SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 2, 1992
Expression Cloning and Characterization of a Functional Thrombin Receptor Reveals a Novel Proteolytic Mechanism of Receptor Activation
Elucidation of the mechanism by which thrombin activates platelets and other cells has been a longstanding problem in thrombosis research. This problem was tantalizing from both clinical and basic science perspectives. From the clinical view, thrombin elicits a host of cellular activities critical for both normal responses to wounding and for pathologic vascular events. Defining the mechanism of thrombin-induced cell activation promised the advent of new therapeutic agents and reagents for identifying thrombin's role in pathophysiology. From the basic view, the fact that thrombin is a protease held out the possibility that a novel mechanism of receptor activation might be uncovered. In this article we briefly review the recent cloning and characterization of the platelet thrombin receptor by our laboratory.1 As will be discussed, identification of the thrombin receptor revealed a unique proteolytic mechanism of receptor activation and led to the development of a novel agonist peptide capable of activating the thrombin receptor independent of thrombin and thrombin's protease activity. It is our hope that having the cloned receptor in hand will allow the development of thrombin receptor antagonists. Such reagents will define the role of thrombin receptor activation in vivo and may provide the basis for a new class of antithrombotic or antiproliferative pharmaceuticals.
From the Cardiovascular Research Institute, and Department of Medicine, University of California, San Francisco, San Francisco, California. Reprint requests: Dr. Coughlin, Box 0524, University of California, San Francisco, San Francisco, CA 94143.
BACKGROUND Pathophysiology: Cell Activating Functions of Thrombin In addition to its well-known role in cleaving fibrinogen to fibrin, thrombin has a number of important "cell activating" functions (Fig. 1) which are reviewed by Shuman.2 First and foremost, thrombin is the most potent stimulator of platelet aggregation.3 Thrombin is also chemotactic for monocytes,4 mitogenic for lymphocytes and for mesenchymal cells, including vascular smooth muscle cells,5,6 and has a number of effects on the vascular endothelium. These include stimulating endothelial production of prostacyclin,7 platelet-activating factor,8 plasminogen activator-inhibitor,9 and the potent smooth muscle cell mitogen platelet-derived growth factor.10 Thrombin also induces neutrophil adherence to the vessel wall by an endothelial-dependent mechanism,11 probably by inducing surface expression of granule membrane protein (GMP)-140.12 Teleologically, these multiple cell activating functions of thrombin may be viewed as orchestrating the response to vascular injury, mediating not only hemostatic but perhaps inflammatory and proliferative or reparative responses. Thrombin activities that would not necessarily be associated with vascular injury have also been described. The surprising effects of thrombin on neurite outgrowth13,14 conjure hypotheses regarding potential roles for thrombin or a related protease in neuronal plasticity or development, hypotheses that remain to be explored. Testing the role of thrombin itself in various in vivo processes has recently become possible with the development of potent thrombin inhibitors such as recombinant hirudin and D-phenylalanyl-prolyl-arginine-chloro-
Copyright © 1992 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
161
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
SHAUN R. COUGHLIN, M.D., Ph.D., THIEN-KHAI H. VU, DAVID T. HUNG, M.D., and VIRGINIA I. WHEATON
SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 2, 1992
FIG. 1. Cellular events stimulated by thrombin. Cellular targets for thrombin are depicted in the context of a blood vessel. Thrombin is the most potent stimulator of platelet aggregation, is chemotactic for monocytes, and is mitogenic for lymphocytes and mesenchymal cells. Thrombin also has a number of effects on the vascular endothelium, including stimulating expression of the neutrophil adhesive protein GMP-140 on the endothelial surface, as well as the production of the potent smooth muscle cell mitogen, platelet-derived growth factor (PDGF). As discussed in the text, these multiple activating functions of thrombin may be viewed teleoloically as orchestrators of hemostatic, inflammatory, and reparative responses to vascular injury. (Reprinted with permission from J. Clin. Invest.)
methyl-ketone. Animal studies with these reagents do suggest a critical role for thrombin in a platelet-dependent model of arterial thrombosis.15,16 The importance of thrombin in mediating other cellular responses, including inflammatory and proliferative responses to vascular injury, remains to be defined. One recent study does suggest a role for thrombin in mediating restenosis after angioplasty in an animal model.17 However, available antithrombin agents do not allow one to distinguish which of the many actions of thrombin are important in mediating a particular response. Currently available potent antithrombins function by interacting with thrombin and blocking not only thrombin's ability to activate its receptor but all active site-dependent thrombin functions, including thrombin's ability to cleave fibrinogen and to activate protein C. Because agents that act directly at the thrombin receptor have not been available, evaluation of the importance of thrombin receptor activation independent of thrombin-induced fibrin formation, protein C ac-
tivation, among others, has not been possible. It is hoped that the identification of the receptor (see later) will provide reagents critical for evaluating the importance of thrombin-induced cell activation in various in vivo settings.
Molecular Mechanisms of Thrombin-induced Cell Activation: Receptor Proteolysis or Occupancy Previous studies that address the mechanism by which thrombin activates platelets have been well-reviewed.2,18 Whether thrombin acts via a classic receptor occupancy mechanism, by a more novel mechanism involving proteolytic cleavage of the receptor, or by some combination of the two has been controversial.
