Glypican-1 Is a VEGF165 Binding Proteoglycan That Acts as an Extracellular Chaperone for VEGF165 *

Glypican-1 is a member of a family of glycosylphosphatidylinositol anchored cell surface heparan sulfate proteoglycans implicated in the control of cellular growth and differentiation. The 165-amino acid form of vascular endothelial growth factor (VEGF165) is a mitogen for endothelial cells and a potent angiogenic factor in vivo. Heparin binds to VEGF165 and enhances its binding to VEGF receptors. However, native HSPGs that bind VEGF165 and modulate its receptor binding have not been identified. Among the glypicans, glypican-1 is the only member that is expressed in the vascular system. We have therefore examined whether glypican-1 can interact with VEGF165. Glypican-1 from rat myoblasts binds specifically to VEGF165 but not to VEGF121. The binding has an apparent dissociation constant of 3 × 10−10 m. The binding of glypican-1 to VEGF165 is mediated by the heparan sulfate chains of glypican-1, because heparinase treatment abolishes this interaction. Only an excess of heparin or heparan sulfates but not other types of glycosaminoglycans inhibited this interaction. VEGF165 interacts specifically not only with rat myoblast glypican-1 but also with human endothelial cell-derived glypican-1. The binding of125I-VEGF165 to heparinase-treated human vascular endothelial cells is reduced following heparinase treatment, and addition of glypican-1 restores the binding. Glypican-1 also potentiates the binding of 125I-VEGF165 to a soluble extracellular domain of the VEGF receptor KDR/flk-1. Furthermore, we show that glypican-1 acts as an extracellular chaperone that can restore the receptor binding ability of VEGF165, which has been damaged by oxidation. Taken together, these results suggest that glypican-1 may play an important role in the control of angiogenesis by regulating the activity of VEGF165, a regulation that may be critical under conditions such as wound repair, in which oxidizing agents that can impair the activity of VEGF are produced, and in situations were the concentrations of active VEGF are limiting.

Vascular endothelial growth factors (VEGFs) are mitogens for endothelial cells and are potent angiogenic factors in vivo. Five VEGF isoforms, designated VEGF 121 , VEGF 145 , VEGF 165 , VEGF 189 , and VEGF 206 , are generated via an alternative splicing mechanism from a unique gene (14 -16). The active form of the VEGFs is a homodimer, but active heterodimers have also been observed. Two VEGF tyrosine-kinase receptor types have been characterized. These tyrosine-kinase receptors do not differentiate between the various VEGF forms. The KDR/flk-1 receptor mediates the mitogenic activity of VEGF, whereas flt-1 stimulates VEGF-induced cell migration (17). In addition, endothelial cells express VEGF 165 -specific receptors of unknown function. These receptors do not bind VEGF 121 or VEGF 145 (16,18).
The best characterized VEGF forms are VEGF 121 and VEGF 165 (165-and 121-amino acid-long polypeptide, respectively). VEGF 165 contains the peptide encoded by exon-7 of the VEGF gene, whereas VEGF 121 lacks this peptide. The presence of exon-7 confers on VEGF 165 the ability to bind heparin-like molecules. Removal of heparan sulfates (HS) from the surface of endothelial cells by heparinase digestion reduces the binding of 125 I-VEGF 165 to its receptors, and addition of heparin re-stores 125 I-VEGF 165 binding. We have recently shown that part of the effect of heparin involves restoration of 125 I-VEGF 165 binding to KDR/flk-1 by a chaperone-like effect that recuperates the biological activity of VEGF 165 that was damaged by oxidizing agents (18). Such a role of heparin-like molecules may be important under conditions like wound repair, hypoxiainduced angiogenesis, or inflammation, in which oxidants or free radicals are produced and may damage VEGFs (19 -21). Unlike VEGF 165 , VEGF 121 is irreversibly inactivated by oxidizing reagents, and heparin cannot restore its receptor binding ability (18).
Recent evidence strongly suggests that VEGFs play a critical role in the process of tumor angiogenesis. This process is essential for tumor progression and for the subsequent process of tumor metastasis. A number of molecules that display some structural similarity to heparin such as suramin and pentosan sulfate exert an anti-angiogenic effect. These observations, together with the findings that HS-degrading enzymes can inhibit tumor angiogenesis (22), suggest that HSPGs play an important role in the angiogenic process. So far, HSPGs that can bind VEGFs and modulate their biological activities have not been identified. Among glypicans, glypican-1 is the only member expressed in the vascular system (23,24). We have therefore examined whether glypican-1 can interact with VEGFs and modulate their interaction with VEGF receptors. We show here that purified glypican-1 binds VEGF 165 with high affinity and supports its binding to heparinase-treated endothelial cells. Furthermore, glypican-1 can also restore the receptor binding ability of oxidized VEGF 165 and can therefore be viewed as a native, cell surface-localized proteoglycan that displays a chaperone-like activity.

EXPERIMENTAL PROCEDURES
Materials-Human recombinant VEGF 165 and VEGF 121 were produced and purified from Sf-9 insect cells as described previously (25)(26)(27). Human recombinant FGF-2, FGF-1, PF4, and FGF-7 were produced in bacteria and purified as described previously (28 -30). Bovine brain FGF-1 was purchased from R & D. Heparinase type I and III was kindly provided by IBEX Technologies (Montreal, Canada). Intestinal mucosa-derived heparin and chondroitin sulfates A and C and hyaluronic acid were from Sigma. Heparan sulfate from bovine lung and heparin-sepharose were obtained from Amersham Pharmacia Biotech. Carrier-free Na 125 I and [ 35 S]Na 2 SO 4 were purchased from NEN Life Science Products. Microtiter ELISA plates were from Corning. