- Email: [email protected]
Thrombin-Stimulated Calcium Mobilization Is Inhibited by Thrombospondin via CD36
EXPERIMENTAL CELL RESEARCH ARTICLE NO.
238, 465–471 (1998)
EX973863
Thrombin-Stimulated Calcium Mobilization Is Inhibited by Thrombospondin via CD361 J. Enenstein,2 K. Gupta, G. M. Vercellotti, and R. P. Hebbel Department of Medicine and Hematology, University of Minnesota, Minneapolis, Minnesota 55455
Activation of the G-protein-linked thrombin receptor in endothelial cells normally leads to an increase in free intracellular calcium, [Ca2/]i , which is the proximate stimulus for many important cell functions. We present evidence showing that signals from CD36, the thrombospondin (TSP) receptor, can inhibit this thrombin-mediated calcium response. Human endothelial cells preloaded with Indo-1 exhibited rapid calcium mobilization in response to thrombin. The presence of TSP inhibited the thrombin-stimulated calcium response in CD36-positive microvascular endothelial cells but not in CD36-negative umbilical vein endothelial cells. This TSP effect was mimicked by anti-CD36 antibodies and a TSP peptide (CSVTCG), but not by an alternative CD36 ligand (collagen IV) or an antibody to an alternative TSP receptor (avb3 ). TSP also inhibited the calcium response to the thrombin receptor-tethered ligand peptide, SFLLRN. In addition, TSP and anti-CD36 antibodies inhibited the calcium response of a closely related receptor, the trypsin/SLIGKVD-activated receptor PAR-2. TSP did not indiscriminately inhibit all calcium release pathways, since histamine- or VEGF-stimulated calcium responses were not inhibited by TSP. We conclude that cross-talk from the CD36 receptor influences the responsive state of the endothelial thrombin receptor family and/or its signaling pathway. q 1998 Academic Press
INTRODUCTION
Cells can respond in a modulated manner to multiple environmental cues because there is cross-talk between heterologous cell surface receptor signaling pathways. Endothelial cells respond with carefully regulated changes in cytosolic calcium concentration ([Ca2/]i )3 to 1
This work was supported by NIH Grant HL30160. To whom correspondence and reprint requests should be addressed at Department of Medicine, University of Minnesota, UMHC Box 480, 420 Delaware Street S.E., Minneapolis, MN 55455. Fax: (612) 625-6919. 3 Abbreviations used: [Ca2/]i , intracellular calcium; HUVEC, human umbilical vein endothelial cells; MVEC, microvascular endothelial cells; TSP, thrombospondin; VEGF, vascular endothelial growth factor. 2
stimulation by thrombin, an important hemostatic agent in vivo that also promotes chemotaxis and mitogenesis [1]. Endothelial [Ca2/]i rises within 15 s of stimulation with thrombin [2], due to initial calcium release from intracellular stores and subsequent influx of extracellular calcium through receptor-mediated channels. This increase in [Ca2/]i is proximate to—and essential for—thrombin-stimulated vWF release, surface expression of P-selectin, prostacyclin production, disassembly of peripheral actin fibers, and increased assembly of stress fibers, and it is partially responsible for increased permeability [2, 3]. The thrombin-stimulated calcium response is known to be modulated via negative feedback involving G-protein-linked receptor inactivation and/or internalization, Gaq inactivation, and inositol triphosphate receptor inactivation [4–6]. The present paper describes a novel cross-talk effect: inhibition of thrombin-stimulated calcium mobilization mediated by thrombospondin (TSP) via one of its receptors, CD36. TSP is a 420-kDa trimer which can bind a variety of proteins and be a ligand for a variety of receptors and, consequently, has a wide range of biological activities. Cell surface TSP receptors include CD36, avb3 , a2b1 , a4b1 , a5b1 , heparan sulfate, sulfatides, and the integrin-associated protein, IAP [7– 11]. TSP binds to CD36 in two stages, the second of which involves the CSVTCG sequence within the type 1 repeats of TSP [12]. CD36 is also a multifunctional molecule. In addition to binding TSP, it binds collagen, acetylated or oxidized LDLs, phospholipids, and fatty acid-binding protein [13–16]. In platelets, the phosphorylated form of CD36 binds collagen, while the dephosphorylated form binds TSP [17]. Although little is known about CD36-mediated cell signaling, several src family members reportedly are associated with CD36 in endothelial cells and platelets [18, 19]. CD36 is expressed on microvascular endothelial cells (MVEC) but not on human umbilical vein endothelial cells (HUVEC) [20]. Hence, the present studies also provide new evidence for endothelial heterogeneity: that large and small vessel endothelial cells have distinct mechanisms of regulating intracellular calcium signaling.
465
AID
ECR 3863
/
6i2d$$$581
01-08-98 10:53:51
0014-4827/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
eca
466
ENENSTEIN ET AL.
MATERIALS AND METHODS Cell culture. MVEC were isolated from human foreskin by a panning method [21]. Briefly, primary isolates were cultured for 3 days and then panned with human endothelial cell antibody, EN4 (Monosan, The Netherlands). Cells were grown in medium MCDB 131 containing 20% male human serum, dibutyryl cAMP (246 mg/L), heparin (40 mg/L), hydrocortisone (1 mg/L), L-glutamine (1 mM), with fresh ECGF (50 mg/ml) in flasks coated with 1% gelatin. MVEC used were uniformly and strongly CD36 positive (by FACS analysis). HUVEC were isolated as previously described [22] and were uniformly negative for CD36. Reagents. TSP was prepared in our laboratory from human platelets using described methods [23]. Two monoclonal antibodies to CD36 were generous gifts: OKM5 (Ortho Diagnostics, Raritan, NJ) and 131.7 (N. Tandon, Rockville, MD). The anti-CD36 monoclonal antibody FA6-152 was obtained from Immunotech (Westbrook, ME). The monoclonal antibody (A2BII) to the b1 integrin was a gift from C. Damsky (San Francisco, CA); and a monoclonal antibody to avb3 integrin, LM609, was obtained from Chemicon (Temecula, CA). Vascular endothelial growth factor (VEGF) was a gift from S. Ramakrishnan (Minneapolis, MN). The bioactive TSP peptide CSVTCG and the scrambled control peptide VGCSTC, both with ACM-blocked cysteines, as well as the trypsin receptor peptide SLIGKVD were synthesized locally. The peptide SFLLRNPNDKYEPF, containing the active thrombin receptor-tethered ligand SFLLRN, was purchased (Sigma, St. Louis, MO). Measurement of [Ca2/]i by confocal fluorescence microscopy. MVEC up to passage 7 or HUVEC at passage 2 were plated onto NUNC coverslip tissue culture 2-well chambers which had been coated with 1% gelatin and were grown to confluence. Cells were washed twice [with Hanks’ balanced salt solution (HBSS) with 0.014% CaCl2 , 25 mM Hepes, 0.01% glucose, 0.5% bovine serum albumin (BSA), pH 7.4] and loaded with 5 mM INDO-1, AM (Molecular Probes, Inc., Eugene, OR) for 30–60 min in a 377C incubator with 5% CO2 . Cells were then washed three additional times. The final well volume was 1 ml. Chambers were placed on a 377C slide holder on the ACAS 570 Interactive Laser Cytometer (Meridian Instruments, Okemox, MI), which was focused on the endothelial cells in the plane of maximal INDO fluorescence. Typically, Ç10 cells were in the field of view and were therefore tested as a single experiment. The microscope was set to repeatedly scan the same field and to record the fluorescence emitted, resulting in data acquisition every 50 s. Scanning the entire field required 40 s, and there was a 10-s delay between scans. Although a field of cells is scanned sequentially from top to bottom during the 40 s, the computer records all cell responses from a single scan as if it were a single time point; for example, 0 s Å all cell responses of the first scan, and 50 s Å all cell responses of the second scan. There was no significant difference in maximal responses of cells recorded from the beginning or end of a scan. Reagent samples of 10 ml were added between scans directly over the test cells in the 1-ml well, producing an approximate 10-fold dilution. This effective dilution was confirmed by experiments in which the whole well volume was exchanged and replaced with 1 ml of reagent of a given concentration. (Such tests were impractical for general use.) All reagent concentrations in the text are reported as the nominal concentration assuming this full 10-fold dilution. Wells were scanned for 150 s after thrombin addition. Antagonists, such as TSP, were added 100 s (two scans) prior to the addition of thrombin. [Ca2/]i was detected as a change in the ratio of fluorescence emission intensity at two wavelengths, 405 and 480 nm, corresponding to INDO-1 binding calcium and calcium-free INDO-1, respectively. For a single experiment this ratio was averaged for 5–10 cells/well. Effects of modulating reagents are expressed in terms of calcium response in the presence of reagent as a percentage of the unmodulated thrombin-stimulated ratio increase for each experi-
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
ment. This percentage response was calculated from thrombin response values obtained on the same day, using endothelial cells from the same culture batch. [Ca2/]i was determined from an in vitro calcium working curve which was generated using calcium/EGTA standards, free INDO-1 (Molecular Probes), and the Meridian microscope working curve program. It should be noted that [Ca2/]i from an in vitro curve is typically higher than that generated by in situ determinations [24].
