Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1.

Thrombospondin-1 (TSP-1) is a naturally occurring inhibitor of angiogenesis that limits vessel density in normal tissues and curtails tumor growth. Here, we show that the inhibition of angiogenesis in vitro and in vivo and the induction of apoptosis by thrombospondin-1 all required the sequential activation of CD36, p59fyn, caspase-3 like proteases and p38 mitogen-activated protein kinases. We also detected increased endothelial cell apoptosis in situ at the margins of tumors in mice treated with thrombospondin-1. These results indicate that thrombospondin-1, and possibly other broad-spectrum natural inhibitors of angiogenesis, act in vivo by inducing receptor-mediated apoptosis in activated microvascular endothelial cells.

Differential roles of CD36 and alphavbeta5 integrin in photoreceptor phagocytosis by the retinal pigment epithelium.

Retinal pigment epithelial (RPE) cells employ alphavbeta5 integrin and CD36 receptors to phagocytose photoreceptor outer segment fragments (OS). We explored special properties of RPE phagocytosis to identify the contribution of CD36 to RPE phagocytosis measuring effects of CD36 antibodies on OS binding and internalization kinetics. Early, CD36 antibodies had no effect on OS binding or internalization. Both control and CD36 antibody treated RPE initiated internalization approximately 2 hours after OS challenge. Later, bivalent CD36 IgG accelerated OS engulfment while monovalent Fab fragments inhibited engulfment. Cross-linking Fab fragments restored the accelerating activity of intact IgG. Strikingly, antibodies were effective even if added to OS already bound by RPE. alphavbeta5 blocking antibody reduced OS binding equally well in the presence of CD36 antibodies but CD36 antibodies accelerated internalization of remaining bound OS. Furthermore, CD36 ligation at either apical or basal RPE surface partially substituted for soluble factors that are required for internalization but not for binding of OS at the RPE apical surface. Our results demonstrate that CD36 ligation is necessary and sufficient to activate the OS internalization mechanism of RPE. They suggest that CD36 acts as a signaling molecule in postbinding steps of RPE phagocytosis independently of the OS binding receptor alphavbeta5 integrin.

CD36 gene transfer confers capacity for phagocytosis of cells undergoing apoptosis.

Phagocyte recognition and ingestion of intact cells undergoing apoptosis are key events in this generally important program of cell death. Insufficient phagocyte capacity for apoptotic cells can result in failure to clear dying cells before membrane integrity is lost, resulting in leakage of noxious cell contents and severe tissue damage. However, no means has been available to increase phagocytic clearance of apoptotic cells. We now report that transfection of the macrophage adhesion molecule CD36 into human Bowes melanoma cells specifically conferred greatly increased capacity to ingest apoptotic neutrophils, lymphocytes, and fibroblasts, comparable to that exhibited by macrophages. Furthermore, when CD36 was transfected into another cell type with limited capacity to take up apoptotic bodies, the monkey COS-7 cell, similar effects were observed. Therefore, CD36 gene transfer can confer "professional" capacity to ingest apoptotic cells upon "amateur" phagocytes.

Interleukin 1 induces endothelial cell synthesis of plasminogen activator inhibitor.

Human endothelial cells activated with IL-1 express a surface membrane-oriented procoagulant generating system characterized by increased tissue factor synthesis and decreased thrombomodulin activity. We now report that IL-1 also stimulates endothelial cell synthesis of plasminogen activator inhibitor. This array of IL-1-induced activities shifts the balance at the endothelial cell surface to a prothrombotic influence and may reflect an early response of the blood vessel wall to injury.

Identification of a CD36-related thrombospondin 1-binding domain in HIV-1 envelope glycoprotein gp120: relationship to HIV-1-specific inhibitory factors in human saliva.

Human and non-human primate salivas retard the infectivity of HIV-1 in vitro and in vivo. Because thrombospondin 1 (TSP1), a high molecular weight trimeric glycoprotein, is concentrated in saliva and can inhibit the infectivity of diverse pathogens in vitro, we sought to determine the role of TSP1 in suppression of HIV infectivity. Sequence analysis revealed a TSP1 recognition motif, previously defined for the CD36 gene family of cell adhesion receptors, in conserved regions flanking the disulfide-linked cysteine residues of the V3 loop of HIV envelope glycoprotein gp120, important for HIV binding to its high affinity cellular receptor CD4. Using solid-phase in vitro binding assays, we demonstrate direct binding of radiolabeled TSP1 to immobilized recombinant gp120. Based on peptide blocking experiments, the TSP1-gp120 interaction involves CSVTCG sequences in the type 1 properdin-like repeats of TSP1, the known binding site for CD36. TSP1 and fusion proteins derived from CD36-related TSP1-binding domains were able to compete with radiolabeled soluble CD4 binding to immobilized gp120. In parallel, purified TSP1 inhibited HIV-1 infection of peripheral blood mononuclear cells and transformed T and promonocytic cell lines. Levels of TSP1 required for both viral aggregation and direct blockade of HIV-1 infection were physiologic, and affinity depletion of salivary TSP1 abrogated >70% of the inhibitory effect of whole saliva on HIV infectivity. Characterization of TSP1-gp120 binding specificity suggests a mechanism for direct blockade of HIV infectivity that might be exploited to retard HIV transmission that occurs via mucosal routes.

Tumor necrosis factor-mediated release of platelet-derived growth factor from cultured endothelial cells.

Platelet-derived growth factor (PDGF) is a 30,000-Mr glycoprotein that is chemotactic and mitogenic for vascular smooth muscle cells (SMC). It is also a potent vasoconstrictor. In the present study, we found that the macrophage-derived polypeptide, tumor necrosis factor (TNF), releases a factor from human umbilical vein endothelial cells (EC) that is mitogenic for SMC. Postculture medium from TNF-stimulated EC induced a 90% increase in mitogenesis is compared with controls. This effect was half-maximal at a TNF dose of 114 pM, reflected a 2.5-fold increase in PDGF-specific mRNA synthesis, and peaked at 15 h of TNF stimulation. Mitogenic activity was completely abrogated by preincubation of postculture medium with antibody to platelet PDGF. Stimulation of EC with IL-1 (60-240 pM) led to the release of similar mitogenic activity. Thus, in addition to its effects on the hemostatic and adhesive properties of EC, TNF also promotes release of PDGF, which may serve to modulate proliferation of vascular SMC during wound healing, inflammation, and atherogenesis.

Immature dendritic cells phagocytose apoptotic cells via alphavbeta5 and CD36, and cross-present antigens to cytotoxic T lymphocytes.

Dendritic cells, but not macrophages, efficiently phagocytose apoptotic cells and cross-present viral, tumor, and self-antigens to CD8(+) T cells. This in vitro pathway corresponds to the in vivo phenomena of cross-priming and cross-tolerance. Here, we demonstrate that phagocytosis of apoptotic cells is restricted to the immature stage of dendritic cell (DC) development, and that this process is accompanied by the expression of a unique profile of receptors, in particular the alphavbeta5 integrin and CD36. Upon maturation, these receptors and, in turn, the phagocytic capacity of DCs, are downmodulated. Macrophages engulf apoptotic cells more efficiently than DCs, and although they express many receptors that mediate this uptake, they lack the alphavbeta5 integrin. Furthermore, in contrast to DCs, macrophages fail to cross-present antigenic material contained within the engulfed apoptotic cells. Thus, DCs use unique pathways for the phagocytosis, processing, and presentation of antigen derived from apoptotic cells on class I major histocompatibility complex. We suggest that the alphavbeta5 integrin plays a critical role in the trafficking of exogenous antigen by immature DCs in this cross-priming pathway.

CD36 mediates the In vitro inhibitory effects of thrombospondin-1 on endothelial cells.

Thrombospondin-1 (TSP-1) is a naturally occurring inhibitor of angiogenesis that is able to make normal endothelial cells unresponsive to a wide variety of inducers. Here we use both native TSP-1 and small antiangiogenic peptides derived from it to show that this inhibition is mediated by CD36, a transmembrane glycoprotein found on microvascular endothelial cells. Both IgG antibodies against CD36 and glutathione-S-transferase-CD36 fusion proteins that contain the TSP-1 binding site blocked the ability of intact TSP-1 and its active peptides to inhibit the migration of cultured microvascular endothelial cells. In addition, antiangiogenic TSP-1 peptides inhibited the binding of native TSP-1 to solid phase CD36 and its fusion proteins, as well as to CD36-expressing cells. Additional molecules known to bind CD36, including the IgM anti-CD36 antibody SM, oxidized (but not unoxidized) low density lipoprotein, and human collagen 1, mimicked TSP-1 by inhibiting the migration of human microvascular endothelial cells. Transfection of CD36-deficient human umbilical vein endothelial cells with a CD36 expression plasmid caused them to become sensitive to TSP-1 inhibition of their migration and tube formation. This work demonstrates that endothelial CD36, previously thought to be involved only in adhesion and scavenging activities, may be essential for the inhibition of angiogenesis by thrombospondin-1.

Thrombospondin binds to monocytes-macrophages and mediates platelet-monocyte adhesion.

Thrombospondin (TSP) is a multifunctional platelet glycoprotein synthesized by a variety of cells in culture including monocytes and macrophages. We now report that 125I-TSP binds specifically, saturably, and reversibly to mouse peritoneal macrophages and to cells of the monocyte-like human cell line U937 with dissociation constants of 6.7-14.5 X 10(-8) M and 3-4 X 10(5) binding sites per cell. TSP mediates an adhesive interaction between thrombin-stimulated platelets and both U937 cells and human blood monocytes. Using a sensitive rosetting assay, we found that monocytes were not rosetted by resting platelets whereas greater than 90% were rosetted by thrombin-stimulated platelets. Monoclonal and polyclonal anti-TSP antibodies markedly inhibited rosetting as did TSP itself. Neither control antibodies nor heparin, fibronectin, fibrinogen, nor the fibronectin adhesion tetrapeptide Arg-Gly-Asp-Ser inhibited rosetting. TSP may thus serve as a molecular bridge linking activated platelets with monocytes at sites of early vascular injury. Such interaction may be of critical importance in the regulation of thrombosis and the initiation of atherosclerosis.

Glycoprotein IV mediates thrombospondin-dependent platelet-monocyte and platelet-U937 cell adhesion.

An adhesive interaction between activated platelets and mononuclear phagocytes may contribute to the role these cells play in regulating inflammation, thrombosis, and atherosclerosis. We have previously shown that this adhesive interaction is mediated by the expression of the glycoprotein thrombospondin (TSP) on the surface of activated platelets. We now show that TSP-dependent platelet-monocyte interactions are mediated by glycoprotein IV (GPIV), an intrinsic membrane protein recently identified as a cell surface TSP receptor. Monoclonal antibodies to GPIV bound to cells of the human monocytoid line U937 as assessed by flow cytometry and inhibited the binding of 125I-TSP to the cell surface by 83%. U937 cells preincubated with anti-GPIV were not rosetted by thrombin-stimulated platelets (72% inhibition compared with control anti-monocyte antibodies). In addition, when platelets were stimulated in the presence of saturating concentrations of monoclonal antibodies to GPIV, only 18% of U937 cells were rosetted (78% inhibition). Control antibodies including anti-GPIb did not inhibit rosette formation. These data suggest that TSP can cross-link platelets and monocytes via an interaction with GPIV on the surface of both cells. This molecular bridge may mediate platelet-macrophage communication in various pathophysiologic settings.

Isolation of the thrombospondin membrane receptor.

Thrombospondin (TSP), a 450-kD multifunctional glycoprotein with a broad tissue distribution, is secreted upon platelet stimulation, binds to the activated platelet surface, and supports platelet aggregation. We have identified and isolated an 88-kd membrane glycoprotein present in platelets, endothelial cells, monocytes, and a variety of human tumor cell lines that is the membrane binding site for TSP. Endogenous platelet TSP binding to thrombin- and ionophore-stimulated human platelets was inhibited in the presence of the monoclonal antibody OKM5. TSP binding to C32 melanoma cells and HT1080 fibrosarcoma cells was specific and also inhibitable with OKM5 Mab. Cell labeling followed by specific immunoprecipitation demonstrated biosynthesis of a single 88-kD glycoprotein. Binding of TSP to the isolated membrane protein was specific and saturable. These studies identify an 88-kD membrane glycoprotein that reacts with the monoclonal antibody, OKM5, and may function as the cellular TSP receptor.

Platelet thrombospondin forms a trimolecular complex with plasminogen and histidine-rich glycoprotein.

Thrombospondin (TSP), a multifunctional alpha-granule glycoprotein of human platelets binds fibrinogen, fibronectin, heparin, histidine-rich glycoprotein (HRGP), and plasminogen (Plg), and thus, may play an important role in regulating thrombotic influences at vessel surfaces. In this study we have demonstrated that purified human platelet TSP formed a trimolecular complex with human Plg and HRGP. Complex formation was detected by a specific binding enzyme-linked immunosorbent assay (ELISA) which demonstrated simultaneous binding of fluid-phase Plg and HRGP to TSP adsorbed to microtitration wells. While neither ligand inhibited complex formation of the other with TSP, 10 mM epsilon-amino-n-caproic acid selectively blocked incorporation of Plg into the complex, suggesting that TSP contains independent binding sites for Plg and HRGP. Comparable extent of trimolecular complex formation was also detected when TSP monomer was substituted for whole TSP in the ELISA. HRGP covalently cross-linked to Sepharose 4B simultaneously bound both 125I-TSP and 131I-Plg, confirming trimolecular complex formation. Rocket immunoelectrophoresis of mixtures of the purified radiolabeled proteins into anti-Plg containing agarose also confirmed trimolecular complex formation. The TSP-HRGP-Plg complex bound a similar amount of heparin as the TSP-HRGP complex, demonstrating that the HRGP within the trimolecular complex maintained functional capability. Similarly, using a fluorometric plasmin substrate, the trimolecular complex was shown to be an effective substrate for tissue plasminogen activator. Significant amounts of plasmin were generated from the TSP-HRGP-Plg complex (equivalent to that from the TSP-Plg complex), but the rate of plasmin generation from the trimolecular complex was greater than from the bimolecular complex, suggesting an important interaction of HRGP with Plg when both are complexed to TSP. The macromolecular assembly of these three proteins on cellular surfaces, such as the platelet, may serve important regulatory functions, both prothrombotic at sites of active fibrin deposition and proteolytic in nonfibrin-containing microenvironments.

Binding of Lys-plasminogen to monocytes/macrophages.

The ability of mononuclear phagocytes to assemble and activate components of the fibrinolytic system on their surfaces may be crucial in effecting an efficient inflammatory response. Lys-plasminogen, the plasmin modified form of this zymogen, was found to bind specifically and with high affinity to murine peritoneal macrophages and to cells of the human monocytoid line U937. This modified plasminogen has been shown to be a more efficient substrate for plasminogen activators than native Glu-plasminogen. Binding was lysine binding site dependent, rapid and reversible. In contrast, although native Glu-plasminogen bound specifically to these cells, affinity was low. Lys-plasminogen inhibited the binding of Glu-plasminogen but the opposite was not true. Molecular analysis of the bound ligands indicated that Glu-plasminogen was converted to Lys-plasminogen and Lys-plasminogen to plasmin on the cell surface but not in the supernatant. Peritoneal macrophages from patients with indwelling catheters and tissue macrophages in chronic inflammatory lesions were shown to express immunologically identified Lys-plasminogen on their surfaces. Therefore binding and surface activation of kinetically favored Lys-plasminogen may provide an important physiological mechanism for localizing proteolytic activity on the surface of inflammatory cells.

Complex formation of platelet thrombospondin with plasminogen. Modulation of activation by tissue activator.

Thrombospondin (TSP), a multifunctional alpha-granule glycoprotein of platelets, binds fibrinogen, fibronectin, heparin, and histidine-rich glycoprotein and thus may play an important role in regulating thrombotic influences at vessel surfaces. In this study we have demonstrated that purified human platelet TSP formed a complex with purified human plasminogen (Plg). Complex formation was detected by rocket immunoelectrophoresis of mixtures of the purified radiolabeled proteins. Significant complex formation of fluid-phase Plg with adsorbed TSP was also demonstrated by enzyme-linked immunosorbent assay (ELISA). The complex formation was specific, saturable, and inhibited by excess fluid-phase TSP, with an apparent KD of approximately 35 nM. In both ELISA and rocket immunoelectrophoresis systems, complex formation was inhibited by 10 mM epsilon-amino-n-caproic acid, implying that there is a role for the lysine binding sites of Plg in mediating the interaction. TSP also formed a complex with plasmin as detected by ELISA but did not directly inhibit plasmin activity measured with a synthetic fluorometric substrate or with a 125I-fibrin plate assay. TSP, when incubated with Plg before addition to 125I-fibrin plates significantly inhibited the generation of plasmin activity by tissue plasminogen activator (TPA) in a manner that was calcium dependent. A kinetic study of Plg activation by TPA in the presence of TSP demonstrated that Michaelis-Menten kinetics were followed and that TSP acted as a noncompetitive inhibitor. These studies support the hypothesis that TSP, acting as a multifunctional regulator in focal areas of active hemostasis, could serve as a prothrombotic influence, leading to increased deposition of fibrin.

