; (2) pretreatment of platelets with prostaglandin E1 or genistein resulted in differential inhibition of thromboxane production in response to TRAP compared with either enzyme agonist; (3) an antibody to the thrombin receptor cleavage site that inhibits increases in intracellular [Ca2+] only partially reduced thromboxane production in response to 5 nmol/L thrombin and 15 nmol/L Thrombin Quick I; (4) preincubation with 20 µmol/L TRAP resulted in desensitization to further stimulation by 100 µmol/L TRAP, but not by 100 nmol/L thrombin; and (5) the response to thrombin after TRAP desensitization was completely inhibited by the tyrosine kinase inhibitor genistein and was independent of an intracellular [Ca2+] flux. The cumulative results may be explained by the existence of two proteolytically activated receptors that result in thromboxane production in response to thrombin. One is the thrombin receptor/substrate, PAR-1. Stimulation through the second receptor/substrate depends on a genistein-sensitive step, is independent of an intracellular Ca2+ flux, and is initiated by a thrombin-activated receptor that does not depend on interaction with anion-binding exosite I, as previously indicated by the relative activity of Thrombin Quick I in stimulating platelet aggregation and thromboxane production. The proposed second thrombin receptor on platelets represents an additional member of the class of proteolytically activated receptors.


Key Words: thrombin • thromboxane • platelets • thrombin receptor


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Thrombin, the final serine protease of the blood coagulation pathway, has a number of effects within the vascular system (reviewed in Reference 11 ), including the prothrombotic functions of fibrin formation and platelet activation. Although it has long been recognized that catalytically active thrombin is required for stimulation of platelet aggregation,2 increased interest in the effects of thrombin in cellular stimulation resulted from the identification of a seven-transmembrane domain G-protein–coupled receptor/substrate for thrombin.3 When this receptor, PAR-1,4 is proteolytically cleaved, the new amino terminus functions as a tethered ligand to initiate signal transduction, and peptides derived from the new amino terminus, thrombin receptor agonist peptides (TRAPs), function as agonists for the uncleaved receptor. In addition to platelet aggregation and release, PAR-1 mediates several other cellular responses.5 6 7 8 9 Subsequent to characterization of this receptor, other investigators have identified two additional proteolytically activated receptors. The first of these, PAR-2, is homologous to the thrombin receptor, but is activated by trypsin and not by thrombin.10 In other studies, platelets from mice lacking the gene corresponding to human PAR-1 still respond to thrombin, indicating that another receptor is present in mouse platelets.11 It was observed earlier that mouse and rat platelets are unresponsive to TRAP.12 More recently, a second thrombin-responsive receptor, PAR-3, has been identified by homology cloning, and the m-RNA has been identified in mouse megakaryocytes as well as in human bone marrow cells.4 Although its functional role has not been clearly defined, it is a good candidate for the thrombin-responsive receptor in mouse platelets. PAR-3 differs from PAR-1 in that it is not responsive to peptides derived from its new amino terminus. These three distinct proteolytically activated receptors/substrates represent a class of cell-surface proteins involved in information transfer across the cell membrane, where a cell-surface protein acts as a substrate for the agonist rather than participating directly in agonist-receptor binding.

For human platelets, TRAP-stimulated responses appear to be equivalent in some cases to those stimulated by thrombin.3 13 14 However in other instances, the responses to TRAP are not equivalent to those obtained with thrombin,14 15 16 raising a question of whether activation of the cloned receptor can fully explain all of the effects of thrombin on human platelets.

In our earlier study with the mutant thrombin, Thrombin Quick I, it was concluded that thrombin stimulation of platelet aggregation and thromboxane production were mediated by different thrombin-platelet interactions.17 This conclusion was derived from the different relative activities of Thrombin Quick I compared with thrombin. Thrombin Quick I, which contains the mutation Arg32(67)§->Cys within anion binding exosite I (the fibrinogen-binding exosite),18 19 has <2% of normal thrombin activity in catalyzing the release of fibrinopeptide A, in the stimulation of prostacyclin production by human umbilical vein endothelial cells, or in stimulating platelet aggregation, when the lag periods to initiation of aggregation are compared. In contrast, by comparing concentrations required to produce a half-maximal response, Thrombin Quick I was 30% as effective as thrombin in stimulating thromboxane production by platelets.17 20

It is now known that PAR-1 interacts with anion binding exosite I of thrombin,22 the site of the mutation in Thrombin Quick I, accounting for the low platelet-aggregating activity of this mutant thrombin. The relatively greater activity of Thrombin Quick I in stimulating thromboxane production suggested that anion binding exosite I is not involved in interaction of thrombin with a second, unidentified receptor/substrate. Since it has now been reported that the hexapeptide derived from the new amino terminus of PAR-1 (TRAP-6) can stimulate thromboxane production by human platelets,12 our current hypothesis is that thromboxane production results from stimulation through both PAR-1 and a second, unidentified thrombin-platelet interaction. The properties of PAR-3 indicate that it is not a likely candidate for this second thrombin-platelet interaction in human platelets. PAR-3 contains a hirudin-like, anionic motif that is expected to interact with anion binding exosite I of thrombin, and the 100-fold difference in sensitivity of this receptor to cleavage by {alpha}- and {gamma}-thrombin indicates that the exosite contributes an important specificity determinant for hydrolysis of PAR-3.4

These studies, which extend our earlier work, were undertaken to further differentiate the two proposed pathways leading to thromboxane production by human platelets in response to thrombin. The results demonstrate differences in thromboxane production and distinguish a TRAP-responsive pathway from an additional thrombin-dependent response. This work has been presented in part at the American Society for Hematology Annual Meeting, Seattle, Washington, December 4, 1995.23


*    Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Methods
down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Materials
Genistein, bovine serum albumin, nonimmune rabbit IgG, prostaglandin E1 (PGE1), and DMSO were purchased from Sigma Chemical Company. Stock solutions of 20 mmol/L genistein and 1 mmol/L PGE1 prepared in DMSO were stored at -20°C. A lyophilized preparation of SFLLRNPNDKYEPF (TRAP-14) corresponding to residues 42 to 55 of the human thrombin receptor3 was a generous gift from William R. Church and Laurie Ouellette (Department of Biochemistry, University of Vermont); the peptide SFLLRN (TRAP-6) was obtained from Bachem. Stock solutions of peptides were dissolved in water and stored at -80°C. Peptide concentrations were determined by amino acid analysis. Human {alpha}-thrombin (thrombin) was prepared as described24 and stored at -80°C as an 11.8 µmol/L solution in 250 mmol/L NaCl, 50 mmol/L Tris-HCl, pH 7.5. Thrombin Quick I and Thrombin Quick II were prepared and stored at -80°C as described previously.25 Rabbit polyclonal antibody directed against the platelet thrombin receptor/substrate residues 32 to 46 was prepared and characterized previously.13 This IgG fraction and control, nonimmune rabbit IgG were dialyzed against 10 mmol/L HEPES, 137 mmol/L NaCl, 2.7 mmol/L KCl, 0.36 mmol/L NaH2PO4, 1 mmol/L MgCl2, 5.6 mmol/L dextrose, pH 7.4 (HEPES-Tyrode's buffer) containing 1 mmol/L Ca2+. Final protein concentrations were determined from the absorbance at 280 nm, after correcting for light scattering determined at 320 nm, with extinction coefficients of 1.5 mL · mg-1 · cm-1,26 molecular weight 150 000 for rabbit IgG, and 1.74 mL · mg-1 · cm-1,27 molecular weight 36 000 for thrombin and the mutant thrombins.

Platelet Preparation
Blood was obtained by the two-syringe technique after obtaining informed consent, from healthy, nonsmoking adults denying use of antiplatelet medication for 10 days before phlebotomy. Whole blood, 6 vol, was anticoagulated with 1 vol 88 mmol/L sodium citrate, 64 mmol/L citric acid, 111 mmol/L dextrose (ACD). These studies were approved by the East Carolina University Policy and Review Committee on Human Research, and all procedures were in accordance with institutional guidelines. Washed platelets were prepared by differential centrifugation essentially as described25 28 with minor modifications. Platelet-rich plasma was obtained by differential centrifugation at 160g for 20 minutes, 22°C, and recentrifuged at 160g to remove additional red blood cells. Platelets were then sedimented by centrifugation at 2000g for 20 minutes, 22°C, resuspended, and washed three times with a solution consisting of 25 vol HEPES-Tyrode's buffer containing 3.5 mg/mL bovine serum albumin (HTA) and 1 vol ACD29 using the same centrifugation conditions. Washed platelets were suspended in HTA at one tenth of the initial blood volume, and their concentration was determined using a Coulter counter. The cellular content of this preparation is >99% platelets, as determined by phase-contrast microscopy. Experiments were performed at a final platelet count of 3.6 to 4.0x108 platelets per milliliter in the presence of 1 mmol/L Ca2+. Determination of platelet [Ca2+]i flux was performed as described.25

Thromboxane Production
In a typical experiment, 85 µL aliquots of a washed platelet suspension (4.2 to 5.0x108 platelets per milliliter) were transferred to glass aggregometer cuvettes (ChronoLog Corporation) and 5 µL of the appropriate inhibitor solution or first agonist diluted into HTA was added. When a solution of effector containing DMSO was used, the concentration of DMSO did not exceed 0.5% (vol/vol). Controls contained DMSO at the same concentration as experimental samples. For experiments using immune or nonimmune IgG, a concentrated platelet sample was diluted into the dialyzed IgG and HTA was added to give the appropriate final concentrations of platelets and IgG. Platelets were incubated without or with stirring (1000 rpm) for 2 to 10 minutes, depending on the particular experiment. Then 10 µL of agonist was added and platelets were stirred for another 1 minute. For investigation of genistein inhibition of the desensitized response, platelets were first incubated without stirring with 20 µmol/L TRAP-14 for 2 minutes followed by addition of either HTA or 60 µmol/L genistein for 2 minutes, after which thrombin was added and platelets were stirred for 1 minute. The platelet suspension was then transferred to a microcentrifuge tube and centrifuged for 1 minute at 16 000g. The supernatant was collected and stored at -80°C before assay. Thromboxane production was determined from the concentration of the stable degradation product thromboxane B2 in the platelet supernatants. Assays were performed in duplicate by competitive enzyme-linked immunosorbent assay (ELISA) using a commercial kit (Neogen Corp). Data are presented without correction for spontaneous thromboxane production, which was approximately 1% of the maximum response.

Statistical Analysis
Experiments were performed three or more times with different platelet donors. Results are reported as mean±SEM. Determination of significance was by Student's t test with P<.05 in a one-tailed test indicating a significant difference.


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 *Results
down arrowDiscussion
 down arrowReferences
 
Dose-Response Effects in Production of Thromboxane
Dose-response curves obtained for thromboxane production by human platelets in response to thrombin, Thrombin Quick I, and TRAP-14 are shown in Fig 1ADown ({circ}). For these experiments, the maximum response obtained for thrombin (94 to 212 ng thromboxane per 108 platelets, n=5) was assigned a value of 100%. The results for 500 nmol/L Thrombin Quick I showed a marginally increased maximal response of 110±3%. Comparison of enzyme concentrations required to produce a half-maximal response in these experiments indicated that Thrombin Quick I was 44% as active as thrombin in stimulating thromboxane production, consistent with our earlier results.17 In contrast, the maximum response obtained for TRAP-14 was only 38±2% of that obtained for thrombin and was essentially unchanged at concentrations ranging from 46 to 147 µmol/L. In separate experiments in which platelets were treated with thrombin or TRAP-6, the maximum response to TRAP-6 occurred at 250 µmol/L and was 51±15% of that seen with 250 nmol/L thrombin. These results are similar to those obtained earlier in which thrombin- and TRAP-stimulated release of labeled arachidonic acid30 31 or of thromboxane from human platelets was determined.15 When the time of incubation with agonist was increased from 1 to 10 minutes, there was no increase in thromboxane production in response to TRAP-14 (data not shown). These results indicate that the maximum thromboxane produced in response to TRAP is significantly less than that produced in response to thrombin or Thrombin Quick I and suggest that TRAP-independent and TRAP-dependent pathways for thromboxane production are present in the platelet.



View larger version (27K):
[in this window]
[in a new window]
  		Figure 1. Dose-response curves for thromboxane (TxB2) production by thrombin-related agonists in the presence and absence of prostaglandin E1 (PGE1). Platelets were stirred for 2 minutes with 1 µmol/L PGE1 or vehicle before addition of agonist, after which stirring was continued for 1 minute. The thromboxane B2 concentration in the supernatant was determined by enzyme-linked immunosorbent assay. The maximum response to thrombin was designated as 100% thromboxane release in each experiment, and other results were compared with this value. The results shown are mean±SEM for five experiments for thrombin and thrombin receptor agonist peptide (TRAP)-14 controls and three experiments for the PGE1 and Thrombin Quick I results. For these experiments, there were five different platelet donors. A, Stimulation by agonist in the presence of vehicle ({circ}); stimulation by agonist in the presence of PGE1 ({bullet}). B, Difference between control and inhibited results obtained for each agonist as percentage inhibition of thromboxane B2 production at each agonist concentration. The lines represent regression fits of the data.

This conclusion was supported further by the differential effects of PGE1 on TRAP versus protease-induced thromboxane production. Stimulation of platelets through PAR-1 is linked to the Gi protein32 33 and results in inhibition of adenylate cyclase with enhancement of thromboxane production. PGE1 raises intracellular cAMP levels and has long been recognized as an inhibitor of thrombin-induced prostaglandin synthesis by platelets.34 The effect of 1 µmol/L PGE1 on thromboxane production in response to the three thrombin-related agonists was determined (Fig 1AUp, {bullet}). When calculated as percent inhibition (Fig 1BUp), the data indicated that PGE1 continued to inhibit TRAP-14–induced thromboxane production with increasing TRAP-14 concentrations, while for thrombin and Thrombin Quick I, the inhibitory effect was markedly decreased at higher agonist concentrations. The TRAP-induced response was significantly inhibited (P<.02, at 91 µmol/L TRAP-14) by elevation of cAMP, which exerts its effects through stimulation of protein kinase A. At 200 nmol/L thrombin and 500 nmol/L Thrombin Quick I, where maximum thromboxane production was observed in the absence of PGE1, the inhibition by PGE1 was not significant (P>.05). These results demonstrate a distinct difference between the effect of PGE1 on TRAP-14–induced stimulation of thromboxane synthesis and that observed with either thrombin or Thrombin Quick I as agonist. Because the TRAP-induced response was subject to greater inhibition by elevated cAMP levels, these results suggest that suppression of thromboxane production in response to this mediator involves inhibition of the PAR-1 signal-transduction pathway.

Proteolysis in Stimulation of Thromboxane Production
The relationship of thromboxane production to binding and proteolysis by thrombin and Thrombin Quick I was further delineated using another dysthrombin, Thrombin Quick II. The latter is a dysthrombin derived from the dysprothrombin Gly558(226)->Val, a mutation in the primary substrate binding pocket of thrombin that results in a loss of activity toward thrombin substrates20 35 without substantially impairing binding to the high-affinity binding site on platelets.25 Treatment of platelets with concentrations up to 600 nmol/L Thrombin Quick II produced no measurable thromboxane (data not shown). This result indicates that proteolysis and not high-affinity binding is required for thromboxane production. Because Thrombin Quick I does not bind platelets with high affinity, the increased thromboxane production compared with the stimulation of platelet aggregation induced by Thrombin Quick I is not simply a consequence of ligand interaction at a high-affinity binding site.17 25 Thus, platelet activation that results in thromboxane production through both PAR-1 and the proposed second thrombin-platelet surface interaction requires proteolysis. Furthermore, platelet activation initiated through PAR-1 utilizes the additional specificity of anion-binding exosite I of thrombin, consistent with the low platelet-aggregating activity of Thrombin Quick I.17

Inhibition by Thrombin Receptor Antibody
If both PAR-1 and a second thrombin-platelet interaction participate in thromboxane production, then blocking access of thrombin to PAR-1 should partially inhibit this response. If a second pathway predominates in Thrombin Quick I–stimulated thromboxane production, blocking PAR-1 would be expected to have less effect on the Thrombin Quick I–induced response than on the thrombin-induced response. An antibody generated to the thrombin cleavage site, residues 32 to 46, of PAR-1 was shown to inhibit similarly the intracellular Ca2+ flux induced by 0.5 nmol/L thrombin or 5.0 nmol/L Thrombin Quick I in human platelets. For stimulation with 0.5 nmol/L thrombin, a maximum inhibition of 97% for the intracellular Ca2+ flux was observed at 2.0 µmol/L IgG.13 This antibody, which also inhibited thrombin-induced platelet aggregation, was used to investigate thromboxane production in response to thrombin and Thrombin Quick I. To minimize possible competition13 of enzyme for the receptor in these experiments, a low concentration of thrombin (5.0 nmol/L) or Thrombin Quick I (15.0 nmol/L) that was sufficient to give a clearly measurable thromboxane response in the absence of antibody was selected (see Fig 1AUp). Stirred platelets were preincubated for 2 minutes with 15 µmol/L immune or nonimmune IgG or with HTA followed by addition of thrombin or Thrombin Quick I. In the presence of receptor antibody, the response to thrombin was reduced to 25±3% of that obtained with nonimmune IgG as a control (P<.005, n=7), indicating that at 5 nmol/L thrombin, a large part, but not all, of thrombin-induced thromboxane production is mediated through PAR-1. The response to Thrombin Quick I was 55±7% of the IgG control, which was also significantly reduced relative to the control (P<.025, n=3). Further, the difference between the thrombin- and Thrombin Quick I–induced responses in the presence of antibody was statistically significant (P<.005). These results indicate that at the enzyme concentrations studied, the thrombin-induced response is predominantly, but not totally, mediated through PAR-1. The Thrombin Quick I response is also apparently mediated through two receptors, with the second pathway accounting for a larger portion of the response than is the case for thrombin. Results obtained without added IgG were not significantly different from the results obtained with nonimmune IgG. In other experiments, it was shown that maximal inhibition of thromboxane production in response to 5 or 10 nmol/L thrombin could be achieved with as little as 2 µmol/L IgG, but at this antibody concentration, inhibition of thromboxane production in response to 25 to 50 nmol/L thrombin was not reproducibly observed (data not shown). Thus, these results, which show partial inhibition of thromboxane production in response to low concentrations of thrombin and Thrombin Quick I, suggest that this response is stimulated through two signal-transduction pathways and that the proposed second pathway may not be dependent on a [Ca2+]i flux. These results are also consistent with the observed partial response of platelets to TRAP, which acts only through PAR-1.

Platelet Receptor Desensitization
Agonist-induced receptor desensitization is used as a tool to assess whether related agonists act through the same receptor. To further differentiate PAR-1 from the proposed second receptor participating in thrombin-induced thromboxane production, additional experiments were performed to examine both homologous and heterologous desensitization with thrombin and TRAP-14. Platelets were pretreated with a low concentration of thrombin, TRAP-14, or buffer as control for 2 or 10 minutes, followed by addition of a second agonist at a concentration sufficient to yield maximal production of thromboxane. Fig 2Down shows the results of these experiments, with the first agonist indicated in each panel. With buffer as the first agonist, shown in panel A, the results indicate the amount of thromboxane produced in response to a high concentration of the second agonist alone. The amounts of thromboxane produced are consistent with those shown in Fig 1AUp, again indicating that TRAP-14 is less effective than thrombin as an agonist for production of thromboxane. In panel B, with 5 nmol/L thrombin as the first agonist, further response to TRAP-14 was eliminated, indicating no further stimulation through PAR-1. Compared with preincubation with buffer alone, the response to 100 nmol/L thrombin as second agonist was decreased significantly after preincubation with 5 nmol/L thrombin for either 2 or 10 minutes (P<.05), indicating that homologous desensitization occurs slowly. Thus, despite desensitization of the TRAP-14–dependent response through PAR-1, thrombin stimulated additional thromboxane production. With 20 µmol/L TRAP-14 as the first agonist, it was predicted that the response stimulated through PAR-1 would again be desensitized. The results, shown in Panel C, indicate that this is the case with no further stimulation of the thromboxane response by TRAP-14. The responses to 100 µmol/L TRAP-14 were significantly decreased at both 2 and 10 minutes compared with the results with buffer as the first agonist (P<.05). These results are consistent with the results shown in Panel B. The rapid occurrence of homologous desensitization to TRAP-14 distinguishes TRAP desensitization from the slower homologous desensitization that occurs in response to thrombin. In contrast, when 100 nmol/L thrombin followed 20 µmol/L TRAP-14 (Panel C), there was no significant desensitization of the response to thrombin even after 10 minutes of treatment with TRAP-14 (P>.05). Thus, in platelets treated with 20 µmol/L TRAP-14, thrombin can still elicit a maximal response, even though no response to additional TRAP-14 is observed. These results provide strong evidence for two thrombin-stimulated pathways for thromboxane production. TRAP stimulation through PAR-1 results in rapid homologous desensitization but does not result in desensitization of the response to thrombin. When platelets are treated first with thrombin, homologous desensitization occurs more slowly, while desensitization of the TRAP response is rapid, as observed when TRAP was the first agonist.



View larger version (20K):
[in this window]
[in a new window]
  		Figure 2. Receptor desensitization in response to repeated agonists. Platelets were treated without stirring with the first agonist, identified in each panel, for 2 or 10 minutes, as indicated. Then, the second agonist, as indicated by shading of the bars, was added and platelets were stirred for 1 minute before centrifugation. The thromboxane B2 (TxB2) concentration in the supernatant was determined by enzyme-linked immunosorbent assay. Results shown are mean±SEM for three experiments with different platelet donors. TRAP indicates thrombin receptor agonist peptide.

The previous observation that treatment of platelets with 2.3 nmol/L thrombin for 5 minutes results in a >90% decrease in antibody binding to the PAR-1 cleavage site36 suggests that the functional PAR-1 and is consistent with our conclusion that thrombin stimulation of platelets can occur through two receptors. To further characterize the response to thrombin after stimulation with 5 nmol/L thrombin, [Ca2+]i flux, known to be stimulated by thrombin through PAR-1,31 was monitored in platelets loaded with fura 2. When these platelets were treated with 5 nmol/L thrombin for 2 minutes, the addition of 100 nmol/L would not be present after the initial treatment with thrombin (Fig 2BUp) and resulted in no additional [Ca2+]i flux. Thus, maximal thromboxane production is observed in response to thrombin in the absence of any additional [Ca2+]i flux, further differentiating two platelet responses to thrombin.