RESULTS
Thrombospondin inhibits thrombin-stimulated calcium responses. The [Ca2/]i response of our MVEC to a saturating concentration of 10 U/ml thrombin was 1 mM, consistent with responses previously reported for endothelial cells [25]. In order to monitor the effect of antagonists on thrombin-mediated calcium release in MVEC we selected a control thrombin concentration, 1 U/ml, which produced a large, but slightly submaximal, rise in [Ca2/]i . Figures 1A and 1B display a representative pair of experiments. MVEC responded strongly to 1 U/ml thrombin within 15 s of stimulation (Fig. 1A). However, a 100-s pretreatment of MVEC with 10 mg/ ml TSP significantly inhibited the thrombin-stimulated rise in [Ca2/]i (Fig. 1B). This represents a maximally inhibitory TSP concentration since 50 mg/ml was no more inhibitory than 10 mg/ml TSP. Serum TSP levels are normally 15–30 mg/ml [26]. Exposure to TSP alone had no effect on [Ca2/]i . When MVEC in several separate experiments were pretreated with 10–50 mg/ml TSP, the average response to thrombin was only 60% of control levels (Fig. 1C). This is also shown as individual pairs of experiments (Fig. 1D) to illustrate that much of the large standard deviation in TSP inhibition experiments was due to variability in the thrombin responsiveness itself. Paired t tests showed that the inhibitory effect was highly significant (P õ 0.003). In contrast, when HUVEC were treated with TSP, there was no inhibition of thrombin-stimulated calcium response (Fig. 1C). TSP inhibition occurs via CD36. Since MVEC but not HUVEC express CD36, these first studies suggested that TSP was acting via the CD36 receptor to inhibit thrombin-mediated calcium responses. We used several types of experiments to ascertain the specificity of the observed inhibitory effect of TSP as ligand and CD36 as receptor. The TSP synthetic peptide, CSVTCG, which specifically binds to CD36 [27], significantly inhibited thrombin-mediated calcium responses in MVEC, while the scrambled control peptide, VGCSTC, had no effect (Fig. 2). An alternate CD36 ligand, collagen IV, did not significantly inhibit thrombin-mediated calcium responses (Fig. 2). These results argue for specificity to TSP in inhibiting thrombin signaling. Pretreatment of MVEC with either OKM5 or FA6152 (which inhibit TSP binding to CD36 [28]) dimin-
eca
TSP INHIBITS THROMBIN/Ca2/ SIGNALING VIA CD36
467
FIG. 1. TSP-mediated inhibition of thrombin stimulated increases in [Ca2/]i . Representative experimental pair (A / B): response to submaximal thrombin stimulation. (A) MVEC (one field of 10 cells) were stimulated with 1 U/ml thrombin at 50 s (arrow), which typically produced a large, submaximal release of calcium. Each symbol indicates the signal from a single cell. (B) MVEC were treated with 10 mg/ ml TSP 100 s prior to thrombin stimulation. Ten cells show blunted response. (C) MVEC or HUVEC were treated with 10–50 mg/ml TSP 100 s prior to thrombin stimulation. The calcium response data are expressed as percentage of response in the absence of TSP (mean { SD). Each histogram represents several tests on at least 3 days of 5–10 cells per test. There were 10 paired MVEC tests ({ TSP) and 5 paired HUVEC tests. An asterisk indicates P õ 0.003 vs control as determined by paired t test. (D) Experimental pairs { thrombin from 6 separate days. Data recorded as the ratio of INDO with calcium bound to free INDO were averaged for 5–10 cells/field.
ished thrombin-mediated calcium responses, while the monoclonal antibody 131.7 (which recognizes a distinct collagen binding site [29]) had no inhibitory effect and, in fact, significantly enhanced the rise in [Ca2/]i (Fig. 2). Antibodies to other TSP receptors avb3 (LM609) (Fig. 2) and a4b1 (anti-b1) (not shown) did not inhibit thrombin-mediated increases in [Ca2/]i . These results argue for specificity to CD36. TSP-mediated inhibition specifically targets certain G-protein-linked receptor calcium release pathways.
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
Thrombin stimulates endothelial cells to release calcium via its G-protein-linked receptor and the PLC-b1dependent path of inositol triphosphate hydrolysis. In order to define the specificity of TSP inhibition of calcium release, it was necessary to determine (1) whether TSP inhibits the thrombin receptor pathway rather than binding thrombin itself, (2) whether TSP inhibited other G-protein-linked receptor pathways, and (3) whether both PLC-g and PLC-b pathways were inhibited by TSP.
eca
468
ENENSTEIN ET AL.
FIG. 2. CD36-mediated inhibition of thrombin signaling. MVEC were treated with a variety of CD36-binding peptides, monoclonal antibodies, or proteins for 100 s prior to thrombin stimulation. Percentage calcium response to thrombin percent is expressed as means { SD. 100% thrombin response is indicated by the dotted line. Antagonist concentrations: CSVTCG and VGCSTC, 100 mg/ml; monoclonal antibodies, 10–20 mg/ml; collagen IV, 10 mg/ml. All samples represent at least seven tests of 5–10 cells per test, paired with thrombin controls. For P values vs control, an asterisk indicates P õ 0.01 and c indicates P õ 0.02.
Thrombin cleaves its G-protein receptor, uncovering a ‘‘tethered ligand,’’ which then binds and activates the thrombin receptor. Peptides containing this tethered ligand sequence, SFLLRN, are capable of increasing [Ca2/]i in endothelial cells. As for thrombin itself, the calcium response to SFLLRN was successfully inhibited by pretreatment with TSP (Fig. 3). Since TSP-mediated inhibition was identical for SFLLRN and authentic thrombin, it appears that TSP is acting on the receptor or the pathway signaling, rather than by interacting with thrombin itself. PAR-2 (protease-activated receptor-2) is closely related to the thrombin receptor but is responsive to trypsin rather than thrombin [30]. Trypsin cleaves the human PAR-2, uncovering the tethered ligand SLIGKVD [31]. Either trypsin or SLIGKVD causes the rapid mobilization of [Ca2/]i . When MVEC were pretreated with TSP or anti-CD36 antibodies, the PAR-2-mediated [Ca2/]i increase was inhibited (Fig. 3). Therefore, TSP inhibits signaling of two members of the tethered ligand receptor family: the thrombin receptor and the trypsin receptor. The histamine receptor, H1, is a G-protein-linked receptor which, like the thrombin receptor, also is coupled to PLC-b. Histamine can trigger increases in [Ca2/]i via either Gaq [32] or Gi [33]. Histamine (1005 M) mobilized calcium in MVEC to an extent similar to that produced by 1 U/ml thrombin. TSP pretreatment
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
FIG. 3. Specificity of TSP-mediated inhibition. MVEC were stimulated by several molecules which cause the release of calcium into the cytosol: the thrombin receptor peptide 14-mer containing SFLLRN (100 mg/ml–40 mM), trypsin (2.5 mg/ml), the trypsin receptor peptide SLIGKVD (20 mg/ml), histamine (1005 M), and VEGF (20– 100 ng/ml). Data are shown as percentage of control (mean { SD), represented by the dashed line of 100% agonist-stimulated calcium release. TSP (10–20 mg/ml) or anti-CD36 (FA6-152) (20 mg/ml) pretreatment 100 s prior to agonist stimulation produced the above levels of inhibition. All samples represent at least four tests of 5–10 cells per test. An asterisk indicates P vs control Å 0.003 for SFLLRN, P Å 0.016 for trypsin, and P Å 0.036 for SLIGKVD.
eca
TSP INHIBITS THROMBIN/Ca2/ SIGNALING VIA CD36
of MVEC did not inhibit this histamine-stimulated calcium response (Fig. 3). This indicates that TSP-mediated inhibition is limited to a subset of G-protein-linked receptors. The VEGF receptor is a member of the tyrosine kinase family of growth factor receptors which signal the release of calcium via the PLC-g pathway [34]. When endothelial cells were pretreated with TSP, VEGFstimulated increases in calcium response were not inhibited (Fig. 3). This suggests that TSP does not affect the PLC-g pathway. However, since the [Ca2/]i response to VEGF is completely dependent on influx of extracellular calcium, the differential response to TSP pretreatment in VEGF- and thrombin-treated cells may reflect this use of different calcium stores [35]. The inhibitory effect of TSP is not the result of TGFb contamination. Recent studies have shown that TGFb contamination of TSP preparations may account for some of the biological properties attributed to TSP [36]. Furthermore, TSP can activate latent TGFb [37]. Western blots of our TSP showed no TGFb contamination (not shown). In addition, TGFb did not inhibit thrombin-mediated rises in [Ca2/]i , and antibodies to TGFb had no effect on TSP-mediated inhibition of thrombin signaling (not shown). Thus, we are confident that the results we observed are not due to TGFb contamination. DISCUSSION
Cells are constantly monitoring their external conditions. Cross-talk between receptor signaling systems permits one cell surface receptor to fine tune the responsive state of another receptor to receive and/or transmit signals. We have described the ability of one TSP receptor, the tyrosine kinase-associated CD36, to inhibit the signaling of the G-protein-linked receptor for thrombin. When CD36 was activated by TSP, by CSVTCG, or by antibodies to TSP-binding sites on CD36, thrombin-stimulated increases in [Ca2/]i were inhibited. TSP also inhibited calcium signaling by PAR-2, a related member of the tethered ligand receptor family. TSP did not inhibit the activity of several other receptors that mobilize calcium, suggesting that the CD36-mediated inhibition of calcium response was limited to a subset of G-protein-linked receptors or PLC signaling pathways. The ability of CSVTCG to inhibit thrombin-mediated calcium mobilization corroborates both the TSP and the CD36 studies. It pinpoints the action of TSP at CD36 and it excludes the possibility that antibodies are just acting via an Fc receptor. It also may provide a clue as to the mechanism of inhibition. Aggregation is a common requirement for intracellular signaling of various receptors such as the growth factor and inte-
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
469
grin families of receptors and of the nonreceptor tyrosine kinases [38–40]. The src family of tyrosine kinases may be activated by clustering [41]. Yet, there is conflicting evidence on the need for CD36 receptor aggregation prior to oxidative burst production in monocytes, and there is some evidence that receptor aggregation is needed for platelet activation [42]. CSVTCG by itself is not large or complex enough to cause CD36 dimerization. Therefore, either CD36-dependent inhibitory signaling does not require receptor aggregation or CD36 receptors exist in an already aggregated form on the cell surface and only need CSVTCG for activation. Since thrombin can form complexes with TSP [43], it was theoretically possible that TSP was directly inhibiting the thrombin molecule rather than inhibiting the receptor pathway. However, thrombin–TSP complex formation requires a 15-min half-time, which is much longer than the 3-min interaction in the present studies. In addition, TSP is unable to inhibit the amidolytic activity of thrombin [44]. TSP-mediated inhibition of calcium-releasing activity by the thrombin receptor peptide SFLLRN and the ability of CSVTCG to inhibit the response to authentic thrombin both argue that TSP acts on the receptor or its pathway, rather than upon thrombin itself. Pinpointing the mechanism of TSP/CD36 effects is difficult because both TSP and CD36 are multifunctional molecules. Moreover, their ionic environment, phosphorylation state, or ligand binding may alter their ability to convey a given signal. For example, platelets bind collagen when CD36 is phosphorylated, but bind TSP when CD36 is dephosphorylated [17]. Our results showed that the 131.7 antibody to a collagen binding epitope of CD36 stimulated thrombin-induced [Ca2/]i increases, while collagen IV itself was neither stimulatory nor inhibitory. Although the MVEC CD36 receptors apparently are unphosphorylated and unable to respond to collagen IV, perhaps the 131.7 antibody does not require the dephosphorylated state for binding and receptor activation. Little is known about how CD36 transmits signals. In platelets, one monoclonal antibody to CD36, NLO7, is reported to cause a small calcium response (but only in the presence of complement) which is inhibitable by pretreatment with OKM5 [45]. Pretreatment of PMNs with TSP is reported to inhibit pertussis toxin-dependent ribosylation of a G protein stimulated by FMLP [46], but this is not a CD36-dependent mechanism [47]. In MVEC, the src family members (fyn, c-src, and yes) are associated with CD36 [18]. Treatment of platelets with antibodies to CD36 causes dissociation of src family members from CD36 [48]. This suggests that tyrosine phosphorylation/dephosphorylation may be important in CD36 signaling. There is a precedent for tyrosine kinase involvement in and modulation of early G-protein-linked receptor
eca
470
ENENSTEIN ET AL.
signaling. Tyrosine phosphorylation of several platelet proteins occurs within 5 s of thrombin stimulation [49, 50]. Gaq , a major mediator of thrombin-stimulated calcium responses [51], has seven tyrosine residues available for phosphorylation, and several other Ga proteins are phosphorylated by src [52]. Future studies will elucidate which part or parts of the calcium release pathway are blocked by TSP/CD36 signaling and whether this inhibition involves tyrosine phosphorylation/dephosphorylation.
15.
16.
17.
REFERENCES 18. 1. Coughlin, S. R. (1993). Thrombin receptor structure and function. Thromb. Haemostasis 66, 184–187. 2. Garcia, J. G. N. (1992). Molecular mechanisms of thrombin-induced human and bovine endothelial cell activation. J. Lab. Clin. Med. 120, 513–519. 3. Hattori, R., Hamilton, K. K., Fugates, R. D., McEver, R. P., and Sims, P. J. (1989). Stimulated secretion of endothelial von Willebrand factor is accompanied by rapid redistribution to the cell surface of the intracellular granule membrane protein GMP140. J. Biol. Chem. 264, 7768–7771. 4. Ishii, K., Chen, J., Ishii, M., Koch, W. J., Freedman, N. J., Lefkowitz, R. J., and Coughlin, S. R. (1994). Inhibition of thrombin receptor signalling by a G-protein coupled receptor kinase. J. Biol. Chem. 269, 1125–1130.
19.
20.
21.
5. Hoxie, J. A., Ahuja, M., Belmonte, E., Pizarro, S., Parton, R., and Brass, L. F. (1993). Internalization and recycling of activated thrombin receptors. J. Biol. Chem. 268, 13756–13763. 6. Hein, L., Ishii, K., Coughlin, S. R., and Kobilka, B. K. (1994). Intracellular targeting and trafficking of thrombin receptors. J. Biol. Chem. 269, 1–8. 7. Bornstein, P. (1995). Diversity of function is inherent in matricellular proteins: An appraisal of thrombospondin 1. J. Cell Biol. 130, 503–506.
22.
23.
8. Gao, A. G., Lindberg, F. P., Finn, M. B., Blystone, S. D., Brown, E. J., and Frazier, W. A. (1996). Integrin-associated protein is a receptor for the C-terminal domain of thrombospondin. J. Biol. Chem. 271, 21–24.
24.
9. Mikhailenko, I., Kounnas, M. Z., and Strickland, D. K. (1995). Low density lipoprotein receptor-related protein/alpha 2-macroglobulin receptor mediates the cellular internalization and degradation of thrombospondin. A process facilitated by cell-surface proteoglycans. J. Biol. Chem. 270, 9543–9549.
25.
26.
10. Joneckis, C. C., Ackley, R. L., Orringer, E. P., Wayner, E. A., and Parise, L. V. (1993). Integrin alpha 4 beta 1 and glycoprotein IV (CD36) are expressed on circulating reticulocytes in sickle cell anemia. Blood 82, 3548–3555.
27.
11. Godyna, S., Liau, G., Popa, I., Stefansson, S., and Argraves, W. S. (1995). Identification of the low density lipoprotein receptor-related protein (LRP) as an endocytic receptor for thrombospondin-1. J. Cell Biol. 129, 1403–1410.
28.
12. Li, W. X., Howard, R. J., and Leung, L. L. (1993). Identification of SVTCG in thrombospondin as the conformation-dependent, high affinity binding site for its receptor, CD36. J. Biol. Chem. 268, 16179–16184. 13. Tandon, N. N., Lipsky, R. H., Burgess, W. H., and Jamieson, G. A. (1989). Isolation and characterization of platelet glycoprotein IV (CD36). J. Biol. Chem. 264, 7570–7575. 14. Acton, S. L., Scherer, P. E., Lodish, H. F., and Krieger, M.
AID
ECR 3863
/
6i2d$$$583
01-08-98 10:53:51
29.
30.
(1994). Expression cloning of SR-BI, a CD36-related class B scavenger receptor. J. Biol. Chem. 269, 21003–21009. Rigotti, A., Acton, S. L., and Krieger, M. (1995). The class B scavenger receptors SR-BI and CD36 are receptors for anionic phospholipids. J. Biol. Chem. 270, 16221–16224. Spitsberg, V. L., Matitashvili, E., and Gorewit, R. C. (1995). Association and coexpression of fatty-acid-binding protein and glycoprotein CD36 in the bovine mammary gland. Eur. J. Biochem. 230, 872–878. Asch, A. S., Liu, I., Briccetti, F. M., Barnwell, J. W., KwakyeBerko, F., Dokun, A., Goldberger, J., and Pernambuco, M. (1993). Analysis of CD36 binding domains: Ligand specificity controlled by dephosphorylation of an ectodomain. Science 262, 1436–1440. Bull, H. A., Brickell, P. M., and Dowd, P. M. (1994). Src-related protein tyrosine kinases are physically associated with the surface antigen CD36 in human dermal microvascular endothelial cell. FEBS Lett. 351, 41–44. Huang, M.-M., Bolen, J. B., Barnwell, J. W., Shattil, S. J., and Brugge, J. S. (1991). Membrane glycoprotein IV (CD36) is physically associated with the fyn, lyn, and yes protein tyrosine kinases in human platelets. Proc. Natl. Acad. Sci. USA 88, 7844– 7848. Swerlick, R. A., Lee, K. H., Wick, T. M., and Lawley, T. J. (1992). Human dermal microvascular endothelium but not human umbilical vein endothelial cells express CD36 in vivo and in vitro. J. Immunol. 148, 78–83. Gupta, K., Ramakrishnan, S., Browne, P. V., Solovey, A., Hebbel, R. P. (1997). A novel technique for culture of human dermal microvascular endothelial cells under either serum-free or serum-supplemented conditions: Isolation by panning and stimulation with vascular endothelial growth factor. Exp. Cell Res. 230, 244–251. Balla, J., Jacob, H. S., Balla, G., Nath, K., Eaton, J. W., and Vercellotti, G. M. (1993). Endothelial-cell heme uptake from heme proteins: Induction of sensitization and desensitization to oxidant damage. Proc. Natl. Acad. Sci. USA 90, 9285–9289. Murphy-Ullrich, J. E., and Mosher, D. F. (1985). Localization of thrombospondin in clots formed in situ. Blood 66, 1098–1104. Wahl, M., Lucherini, J., and Gruenstein, E. (1990). Intracellular calcium measurement with Indo-1 in substrate attached cells: Advantages and special consideration. Cell Calcium 11, 487– 500. Lerner, R. (1994). Changes of cytosolic calcium concentrations in human endothelial cells in response to thrombin, platelet activating factor and leukotriene B4. J. Lab. Clin. Med. 124, 723–729. Slayter, H. S. (1989). Secretion of thrombospondin from human blood platelets. Methods Enzymol. 169, 251–268. Asch, A. S., Silbiger, S., Heimer, E., and Nachman, R. L. (1992). Thrombospondin sequence motif (CSVTCG) is responsible for CD36 binding. Biochem. Biophys. Res. Commun. 182, 1208– 1217. Daviet, L., Buckland, R., Puente Navazo, M. D., and McGregor, J. L. (1995). Identification of an immunodominant functional domain on human CD36 antigen using human–mouse chimaeric proteins and homologue-replacement mutagenesis. Biochem. J. 305, 221–224. Matsuno, K., Diaz-Ricart, M., Montgomery, R. R., Aster, R. H., Jamieson, G. A., and Tandon, N. N. (1996). Inhibition of platelet adhesion to collagen by monoclonal anti-CD36 antibodies. Br. J. Haematol. 92, 960–967. Blackhart, D., Emilsson, K., Nguyen, D., Teng, W., Martelli, A. J., Nystedt, S., Sundelin, J., and Scarborough, R. M. (1996).
eca
TSP INHIBITS THROMBIN/Ca2/ SIGNALING VIA CD36 Ligand cross-reactivity within the protease-activated receptor family. J. Biol. Chem. 271, 16466–16471. 31. Bo¨hm, S. K., Kong, W., Bromme, D., Smeekens, S. D., Anderson, D. C., Connolly, A., Kahn, M., Nelken, N. A., Coughlin, S. R., Payan, D. G., and Bunnett, N. W. (1996). Molecular cloning expression and potential functions of the human proteinase-activated receptor-2. Biochem. J. 314, 1009–1016. 32. Leurs, R., Smit, M. J., Menge, W. M., and Timmerman, H. (1994). Pharmacological characterization of the human histamine H2 receptor stably expressed in Chinese hamster ovary cells. Br. J. Pharmacol. 112, 847–854. 33. Seifert, R., Hageluken, A., Hoer, A., Hoer, D., Grunbaum, L., Offermanns, S., Schwaner, I., Zingel, V., Schunack, W., and Schultz, G. (1994). The H1 receptor agonist 2-(3-chlorophenyl) histamine activates Gi proteins in HL-60 cells through a mechanism that is independent of known histamine receptor subtypes. Mol. Pharmacol. 45, 578–586. 34. Brock, T. A., Dvorak, H. F., and Senger, D. R. (1991). Tumorsecreted vascular permeability factor increases cytosolic Ca2/ and von Willebrand factor release in human endothelial cells. Am. J. Pathol. 138, 213–221. 35. Seymour, S. W., Shoaibi, M. A., Martin, A., Ahmed, A., Elvin, P., Kerr, D. J., and Wakelam, M. J. O. (1996). Vascular endothelial growth factor stimulates protein kinase C-dependent phospholipase D activity in endothelial cells. Lab. Invest. 75, 427–437. 36. Murphy-Ullrich, J. E., Schultz-Cherry, S., and Hook, M. (1992). Transforming growth factor-b complexes with thrombospondin. Mol. Biol. Cell 3, 181–188. 37. Schultz-Cherry, S., Ribeiro, S., Gentry, L., and Murphy-Ullrich, J. E. (1994). Thrombospondin binds and activates the small and large forms of latent transforming growth factor-b in a chemically defined system. J. Biol. Chem. 269, 26775–26782. 38. Schlessinger, J. (1988). Signal transduction by allosteric receptor oligomerization. Trends Biochem. Sci. 13, 443–447. 39. Miyamoto, S., Akiyama, S. K., and Yamada, K. M. (1995). Synergistic roles for receptor occupancy and aggregation in integrin transmembrane function. Science 267, 883–885. 40. Seed, B. (1995). Initiation of signal transduction by receptor aggregation: Role of nonreceptor tyrosine kinases. Semin. Immunol. 7, 3–11.
41. Bolen, J. B. (1991). Signal transduction by the SRC family of tyrosine protein kinases in hemopoietic cells. Cell Growth Differ. 2, 409–414. 42. Greenwalt, D. E., Lipsky, R. H., Ockenhouse, C. F., Ikeda, H., Tandon, N. N., and Jamieson, G. A. (1992). Membrane glycoprotein CD36: A review of its roles in adherence, signal transduction, and transfusion medicine. Blood 80, 1105–1115. 43. Browne, P. C., Miller, J. J., and Detwiler, T. C. (1988). Kinetics of the formation of thrombin-thrombospondin complexes. Involvement of a 77-kDa intermediate. Arch. Biochem. Biophys. 265, 534–538. 44. Hogg, P. J., Stenflo, J., and Mosher, D. F. (1992). Thrombospondin is a slow tight-binding inhibitor of plasmin. Biochemistry 31, 265–269. 45. Alessio, M., Greco, N. J., Primo, L., Ghigo, D., Bosia, A., Tandon, N. N., Ockenhouse, C. F., Jamieson, G. A., and Malavasi, F. (1993). Platelet activation and inhibition of malarial cytoadherence by the antiCD36 IgM monoclonal antibody NL07. Blood 82, 3637–3647. 46. Suchard, S. J., and Mansfield, P. J. (1994). Thrombospondin receptors are linked to G-proteins in human neutrophils. Molec. Biol. Cell 5, 64a. 47. Mansfield, P. J., and Suchard, S. J. (1994). Thrombospondin promotes chemotaxis and haptotaxis of human peripheral blood monocytes. J. Immunol. 153, 4219–4229. 48. Huang, M.-M., Indik, Z., Brass, L. F., Hoxie, J. A., Schreiber, A. D., and Brugge, J. S. (1992). Activation of FCgRII induces tyrosine phosphorylation of multiple proteins including FCgRII. J. Biol. Chem. 267, 5467–5473. 49. Nakamura, S., and Yamamura, H. (1989). Duality of plasmin effect on cytosolic free calcium in human platelets. J. Biol. Chem. 264, 7089–7091. 50. Golden, A., and Brugge, J. S. (1989). Thrombin treatment induces rapid change in tyrosine phosphorylation in platelets. Proc. Natl. Acad. Sci. USA 86, 901–905. 51. Clapham, D. E., and Neer, E. J. (1993). New roles for G-protein bg-dimers in transmembrane signalling. Nature 365, 403–406. 52. Hausdorff, W. P., Pitcher, J. A., Luttrell, D. K., Linder, M. E., Kurose, H., Parsons, S. J., Caron, M. G., and Lefkowitz, R. J. (1992). Tyrosine phosphorylation of G protein alpha subunits by pp60c-src. Proc. Natl. Acad. Sci. USA 89, 5720–5724.
Received August 20, 1997
AID
ECR 3863
/
6i2d$$$583
01-08-98 10:53:51
471
eca
238, 465–471 (1998)
EX973863
Thrombin-Stimulated Calcium Mobilization Is Inhibited by Thrombospondin via CD361 J. Enenstein,2 K. Gupta, G. M. Vercellotti, and R. P. Hebbel Department of Medicine and Hematology, University of Minnesota, Minneapolis, Minnesota 55455
Activation of the G-protein-linked thrombin receptor in endothelial cells normally leads to an increase in free intracellular calcium, [Ca2/]i , which is the proximate stimulus for many important cell functions. We present evidence showing that signals from CD36, the thrombospondin (TSP) receptor, can inhibit this thrombin-mediated calcium response. Human endothelial cells preloaded with Indo-1 exhibited rapid calcium mobilization in response to thrombin. The presence of TSP inhibited the thrombin-stimulated calcium response in CD36-positive microvascular endothelial cells but not in CD36-negative umbilical vein endothelial cells. This TSP effect was mimicked by anti-CD36 antibodies and a TSP peptide (CSVTCG), but not by an alternative CD36 ligand (collagen IV) or an antibody to an alternative TSP receptor (avb3 ). TSP also inhibited the calcium response to the thrombin receptor-tethered ligand peptide, SFLLRN. In addition, TSP and anti-CD36 antibodies inhibited the calcium response of a closely related receptor, the trypsin/SLIGKVD-activated receptor PAR-2. TSP did not indiscriminately inhibit all calcium release pathways, since histamine- or VEGF-stimulated calcium responses were not inhibited by TSP. We conclude that cross-talk from the CD36 receptor influences the responsive state of the endothelial thrombin receptor family and/or its signaling pathway. q 1998 Academic Press
INTRODUCTION
Cells can respond in a modulated manner to multiple environmental cues because there is cross-talk between heterologous cell surface receptor signaling pathways. Endothelial cells respond with carefully regulated changes in cytosolic calcium concentration ([Ca2/]i )3 to 1
This work was supported by NIH Grant HL30160. To whom correspondence and reprint requests should be addressed at Department of Medicine, University of Minnesota, UMHC Box 480, 420 Delaware Street S.E., Minneapolis, MN 55455. Fax: (612) 625-6919. 3 Abbreviations used: [Ca2/]i , intracellular calcium; HUVEC, human umbilical vein endothelial cells; MVEC, microvascular endothelial cells; TSP, thrombospondin; VEGF, vascular endothelial growth factor. 2
stimulation by thrombin, an important hemostatic agent in vivo that also promotes chemotaxis and mitogenesis [1]. Endothelial [Ca2/]i rises within 15 s of stimulation with thrombin [2], due to initial calcium release from intracellular stores and subsequent influx of extracellular calcium through receptor-mediated channels. This increase in [Ca2/]i is proximate to—and essential for—thrombin-stimulated vWF release, surface expression of P-selectin, prostacyclin production, disassembly of peripheral actin fibers, and increased assembly of stress fibers, and it is partially responsible for increased permeability [2, 3]. The thrombin-stimulated calcium response is known to be modulated via negative feedback involving G-protein-linked receptor inactivation and/or internalization, Gaq inactivation, and inositol triphosphate receptor inactivation [4–6]. The present paper describes a novel cross-talk effect: inhibition of thrombin-stimulated calcium mobilization mediated by thrombospondin (TSP) via one of its receptors, CD36. TSP is a 420-kDa trimer which can bind a variety of proteins and be a ligand for a variety of receptors and, consequently, has a wide range of biological activities. Cell surface TSP receptors include CD36, avb3 , a2b1 , a4b1 , a5b1 , heparan sulfate, sulfatides, and the integrin-associated protein, IAP [7– 11]. TSP binds to CD36 in two stages, the second of which involves the CSVTCG sequence within the type 1 repeats of TSP [12]. CD36 is also a multifunctional molecule. In addition to binding TSP, it binds collagen, acetylated or oxidized LDLs, phospholipids, and fatty acid-binding protein [13–16]. In platelets, the phosphorylated form of CD36 binds collagen, while the dephosphorylated form binds TSP [17]. Although little is known about CD36-mediated cell signaling, several src family members reportedly are associated with CD36 in endothelial cells and platelets [18, 19]. CD36 is expressed on microvascular endothelial cells (MVEC) but not on human umbilical vein endothelial cells (HUVEC) [20]. Hence, the present studies also provide new evidence for endothelial heterogeneity: that large and small vessel endothelial cells have distinct mechanisms of regulating intracellular calcium signaling.
465
AID
ECR 3863
/
6i2d$$$581
01-08-98 10:53:51
0014-4827/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
eca
466
ENENSTEIN ET AL.
MATERIALS AND METHODS Cell culture. MVEC were isolated from human foreskin by a panning method [21]. Briefly, primary isolates were cultured for 3 days and then panned with human endothelial cell antibody, EN4 (Monosan, The Netherlands). Cells were grown in medium MCDB 131 containing 20% male human serum, dibutyryl cAMP (246 mg/L), heparin (40 mg/L), hydrocortisone (1 mg/L), L-glutamine (1 mM), with fresh ECGF (50 mg/ml) in flasks coated with 1% gelatin. MVEC used were uniformly and strongly CD36 positive (by FACS analysis). HUVEC were isolated as previously described [22] and were uniformly negative for CD36. Reagents. TSP was prepared in our laboratory from human platelets using described methods [23]. Two monoclonal antibodies to CD36 were generous gifts: OKM5 (Ortho Diagnostics, Raritan, NJ) and 131.7 (N. Tandon, Rockville, MD). The anti-CD36 monoclonal antibody FA6-152 was obtained from Immunotech (Westbrook, ME). The monoclonal antibody (A2BII) to the b1 integrin was a gift from C. Damsky (San Francisco, CA); and a monoclonal antibody to avb3 integrin, LM609, was obtained from Chemicon (Temecula, CA). Vascular endothelial growth factor (VEGF) was a gift from S. Ramakrishnan (Minneapolis, MN). The bioactive TSP peptide CSVTCG and the scrambled control peptide VGCSTC, both with ACM-blocked cysteines, as well as the trypsin receptor peptide SLIGKVD were synthesized locally. The peptide SFLLRNPNDKYEPF, containing the active thrombin receptor-tethered ligand SFLLRN, was purchased (Sigma, St. Louis, MO). Measurement of [Ca2/]i by confocal fluorescence microscopy. MVEC up to passage 7 or HUVEC at passage 2 were plated onto NUNC coverslip tissue culture 2-well chambers which had been coated with 1% gelatin and were grown to confluence. Cells were washed twice [with Hanks’ balanced salt solution (HBSS) with 0.014% CaCl2 , 25 mM Hepes, 0.01% glucose, 0.5% bovine serum albumin (BSA), pH 7.4] and loaded with 5 mM INDO-1, AM (Molecular Probes, Inc., Eugene, OR) for 30–60 min in a 377C incubator with 5% CO2 . Cells were then washed three additional times. The final well volume was 1 ml. Chambers were placed on a 377C slide holder on the ACAS 570 Interactive Laser Cytometer (Meridian Instruments, Okemox, MI), which was focused on the endothelial cells in the plane of maximal INDO fluorescence. Typically, Ç10 cells were in the field of view and were therefore tested as a single experiment. The microscope was set to repeatedly scan the same field and to record the fluorescence emitted, resulting in data acquisition every 50 s. Scanning the entire field required 40 s, and there was a 10-s delay between scans. Although a field of cells is scanned sequentially from top to bottom during the 40 s, the computer records all cell responses from a single scan as if it were a single time point; for example, 0 s Å all cell responses of the first scan, and 50 s Å all cell responses of the second scan. There was no significant difference in maximal responses of cells recorded from the beginning or end of a scan. Reagent samples of 10 ml were added between scans directly over the test cells in the 1-ml well, producing an approximate 10-fold dilution. This effective dilution was confirmed by experiments in which the whole well volume was exchanged and replaced with 1 ml of reagent of a given concentration. (Such tests were impractical for general use.) All reagent concentrations in the text are reported as the nominal concentration assuming this full 10-fold dilution. Wells were scanned for 150 s after thrombin addition. Antagonists, such as TSP, were added 100 s (two scans) prior to the addition of thrombin. [Ca2/]i was detected as a change in the ratio of fluorescence emission intensity at two wavelengths, 405 and 480 nm, corresponding to INDO-1 binding calcium and calcium-free INDO-1, respectively. For a single experiment this ratio was averaged for 5–10 cells/well. Effects of modulating reagents are expressed in terms of calcium response in the presence of reagent as a percentage of the unmodulated thrombin-stimulated ratio increase for each experi-
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
ment. This percentage response was calculated from thrombin response values obtained on the same day, using endothelial cells from the same culture batch. [Ca2/]i was determined from an in vitro calcium working curve which was generated using calcium/EGTA standards, free INDO-1 (Molecular Probes), and the Meridian microscope working curve program. It should be noted that [Ca2/]i from an in vitro curve is typically higher than that generated by in situ determinations [24].
RESULTS
Thrombospondin inhibits thrombin-stimulated calcium responses. The [Ca2/]i response of our MVEC to a saturating concentration of 10 U/ml thrombin was 1 mM, consistent with responses previously reported for endothelial cells [25]. In order to monitor the effect of antagonists on thrombin-mediated calcium release in MVEC we selected a control thrombin concentration, 1 U/ml, which produced a large, but slightly submaximal, rise in [Ca2/]i . Figures 1A and 1B display a representative pair of experiments. MVEC responded strongly to 1 U/ml thrombin within 15 s of stimulation (Fig. 1A). However, a 100-s pretreatment of MVEC with 10 mg/ ml TSP significantly inhibited the thrombin-stimulated rise in [Ca2/]i (Fig. 1B). This represents a maximally inhibitory TSP concentration since 50 mg/ml was no more inhibitory than 10 mg/ml TSP. Serum TSP levels are normally 15–30 mg/ml [26]. Exposure to TSP alone had no effect on [Ca2/]i . When MVEC in several separate experiments were pretreated with 10–50 mg/ml TSP, the average response to thrombin was only 60% of control levels (Fig. 1C). This is also shown as individual pairs of experiments (Fig. 1D) to illustrate that much of the large standard deviation in TSP inhibition experiments was due to variability in the thrombin responsiveness itself. Paired t tests showed that the inhibitory effect was highly significant (P õ 0.003). In contrast, when HUVEC were treated with TSP, there was no inhibition of thrombin-stimulated calcium response (Fig. 1C). TSP inhibition occurs via CD36. Since MVEC but not HUVEC express CD36, these first studies suggested that TSP was acting via the CD36 receptor to inhibit thrombin-mediated calcium responses. We used several types of experiments to ascertain the specificity of the observed inhibitory effect of TSP as ligand and CD36 as receptor. The TSP synthetic peptide, CSVTCG, which specifically binds to CD36 [27], significantly inhibited thrombin-mediated calcium responses in MVEC, while the scrambled control peptide, VGCSTC, had no effect (Fig. 2). An alternate CD36 ligand, collagen IV, did not significantly inhibit thrombin-mediated calcium responses (Fig. 2). These results argue for specificity to TSP in inhibiting thrombin signaling. Pretreatment of MVEC with either OKM5 or FA6152 (which inhibit TSP binding to CD36 [28]) dimin-
eca
TSP INHIBITS THROMBIN/Ca2/ SIGNALING VIA CD36
467
FIG. 1. TSP-mediated inhibition of thrombin stimulated increases in [Ca2/]i . Representative experimental pair (A / B): response to submaximal thrombin stimulation. (A) MVEC (one field of 10 cells) were stimulated with 1 U/ml thrombin at 50 s (arrow), which typically produced a large, submaximal release of calcium. Each symbol indicates the signal from a single cell. (B) MVEC were treated with 10 mg/ ml TSP 100 s prior to thrombin stimulation. Ten cells show blunted response. (C) MVEC or HUVEC were treated with 10–50 mg/ml TSP 100 s prior to thrombin stimulation. The calcium response data are expressed as percentage of response in the absence of TSP (mean { SD). Each histogram represents several tests on at least 3 days of 5–10 cells per test. There were 10 paired MVEC tests ({ TSP) and 5 paired HUVEC tests. An asterisk indicates P õ 0.003 vs control as determined by paired t test. (D) Experimental pairs { thrombin from 6 separate days. Data recorded as the ratio of INDO with calcium bound to free INDO were averaged for 5–10 cells/field.
ished thrombin-mediated calcium responses, while the monoclonal antibody 131.7 (which recognizes a distinct collagen binding site [29]) had no inhibitory effect and, in fact, significantly enhanced the rise in [Ca2/]i (Fig. 2). Antibodies to other TSP receptors avb3 (LM609) (Fig. 2) and a4b1 (anti-b1) (not shown) did not inhibit thrombin-mediated increases in [Ca2/]i . These results argue for specificity to CD36. TSP-mediated inhibition specifically targets certain G-protein-linked receptor calcium release pathways.
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
Thrombin stimulates endothelial cells to release calcium via its G-protein-linked receptor and the PLC-b1dependent path of inositol triphosphate hydrolysis. In order to define the specificity of TSP inhibition of calcium release, it was necessary to determine (1) whether TSP inhibits the thrombin receptor pathway rather than binding thrombin itself, (2) whether TSP inhibited other G-protein-linked receptor pathways, and (3) whether both PLC-g and PLC-b pathways were inhibited by TSP.
eca
468
ENENSTEIN ET AL.
FIG. 2. CD36-mediated inhibition of thrombin signaling. MVEC were treated with a variety of CD36-binding peptides, monoclonal antibodies, or proteins for 100 s prior to thrombin stimulation. Percentage calcium response to thrombin percent is expressed as means { SD. 100% thrombin response is indicated by the dotted line. Antagonist concentrations: CSVTCG and VGCSTC, 100 mg/ml; monoclonal antibodies, 10–20 mg/ml; collagen IV, 10 mg/ml. All samples represent at least seven tests of 5–10 cells per test, paired with thrombin controls. For P values vs control, an asterisk indicates P õ 0.01 and c indicates P õ 0.02.
Thrombin cleaves its G-protein receptor, uncovering a ‘‘tethered ligand,’’ which then binds and activates the thrombin receptor. Peptides containing this tethered ligand sequence, SFLLRN, are capable of increasing [Ca2/]i in endothelial cells. As for thrombin itself, the calcium response to SFLLRN was successfully inhibited by pretreatment with TSP (Fig. 3). Since TSP-mediated inhibition was identical for SFLLRN and authentic thrombin, it appears that TSP is acting on the receptor or the pathway signaling, rather than by interacting with thrombin itself. PAR-2 (protease-activated receptor-2) is closely related to the thrombin receptor but is responsive to trypsin rather than thrombin [30]. Trypsin cleaves the human PAR-2, uncovering the tethered ligand SLIGKVD [31]. Either trypsin or SLIGKVD causes the rapid mobilization of [Ca2/]i . When MVEC were pretreated with TSP or anti-CD36 antibodies, the PAR-2-mediated [Ca2/]i increase was inhibited (Fig. 3). Therefore, TSP inhibits signaling of two members of the tethered ligand receptor family: the thrombin receptor and the trypsin receptor. The histamine receptor, H1, is a G-protein-linked receptor which, like the thrombin receptor, also is coupled to PLC-b. Histamine can trigger increases in [Ca2/]i via either Gaq [32] or Gi [33]. Histamine (1005 M) mobilized calcium in MVEC to an extent similar to that produced by 1 U/ml thrombin. TSP pretreatment
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
FIG. 3. Specificity of TSP-mediated inhibition. MVEC were stimulated by several molecules which cause the release of calcium into the cytosol: the thrombin receptor peptide 14-mer containing SFLLRN (100 mg/ml–40 mM), trypsin (2.5 mg/ml), the trypsin receptor peptide SLIGKVD (20 mg/ml), histamine (1005 M), and VEGF (20– 100 ng/ml). Data are shown as percentage of control (mean { SD), represented by the dashed line of 100% agonist-stimulated calcium release. TSP (10–20 mg/ml) or anti-CD36 (FA6-152) (20 mg/ml) pretreatment 100 s prior to agonist stimulation produced the above levels of inhibition. All samples represent at least four tests of 5–10 cells per test. An asterisk indicates P vs control Å 0.003 for SFLLRN, P Å 0.016 for trypsin, and P Å 0.036 for SLIGKVD.
eca
TSP INHIBITS THROMBIN/Ca2/ SIGNALING VIA CD36
of MVEC did not inhibit this histamine-stimulated calcium response (Fig. 3). This indicates that TSP-mediated inhibition is limited to a subset of G-protein-linked receptors. The VEGF receptor is a member of the tyrosine kinase family of growth factor receptors which signal the release of calcium via the PLC-g pathway [34]. When endothelial cells were pretreated with TSP, VEGFstimulated increases in calcium response were not inhibited (Fig. 3). This suggests that TSP does not affect the PLC-g pathway. However, since the [Ca2/]i response to VEGF is completely dependent on influx of extracellular calcium, the differential response to TSP pretreatment in VEGF- and thrombin-treated cells may reflect this use of different calcium stores [35]. The inhibitory effect of TSP is not the result of TGFb contamination. Recent studies have shown that TGFb contamination of TSP preparations may account for some of the biological properties attributed to TSP [36]. Furthermore, TSP can activate latent TGFb [37]. Western blots of our TSP showed no TGFb contamination (not shown). In addition, TGFb did not inhibit thrombin-mediated rises in [Ca2/]i , and antibodies to TGFb had no effect on TSP-mediated inhibition of thrombin signaling (not shown). Thus, we are confident that the results we observed are not due to TGFb contamination. DISCUSSION
Cells are constantly monitoring their external conditions. Cross-talk between receptor signaling systems permits one cell surface receptor to fine tune the responsive state of another receptor to receive and/or transmit signals. We have described the ability of one TSP receptor, the tyrosine kinase-associated CD36, to inhibit the signaling of the G-protein-linked receptor for thrombin. When CD36 was activated by TSP, by CSVTCG, or by antibodies to TSP-binding sites on CD36, thrombin-stimulated increases in [Ca2/]i were inhibited. TSP also inhibited calcium signaling by PAR-2, a related member of the tethered ligand receptor family. TSP did not inhibit the activity of several other receptors that mobilize calcium, suggesting that the CD36-mediated inhibition of calcium response was limited to a subset of G-protein-linked receptors or PLC signaling pathways. The ability of CSVTCG to inhibit thrombin-mediated calcium mobilization corroborates both the TSP and the CD36 studies. It pinpoints the action of TSP at CD36 and it excludes the possibility that antibodies are just acting via an Fc receptor. It also may provide a clue as to the mechanism of inhibition. Aggregation is a common requirement for intracellular signaling of various receptors such as the growth factor and inte-
AID
ECR 3863
/
6i2d$$$582
01-08-98 10:53:51
469
grin families of receptors and of the nonreceptor tyrosine kinases [38–40]. The src family of tyrosine kinases may be activated by clustering [41]. Yet, there is conflicting evidence on the need for CD36 receptor aggregation prior to oxidative burst production in monocytes, and there is some evidence that receptor aggregation is needed for platelet activation [42]. CSVTCG by itself is not large or complex enough to cause CD36 dimerization. Therefore, either CD36-dependent inhibitory signaling does not require receptor aggregation or CD36 receptors exist in an already aggregated form on the cell surface and only need CSVTCG for activation. Since thrombin can form complexes with TSP [43], it was theoretically possible that TSP was directly inhibiting the thrombin molecule rather than inhibiting the receptor pathway. However, thrombin–TSP complex formation requires a 15-min half-time, which is much longer than the 3-min interaction in the present studies. In addition, TSP is unable to inhibit the amidolytic activity of thrombin [44]. TSP-mediated inhibition of calcium-releasing activity by the thrombin receptor peptide SFLLRN and the ability of CSVTCG to inhibit the response to authentic thrombin both argue that TSP acts on the receptor or its pathway, rather than upon thrombin itself. Pinpointing the mechanism of TSP/CD36 effects is difficult because both TSP and CD36 are multifunctional molecules. Moreover, their ionic environment, phosphorylation state, or ligand binding may alter their ability to convey a given signal. For example, platelets bind collagen when CD36 is phosphorylated, but bind TSP when CD36 is dephosphorylated [17]. Our results showed that the 131.7 antibody to a collagen binding epitope of CD36 stimulated thrombin-induced [Ca2/]i increases, while collagen IV itself was neither stimulatory nor inhibitory. Although the MVEC CD36 receptors apparently are unphosphorylated and unable to respond to collagen IV, perhaps the 131.7 antibody does not require the dephosphorylated state for binding and receptor activation. Little is known about how CD36 transmits signals. In platelets, one monoclonal antibody to CD36, NLO7, is reported to cause a small calcium response (but only in the presence of complement) which is inhibitable by pretreatment with OKM5 [45]. Pretreatment of PMNs with TSP is reported to inhibit pertussis toxin-dependent ribosylation of a G protein stimulated by FMLP [46], but this is not a CD36-dependent mechanism [47]. In MVEC, the src family members (fyn, c-src, and yes) are associated with CD36 [18]. Treatment of platelets with antibodies to CD36 causes dissociation of src family members from CD36 [48]. This suggests that tyrosine phosphorylation/dephosphorylation may be important in CD36 signaling. There is a precedent for tyrosine kinase involvement in and modulation of early G-protein-linked receptor
eca
470
ENENSTEIN ET AL.
signaling. Tyrosine phosphorylation of several platelet proteins occurs within 5 s of thrombin stimulation [49, 50]. Gaq , a major mediator of thrombin-stimulated calcium responses [51], has seven tyrosine residues available for phosphorylation, and several other Ga proteins are phosphorylated by src [52]. Future studies will elucidate which part or parts of the calcium release pathway are blocked by TSP/CD36 signaling and whether this inhibition involves tyrosine phosphorylation/dephosphorylation.
15.
16.
17.
REFERENCES 18. 1. Coughlin, S. R. (1993). Thrombin receptor structure and function. Thromb. Haemostasis 66, 184–187. 2. Garcia, J. G. N. (1992). Molecular mechanisms of thrombin-induced human and bovine endothelial cell activation. J. Lab. Clin. Med. 120, 513–519. 3. Hattori, R., Hamilton, K. K., Fugates, R. D., McEver, R. P., and Sims, P. J. (1989). Stimulated secretion of endothelial von Willebrand factor is accompanied by rapid redistribution to the cell surface of the intracellular granule membrane protein GMP140. J. Biol. Chem. 264, 7768–7771. 4. Ishii, K., Chen, J., Ishii, M., Koch, W. J., Freedman, N. J., Lefkowitz, R. J., and Coughlin, S. R. (1994). Inhibition of thrombin receptor signalling by a G-protein coupled receptor kinase. J. Biol. Chem. 269, 1125–1130.
19.
20.
21.
5. Hoxie, J. A., Ahuja, M., Belmonte, E., Pizarro, S., Parton, R., and Brass, L. F. (1993). Internalization and recycling of activated thrombin receptors. J. Biol. Chem. 268, 13756–13763. 6. Hein, L., Ishii, K., Coughlin, S. R., and Kobilka, B. K. (1994). Intracellular targeting and trafficking of thrombin receptors. J. Biol. Chem. 269, 1–8. 7. Bornstein, P. (1995). Diversity of function is inherent in matricellular proteins: An appraisal of thrombospondin 1. J. Cell Biol. 130, 503–506.
22.
23.
8. Gao, A. G., Lindberg, F. P., Finn, M. B., Blystone, S. D., Brown, E. J., and Frazier, W. A. (1996). Integrin-associated protein is a receptor for the C-terminal domain of thrombospondin. J. Biol. Chem. 271, 21–24.
24.
9. Mikhailenko, I., Kounnas, M. Z., and Strickland, D. K. (1995). Low density lipoprotein receptor-related protein/alpha 2-macroglobulin receptor mediates the cellular internalization and degradation of thrombospondin. A process facilitated by cell-surface proteoglycans. J. Biol. Chem. 270, 9543–9549.
25.
26.
10. Joneckis, C. C., Ackley, R. L., Orringer, E. P., Wayner, E. A., and Parise, L. V. (1993). Integrin alpha 4 beta 1 and glycoprotein IV (CD36) are expressed on circulating reticulocytes in sickle cell anemia. Blood 82, 3548–3555.
27.
11. Godyna, S., Liau, G., Popa, I., Stefansson, S., and Argraves, W. S. (1995). Identification of the low density lipoprotein receptor-related protein (LRP) as an endocytic receptor for thrombospondin-1. J. Cell Biol. 129, 1403–1410.
28.
12. Li, W. X., Howard, R. J., and Leung, L. L. (1993). Identification of SVTCG in thrombospondin as the conformation-dependent, high affinity binding site for its receptor, CD36. J. Biol. Chem. 268, 16179–16184. 13. Tandon, N. N., Lipsky, R. H., Burgess, W. H., and Jamieson, G. A. (1989). Isolation and characterization of platelet glycoprotein IV (CD36). J. Biol. Chem. 264, 7570–7575. 14. Acton, S. L., Scherer, P. E., Lodish, H. F., and Krieger, M.
AID
ECR 3863
/
6i2d$$$583
01-08-98 10:53:51
29.
30.
(1994). Expression cloning of SR-BI, a CD36-related class B scavenger receptor. J. Biol. Chem. 269, 21003–21009. Rigotti, A., Acton, S. L., and Krieger, M. (1995). The class B scavenger receptors SR-BI and CD36 are receptors for anionic phospholipids. J. Biol. Chem. 270, 16221–16224. Spitsberg, V. L., Matitashvili, E., and Gorewit, R. C. (1995). Association and coexpression of fatty-acid-binding protein and glycoprotein CD36 in the bovine mammary gland. Eur. J. Biochem. 230, 872–878. Asch, A. S., Liu, I., Briccetti, F. M., Barnwell, J. W., KwakyeBerko, F., Dokun, A., Goldberger, J., and Pernambuco, M. (1993). Analysis of CD36 binding domains: Ligand specificity controlled by dephosphorylation of an ectodomain. Science 262, 1436–1440. Bull, H. A., Brickell, P. M., and Dowd, P. M. (1994). Src-related protein tyrosine kinases are physically associated with the surface antigen CD36 in human dermal microvascular endothelial cell. FEBS Lett. 351, 41–44. Huang, M.-M., Bolen, J. B., Barnwell, J. W., Shattil, S. J., and Brugge, J. S. (1991). Membrane glycoprotein IV (CD36) is physically associated with the fyn, lyn, and yes protein tyrosine kinases in human platelets. Proc. Natl. Acad. Sci. USA 88, 7844– 7848. Swerlick, R. A., Lee, K. H., Wick, T. M., and Lawley, T. J. (1992). Human dermal microvascular endothelium but not human umbilical vein endothelial cells express CD36 in vivo and in vitro. J. Immunol. 148, 78–83. Gupta, K., Ramakrishnan, S., Browne, P. V., Solovey, A., Hebbel, R. P. (1997). A novel technique for culture of human dermal microvascular endothelial cells under either serum-free or serum-supplemented conditions: Isolation by panning and stimulation with vascular endothelial growth factor. Exp. Cell Res. 230, 244–251. Balla, J., Jacob, H. S., Balla, G., Nath, K., Eaton, J. W., and Vercellotti, G. M. (1993). Endothelial-cell heme uptake from heme proteins: Induction of sensitization and desensitization to oxidant damage. Proc. Natl. Acad. Sci. USA 90, 9285–9289. Murphy-Ullrich, J. E., and Mosher, D. F. (1985). Localization of thrombospondin in clots formed in situ. Blood 66, 1098–1104. Wahl, M., Lucherini, J., and Gruenstein, E. (1990). Intracellular calcium measurement with Indo-1 in substrate attached cells: Advantages and special consideration. Cell Calcium 11, 487– 500. Lerner, R. (1994). Changes of cytosolic calcium concentrations in human endothelial cells in response to thrombin, platelet activating factor and leukotriene B4. J. Lab. Clin. Med. 124, 723–729. Slayter, H. S. (1989). Secretion of thrombospondin from human blood platelets. Methods Enzymol. 169, 251–268. Asch, A. S., Silbiger, S., Heimer, E., and Nachman, R. L. (1992). Thrombospondin sequence motif (CSVTCG) is responsible for CD36 binding. Biochem. Biophys. Res. Commun. 182, 1208– 1217. Daviet, L., Buckland, R., Puente Navazo, M. D., and McGregor, J. L. (1995). Identification of an immunodominant functional domain on human CD36 antigen using human–mouse chimaeric proteins and homologue-replacement mutagenesis. Biochem. J. 305, 221–224. Matsuno, K., Diaz-Ricart, M., Montgomery, R. R., Aster, R. H., Jamieson, G. A., and Tandon, N. N. (1996). Inhibition of platelet adhesion to collagen by monoclonal anti-CD36 antibodies. Br. J. Haematol. 92, 960–967. Blackhart, D., Emilsson, K., Nguyen, D., Teng, W., Martelli, A. J., Nystedt, S., Sundelin, J., and Scarborough, R. M. (1996).
eca
TSP INHIBITS THROMBIN/Ca2/ SIGNALING VIA CD36 Ligand cross-reactivity within the protease-activated receptor family. J. Biol. Chem. 271, 16466–16471. 31. Bo¨hm, S. K., Kong, W., Bromme, D., Smeekens, S. D., Anderson, D. C., Connolly, A., Kahn, M., Nelken, N. A., Coughlin, S. R., Payan, D. G., and Bunnett, N. W. (1996). Molecular cloning expression and potential functions of the human proteinase-activated receptor-2. Biochem. J. 314, 1009–1016. 32. Leurs, R., Smit, M. J., Menge, W. M., and Timmerman, H. (1994). Pharmacological characterization of the human histamine H2 receptor stably expressed in Chinese hamster ovary cells. Br. J. Pharmacol. 112, 847–854. 33. Seifert, R., Hageluken, A., Hoer, A., Hoer, D., Grunbaum, L., Offermanns, S., Schwaner, I., Zingel, V., Schunack, W., and Schultz, G. (1994). The H1 receptor agonist 2-(3-chlorophenyl) histamine activates Gi proteins in HL-60 cells through a mechanism that is independent of known histamine receptor subtypes. Mol. Pharmacol. 45, 578–586. 34. Brock, T. A., Dvorak, H. F., and Senger, D. R. (1991). Tumorsecreted vascular permeability factor increases cytosolic Ca2/ and von Willebrand factor release in human endothelial cells. Am. J. Pathol. 138, 213–221. 35. Seymour, S. W., Shoaibi, M. A., Martin, A., Ahmed, A., Elvin, P., Kerr, D. J., and Wakelam, M. J. O. (1996). Vascular endothelial growth factor stimulates protein kinase C-dependent phospholipase D activity in endothelial cells. Lab. Invest. 75, 427–437. 36. Murphy-Ullrich, J. E., Schultz-Cherry, S., and Hook, M. (1992). Transforming growth factor-b complexes with thrombospondin. Mol. Biol. Cell 3, 181–188. 37. Schultz-Cherry, S., Ribeiro, S., Gentry, L., and Murphy-Ullrich, J. E. (1994). Thrombospondin binds and activates the small and large forms of latent transforming growth factor-b in a chemically defined system. J. Biol. Chem. 269, 26775–26782. 38. Schlessinger, J. (1988). Signal transduction by allosteric receptor oligomerization. Trends Biochem. Sci. 13, 443–447. 39. Miyamoto, S., Akiyama, S. K., and Yamada, K. M. (1995). Synergistic roles for receptor occupancy and aggregation in integrin transmembrane function. Science 267, 883–885. 40. Seed, B. (1995). Initiation of signal transduction by receptor aggregation: Role of nonreceptor tyrosine kinases. Semin. Immunol. 7, 3–11.
41. Bolen, J. B. (1991). Signal transduction by the SRC family of tyrosine protein kinases in hemopoietic cells. Cell Growth Differ. 2, 409–414. 42. Greenwalt, D. E., Lipsky, R. H., Ockenhouse, C. F., Ikeda, H., Tandon, N. N., and Jamieson, G. A. (1992). Membrane glycoprotein CD36: A review of its roles in adherence, signal transduction, and transfusion medicine. Blood 80, 1105–1115. 43. Browne, P. C., Miller, J. J., and Detwiler, T. C. (1988). Kinetics of the formation of thrombin-thrombospondin complexes. Involvement of a 77-kDa intermediate. Arch. Biochem. Biophys. 265, 534–538. 44. Hogg, P. J., Stenflo, J., and Mosher, D. F. (1992). Thrombospondin is a slow tight-binding inhibitor of plasmin. Biochemistry 31, 265–269. 45. Alessio, M., Greco, N. J., Primo, L., Ghigo, D., Bosia, A., Tandon, N. N., Ockenhouse, C. F., Jamieson, G. A., and Malavasi, F. (1993). Platelet activation and inhibition of malarial cytoadherence by the antiCD36 IgM monoclonal antibody NL07. Blood 82, 3637–3647. 46. Suchard, S. J., and Mansfield, P. J. (1994). Thrombospondin receptors are linked to G-proteins in human neutrophils. Molec. Biol. Cell 5, 64a. 47. Mansfield, P. J., and Suchard, S. J. (1994). Thrombospondin promotes chemotaxis and haptotaxis of human peripheral blood monocytes. J. Immunol. 153, 4219–4229. 48. Huang, M.-M., Indik, Z., Brass, L. F., Hoxie, J. A., Schreiber, A. D., and Brugge, J. S. (1992). Activation of FCgRII induces tyrosine phosphorylation of multiple proteins including FCgRII. J. Biol. Chem. 267, 5467–5473. 49. Nakamura, S., and Yamamura, H. (1989). Duality of plasmin effect on cytosolic free calcium in human platelets. J. Biol. Chem. 264, 7089–7091. 50. Golden, A., and Brugge, J. S. (1989). Thrombin treatment induces rapid change in tyrosine phosphorylation in platelets. Proc. Natl. Acad. Sci. USA 86, 901–905. 51. Clapham, D. E., and Neer, E. J. (1993). New roles for G-protein bg-dimers in transmembrane signalling. Nature 365, 403–406. 52. Hausdorff, W. P., Pitcher, J. A., Luttrell, D. K., Linder, M. E., Kurose, H., Parsons, S. J., Caron, M. G., and Lefkowitz, R. J. (1992). Tyrosine phosphorylation of G protein alpha subunits by pp60c-src. Proc. Natl. Acad. Sci. USA 89, 5720–5724.
Received August 20, 1997
AID
ECR 3863
/
6i2d$$$583
01-08-98 10:53:51
471
eca