Histidine-rich glycoprotein inhibits the antiangiogenic effect of thrombospondin-1.

Angiogenesis is critical for the growth and proliferation of tumors as well as for normal development. We now describe a novel role for histidine-rich glycoprotein (HRGP) in the modulation of angiogenesis. HRGP is a plasma protein that circulates in relatively high concentrations (1.5 microM), but has no known function in vivo. We have shown previously that HRGP binds with high affinity to thrombospondin-1 (TSP-1), a homotrimeric glycoprotein that is a potent inhibitor of angiogenesis. The antiangiogenic activity of TSP-1 is mediated by the binding of properdin-like type I repeats to the receptor CD36. We found that binding of HRGP to TSP-1 was similarly mediated by TSP type I repeats. HRGP colocalized with TSP-1 in the stroma of human breast cancer specimens, and this interaction masked the antiangiogenic epitope of TSP-1. In assays performed in vitro of endothelial cell migration and tube formation, and in vivo corneal angiogenesis assays, HRGP inhibited the antiangiogenic effect of TSP-1. These studies suggest that HRGP can modulate the antiangiogenic activity of TSP-1, and identify a potential mechanism of resistance to the antiangiogenic effect of TSP-1.

Targeted disruption of the class B scavenger receptor CD36 protects against atherosclerotic lesion development in mice.

Macrophage scavenger receptors have been implicated as key players in the pathogenesis of atherosclerosis. To assess the role of the class B scavenger receptor CD36 in atherogenesis, we crossed a CD36-null strain with the atherogenic apo E-null strain and quantified lesion development. There was a 76.5% decrease in aortic tree lesion area (Western diet) and a 45% decrease in aortic sinus lesion area (normal chow) in the CD36-apo E double-null mice when compared with controls, despite alterations in lipoprotein profiles that often correlate with increased atherogenicity. Macrophages derived from CD36-apo E double-null mice bound and internalized more than 60% less copper-oxidized LDL and LDL modified by monocyte-generated reactive nitrogen species. A similar inhibition of in vitro lipid accumulation and foam cell formation after exposure to these ligands was seen. These results support a major role for CD36 in atherosclerotic lesion development in vivo and suggest that blockade of CD36 can be protective even in more extreme proatherogenic circumstances.

Macrophage scavenger receptor CD36 is the major receptor for LDL modified by monocyte-generated reactive nitrogen species.

The oxidative conversion of LDL into an atherogenic form is considered a pivotal event in the development of cardiovascular disease. Recent studies have identified reactive nitrogen species generated by monocytes by way of the myeloperoxidase-hydrogen peroxide-nitrite (MPO-H(2)O(2)-NO(2)(-)) system as a novel mechanism for converting LDL into a high-uptake form (NO(2)-LDL) for macrophages. We now identify the scavenger receptor CD36 as the major receptor responsible for high-affinity and saturable cellular recognition of NO(2)-LDL by murine and human macrophages. Using cells stably transfected with CD36, CD36-specific blocking mAbs, and CD36-null macrophages, we demonstrated CD36-dependent binding, cholesterol loading, and macrophage foam cell formation after exposure to NO(2)-LDL. Modification of LDL by the MPO-H(2)O(2)-NO(2)(-) system in the presence of up to 80% lipoprotein-deficient serum (LPDS) still resulted in the conversion of the lipoprotein into a high-uptake form for macrophages, whereas addition of less than 5% LPDS totally blocked Cu(2+)-catalyzed LDL oxidation and conversion into a ligand for CD36. Competition studies demonstrated that lipid oxidation products derived from 1-palmitoyl-2-arachidonyl-sn-glycero-3-phosphocholine can serve as essential moieties on NO(2)-LDL recognized by CD36. Collectively, these results suggest that MPO-dependent conversion of LDL into a ligand for CD36 is a likely pathway for generating foam cells in vivo. MPO secreted from activated phagocytes may also tag phospholipid-containing targets for removal by CD36-positive cells.

Activation of cultured vascular endothelial cells by antiphospholipid antibodies.

Circulating antiphospholipid antibodies (aPL) are associated with a syndrome of thrombosis, recurrent fetal loss, and thrombocytopenia. We have demonstrated the activation of cultured human umbilical vein endothelial cells (HUVEC) by IgG from patients with anticardiolipin antibodies (aCL). Incubation of HUVEC for 4 h with purified IgG (100 micrograms/ml) from patients with high-titer aCL induced a 2.3-fold increase in monocyte adhesion over that seen in HUVEC incubated with IgG's from normal subjects. The effect of aCL was not attributable to LPS contamination, Fc receptors, or immune complexes. Monocyte adhesion was not induced when the aCL were added in serum-free media but was restored by the addition of purified beta 2GP1, previously described as a necessary cofactor for aCL reactivity. Purified rabbit polyclonal IgG raised against beta 2GP1 also induced monocyte adhesion when incubated with HUVEC. Preadsorption of patient serum with cardiolipin reduced monocyte adhesion by 60%. Immunofluorescent microscopy demonstrated that endothelial cells incubated with patient IgG expressed cell adhesion molecules, including E-selectin, vascular cell adhesion molecule-1, and intracellular adhesion molecule-1. These data support the hypothesis that aPL activate vascular endothelial cells, thereby leading to a pro-thrombotic state.

CD36 mediates the cardiovascular action of growth hormone-releasing peptides in the heart.

Growth hormone-releasing peptides (GHRPs) are known as potent growth hormone secretagogues whose actions are mediated by the ghrelin receptor, a G protein-coupled receptor cloned from pituitary libraries. Hexarelin, a hexapeptide of the GHRP family, has reported cardiovascular activity. To identify the molecular target mediating this activity, rat cardiac membranes were labeled with a radioactive photoactivatable derivative of hexarelin and purified using lectin affinity chromatography and preparative gel electrophoresis. A binding protein of M(r) 84 000 was identified. The N-terminal sequence determination of the deglycosylated protein was identical to rat CD36, a multifunctional glycoprotein, which was expressed in cardiomyocytes and microvascular endothelial cells. Activation of CD36 in perfused hearts by hexarelin was shown to elicit an increase in coronary perfusion pressure in a dose-dependent manner. This effect was lacking in hearts from CD36-null mice and hearts from spontaneous hypertensive rats genetically deficient in CD36. The coronary vasoconstrictive response correlated with expression of CD36 as assessed by immunoblotting and covalent binding with hexarelin. These data suggest that CD36 may mediate the coronary vasospasm seen in hypercholesterolemia and atherosclerosis.

The antiangiogenic effect of thrombospondin-2 is mediated by CD36 and modulated by histidine-rich glycoprotein.

Thrombospondins-1 and -2 (TSP-1, TSP-2) are matricellular glycoproteins with potent antiangiogenic activity. We have previously shown that the antiangiogenic activity of TSP-1 is mediated by the interaction of the type I repeats (TSR) with the receptor CD36, although other domains of TSP-1 have also been implicated. We now show that the antiangiogenic activity of TSP-2, which contains three TSRs but, unlike TSP-1, lacks the capacity to activate TGF-beta, is similarly dependent on CD36. Using the corneal pocket assay we found that TSP-2 did not inhibit bFGF-induced angiogenesis in CD36 null mice. We then demonstrated that (125)[I]-TSP-2 bound to murine macrophages and that binding was diminished by 70% by anti-CD36 antibody or by using cells from CD36 null animals. Solid-phase binding studies revealed that (125)[I]-TSP-2 bound to CD36/glutathione-S-transferase (GST) fusion proteins encoding the region spanning amino acids 93-120, but not amino acids 298-439. This 93-120 amino acid region, previously identified as the TSP-1 binding site, is homologous to domains on other TSP binding proteins, such as LIMP-2 and histidine-rich glycoprotein (HRGP). Finally, we showed with an immunoabsorbent binding assay that TSP-2 bound HRGP with high affinity and that HRGP blocked the antiangiogenic activity of TSP-2, acting like a "decoy" receptor. These data suggest that modulation of the TSR/CD36 system may play an important role in the regulation of the angiogenic "switch," and may provide a target for therapeutic interventions.