Effects of the Protein Kinase Inhibitor Genistein
Genistein, an inhibitor of tyrosine kinase activity, has been reported to inhibit arachidonate release from platelets.37 Therefore, we examined the effect of this inhibitor on the relative responses of the thrombin-related agonists in production of thromboxane. The results are shown in Fig 3Down. In all cases, thromboxane production was decreased, but the inhibitory effect of genistein on the response to TRAP-14 was less than that seen for the two thrombins. These results indicate that tyrosine kinases are most probably involved in the response to all three agonists, but genistein is a more potent inhibitor of the responses induced by thrombin and Thrombin Quick I at all agonist concentrations. At low agonist concentrations that yield approximately equivalent amounts of thromboxane (26% to 34%), the percentage inhibition is significantly lower (P<.005) for TRAP-14 than for either thrombin or Thrombin Quick I. Thus, the inhibition due to genistein appears to be principally associated with the TRAP-independent response. These results are in agreement with previous reports that genistein inhibits release of the arachidonate metabolites, thromboxane37 and 12-hydroxy-5,8,10-heptadecatrienoic acid.38



View larger version (14K):
[in this window]
[in a new window]
  		Figure 3. Dose-response curves for thromboxane (TxB2) production by thrombin-related agonists in the presence and absence of genistein. Platelets were stirred for 2 minutes with 60 µmol/L genistein or vehicle before addition of agonist, after which stirring was continued for 1 minute. The thromboxane B2 concentration in the supernatant was determined by enzyme-linked immunosorbent assay. The results shown are mean±SEM for three experiments with different platelet donors. The maximum response in each experiment was designated as 100% thromboxane release, and other results were compared with this value. {circ} represents stimulation by agonist in the presence of vehicle; {bullet}, stimulation by agonist in the presence of genistein. The lines represent regression fits of the data. TRAP indicates thrombin receptor agonist peptide.

Effect of Genistein After Desensitization by TRAP
To further explore the properties of the proposed second pathway, the effects of genistein on thromboxane production in response to thrombin were determined in TRAP-14–desensitized platelets. The results of genistein inhibition of thromboxane production in response to thrombin, shown in Fig 4ADown, serve as a control for the results shown in Fig 4BDown and are similar to results shown in Fig 3Up. The thrombin concentration dependence for thromboxane production after a 2.0-minute desensitization by 20 µmol/L TRAP-14 in the presence of genistein, Fig 4BDown, demonstrated that there is no additional thromboxane produced in response to thrombin. These results clearly show that after desensitization by TRAP, the TRAP-independent component of thromboxane production requires participation of a genistein-sensitive intermediate, consistent with the results shown in Fig 3Up for the nondesensitized response. This observation raises the interesting possibility that the second pathway may be initiated through a receptor tyrosine kinase or a receptor that is coupled to a tyrosine kinase.



View larger version (16K):
[in this window]
[in a new window]
  		Figure 4. Effect of genistein after desensitization by thrombin receptor agonist peptide (TRAP). Platelets were preincubated for 2 minutes with HTA in A or with 20 µmol/L TRAP-14 in B. For both panels, this step was followed by incubation with either HTA ({circ}) or 60 µmol/L genistein ({bullet}) for 2 minutes, after which platelets were stirred with the indicated concentration of thrombin for 1 minute. The dashed line in B indicates the amount of thromboxane (TxB2) produced by 20 µmol/L TRAP-14 only. All results are expressed as percent of the maximum response to thrombin in the absence of genistein before desensitization. The lines represent regression fits of the data.


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 *Discussion
down arrowReferences
 
The cumulative results presented support the conclusion that thrombin stimulates thromboxane production through two proteolytically activated receptors on the platelet surface. One pathway, which does not result in maximal thromboxane production, is stimulated and downregulated by TRAP through PAR-1.3 The second pathway, which is inhibited by genistein, apparently requires at least one tyrosine kinase but does not depend on a [Ca2+]i flux. From the results obtained with Thrombin Quick I, we have concluded that the second pathway does not involve participation of thrombin anion-binding exosite I.17 From studies of platelet aggregation in response to {alpha}-thrombin, {gamma}-thrombin, or TRAP, Seiler and coworkers39 have also concluded that thrombin stimulates platelet activation through multiple receptors, one of which does not involve the specificity determinants associated with the anion-binding exosite of thrombin.

A diagram indicating our current working hypothesis for these thrombin-platelet surface interactions is shown in Fig 5Down. In this model, thrombin and Thrombin Quick I both act through these two distinct receptors, but with differing concentration dependence. TRAP stimulates only PAR-1. These pathways are further defined by their differential sensitivity to the inhibitors of thromboxane production, PGE1 and genistein. Whether these two pathways converge at phospholipase A2 is not known. There is, however, evidence indicating that more than one form of phospholipase A2 may participate in arachidonate release in platelets.40 41 Results presented here supporting the conclusion that there are two receptors for thrombin-induced platelet stimulation are (1) the difference in maximal thromboxane production in response to thrombin and TRAP; (2) the ability of PGE1 and genistein to differentially inhibit thrombin- and TRAP-induced thromboxane production; (3) the partial and differential inhibition of thrombin- and Thrombin Quick I–induced thromboxane production by an antibody to the PAR-1 cleavage site; (4) that TRAP-14, which activates only one receptor type, does not cause desensitization of thromboxane production in response to thrombin; and (5) the requirement of a genistein-sensitive step for stimulation of the desensitized response, which is not dependent on a [Ca2+]i flux. Further distinguishing the TRAP and thrombin responses, it may be noted that the TRAP-14 concentration required to stimulate thromboxane production is similar to that required to stimulate platelet aggregation,3 the assembly of the prothrombinase complex,13 or [Ca2+]i flux (Bouchard and Tracy, unpublished observations, 1994), responses in which thrombin and TRAP-14 have approximately equivalent efficacy. However, the concentration of thrombin required to give maximal thromboxane production is more than an order of magnitude greater than that required to stimulate maximal [Ca2+]i flux or platelet aggregation. Although other investigators have concluded that there are two thrombin receptors present on platelets, this is the first demonstration that thromboxane production in response to thrombin and Thrombin Quick I are only partially inhibited by an antibody to the PAR-1 cleavage site, that the responses to thrombin and TRAP are differentially desensitized, and that the response after TRAP desensitization is both completely inhibited by 60 µmol/L genistein and independent of [Ca2+]i.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 5. Schematic diagram indicating characteristics of two pathways for thromboxane (TxA2) production. The results of the dose response, antibody inhibition, and receptor desensitization studies, as well as the differential responses to prostaglandin E1 and genistein, support the conclusion that two receptors are involved in thrombin stimulation of platelets, both of which may result in thromboxane production. PAR-1, shown as the G-protein–coupled receptor, possesses specificity determinants that interact with anion-binding exosite I of thrombin and the second receptor with an undefined signaling mechanism that apparently depends only on the specificity of the primary substrate binding site. The inhibition indicated for protein kinase A (PKA) and genistein are the predominant, but not necessarily exclusive, effects on the indicated pathway. Whether these pathways involve one or multiple phospholipase A2 (PLA2) activities is not known.

