Identification of cell adhesive active sites in the N-terminal domain of thrombospondin-1

The carboxyl-terminal TSP1-homology domain is the biologically active effector peptide of matricellular protein CCN5 that counteracts profibrotic CCN2

Cellular Communication Network (CCN) proteins have multimodular structures important for their roles in cellular responses associated with organ development and tissue homeostasis. CCN2 has previously been reported to be secreted as a preproprotein that requires proteolytic activation to release its bioactive carboxyl-terminal fragment. Here, our goal was to resolve whether CCN5, a divergent member of the CCN family with converse functions relative to CCN2, releases the TSP1 homology domain as its bioactive signaling entity. The recombinant CCN5 or CCN3 TSP1 homology domains were produced in ExpiCHO-S or DG44 CHO cells as secretory fusion proteins appended to the carboxyl-terminal end of His-Halo-Sumo or amino-terminal end of human albumin and purified from the cell culture medium. We tested these fusion proteins in various phosphokinase signaling pathways or cell physiologic assays. Fusion proteins with the CCN5 TSP1 domain inhibited key signaling pathways previously reported to be stimulated by CCN2, irrespective of fusion partner. The fusion proteins also efficiently inhibited CCN1/2-stimulated cell migration and gap closure following scratch wound of fibroblasts. Fusion protein with the CCN3 TSP1 domain inhibited these functions with similar efficacy and potency as that of the CCN5 TSP1 domain. The CCN5 TSP1 domain also recapitulated a positive regulatory function previously assigned to full-length CCN5, that is, induction of estrogen receptor-α mRNA expression in triple negative MDA-MB-231 mammary adenocarcinoma cells and inhibited epithelial-to-mesenchymal transition and CCN2-induced mammosphere formation of MCF-7 adenocarcinoma cells. In conclusion, the CCN5 TSP1 domain is the bioactive entity that confers the biologic functions of unprocessed CCN5.

Thbs1 induces lethal cardiac atrophy through PERK-ATF4 regulated autophagy

The thrombospondin (Thbs) family of secreted matricellular proteins are stress- and injury-induced mediators of cellular attachment dynamics and extracellular matrix protein production. Here we show that Thbs1, but not Thbs2, Thbs3 or Thbs4, induces lethal cardiac atrophy when overexpressed. Mechanistically, Thbs1 binds and activates the endoplasmic reticulum stress effector PERK, inducing its downstream transcription factor ATF4 and causing lethal autophagy-mediated cardiac atrophy. Antithetically, Thbs1-/- mice develop greater cardiac hypertrophy with pressure overload stimulation and show reduced fasting-induced atrophy. Deletion of Thbs1 effectors/receptors, including ATF6α, CD36 or CD47 does not diminish Thbs1-dependent cardiac atrophy. However, deletion of the gene encoding PERK in Thbs1 transgenic mice blunts the induction of ATF4 and autophagy, and largely corrects the lethal cardiac atrophy. Finally, overexpression of PERK or ATF4 using AAV9 gene-transfer similarly promotes cardiac atrophy and lethality. Hence, we identified Thbs1-mediated PERK-eIF2α-ATF4-induced autophagy as a critical regulator of cardiomyocyte size in the stressed heart.

Endothelial Cell Behavior Is Determined by Receptor Clustering Induced by Thrombospondin-1

The thrombospondins (TSPs) are a family of multimeric extracellular matrix proteins that dynamically regulate cellular behavior and response to stimuli. In so doing, the TSPs directly and indirectly affect biological processes such as embryonic development, wound healing, immune response, angiogenesis, and cancer progression. Many of the direct effects of Thrombospondin 1 (TSP-1) result from the engagement of a wide range of cell surface receptors including syndecans, low density lipoprotein receptor-related protein 1 (LRP1), CD36, integrins, and CD47. Different or even opposing outcomes of TSP-1 actions in certain pathologic contexts may occur, depending on the structural/functional domain involved. To expedite response to external stimuli, these receptors, along with vascular endothelial growth factor receptor 2 (VEGFR2) and Src family kinases, are present in specific membrane microdomains, such as lipid rafts or tetraspanin-enriched microdomains. The molecular organization of these membrane microdomains and their constituents is modulated by TSP-1. In this review, we will describe how the presence of TSP-1 at the plasma membrane affects endothelial cell signal transduction and angiogenesis.

The thrombospondin module 1 domain of the matricellular protein CCN3 shows an atypical disulfide pattern and incomplete CWR layers

The members of the CCN (Cyr61/CTGF/Nov) family are a group of matricellular regulatory proteins that are essential to a wide range of functional pathways in cell signalling. Through interacting with extracellular matrix components and growth factors via one of their four domains, the CCN proteins are involved in critical biological processes such as angiogenesis, cell proliferation, bone development, fibrogenesis and tumorigenesis. Here, the crystal structure of the thrombospondin module 1 (TSP1) domain of CCN3 (previously known as Nov) is presented, which shares a similar three-stranded fold with the thrombospondin type 1 repeats of thrombospondin-1 and spondin-1, but with variations in the disulfide connectivity. Moreover, the CCN3 TSP1 domain lacks the typical π-stacked ladder of charged and aromatic residues on one side of the domain that is seen in other TSP1 domains. Using conservation analysis among orthologous domains, it is shown that a charged cluster in the centre of the domain is the most conserved site and this cluster is predicted to be a potential functional epitope for heparan sulfate binding. This variant TSP1 domain has also been used to revise the sequence determinants of TSP1 domains and to derive improved Pfam sequence profiles for the identification of novel TSP1 domains in more than 10 000 proteins across diverse phyla.

Phosphatidylserine-positive erythrocytes bind to immobilized and soluble thrombospondin-1 via its heparin-binding domain

Phosphatidylserine (PS)-dependent erythrocyte adhesion to endothelium and subendothelial matrix components is mediated in part via thrombospondin (TSP). Although TSP exhibits multiple cell-binding domains, the PS-binding site on TSP is unknown. Because a cell-binding domain for anionic heparin is located at the amino-terminus, we hypothesized that PS-positive red blood cells (PS(+ve)-RBCs) bind to this domain. We demonstrate that both heparin and its low-molecular-weight derivative enoxaparin (0.5-50 u/mL) inhibited PS(+ve)-RBC adhesion to immobilized TSP in a concentration-dependent manner (21% to 77% inhibition, P < 0.05). Preincubation of immobilized TSP with an antibody against the heparin-binding domain blocked PS(+ve)-RBC adhesion to TSP. Antibodies that recognize the collagen- and the carboxy-terminal CD47-binding domain on TSP had no effect on this process. Although preincubation of PS(+ve)-RBCs with TSP peptides from the heparin-binding domain that contained the specific heparin-binding motif KKTRG inhibited PS(+ve)-erythrocyte adhesion to matrix TSP (P < 0.001), these peptides in the immobilized form supported PS-mediated erythrocyte adhesion. A TSP-peptide that lacks the binding motif neither inhibited nor supported PS(+ve)-RBC adhesion. Additional experiments show that soluble TSP also interacted with PS(+ve)-RBCs via its heparin-binding domain. Our results demonstrate that PS-positive erythrocytes bind to both immobilized and soluble TSP via its heparin-binding domain and that both heparin and enoxaparin, at clinically relevant concentrations, block this interaction. Other studies have shown that heparin inhibited P-selectin- and soluble-TSP-mediated sickle erythrocyte adhesion to endothelial cells. Our results, taken together with the previously documented findings, provide a rational basis for clinical use of heparin or its low-molecular-weight derivatives as therapeutic agents in treating vaso-occlusive pain in patients with sickle cell disease.

Purification and analysis of thrombospondin-1

Thromboapondin 1 (TSP-1) is a trimeric matricellular protein that is expressed by many cells. It contains several different domains that allow it to participate in cell adhesion, cell migration, and cell signaling. Recently TSP-1 has been shown to activate transforming growth factor beta (TGF-beta) and to inhibit both angiogenesis and tumor growth. This unit contains protocols for the purification of TSP-1 from platelet-rich plasma and the purification of TSP-1 proteolytic fragments.

Thrombospondin-1 (TSP-1) and Survivin (S) expression in non-Hogkin's lymphomas

Survivin (S) is a member of inhibitor of apoptosis family (IAP) and is expressed in the majority of malignant tumors but undetectable in normal differentiated adult tissues. S is an encouraging target for cancer therapy. TSP-1 is a multifunctional protein regulating cell growth, motility and apoptosis in both physiological and pathological conditions. The role of TSP-1 in cancer progression remains controversial. We aimed to determine the pathogenetic and prognostic role of TSP-1 and S in non-Hodgkin's lymphomas (NHL). S and TSP-1 expressions were looked for in 177 cases with NHL. S was found to be positive in 94 of the cases (53%). TSP-1 was found to be positive in 31 of the cases (17.5%). There was a strong association between S and TSP-1 and also aggressive histology with S and TSP-1. The overall survival (OS) times were longer in cases without S expression than cases with S expression (p=0.0514). Although the OS was shorter in TSP-1 expressing cases as compared with TSP-1 (-) cases, difference was not significant (p=0.2428). In conclusion, S and TSP-1 expressions were detected in 53 and 17.5% of the cases with NHL, and are associated with aggressive histology and shorter OS.

Syndecan-4 contributes to endothelial tubulogenesis through interactions with two motifs inside the pro-angiogenic N-terminal domain of thrombospondin-1

Thrombospondin-1 (TSP-1) is an extracellular matrix protein that modulates focal adhesion in mammalian cells and exhibits dual roles in angiogenesis. In a previous work, we showed that a recombinant 18 kDa protein encompassing the N-terminal residues 1-174 of human TSP-1 (TSP18) induced tubulogenesis of human umbilical vein endothelial cells and protected them from apoptosis. Our results indicated that these effects were possibly mediated by syndecan-4 proteoglycan, since binding of TSP18 to endothelial extracts was inhibited by anti-syndecan-4 antibody. Syndecan-4 is a heparan-sulfate proteoglycan that regulates cell-matrix interactions and is the only member of its family present in focal adhesions. In this report, we demonstrate that a monoclonal antibody against syndecan-4 blocks TSP18-induced tubulogenesis. Furthermore, through 2D adhesion and 3D angiogenic assays, we demonstrate that two sequences, TSP Hep I and II, retain the major pro-angiogenic activity of TSP18. These TSP-1 motifs also compete with the fibronectin Hep II domain for binding to syndecan-4 on endothelial cell surface, indicating that they may exert their effects by interfering with the recognition of fibronectin by syndecan-4. Additionally, TSP18 and its derived peptides activate the PKC-dependent Akt-PKB signaling pathway. Blockage of PKC activation prevented HUVEC spreading when seeded on TSP18 fragment, and on TSP Hep I and TSP Hep II peptides, but not on gelatin-coated substrates. Our results identify syndecan-4 as a novel receptor for the N-terminus of TSP-1 and suggest that TSP-1 N-terminal pro-angiogenic activity is linked to its capacity of interfering with syndecan-4 functions in the course of cell adhesion.

Proteomic analysis of amniotic membrane prepared for human transplantation: characterization of proteins and clinical implications

Amniotic membrane is commonly exploited in several surgical procedures. Despite a freeze preservation period, it is reported to retain wound healing, anti-angiogenic, antiinflammatory and anti-scarring properties; however, little is known about the active protein content. 2-DE analysis of transplant-ready amniotic membrane (TRAM) was performed. The effects of preservation and processing on amnion proteome were investigated, and the major proteins in the TRAM characterized using mass spectrometry and immunoblotting. This identified a spectrum of proteins including thrombospondin, mimecan, BIG-H3, and integrin alpha 6. Preservation compromises cellular viability resulting in selective elution of soluble cellular proteins, leaving behind extracellular matrix-associated and cell structural proteins. A number of key architectural proteins common to the architecture of the ocular surface were demonstrated in AM, which are involved in homeostasis and wound healing. Handling procedures alter the protein composition of amniotic membrane prepared for transplantation. Without standardization, there will be inter-membrane variation, which may compromise the desired therapeutic effect of transplant ready amniotic membrane.

Nell-1-induced bone regeneration in calvarial defects

Many craniofacial birth defects contain skeletal components requiring bone grafting. We previously identified the novel secreted osteogenic molecule NELL-1, first noted to be overexpressed during premature bone formation in calvarial sutures of craniosynostosis patients. Nell-1 overexpression significantly increases differentiation and mineralization selectively in osteoblasts, while newborn Nell-1 transgenic mice significantly increase premature bone formation in calvarial sutures. In the current study, cultured calvarial explants isolated from Nell-1 transgenic newborn mice (with mild sagittal synostosis) demonstrated continuous bone growth and overlapping sagittal sutures. Further investigation into gene expression cascades revealed that fibroblast growth factor-2 and transforming growth factor-beta1 stimulated Nell-1 expression, whereas bone morphogenetic protein (BMP)-2 had no direct effect. Additionally, Nell-1-induced osteogenesis in MC3T3-E1 osteoblasts through reduction in the expression of early up-regulated osteogenic regulators (OSX and ALP) but induction of later markers (OPN and OCN). Grafting Nell-1 protein-coated PLGA scaffolds into rat calvarial defects revealed the osteogenic potential of Nell-1 to induce bone regeneration equivalent to BMP-2, whereas immunohistochemistry indicated that Nell-1 reduced osterix-producing cells and increased bone sialoprotein, osteocalcin, and BMP-7 expression. Insights into Nell-1-regulated osteogenesis coupled with its ability to stimulate bone regeneration revealed a potential therapeutic role and an alternative to the currently accepted techniques for bone regeneration.

Internalization but not binding of thrombospondin-1 to low density lipoprotein receptor-related protein-1 requires heparan sulfate proteoglycans

The amino-terminal domain of the extracellular matrix (ECM) protein thrombospondin-1 (TSP-1) mediates binding to cell surface heparan sulfate proteoglycans (HSPG) as well as binding to the endocytic receptor, low density lipoprotein-related protein (LRP-1). We previously found that recombinant TSP-1 containing the amino-terminal residues 1-214, retained both of these interactions (Mikhailenko et al. [1997]: J Biol Chem 272:6784-6791). Here, we examined the activity of a recombinant protein containing amino-terminal residues 1-90 of TSP-1 and found that this domain did not retain high-affinity heparin-binding. The loss of heparin-binding correlated with decreased binding to the fibroblast cell surface. However, both ligand blotting and solid phase binding studies indicate that this truncated fragment of TSP-1 retained high-affinity binding to LRP-1. Consistent with this, it also retained the ability to block the uptake and degradation of (125)I-TSP-1. However, TSP-1(1-90) itself was poorly endocytosed and this truncated amino-terminal domain was considerably more effective than the full-length heparin-binding domain (HBD) of TSP-1 in blocking the catabolism of endogenously expressed TSP-1. These results indicate that TSP-1 binding to LRP-1 does not require prior or concomitant interaction with cell surface HSPG but suggest subsequent endocytosis requires high-affinity heparin-binding.

The N-terminus of thrombospondin: the domain stands apart

Thrombospondin 1 (TSP1) was first recognized as a thrombin-sensitive protein associated with platelet membranes. It is secreted by numerous cell types and its expression is predominant in areas of active tissue remodeling. Thrombospondins 1 and 2 are large, trimeric, matricellular proteins, composed of multiple structural motifs which interact with a diverse array of receptors and molecules. Thrombospondin's capacity to bind multiple receptors renders it multifunctional. The functions of its isolated domains can be overlapping or contradictory. In this review, we focus on the N-terminus of the molecule, first recognized for its strong heparin binding properties and characterized by its susceptibility to proteolytic cleavage from the stalk region of thrombospondin. The N-terminus, called the heparin binding domain (HBD), interacts with a variety of macromolecules including heparan sulfate proteoglycans at the membrane and in the matrix, LDL receptor-related protein (LRP), sulfated glycolipids, calreticulin, and integrins. The HBD mediates endocytosis of thrombospondin. It functions both as a soluble and an insoluble modulator of cell adhesion and motility. In contrast to thrombospondin, the HBD has pro-angiogenic activity. We propose that the HBD of thrombospondins 1 and 2 are found primarily in the cellular microenvironment in conditions of cellular injury, stress and tissue remodeling and that the HBD conveys multiple signals involved in cellular adaptation to injury.

Expression of the type-1 repeats of thrombospondin-1 inhibits tumor growth through activation of transforming growth factor-beta

In the present study, the type-1 repeats of thrombospondin-1 (TSP-1) were transfected into A431 cells. Expression of all three type-1 repeats (3TSR) and expression of just the second type-1 repeat containing the transforming growth factor (TGF)-beta activating sequence KRFK (TSR2 + KRFK) significantly inhibited in vivo tumor angiogenesis and growth in nude mice. These tumors expressed increased levels of both active and total TGF-beta. A431 cells expressing the second type-1 repeat without the KRFK sequence (TSR2 - KRFK) produced tumors that were slightly larger than the 3TSR and TSR2 + KRFK tumors. These tumors expressed elevated levels of active TGF-beta but levels of total TGF-beta were not different from control tumors. Injection of the peptide, LSKL, which blocks TSP-1 activation of TGF-beta, reversed the growth inhibition observed with cells expressing TSR2 + KRFK to a level comparable to controls. Various residues in the WSHWSPW region and the VTCG sequence of both TSR2+/- KRFK were mutated. Although mutation of the VTCG sequence had no significant effect on tumor growth, mutation of the WSHWSPW sequence reduced inhibition of tumor growth. These findings suggest that the inhibition of tumor angiogenesis and growth by endogenous TSP-1 involves regulation of both active and total TGF-beta and the sequences KRFK and WSHWSPW in the second type-1 repeat.

Integrin alpha(v)beta3 expression confers on tumor cells a greater propensity to metastasize to bone

The reasons why tumor cells metastasize to bone remain obscure. There is some evidence to support the theory that integrins (acting as cell surface adhesion receptors) play a role in mediating metastasis in certain organs. Here, we report that overexpression of a functionally active integrin alpha(v)b3 in Chinese hamster ovary (CHO) tumor cells drastically increased the incidence, number, and area of bone metastases in nude mice compared with those observed in mock-transfected CHO cells (CHO dhfr+) or in CHO cells expressing a functionally inactive integrin alpha(v)b3 (CHO beta3Delta744). Moreover, a breast cancer cell line (B02) established from bone metastases caused by MDA-MB-231 cells constitutively overexpressed integrin alpha(v)b3, whereas the cell surface expression level of other integrins remained unchanged. In vivo, the extent of bone metastases in B02-bearing mice was significantly increased compared with that of MDA-MB-231-bearing mice. In vitro, B02 cells and CHO cells expressing a functionally active integrin alpha(v)b3 exhibited substantially increased invasion of and adhesion to mineralized bone, bone sialoprotein, and collagen compared with those found with MDA-MB-231, CHO dhfr+, and CHO beta3Delta744 cells, respectively. Overall, our findings suggest that integrin alpha(v)b3 expression in tumor cells accelerates the development of osteolytic lesions, presumably through increased invasion of and adhesion to bone.

A recombinant NH(2)-terminal heparin-binding domain of the adhesive glycoprotein, thrombospondin-1, promotes endothelial tube formation and cell survival: a possible role for syndecan-4 proteoglycan

Thrombospondin-1 (TSP-1) is a multifunctional protein known to modulate angiogenesis, endothelial cell adhesion and apoptosis. In this study, we have demonstrated that TSP18, a recombinant 18 kDa protein encompassing the N-terminal residues 1-174 of human TSP-1, accelerated the process of tube-like structures formation by human umbilical vein endothelial cells (HUVECs) when included in fibrin matrices at 0.55-2.2 microM concentrations, for times ranging from 24 to 72 h. This effect was specifically inhibited by V58A4, a Mab raised against TSP18. Whole TSP-1 showed a dual effect, weakly enhancing tube formation at 22 nM (10 microg/ml), but causing inhibition at 45 and 90 nM (20 and 40 microg/ml, respectively). In order to investigate the possible effects of TSP18 on cell adhesion and viability, we performed adhesion assays on different protein supports. HUVECs adhered more weakly on TSP-1-coated surfaces, remaining round-shaped, as compared to the well-spread phenotype displayed on fibronectin and gelatin. Cells adhering on TSP18-coated surfaces displayed a well spread phenotype, with this adhesion strongly inhibited by heparin. The binding of TSP18 to endothelial membrane extracts was blocked by a monoclonal IgG directed against the cell surface proteoglycan syndecan-4. The DNA fragmentation patterns and the nuclear morphology were comparable for HUVECs adhering on all proteins, including TSP18, showing minimal cell apoptosis. Our results indicate that the N-terminal region of TSP-1 constitutes a suitable adhesive support for HUVECs, protecting them from apoptosis, possibly mediated by syndecan-4 proteoglycan.

The C-terminal domain of thrombospondin-1 induces vascular smooth muscle cell chemotaxis

OBJECTIVE

Thrombospondin-1 (TSP-1), an acute-phase reactant implicated in vascular disease, is a 420-kd multifunctional glycoprotein chemotactic for vascular smooth muscle cells (VSMCs). TSP-1 has six domains of repeating homologous amino acid sequences: N-terminal, procollagen homology, type 1 repeat, type 2 repeat, type 3 repeat/RGD (T3), and C-terminal (COOH). The purpose of this experiment was to determine which domains of TSP-1 induce VSMC chemotaxis.

METHODS

A modified Boyden Chamber chemotaxis assay was used to assess VSMC migration. Serum-free medium, TSP-1, or each of the fusion proteins (10 and 20 microg/mL) synthesized for the different domains were placed in the bottom wells. Quiescent bovine aortic VSMCs (50,000) were placed in the top wells. After 4 hours at 37 degrees C, migrated VSMCs were recorded as cells per five fields (400x) and analyzed with the paired t test. To verify the fusion protein data, we performed chemotaxis assays with antibodies to each of the domains (25 microg/mL) combined with TSP-1 (20 microg/mL) in the bottom wells and VSMCs in the top wells.

RESULTS

The COOH domain significantly stimulated VSMC chemotaxis (P = <.001). To a lesser extent, the N-terminal and T3 domains also induced chemotaxis (P <.05). However, only the anti-COOH antibody (C6.7) and the anti-integrin-associated protein portion of COOH antibody (D4.6) significantly inhibited TSP-1-induced VSMC chemotaxis (by 85% and 92%, respectively).

CONCLUSIONS

These results implicate the COOH domain as the portion of the TSP-1 molecule primarily responsible for VSMC chemotaxis. This experiment suggests that future strategies in the prevention of VSMC migration, an initial step in the development of vascular lesions, may involve selective inhibition of the COOH domain of TSP-1.

Platelet-osteosarcoma cell interaction is mediated through a specific fibrinogen-binding sequence located within the N-terminal domain of thrombospondin 1

Approximately 20% of patients with osteosarcoma have metastatic disease in lungs or bones at diagnosis. The requirement of platelets in hematogenous dissemination of metastatic cells is now well established. Tumor cells interact with platelets and induce platelet aggregation. In this respect, metastatic potential of tumor cells correlates with their capacity to aggregate platelets in vitro. We have previously shown that thrombospondin 1 (TSP-1) is synthesized and expressed on the surface of MG-63 osteosarcoma cells and mediates platelet-osteosarcoma cell interaction. However, active sites mimicking the function of TSP-1 during platelet-osteosarcoma cell interaction are not known. In this study, a panel of antibodies directed against the N-terminal and C-terminal domains and type 1, type 2, and type 3 repeats of TSP-1 were first used to delineate the structural requirement for the binding of osteosarcoma cell surface-associated TSP-1 to platelets. A drastic inhibition of the platelet-aggregating activity of MG-63 cells was obtained in the presence of a monoclonal antibody directed against the N-terminal domain of TSP-1. Among a series of 16 synthetic peptides spanning the whole N-terminal domain of TSP-1, only synthetic peptide N12/I encompassing amino acid residues 151-164 of the N-terminal domain of TSP-1 inhibited the platelet-aggregating activity of MG-63 cells. Electron microscopy studies showed that peptide N12/I strongly inhibited platelet-osteosarcoma cell interaction. A polyclonal antibody directed against peptide N12/I specifically bound to the surface of MG-63 cells, recognized TSP-1 and drastically inhibited the platelet-aggregating activity of MG-63 cells. In addition, peptide N12/I specifically bound to fibrinogen and inhibited TSP-1/fibrinogen interaction. Overall, our results provide evidence that a fibrinogen-binding sequence located within the N-terminal domain of TSP-1 mediates the binding of osteosarcoma cell surface-associated TSP-1 to platelet-bound fibrinogen.

Identification of an alpha(3)beta(1) integrin recognition sequence in thrombospondin-1

A synthetic peptide containing amino acid residues 190-201 of thrombospondin-1 (TSP1) promoted adhesion of MDA-MB-435 breast carcinoma cells when immobilized and inhibited adhesion of the same cells to TSP1 when added in solution. Adhesion to this peptide was enhanced by a beta(1) integrin-activating antibody, Mn(2+), and insulin-like growth factor I and was inhibited by an alpha(3)beta(1) integrin function-blocking antibody. The soluble peptide inhibited adhesion of cells to the immobilized TSP1 peptide or spreading on intact TSP1 but at the same concentrations did not inhibit attachment or spreading on type IV collagen or fibronectin. Substitution of several residues in the TSP1 peptide with Ala residues abolished or diminished the inhibitory activity of the peptide in solution, but only substitution of Arg-198 completely inactivated the adhesive activity of the immobilized peptide. The essential residues for activity of the peptide as a soluble inhibitor are Asn-196, Val-197, and Arg-198, but flanking residues enhance the inhibitory activity of this core sequence, either by altering the conformation of the active sequence or by interacting with the integrin. This functional sequence is conserved in all known mammalian TSP1 sequences and in TSP1 from Xenopus laevis. The TSP1 peptide also inhibited adhesion of MDA-MB-435 cells to the laminin-1 peptide GD6, which contains a potential integrin-recognition sequence Asn-Leu-Arg and is derived from a similar position in a pentraxin module. Adhesion studies using recombinant TSP1 fragments also localized beta1 integrin-dependent adhesion to residues 175-242 of this region, which contain the active sequence.

Cooperation between thrombospondin-1 type 1 repeat peptides and alpha(v)beta(3) integrin ligands to promote melanoma cell spreading and focal adhesion kinase phosphorylation

CD47-binding sequences from the carboxyl-terminal domain of thrombospondin-1 (TSP1) are known to regulate activity of the alpha(v)beta(3) integrin (Gao, G., Lindberg, F. P., Dimitry, J. M., Brown, E. J., and Frazier, W. A. (1996) J. Cell Biol. 135, 533-544). Here we show that peptides from the type 1 repeats of TSP1 also stimulate alpha(v)beta(3) integrin function in melanoma cells. Addition of soluble peptide 246 (KRFKQDGGWSHWSPWSS) enhances spreading of A2058 melanoma cells on several alpha(v)beta(3) integrin ligands, including vitronectin, recombinant TSP1 fragments containing the Arg-Gly-Asp sequence, and native TSP1. This activity requires the Trp residues and is independent of CD36-binding sequences in the type 1 repeats. Recombinant type 1 repeats expressed as a glutathione S-transferase fusion protein also enhance spreading on vitronectin and TSP1. Activation of alpha(v)beta(3) integrin by the soluble peptide 246 stimulates organization of F-actin and increases tyrosine phosphorylation of focal adhesion kinase. In contrast, direct adhesion of melanoma cells on immobilized peptide 246 inhibits tyrosine phosphorylation of focal adhesion kinase. Stimulation of alpha(v)beta(3) integrin function by the type 1 repeat peptide differs from that induced by CD47-binding TSP1 peptides in that heparan sulfate proteoglycans are required and pertussis toxin does not inhibit the former activity. Thus, the type 1 repeats contain a second sequence of TSP1 that can enhance alpha(v)beta(3) integrin signaling, and these two sequences stimulate recognition of both vitronectin and TSP1 by the alpha(v)beta(3) integrin.

Mechanism of interaction of thrombospondin with human endothelium and inhibition of sickle erythrocyte adhesion to human endothelial cells by heparin

Thrombospondin (TSP) mediates sickle erythrocyte adhesion to endothelium, but the mechanism remains unknown. Since TSP is comprised of heterogeneously distinct domains, this adhesion may depend on the interaction of specific regions of TSP with different cell surface receptors. To examine the mechanisms of interaction of TSP with human umbilical vein endothelial cells (HUVEC), we performed binding studies using soluble [125I]TSP. Our data showed that (i) monoclonal antibodies (MoAbs) against cell surface heparan sulfate (HS) or the heparin-binding domain of TSP, or cleavage of HS on HUVEC by heparitinase reduced TSP binding by 28-40%, (ii) the RGD peptide or MoAbs against integrin alpha v beta 3 or the calcium binding region of TSP inhibited binding by 18-28%, and (iii) a MoAb against the cell-binding domain of TSP inhibited binding by 36%. Unmodified heparin inhibited the binding of TSP to endothelial cells by 70% and did so far more effectively than selectively desulfated heparins, HS or chondroitin sulfate. Heparin inhibited TSP binding to HUVEC at much lower concentrations than were required to inhibit TSP binding to sickle erythrocytes. Unmodified heparin effectively inhibited the TSP-mediated adhesion of sickle erythrocytes to HUVEC. These data imply that cell surface HS-mediated mechanisms play a key role in TSP-mediated sickle erythrocyte adhesion to endothelium, and heparin may be of use for inhibition of this adhesion.

Interaction of recombinant procollagen and properdin modules of thrombospondin-1 with heparin and fibrinogen/fibrin

Many properties have been assigned to the procollagen and properdin (Type I) modules of thrombospondin-1 (TSP1) based on activities of large proteolytic fragments of TSP1 or peptides containing TSP1-derived sequences. To examine the activities of the modules more exactly, we expressed the first properdin module (P1); the third properdin module (P3); the first and second properdin modules (P12); the first, second, and third properdin modules (P123); and the procollagen module with the first, second, and third properdin modules (CP123) in the GELEX expression vector (GE1) using the baculovirus system. GE1 encodes the pre-pro sequence, the transglutaminase cross-linking site(s), the protease-sensitive site, and the gelatin binding domain from the amino terminus of rat fibronectin. All five recombinant proteins were expressed by insect cells, secreted into the culture medium, and purified by gelatin-agarose affinity chromatography. P123 shared with TSP1 a resistance to trypsin unless reduced and alkylated. P12/GE1, P123/GE1, and CP123/GE1 bound poorly to heparin-agarose except in the absence of sodium chloride, whereas peptides based on P2 are known to bind to heparin in up to 150 mM sodium chloride. In cross-linking experiments employing activated recombinant factor XIII and the transglutaminase cross-linking site in the fibronectin-derived sequence, P12/GE1, P123/GE1, CP123/GE1, and P3/GE1 but not P1/GE1 became incorporated into a fibrin clot more than GE1 alone. Analysis of the complex indicated that cross-linking was to the portion of the fibrin alpha-chain remaining in the D-dimer of plasmin digests. P123 also cross-linked to the Aalpha-chain of unclotted fibrinogen. P123 competed for 125I-TSP1 incorporation into the fibrin clot. P123 did not cross-link to plasminogen, histidine-rich glycoprotein, fibronectin, or plasma globulins other than fibrinogen/fibrin. These results indicate that the properdin modules of TSP1 specifically interact with fibrinogen/fibrin but not with heparin under physiologic conditions.

CD36 mediates binding of soluble thrombospondin-1 but not cell adhesion and haptotaxis on immobilized thrombospondin-1

In this study, we examined the binding of soluble TSP1 (and ox-LDL) to CD36-transfected cells and the mechanisms by which immobilized TSP1 mediated attachment and haptotaxis (cell migration towards a substratum-bound ligand) of these transfected cells. CD36 cDNA transfection of NIH 3T3 cells clearly induced a dramatic increase in binding of both soluble [125I]-TSP1 and [125I]-ox-LDL to the surface of CD36-transfected cells, indicating that there was a gain of function with CD36 transfection in NIH 3T3 cells. Despite this gain of function, mock- and CD36-transfected NIH 3T3 cells attached and migrated to a similar extent on immobilized TSP1. An anti-TSP1 oligoclonal antibody inhibited CD36-transfected cell attachment to TSP1 while function blocking anti-CD36 antibodies, alone or in combination with heparin, did not. A series of fusion proteins encompassing cell-recognition domains of TSP1 was then used to delineate mechanisms by which NIH 3T3 cells adhere to TSP1. Although CD36 binds soluble TSP1 through a CSVTCG sequence located within type 1 repeats, 18,19CD36-transfected NIH 3T3 cells did not attach to immobilized type 1 repeats while they did adhere to the N-terminal, type 3 repeats (in an RGD-dependent manner) and the C-terminal domain of TSP1. Conversely, Bowes melanoma cells attached to type 1 repeats and the N- and C-terminal domains of TSP1. However, CD36cDNA transfection of Bowes cells did not increase cell attachment to type 1 repeats compared to that observed with mock-transfected Bowes cells. Moreover, a function blocking anti-CSVTCG peptide antibody did not inhibit the attachment of mock- and CD36-transfected Bowes cells to type 1 repeats. It is suggested that CD36/TSP1 interaction does not occur upon cell-matrix adhesion and haptotaxis because TSP1 undergoes conformational changes that do not allow the exposure of the CD36 binding site.