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
162
EXPRESSION CLONING—COUGHLIN ET AL
Choice of Strategy Traditional ligand binding approaches to identify a functional thrombin receptor have succeeded in identifying a number of thrombin binding proteins19,20 but have not led to identification of a signal transduction molecule. Moreover, because modified thrombins that function neither as thrombin agonists nor as thrombin antagonists do bind to platelets in a manner indistinguishable from wild type thrombin binding,21-25 it has not been possible to conclude that the sites identified in binding studies are related to the functional thrombin receptor. For this reason, we adopted an expression cloning approach that followed thrombin-induced responses in Xenopus oocytes expressing exogenous mRNA to isolate a functional human thrombin receptor cDNA.1 We were encouraged by the work of Nakanishi and colleagues who had successfully used expression cloning in Xenopus oocytes to isolate a cDNA for the substance K receptor.26
Developing an Assay for an mRNA Encoding a Functional Thrombin Receptor These studies have been described in detail elsewhere.1 We first sought a cellular source of thrombin receptor mRNA by screening a number of cell lines for thrombin-responsiveness. Fluorescence-activated cell sorter analysis with Indo-I was used to measure thrombin-induced increases in cytoplasmic calcium, a normal second messenger response to thrombin. Among the best responders were Dami and HEL cells, which have megakaryocyte-like properties. mRNA was prepared from these cells and microinjected into Xenopus oocytes. Oocytes are known to translate exogenous mRNAs efficiently and, in general, to process their protein products appropriately. Oocytes microinjected with mRNA from Dami or HEL cells became responsive to thrombin, whereas uninjected oocytes did not. The endpoints measured in the oocytes again reflected thrombin-induced calcium mobilization; they were thrombin-induced 45Ca release from radiolabeled oocytes and thrombin-stimulated calcium-dependent inward chloride current in voltage-clamped oocytes. The ability to confer thrombin responsiveness to oocytes constituted an assay for the mRNA encoding a functional thrombin receptor.
Mechanics of Cloning mRNA from Dami cells was size-fractionated to enrich for the mRNA of interest and to test, to the extent possible, whether the receptor was encoded by a single
mRNA species. A cDNA library was made from the active fraction, using a plasmid cloning vector that allowed the production of cRNA by in vitro transcription. The library was initially plated in 50 pools at a complexity of 20,000 clones per pool. cRNA transcribed from each pool's plasmid DNA was tested in the oocyte system for thrombin receptor activity. Once a positive pool was identified, it was plated in pools of lower complexity. This process was repeated until a single cDNA clone was isolated.
CHARACTERIZATION OF THE THROMBIN RECEPTOR CLONE General Observations The thrombin receptor encoded by this cDNA clone was found to be expressed by both human platelets and endothelial cells. The deduced amino acid sequence of this clone revealed a novel member of the seven transmembrane receptor family. Of note, closest relatives include receptors for small peptides, substance P, and substance K.1
Specific Receptor Amino Acid Sequences Suggest a Model Thrombin is known to prefer to cleave peptides after arginines.27 We therefore searched for arginines predicted to be extracellular by the receptor sequence. We noted that the receptor's relatively long (100 residue) extracellular amino-terminal extension contained a putative thrombin cleavage site (LDPR/SFLL; / representing the point of cleavage after arginine 41 in the receptor). This site resembled the known thrombin cleavage site found in the thrombin-activated zymogen protein C (LDPR/I).28 At the carboxyl end of this putative cleavage site was a receptor domain resembling the carboxyl tail of the polypeptide hirudin, a leech-derived anticoagulant that binds thrombin with remarkable avidity. To put these sequence observations into context, thrombin has an extended substrate binding surface that recognizes residues both amino and carboxyl to a substrate's cleavage site29 (Fig. 2). Part of this extended substrate binding surface, the anion-binding exosite (Fig. 2), is important for thrombin's ability to activate its receptor.1,30 The carboxyl tail of hirudin is known to interact with thrombin's anion-binding exosite.31 The presence of the hirudin-like sequence in the carboxyl direction of the putative thrombin cleavage site within the thrombin receptor's amino terminal extension suggested that receptor proteolysis at this cleavage site might be important in receptor activation.
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
CLONING THE THROMBIN RECEPTOR
163
SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 2, 1992
FIG. 2. Thrombin's extended binding surface and its proposed interaction with the thrombin receptor. As discussed in the text, thrombin has an extended substrate binding surface that recognizes residues both amino and carboxyl to its substrate's cleavage site. Sequence analysis of the cloned thrombin receptor 1 and recent studies 34,35 suggest that the receptor's hirudin-like domain (YEPFWEDEE) interacts with thrombin's anion-binding exosite, whereas its cleavage site (LDPR/SFLL) interacts with thrombin's S1-S4 subsites. This model has important implications for the development of blocking monoclonal antibodies and receptorpeptide-based thrombin inhibitors. (Reprinted with permission from Nature, Ref. 34.)
Testing the Importance of the Cleavage Site We first changed the arginines in the receptor's amino terminal extension to alanines, a maneuver designed to render each site uncleavable by thrombin. Mutant receptors were expressed in Xenopus oocytes and their ability to respond to thrombin assessed. The putative cleavage site mutant (R41A) could not be activated by thrombin. The other mutants were fully activatable. As an independent means of rendering the putative thrombin cleavage site uncleavable, the serine at the site's P 1 ' position was changed to a proline. Like the R41A mutant, this S42P mutant receptor could not be activated by thrombin. These data strongly suggested that the receptor's putative thrombin cleavage site was critical for thrombin-induced receptor action.1 How might proteolysis at this site lead to receptor activation? A number of alternative hypotheses were generated.1 The favored hypothesis was that proteolysis at this site would unmask a new amino terminus that might serve as a peptide ligand for the receptor. Precedent for proteolytic unmasking of a binding site or ligand existed in trypsinogen activation to trypsin32 and in fibrinogen cleavage to fibrin monomer.33 Indeed, a peptide mimicking the new receptor amino terminus that would be revealed on receptor cleavage at the R41 position was a full agonist for the cloned receptor expressed in oocytes. Not only could the wild type receptor be activated, the R41A
receptor, unactivatable by thrombin, was fully responsive to the new amino terminus "agonist peptide." Thus, the agonist peptide could bypass the requirement for proteolysis and activate the receptor directly.1
MODEL FOR THROMBIN RECEPTOR ACTIVATION The studies just described suggest that thrombin receptor activation proceeds by the novel mechanism depicted in Figure 3. Thrombin cleaves its receptor after arginine 41 in the receptor's amino-terminal extension, exposing a new amino-terminus that functions as a tethered peptide ligand for the receptor. The new aminoterminus then binds to an as yet undefined receptor site, effecting receptor activation.1 Recent studies buttress this model. 34,35 Replacing the receptor's thrombin cleavage site (LDPR/S) with that for enterokinase (DDDDK/S) completely switched the receptor's specificity. A functional "enterokinase receptor" that was fully activatable by enterokinase but unresponsive to thrombin was created. Consistent with the ability of the agonist peptide to activate the receptor, these results prove that receptor proteolysis with exposure of the amino-terminus beginning at position 42 (SFLL . . .) is sufficient for receptor activation.34 Additional studies have shown the receptor's hirudin-like sequence can indeed bind thrombin's anion-binding ex-
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
164
165
FIG. 3. Model of thrombin receptor activation. Thrombin cleaves the receptor at the LDPR/S cleavage site (junction between open and filled receptor segments), releasing an inactive fragment of the receptor's amino-terminus (open fragment) and exposing a new amino-terminus. This newly unmasked amino-terminus then functions as a tethered peptide ligand, binding to an as yet undefined pocket, and effecting receptor activation1. (Reprinted with permission from Nature, Ref. 34.)
osite, 34,35 supporting the model of thrombin-receptor interaction shown in Figure 2.
UNRESOLVED ISSUES Our studies raise many provocative questions and speculations. The unusual activation mechanism of thrombin receptor activation in essence requires a peptide receptor that contains its own ligand. Where did this mechanism come from? Was it an evolutionary specialization of previously peptide receptors? Or is it representative of a more primitive heritage? Will this receptor sire a family of thrombin or protease receptors? And how many of the assorted cell activation functions of thrombin does this receptor account for? The agonist peptide described herein, which activates the cloned thrombin receptor directly and independent of thrombin and its protease activity, provides a useful tool in this regard. Specifically, thrombin responses mimicked by the agonist peptide are likely mediated by the unusual mechanism described before via the cloned receptor or a highly related receptor. Thrombin responses not mimicked by the agonist peptide may be mediated by a different mechanism. The unusual mechanism of receptor activation displayed by the cloned thrombin receptor provokes still more questions. The model predicts that receptor's ligand is tethered and cannot diffuse away. Once cleaved and activated, the receptor should be "on" until turned off by additional specific machinery. What turns it off? We have recently shown that the receptor can be desensitized with agonist peptide, suggesting that a mechanism for
receptor desensitization independent of thrombin-mediated proteolysis exists. A beta-adrenergic receptor kinase-like enzyme36 is one good candidate for the molecule responsible for turning off the activated thrombin receptor. What is the relationship of known thrombin binding proteins (for example, platelet glycoprotein Ib [GPIb]) to thrombin signaling and to this receptor? Is GPIb a sink? Perhaps, at low thrombin concentrations, GPIb might bind thrombin long enough to allow inactivation by antithrombin III, thereby preventing interaction with the receptor and receptor desensitization by low concentrations of thrombin. Or does GPIb promote activation of the cloned receptor by binding thrombin20? Or does thrombin binding to GPIb serve an independent, as yet undefined, function? Additional issues regarding how this receptor accounts for the kinetics of thrombin responses, and how it couples to those responses, also remain to be explored.
PATHS TO NEW REAGENTS These studies suggest a number of paths to novel reagents. Most useful would be thrombin receptor blockers. Blocking monoclonal antibodies might be obtained in a number of ways. One might predict that antibodies to the receptor's hirudin-like sequence or thrombin cleavage site would inhibit thrombin-receptor interaction and receptor proteolysis. Antibodies to these regions or to the agonist peptide domain itself might cause steric interference with the "liganding" of the receptor after proteolysis. A similar function would be served by antibodies to
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
EXPRESSION CLONING—COUGHLIN ET AL
SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 2, 1992
receptor regions comprising or affecting the as yet undefined binding site for the agonist peptide. Whether nonantibody receptor blockers, that is, competitive or noncompetitive antagonists of the agonist peptide, can be obtained remains to be explored. Such reagents will define the role of thrombin receptor activation in vivo and may lead to the development of a new class of pharmaceuticals.
REFERENCES 1. Vu T-KH, DT Hung, VI Wheaton, SR Coughlin: Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell 64:1057-1068, 1991. 2. Shuman MA: Thrombin-cellular interactions. Ann NY Acad Sci 485:228-239, 1986. 3. Berndt MC, DR Phillips: Platelet membrane proteins: Composition and receptor function. In: Gordon, JL (Ed): Platelets in Biology and Pathology. Elsevier/North Holland Biomedical Press, Amsterdam, 1981, pp 43-74. 4. Bar-Shavit R, A Kahn, GD Wilner, JW Fenton II: Monocyte chemotaxis: stimulation by specific exosite region in thrombin. Science 220:728-731, 1983. 5. Chen LB, NNH Teng, JM Buchanan: Mitogenicity of thrombin and surface alterations on mouse splenocytes. Exp Cell Res 101:41—46, 1976. 6. Chen LB, JM Buchanan: Mitogenic activity of blood components. I. Thrombin and prothrombin. Proc Natl Acad Sci USA 72:131135, 1975. 7. Weksler BB, CW Ley, EA Jaffe: Stimulation of endothelial cell prostacyclin production by thrombin, trypsin, and the ionophore A23187. J Clin Invest 62:923-930, 1978. 8. Prescott SM, GA Zimmerman, TM McIntyre: Human endothelial cells in culture produce platelet-activating factor when stimulated by thrombin. Proc Natl Acad Sci USA 81:3534-3538, 1984. 9. Gelehrter TD, R Sznycer-Laszyk: Thrombin induction of plasminogen activator-inhibitor in cultured human endothelial cells. J Clin Invest 77:165-169, 1986. 10. Daniel TO, VC Gibbs, DF Milfay, MR Garavoy, LT Williams: Thrombin stimulates c-sis gene expression in microvascular endothelial cells. J Biol Chem 261:9579-9582, 1986. 11. Zimmerman GA, TM Mclntyre, SM Prescott: Thrombin stimulates neutrophil adherence by an endothelial cell-dependent mechanism. Ann NY Acad Sci 485:349-368, 1986. 12. Hattori R, KK Hamilton, RD Fugate, RP McEver, PJ Sims: Stimulated secretion of endothelial vWF is accompanied by rapid redistribution to the cell surface of the intracellular granule membrane protein GMP-140. J Biol Chem 264:7768-7771, 1989. 13. Farmer L, J Sommer, D Monard: Glia-derived nexin potentiates neurite extension in hippocampal pyramidal cells in vitro. Dev Neurosci 12:73-80, 1990. 14. Gurwitz D, DD Cunningham: Thrombin modulates and reversed neuroblastoma neurite outgrowth. Proc Natl Acad Sci USA 85:3440-3444, 1988. 15. Eidt JF, P Allison, S Nobel, J Ashton, P Golino, J McNatt, LM Buja, JT Willerson: Thrombin is an important mediator of platelet aggregation in stenosed canine coronary arteries with endothelial injury. J Clin Invest 84:18-27, 1989. 16. Hansen SR, LA Harker: Interruption of acute platelet-dependent thrombosis by the synthetic antithrombin PPACK. Proc Natl Acad Sci USA 85:3184-3188, 1988.
17. Sarembock IJ, SD Gertz, LW Gimple, RM Owen, ER Powers, WC Roberts: Effectiveness of recombinant desulphatohirudin (CGP39393) in reducing restenosis after balloon angioplasty of atherosclerotic femoral arteries in rabbits. Circulation (in press). 18. Berndt MC, C Gregory, G Dowden, PA Castaldi: Thrombin interactions with platelet membrane proteins. Ann NY Acad Sci 485:374-386, 1985. 19. Gronke RS, BL Bergman, JB Baker: Thrombin interaction with platelets: Influence of a platelet protease nexin. J Biol Chem 262:3030-3036, 1987. 20. Okamura T, M Hasitz, GA Jamieson: Platelet glycocalicin: Interaction with thrombin and role as thrombin receptor on the platelet surface. J Biol Chem 253:3435-3443, 1978. 21. Davey MG, EF Lüscher: Actions of thrombin and other coagulant and proteolytic enzymes on blood platelets. Nature 216:857-858, 1967. 22. Phillips DR: Thrombin interaction with human platelets: Potentiation of thrombin-induced aggregation and release by inactivated thrombin. Thromb Diath Haemorrh 32:207-215, 1974. 23. Martin BM, RD Feinman, TC Detwiler: Platelet stimulation by thrombin and other proteases. Biochemistry 14:1308-1314, 1975. 24. Tollefsen DM, JR Feagler, PW Majerus: The binding of thrombin to the surface of platelets. J Biol Chem 249:2646-2651, 1974. 25. Workman EF, Jr., GC White, II, RI Lundblad: Structure-function relationships in the interaction of alpha-thrombin with blood platelets. J Biol Chem 252:7118-7123, 1977. 26. Masu Y, K Nakayama, H Tamaki, Y Harada, M Kuno, S Nakanishi: cDNA cloning of bovine substance K receptor through oocyte expression system. Nature 329:836-840, 1987. 27. Maszbek L, K Laki: Interaction of thrombin with proteins other than fibrinogen (thrombin-susceptible bonds), In: Machovich R. (Ed): The Thrombin. CRC Press, Boca Raton, FL, 1984, pp 8 3 90. 28. Stenflo J, P Fernlund: Amino acid sequence of the heavy chain of bovine protein C. J Biol Chem 257:12180-12190, 1982. 29. Bode W, I Mayr, U Baumann, R Huber, SR Stone, J Hofsteenge: The refined 1.9A crystal structure of human alpha-thrombin: interaction with D-Phe-Pro-Arg chloromethylketone and significance of the Tyr-Pro-Pro-Trp insertion segment. EMBOJ 8.3467-3475, 1989. 30. Jakubowski JA, JM Maragonore: Inhibition of coagulation and thrombin-induced platelet activities by a synthetic dodecapeptide modeled on the carboxy-terminus of hirudin. Blood 75:399-406, 1990. 31. Rydel TJ, KG Rabichandran, A Tulinsky, W Bode, R Huber, C Roitsch, JW Fenton II: The structure of a complex of recombinant hirudin and human alpha-thrombin. Science 249:277-280, 1990. 32. Bode W, P Schwager, R Huber: The transition of bovine trypsinogen to a trypsin-like state upon strong ligand binding. J Mol Biol 118:99-112, 1978. 33. Hantgan RR, CW Francis, HA Scheraga, VJ Marder: Fibrinogen structure and physiology. In: Colman RW, J Hirsch, VJ Marder, EW Salzman, (Eds). Hemostasis and Thrombosis. JB Lippincott, Philadelphia, 1987, pp 269-288. 34. Vu T-HK, VI Wheaton, DT Hung, I Chao, SR Coughlin: Domains defining thrombin-receptor interaction. Nature 353:674—677, 1991. 35. Liu L-W, T-KH Vu, CT Esmon, SR Coughlin: The region of the thrombin receptor resembling hirudin binds to thrombin and alters enzyme specificity. J Biol Chem (in press). 36. Lohse MJ, RJ Lefkowitz, MG Caron, JL Benovic: Inhibition of beta-adrenergic receptor kinase prevents rapid homologous desensitization of beta2-adrenergic receptors. Proc Natl Acad Sci USA 86:3011-3015, 1989.
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
166
Expression Cloning and Characterization of a Functional Thrombin Receptor Reveals a Novel Proteolytic Mechanism of Receptor Activation
Elucidation of the mechanism by which thrombin activates platelets and other cells has been a longstanding problem in thrombosis research. This problem was tantalizing from both clinical and basic science perspectives. From the clinical view, thrombin elicits a host of cellular activities critical for both normal responses to wounding and for pathologic vascular events. Defining the mechanism of thrombin-induced cell activation promised the advent of new therapeutic agents and reagents for identifying thrombin's role in pathophysiology. From the basic view, the fact that thrombin is a protease held out the possibility that a novel mechanism of receptor activation might be uncovered. In this article we briefly review the recent cloning and characterization of the platelet thrombin receptor by our laboratory.1 As will be discussed, identification of the thrombin receptor revealed a unique proteolytic mechanism of receptor activation and led to the development of a novel agonist peptide capable of activating the thrombin receptor independent of thrombin and thrombin's protease activity. It is our hope that having the cloned receptor in hand will allow the development of thrombin receptor antagonists. Such reagents will define the role of thrombin receptor activation in vivo and may provide the basis for a new class of antithrombotic or antiproliferative pharmaceuticals.
From the Cardiovascular Research Institute, and Department of Medicine, University of California, San Francisco, San Francisco, California. Reprint requests: Dr. Coughlin, Box 0524, University of California, San Francisco, San Francisco, CA 94143.
BACKGROUND Pathophysiology: Cell Activating Functions of Thrombin In addition to its well-known role in cleaving fibrinogen to fibrin, thrombin has a number of important "cell activating" functions (Fig. 1) which are reviewed by Shuman.2 First and foremost, thrombin is the most potent stimulator of platelet aggregation.3 Thrombin is also chemotactic for monocytes,4 mitogenic for lymphocytes and for mesenchymal cells, including vascular smooth muscle cells,5,6 and has a number of effects on the vascular endothelium. These include stimulating endothelial production of prostacyclin,7 platelet-activating factor,8 plasminogen activator-inhibitor,9 and the potent smooth muscle cell mitogen platelet-derived growth factor.10 Thrombin also induces neutrophil adherence to the vessel wall by an endothelial-dependent mechanism,11 probably by inducing surface expression of granule membrane protein (GMP)-140.12 Teleologically, these multiple cell activating functions of thrombin may be viewed as orchestrating the response to vascular injury, mediating not only hemostatic but perhaps inflammatory and proliferative or reparative responses. Thrombin activities that would not necessarily be associated with vascular injury have also been described. The surprising effects of thrombin on neurite outgrowth13,14 conjure hypotheses regarding potential roles for thrombin or a related protease in neuronal plasticity or development, hypotheses that remain to be explored. Testing the role of thrombin itself in various in vivo processes has recently become possible with the development of potent thrombin inhibitors such as recombinant hirudin and D-phenylalanyl-prolyl-arginine-chloro-
Copyright © 1992 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
161
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
SHAUN R. COUGHLIN, M.D., Ph.D., THIEN-KHAI H. VU, DAVID T. HUNG, M.D., and VIRGINIA I. WHEATON
SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 2, 1992
FIG. 1. Cellular events stimulated by thrombin. Cellular targets for thrombin are depicted in the context of a blood vessel. Thrombin is the most potent stimulator of platelet aggregation, is chemotactic for monocytes, and is mitogenic for lymphocytes and mesenchymal cells. Thrombin also has a number of effects on the vascular endothelium, including stimulating expression of the neutrophil adhesive protein GMP-140 on the endothelial surface, as well as the production of the potent smooth muscle cell mitogen, platelet-derived growth factor (PDGF). As discussed in the text, these multiple activating functions of thrombin may be viewed teleoloically as orchestrators of hemostatic, inflammatory, and reparative responses to vascular injury. (Reprinted with permission from J. Clin. Invest.)
methyl-ketone. Animal studies with these reagents do suggest a critical role for thrombin in a platelet-dependent model of arterial thrombosis.15,16 The importance of thrombin in mediating other cellular responses, including inflammatory and proliferative responses to vascular injury, remains to be defined. One recent study does suggest a role for thrombin in mediating restenosis after angioplasty in an animal model.17 However, available antithrombin agents do not allow one to distinguish which of the many actions of thrombin are important in mediating a particular response. Currently available potent antithrombins function by interacting with thrombin and blocking not only thrombin's ability to activate its receptor but all active site-dependent thrombin functions, including thrombin's ability to cleave fibrinogen and to activate protein C. Because agents that act directly at the thrombin receptor have not been available, evaluation of the importance of thrombin receptor activation independent of thrombin-induced fibrin formation, protein C ac-
tivation, among others, has not been possible. It is hoped that the identification of the receptor (see later) will provide reagents critical for evaluating the importance of thrombin-induced cell activation in various in vivo settings.
Molecular Mechanisms of Thrombin-induced Cell Activation: Receptor Proteolysis or Occupancy Previous studies that address the mechanism by which thrombin activates platelets have been well-reviewed.2,18 Whether thrombin acts via a classic receptor occupancy mechanism, by a more novel mechanism involving proteolytic cleavage of the receptor, or by some combination of the two has been controversial.
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
162
EXPRESSION CLONING—COUGHLIN ET AL
Choice of Strategy Traditional ligand binding approaches to identify a functional thrombin receptor have succeeded in identifying a number of thrombin binding proteins19,20 but have not led to identification of a signal transduction molecule. Moreover, because modified thrombins that function neither as thrombin agonists nor as thrombin antagonists do bind to platelets in a manner indistinguishable from wild type thrombin binding,21-25 it has not been possible to conclude that the sites identified in binding studies are related to the functional thrombin receptor. For this reason, we adopted an expression cloning approach that followed thrombin-induced responses in Xenopus oocytes expressing exogenous mRNA to isolate a functional human thrombin receptor cDNA.1 We were encouraged by the work of Nakanishi and colleagues who had successfully used expression cloning in Xenopus oocytes to isolate a cDNA for the substance K receptor.26
Developing an Assay for an mRNA Encoding a Functional Thrombin Receptor These studies have been described in detail elsewhere.1 We first sought a cellular source of thrombin receptor mRNA by screening a number of cell lines for thrombin-responsiveness. Fluorescence-activated cell sorter analysis with Indo-I was used to measure thrombin-induced increases in cytoplasmic calcium, a normal second messenger response to thrombin. Among the best responders were Dami and HEL cells, which have megakaryocyte-like properties. mRNA was prepared from these cells and microinjected into Xenopus oocytes. Oocytes are known to translate exogenous mRNAs efficiently and, in general, to process their protein products appropriately. Oocytes microinjected with mRNA from Dami or HEL cells became responsive to thrombin, whereas uninjected oocytes did not. The endpoints measured in the oocytes again reflected thrombin-induced calcium mobilization; they were thrombin-induced 45Ca release from radiolabeled oocytes and thrombin-stimulated calcium-dependent inward chloride current in voltage-clamped oocytes. The ability to confer thrombin responsiveness to oocytes constituted an assay for the mRNA encoding a functional thrombin receptor.
Mechanics of Cloning mRNA from Dami cells was size-fractionated to enrich for the mRNA of interest and to test, to the extent possible, whether the receptor was encoded by a single
mRNA species. A cDNA library was made from the active fraction, using a plasmid cloning vector that allowed the production of cRNA by in vitro transcription. The library was initially plated in 50 pools at a complexity of 20,000 clones per pool. cRNA transcribed from each pool's plasmid DNA was tested in the oocyte system for thrombin receptor activity. Once a positive pool was identified, it was plated in pools of lower complexity. This process was repeated until a single cDNA clone was isolated.
CHARACTERIZATION OF THE THROMBIN RECEPTOR CLONE General Observations The thrombin receptor encoded by this cDNA clone was found to be expressed by both human platelets and endothelial cells. The deduced amino acid sequence of this clone revealed a novel member of the seven transmembrane receptor family. Of note, closest relatives include receptors for small peptides, substance P, and substance K.1
Specific Receptor Amino Acid Sequences Suggest a Model Thrombin is known to prefer to cleave peptides after arginines.27 We therefore searched for arginines predicted to be extracellular by the receptor sequence. We noted that the receptor's relatively long (100 residue) extracellular amino-terminal extension contained a putative thrombin cleavage site (LDPR/SFLL; / representing the point of cleavage after arginine 41 in the receptor). This site resembled the known thrombin cleavage site found in the thrombin-activated zymogen protein C (LDPR/I).28 At the carboxyl end of this putative cleavage site was a receptor domain resembling the carboxyl tail of the polypeptide hirudin, a leech-derived anticoagulant that binds thrombin with remarkable avidity. To put these sequence observations into context, thrombin has an extended substrate binding surface that recognizes residues both amino and carboxyl to a substrate's cleavage site29 (Fig. 2). Part of this extended substrate binding surface, the anion-binding exosite (Fig. 2), is important for thrombin's ability to activate its receptor.1,30 The carboxyl tail of hirudin is known to interact with thrombin's anion-binding exosite.31 The presence of the hirudin-like sequence in the carboxyl direction of the putative thrombin cleavage site within the thrombin receptor's amino terminal extension suggested that receptor proteolysis at this cleavage site might be important in receptor activation.
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
CLONING THE THROMBIN RECEPTOR
163
SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 2, 1992
FIG. 2. Thrombin's extended binding surface and its proposed interaction with the thrombin receptor. As discussed in the text, thrombin has an extended substrate binding surface that recognizes residues both amino and carboxyl to its substrate's cleavage site. Sequence analysis of the cloned thrombin receptor 1 and recent studies 34,35 suggest that the receptor's hirudin-like domain (YEPFWEDEE) interacts with thrombin's anion-binding exosite, whereas its cleavage site (LDPR/SFLL) interacts with thrombin's S1-S4 subsites. This model has important implications for the development of blocking monoclonal antibodies and receptorpeptide-based thrombin inhibitors. (Reprinted with permission from Nature, Ref. 34.)
Testing the Importance of the Cleavage Site We first changed the arginines in the receptor's amino terminal extension to alanines, a maneuver designed to render each site uncleavable by thrombin. Mutant receptors were expressed in Xenopus oocytes and their ability to respond to thrombin assessed. The putative cleavage site mutant (R41A) could not be activated by thrombin. The other mutants were fully activatable. As an independent means of rendering the putative thrombin cleavage site uncleavable, the serine at the site's P 1 ' position was changed to a proline. Like the R41A mutant, this S42P mutant receptor could not be activated by thrombin. These data strongly suggested that the receptor's putative thrombin cleavage site was critical for thrombin-induced receptor action.1 How might proteolysis at this site lead to receptor activation? A number of alternative hypotheses were generated.1 The favored hypothesis was that proteolysis at this site would unmask a new amino terminus that might serve as a peptide ligand for the receptor. Precedent for proteolytic unmasking of a binding site or ligand existed in trypsinogen activation to trypsin32 and in fibrinogen cleavage to fibrin monomer.33 Indeed, a peptide mimicking the new receptor amino terminus that would be revealed on receptor cleavage at the R41 position was a full agonist for the cloned receptor expressed in oocytes. Not only could the wild type receptor be activated, the R41A
receptor, unactivatable by thrombin, was fully responsive to the new amino terminus "agonist peptide." Thus, the agonist peptide could bypass the requirement for proteolysis and activate the receptor directly.1
MODEL FOR THROMBIN RECEPTOR ACTIVATION The studies just described suggest that thrombin receptor activation proceeds by the novel mechanism depicted in Figure 3. Thrombin cleaves its receptor after arginine 41 in the receptor's amino-terminal extension, exposing a new amino-terminus that functions as a tethered peptide ligand for the receptor. The new aminoterminus then binds to an as yet undefined receptor site, effecting receptor activation.1 Recent studies buttress this model. 34,35 Replacing the receptor's thrombin cleavage site (LDPR/S) with that for enterokinase (DDDDK/S) completely switched the receptor's specificity. A functional "enterokinase receptor" that was fully activatable by enterokinase but unresponsive to thrombin was created. Consistent with the ability of the agonist peptide to activate the receptor, these results prove that receptor proteolysis with exposure of the amino-terminus beginning at position 42 (SFLL . . .) is sufficient for receptor activation.34 Additional studies have shown the receptor's hirudin-like sequence can indeed bind thrombin's anion-binding ex-
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
164
165
FIG. 3. Model of thrombin receptor activation. Thrombin cleaves the receptor at the LDPR/S cleavage site (junction between open and filled receptor segments), releasing an inactive fragment of the receptor's amino-terminus (open fragment) and exposing a new amino-terminus. This newly unmasked amino-terminus then functions as a tethered peptide ligand, binding to an as yet undefined pocket, and effecting receptor activation1. (Reprinted with permission from Nature, Ref. 34.)
osite, 34,35 supporting the model of thrombin-receptor interaction shown in Figure 2.
UNRESOLVED ISSUES Our studies raise many provocative questions and speculations. The unusual activation mechanism of thrombin receptor activation in essence requires a peptide receptor that contains its own ligand. Where did this mechanism come from? Was it an evolutionary specialization of previously peptide receptors? Or is it representative of a more primitive heritage? Will this receptor sire a family of thrombin or protease receptors? And how many of the assorted cell activation functions of thrombin does this receptor account for? The agonist peptide described herein, which activates the cloned thrombin receptor directly and independent of thrombin and its protease activity, provides a useful tool in this regard. Specifically, thrombin responses mimicked by the agonist peptide are likely mediated by the unusual mechanism described before via the cloned receptor or a highly related receptor. Thrombin responses not mimicked by the agonist peptide may be mediated by a different mechanism. The unusual mechanism of receptor activation displayed by the cloned thrombin receptor provokes still more questions. The model predicts that receptor's ligand is tethered and cannot diffuse away. Once cleaved and activated, the receptor should be "on" until turned off by additional specific machinery. What turns it off? We have recently shown that the receptor can be desensitized with agonist peptide, suggesting that a mechanism for
receptor desensitization independent of thrombin-mediated proteolysis exists. A beta-adrenergic receptor kinase-like enzyme36 is one good candidate for the molecule responsible for turning off the activated thrombin receptor. What is the relationship of known thrombin binding proteins (for example, platelet glycoprotein Ib [GPIb]) to thrombin signaling and to this receptor? Is GPIb a sink? Perhaps, at low thrombin concentrations, GPIb might bind thrombin long enough to allow inactivation by antithrombin III, thereby preventing interaction with the receptor and receptor desensitization by low concentrations of thrombin. Or does GPIb promote activation of the cloned receptor by binding thrombin20? Or does thrombin binding to GPIb serve an independent, as yet undefined, function? Additional issues regarding how this receptor accounts for the kinetics of thrombin responses, and how it couples to those responses, also remain to be explored.
PATHS TO NEW REAGENTS These studies suggest a number of paths to novel reagents. Most useful would be thrombin receptor blockers. Blocking monoclonal antibodies might be obtained in a number of ways. One might predict that antibodies to the receptor's hirudin-like sequence or thrombin cleavage site would inhibit thrombin-receptor interaction and receptor proteolysis. Antibodies to these regions or to the agonist peptide domain itself might cause steric interference with the "liganding" of the receptor after proteolysis. A similar function would be served by antibodies to
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
EXPRESSION CLONING—COUGHLIN ET AL
SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 2, 1992
receptor regions comprising or affecting the as yet undefined binding site for the agonist peptide. Whether nonantibody receptor blockers, that is, competitive or noncompetitive antagonists of the agonist peptide, can be obtained remains to be explored. Such reagents will define the role of thrombin receptor activation in vivo and may lead to the development of a new class of pharmaceuticals.
REFERENCES 1. Vu T-KH, DT Hung, VI Wheaton, SR Coughlin: Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell 64:1057-1068, 1991. 2. Shuman MA: Thrombin-cellular interactions. Ann NY Acad Sci 485:228-239, 1986. 3. Berndt MC, DR Phillips: Platelet membrane proteins: Composition and receptor function. In: Gordon, JL (Ed): Platelets in Biology and Pathology. Elsevier/North Holland Biomedical Press, Amsterdam, 1981, pp 43-74. 4. Bar-Shavit R, A Kahn, GD Wilner, JW Fenton II: Monocyte chemotaxis: stimulation by specific exosite region in thrombin. Science 220:728-731, 1983. 5. Chen LB, NNH Teng, JM Buchanan: Mitogenicity of thrombin and surface alterations on mouse splenocytes. Exp Cell Res 101:41—46, 1976. 6. Chen LB, JM Buchanan: Mitogenic activity of blood components. I. Thrombin and prothrombin. Proc Natl Acad Sci USA 72:131135, 1975. 7. Weksler BB, CW Ley, EA Jaffe: Stimulation of endothelial cell prostacyclin production by thrombin, trypsin, and the ionophore A23187. J Clin Invest 62:923-930, 1978. 8. Prescott SM, GA Zimmerman, TM McIntyre: Human endothelial cells in culture produce platelet-activating factor when stimulated by thrombin. Proc Natl Acad Sci USA 81:3534-3538, 1984. 9. Gelehrter TD, R Sznycer-Laszyk: Thrombin induction of plasminogen activator-inhibitor in cultured human endothelial cells. J Clin Invest 77:165-169, 1986. 10. Daniel TO, VC Gibbs, DF Milfay, MR Garavoy, LT Williams: Thrombin stimulates c-sis gene expression in microvascular endothelial cells. J Biol Chem 261:9579-9582, 1986. 11. Zimmerman GA, TM Mclntyre, SM Prescott: Thrombin stimulates neutrophil adherence by an endothelial cell-dependent mechanism. Ann NY Acad Sci 485:349-368, 1986. 12. Hattori R, KK Hamilton, RD Fugate, RP McEver, PJ Sims: Stimulated secretion of endothelial vWF is accompanied by rapid redistribution to the cell surface of the intracellular granule membrane protein GMP-140. J Biol Chem 264:7768-7771, 1989. 13. Farmer L, J Sommer, D Monard: Glia-derived nexin potentiates neurite extension in hippocampal pyramidal cells in vitro. Dev Neurosci 12:73-80, 1990. 14. Gurwitz D, DD Cunningham: Thrombin modulates and reversed neuroblastoma neurite outgrowth. Proc Natl Acad Sci USA 85:3440-3444, 1988. 15. Eidt JF, P Allison, S Nobel, J Ashton, P Golino, J McNatt, LM Buja, JT Willerson: Thrombin is an important mediator of platelet aggregation in stenosed canine coronary arteries with endothelial injury. J Clin Invest 84:18-27, 1989. 16. Hansen SR, LA Harker: Interruption of acute platelet-dependent thrombosis by the synthetic antithrombin PPACK. Proc Natl Acad Sci USA 85:3184-3188, 1988.
17. Sarembock IJ, SD Gertz, LW Gimple, RM Owen, ER Powers, WC Roberts: Effectiveness of recombinant desulphatohirudin (CGP39393) in reducing restenosis after balloon angioplasty of atherosclerotic femoral arteries in rabbits. Circulation (in press). 18. Berndt MC, C Gregory, G Dowden, PA Castaldi: Thrombin interactions with platelet membrane proteins. Ann NY Acad Sci 485:374-386, 1985. 19. Gronke RS, BL Bergman, JB Baker: Thrombin interaction with platelets: Influence of a platelet protease nexin. J Biol Chem 262:3030-3036, 1987. 20. Okamura T, M Hasitz, GA Jamieson: Platelet glycocalicin: Interaction with thrombin and role as thrombin receptor on the platelet surface. J Biol Chem 253:3435-3443, 1978. 21. Davey MG, EF Lüscher: Actions of thrombin and other coagulant and proteolytic enzymes on blood platelets. Nature 216:857-858, 1967. 22. Phillips DR: Thrombin interaction with human platelets: Potentiation of thrombin-induced aggregation and release by inactivated thrombin. Thromb Diath Haemorrh 32:207-215, 1974. 23. Martin BM, RD Feinman, TC Detwiler: Platelet stimulation by thrombin and other proteases. Biochemistry 14:1308-1314, 1975. 24. Tollefsen DM, JR Feagler, PW Majerus: The binding of thrombin to the surface of platelets. J Biol Chem 249:2646-2651, 1974. 25. Workman EF, Jr., GC White, II, RI Lundblad: Structure-function relationships in the interaction of alpha-thrombin with blood platelets. J Biol Chem 252:7118-7123, 1977. 26. Masu Y, K Nakayama, H Tamaki, Y Harada, M Kuno, S Nakanishi: cDNA cloning of bovine substance K receptor through oocyte expression system. Nature 329:836-840, 1987. 27. Maszbek L, K Laki: Interaction of thrombin with proteins other than fibrinogen (thrombin-susceptible bonds), In: Machovich R. (Ed): The Thrombin. CRC Press, Boca Raton, FL, 1984, pp 8 3 90. 28. Stenflo J, P Fernlund: Amino acid sequence of the heavy chain of bovine protein C. J Biol Chem 257:12180-12190, 1982. 29. Bode W, I Mayr, U Baumann, R Huber, SR Stone, J Hofsteenge: The refined 1.9A crystal structure of human alpha-thrombin: interaction with D-Phe-Pro-Arg chloromethylketone and significance of the Tyr-Pro-Pro-Trp insertion segment. EMBOJ 8.3467-3475, 1989. 30. Jakubowski JA, JM Maragonore: Inhibition of coagulation and thrombin-induced platelet activities by a synthetic dodecapeptide modeled on the carboxy-terminus of hirudin. Blood 75:399-406, 1990. 31. Rydel TJ, KG Rabichandran, A Tulinsky, W Bode, R Huber, C Roitsch, JW Fenton II: The structure of a complex of recombinant hirudin and human alpha-thrombin. Science 249:277-280, 1990. 32. Bode W, P Schwager, R Huber: The transition of bovine trypsinogen to a trypsin-like state upon strong ligand binding. J Mol Biol 118:99-112, 1978. 33. Hantgan RR, CW Francis, HA Scheraga, VJ Marder: Fibrinogen structure and physiology. In: Colman RW, J Hirsch, VJ Marder, EW Salzman, (Eds). Hemostasis and Thrombosis. JB Lippincott, Philadelphia, 1987, pp 269-288. 34. Vu T-HK, VI Wheaton, DT Hung, I Chao, SR Coughlin: Domains defining thrombin-receptor interaction. Nature 353:674—677, 1991. 35. Liu L-W, T-KH Vu, CT Esmon, SR Coughlin: The region of the thrombin receptor resembling hirudin binds to thrombin and alters enzyme specificity. J Biol Chem (in press). 36. Lohse MJ, RJ Lefkowitz, MG Caron, JL Benovic: Inhibition of beta-adrenergic receptor kinase prevents rapid homologous desensitization of beta2-adrenergic receptors. Proc Natl Acad Sci USA 86:3011-3015, 1989.
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
166