A possible alternate explanation for the differential effects of thrombin and TRAP in stimulating thromboxane production is that the kinetics of receptor activation differ for these agonists. However, for platelet aggregation, TRAP stimulates a response equivalent to that seen with thrombin.3 For Thrombin Quick I, which would be expected to activate the receptor more slowly than thrombin,20 there is increased efficacy in thromboxane production compared with the stimulation of platelet aggregation.17 Thus, a simple kinetic effect does not appear to explain the relative efficacy of these agonists in thromboxane production.

Another possible interpretation of the results is that TRAP acts as a partial agonist42 for PAR-1. Again, if the three agonists used in this study were acting through one receptor, these agonists would be expected to have similar relative efficacy in different responses, such as platelet aggregation and thromboxane production. This is not the case, because compared with thrombin, Thrombin Quick I has greater efficacy in stimulating thromboxane production than in stimulating platelet aggregation. In contrast, compared with thrombin, the relative efficacy of Thrombin Quick I in producing prostacyclin from endothelial cells is similar to its relative efficacy for platelet aggregation.17 Further, it is observed that the concentration of thrombin required for stimulation of maximal thromboxane production relative to that required for platelet aggregation is increased by about one order of magnitude, while the concentrations of TRAP needed to stimulate maximal aggregation and thromboxane production are quite similar. Partial agonists usually have distinct structures from the full agonist and interact differently with the receptor.42 The TRAP used as an agonist in these studies has the same amino acid sequence as the natural ligand, although the tethered ligand may have additional binding interactions. Compared with thrombin, TRAP stimulates only 40% to 50% of the thromboxane production but stimulates an equivalent extent of platelet aggregation. Further, we have found that TRAP-6 shows a maximal efficacy similar to that of TRAP-14 in stimulating thromboxane production. If partial agonism were present, it might be anticipated that the shorter peptide would be less efficacious because it possesses less of the structure of the physiological agonist and thus has less potential for interacting with the receptor. These relative activities argue against partial agonism of one receptor as an explanation for our observations.

Currently, three protease-activated receptors have been reported, including the seven-transmembranedomain G protein–coupled thrombin receptor PAR-1 and two additional receptors, PAR-2 and PAR-3, identified by homology cloning. Of these, PAR-2 is activated by trypsin or tryptase and like PAR-1 responds to its tethered-ligand–derived peptide.10 43 The third member of this group, PAR-3, when expressed in Cos 7 cells responds to thrombin with stimulation of phosphoinositide hydrolysis. This receptor contains a hirudin-like domain and is quite probably a functional thrombin receptor in murine platelets. Like PAR-1, interaction with anion-binding exosite I of thrombin is an important specificity determinant for hydrolysis of PAR-3, but PAR-3 does not respond to peptides derived from the new amino terminus formed on thrombin hydrolysis.4 The physiological consequences of activation of PAR-3 are not yet well defined. These three receptors, each with distinct features, form a class of proteolytically activated receptors. The proposed second thrombin receptor/substrate that contributes to thromboxane production in human platelets is expected to be still another distinct member of this class of receptors. Because interaction at anion-binding exosite 1 is not required for activation of the proposed receptor by thrombin, it is clearly distinguished from PAR-3. Identification of this additional receptor will provide further insight into mechanisms of protease-stimulated cellular activation and may also suggest another route for intervention in the thrombin-stimulated production of thromboxane, a potent platelet agonist that contributes to thrombotic disease processes.


*    Acknowledgments
 
This study was supported by NIH grant HL 45194 to R.A. Henriksen and HL P01-46703 (Project 4) to P.B. Tracy. R.A. Henriksen is an established investigator of the American Heart Association. TRAP-14 was a generous gift from William R. Church and Laurie Ouellette, Department of Biochemistry, University of Vermont, Burlington. The authors thank Gareth C. Witham for performing the experiments with TRAP-6, Reem Sallam for performing experiments determining [Ca2+]i flux, and C. Kendrick Dunham for preparing Fig 5Up.


*    Footnotes
 
1 Residues of thrombin are numbered according to the prothrombin sequence,21 with the chymotrypsinogen numbering19 in parentheses. Back

Received March 15, 1997; accepted October 9, 1997.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 up arrowDiscussion
 *References
 

  1. Hemker HC. Thrombin generation, an essential step in haemostasis and thrombosis. In: Bloom AL, Forbes CD, Thomas DP, Tuddenham EGD, eds. Haemostasis and Thrombosis. Edinburgh, UK: Churchill Livingstone; 1994:477–490.
  2. Phillips DR. Thrombin interaction with human platelets: potentiation of thrombin-induced aggregation and release by inactivated thrombin. Thromb Diathesis Haemorrhag. 1974;32:207–215.
  3. Vu TH, Hung DT, Wheaton VI, Coughlin SR. Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell. 1991;64:1057–1066.[Medline] [Order article via Infotrieve]
  4. Ishihara H, Connolly AJ, Zeng D, Kahn ML, Zheng YW, Timmons C, Tram T, Coughlin SR. Protease-activated receptor 3 is a second thrombin receptor in humans. Nature. 1997;386:502–506.[Medline] [Order article via Infotrieve]
  5. Huang R, Sorisky A, Church WR, Simons ER, Rittenhouse SE. "Thrombin" receptor-directed ligand accounts for activation by thrombin of platelet phospholipase C and accumulation of 3-phosphorylated phosphoinositides. J Biol Chem. 1991;266:18435–18438.[Abstract/Free Full Text]
  6. Hung DT, Vu TH, Nelken NA, Coughlin SR. Thrombin-induced events in non-platelet cells are mediated by the unique proteolytic mechanism established for the cloned platelet thrombin receptor. J Cell Biol. 1992;116:827–832.[Abstract/Free Full Text]
  7. Pilcher BK, Kim DW, Carnery DH, Tomasek JJ. Thrombin stimulates fibroblast-mediated collagen lattice contraction by its proteolytically activated receptor. Exp Cell Res. 1994;211:368–373.[Medline] [Order article via Infotrieve]
  8. Sower LE, Froelich CJ, Carney DH, Fenton JW II, Klimpel GR. Thrombin induces IL-6 production in fibroblasts and epithelial cells: evidence for the involvement of the seven-transmembrane domain (STD) receptor for {alpha}-thrombin. J Immunol. 1995;155:895–901.[Abstract]
  9. Rabiet M, Plantier J, Rival Y, Genoux Y, Lampugnani M, Del Mar EG. Thrombin-induced increase in endothelial permeability is associated with changes in cell-to-cell junction organization. Arterioscler Thromb Vasc Biol. 1996;16:488–496.[Abstract/Free Full Text]
  10. 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.[Abstract/Free Full Text]
  11. Connolly AJ, Ishihara H, Kahn ML, Farese RV Jr, Coughlin SR. Role of the thrombin receptor in development and evidence for a second receptor. Nature. 1996;381:516–519.[Medline] [Order article via Infotrieve]
  12. Kinlough-Rathbone RL, Rand ML, Packham MA. Rabbit and rat platelets do not respond to thrombin receptor peptides that activate human platelets. Blood. 1993;82:103–106.[Abstract/Free Full Text]
  13. Hayes KL, Leong L, Henriksen RA, Bouchard BA, Ouellette L, Church WR, Tracy PB. {alpha}-Thrombin-induced human platelet activation results solely from formation of a specific enzyme-substrate complex. J Biol Chem. 1994;269:28606–28612.[Abstract/Free Full Text]
  14. Kinlough-Rathbone RL, Perry DW, Guccione MA, Rand ML, Packham MA. Degranulation of human platelets by the thrombin receptor peptide SFLLRN: comparison with degranulation by thrombin. Thromb Haemost. 1993;70:1019–1023.[Medline] [Order article via Infotrieve]
  15. Kramer RM, Roberts EF, Hyslop PA, Utterback BG, Hui KY, Jakubowski JA. Differential activation of cytosolic phospholipase A2(cPLA2) by thrombin and thrombin receptor agonist peptide in human platelets. J Biol Chem. 1995;270:14816–14823.[Abstract/Free Full Text]
  16. Satoh K, Ozaki Y, Asazuma N, Yatomi Y, Ruomei Q, Kuroda K, Yang L, Kume S. Differential mobilization of tyrosine kinases in human platelets stimulated with thrombin or thrombin receptor agonist peptide. Biochem Biophys Res Commun. 1996;225:1084–1089.[Medline] [Order article via Infotrieve]
  17. Henriksen RA, Brotherton AFA. Evidence that activation of platelets and endothelium by thrombin involves distinct sites of interaction. J Biol Chem. 1983;258:13717–13721.[Abstract/Free Full Text]
  18. Henriksen RA, Mann KG. Identification of the primary structural defect in the dysthrombin, Thrombin Quick I: substitution of Cys for Arg-382. Biochemistry. 1988;27:9160–9165.[Medline] [Order article via Infotrieve]
  19. Bode W, Mayr I, Baumann RH, Huber R, Stone SR, Hofsteenge J. The refined 1.9 Å crystal structure of human {alpha}-thrombin: interaction with D-Phe-Pro-Arg-chloromethylketone and significance of the Tyr-Pro-Pro-Trp insertion segment. EMBO J. 1989;8:3467–3475.[Medline] [Order article via Infotrieve]
  20. Henriksen RA, Owen WG. Characterization of the catalytic defect in the dysthrombin, Thrombin Quick. J Biol Chem. 1987;262:4664–4669.[Abstract/Free Full Text]
  21. Degen SJF, MacGillivray RTA, Davie EW. Characterization of the complementary deoxyribonucleic acid and gene coding for human prothrombin. Biochemistry. 1983;22:2087–2097.[Medline] [Order article via Infotrieve]
  22. Ishii K, Gerszten R, Zheng YW, Welsh JR, Turck CW, Coughlin SR. Determinants of thrombin receptor cleavage: receptor domains involved, specificity, and role of the P3 aspartate. J Biol Chem. 1995;270:16435–16440.[Abstract/Free Full Text]
  23. Henriksen RA, Tracy PB. Further evidence for two thrombin receptors on human platelets from studies of thromboxane production. Blood. 1995;86:548a. Abstract.
  24. Henriksen RA, Owen WG, Nesheim ME, Mann KG. Identification of a congenital dysthrombin, Thrombin Quick. J Clin Invest. 1980;66:934–940.
  25. Leong L, Henriksen RA, Kermode JC, Rittenhouse SE, Tracy PB. The thrombin high-affinity binding site on platelets is a negative regulator of thrombin-induced platelet activation. Structure-function studies using two mutant thrombins, Quick I and Quick II. Biochemistry. 1992;31:2567–2576.[Medline] [Order article via Infotrieve]
  26. Little JR, Donahue H. Spectral properties of proteins and small molecules of immunological interest. Methods Immunol Immunochem. 1968;2:343–364.
  27. Fenton JW II, Fasco MJ, Stackrow AB, Aronson DL, Young AM, Finlayson JS. Human thrombins. J Biol Chem. 1977;252:3587–3598.[Abstract/Free Full Text]
  28. Mustard JF, Perry DW, Aralie NG, Packham MA. Preparation of suspensions of washed platelets from humans. Br J Haematol. 1972;22:193–204.[Medline] [Order article via Infotrieve]
  29. Czervionke RL, Hoak JC, Fry GL. Effect of aspirin on thrombin-induced adherence of platelets to cultured cells from the blood vessel wall. J Clin Invest. 1978;62:847–856.
  30. Seiler SM, Peluso M, Michel IM, Goldenberg H, Fenton JW II, Riexingers D, Natarajan S. Inhibition of thrombin and SFLLR-peptide stimulation of platelet aggregation, phospholipase A2 and Na+/H+ exchange by a thrombin receptor antagonist. Biochem Pharmacol. 1995;49:519–528.[Medline] [Order article via Infotrieve]
  31. Lau L, Pumiglia K, Côté YP, Feinstein MB. Thrombin-receptor agonist peptides, in contrast to thrombin itself, are not full agonists for activation and signal transduction in human platelets in the absence of platelet-derived secondary mediators. Biochem J. 1994;303:391–400.
  32. Brass LF, Hoxie JA, Manning DR. Signaling through G proteins and G protein-coupled receptors during platelet activation. Thromb Haemost. 1993;70:217–223.[Medline] [Order article via Infotrieve]
  33. Seiler SM, Michel IM, Fenton JW II. Involvement of the "tethered-ligand" receptor in thrombin inhibition of platelet adenylate cyclase. Biochem Biophys Res Commun. 1992;182:1296–1302.[Medline] [Order article via Infotrieve]
  34. Feinstein MB, Becker EL, Fraser C. Thrombin, collagen and A23187 stimulated endogenous platelet arachidonate metabolism: differential inhibition by PGE1, local anesthetics and a serine-protease inhibitor. Prostaglandins. 1977;14:1075–1093.[Medline] [Order article via Infotrieve]
  35. Henriksen RA, Mann KG. Substitution of valine for glycine-558 in the congenital dysthrombin Thrombin Quick II alters primary substrate specificity. Biochemistry. 1989;28:2078–2082.[Medline] [Order article via Infotrieve]
  36. Molino M, Bainton DF, Hoxie JA, Coughlin SR, Brass LF. Thrombin receptors on human platelets. J Biol Chem. 1997;272:6011–6017.[Abstract/Free Full Text]
  37. Hargreaves PG, Licking EF, Sargeant P, Sage SO, Barnes MJ, Farndale RW. The tyrosine kinase inhibitors, genistein and methyl 2,5-dihydroxycinnamate, inhibit the release of [3H]arachidonate from human platelets stimulated by thrombin or collagen. Thromb Haemost. 1994;72:634–642.[Medline] [Order article via Infotrieve]
  38. Ozaki Y, Yatomi Y, Jinnai Y, Kume S. Effects of genistein, a tyrosine kinase inhibitor, on platelet functions: genistein attenuates thrombin-induced Ca2+ mobilization in human platelets by affecting polyphosphoinositide turnover. Biochem Pharmacol. 1993;46:395–403.[Medline] [Order article via Infotrieve]
  39. Seiler SM, Goldenberg HJ, Michel IM, Hunt JT, Zavoico GB. Multiple pathways of thrombin-induced platelet activation differentiated by desensitization and a thrombin exosite inhibitor. Biochem Biophys Res Commun. 1991;181:636–643.[Medline] [Order article via Infotrieve]
  40. Ballou LR, DeWitt LM, Cheung WY. Substrate-specific forms of human platelet phospholipase A2. J Biol Chem. 1986;261:3107–3111.[Abstract/Free Full Text]
  41. Han X, Gubitosi-Klug RA, Collins BJ, Gross RW. Alterations in individual molecular species of human platelet phospholipids during thrombin stimulation: electrospray ionization mass spectrometry-facilitated identification of the boundary conditions for the magnitude and selectivity of thrombin-induced platelet phospholipid hydrolysis. Biochemistry. 1996;35:5822–5832.[Medline] [Order article via Infotrieve]
  42. Jasper JR, Insel PA. Evolving concepts of partial agonism. Biochem Pharmacol. 1992;43:119–130.[Medline] [Order article via Infotrieve]
  43. Molino M, Barnathan ES, Numerof R, Clark J, Dreyer M, Cumashi A, Hoxie JA, Schechter N, Woolkalis M, Brass LF. Interactions of mast cell tryptase with thrombin receptors and PAR-2. J Biol Chem. 1997;272:4043–4049.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Endocr. Rev.Home page
M. Steinhoff, J. Buddenkotte, V. Shpacovitch, A. Rattenholl, C. Moormann, N. Vergnolle, T. A. Luger, and M. D. Hollenberg
Proteinase-Activated Receptors: Transducers of Proteinase-Mediated Signaling in Inflammation and Immune Response
Endocr. Rev., February 1, 2005; 26(1): 1 - 43.
[Abstract] [Full Text] [PDF]

Home page
 Arterioscler. Thromb. Vasc. Bio.Home page
R. A. Henriksen and V. K. Hanks
PAR-4 Agonist AYPGKF Stimulates Thromboxane Production by Human Platelets
Arterioscler. Thromb. Vasc. Biol., May 1, 2002; 22(5): 861 - 866.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Henriksen, R. A.
Right arrow Articles by Tracy, P. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Henriksen, R. A.
Right arrow Articles by Tracy, P. B.

ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search
Copyright © 1997 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited.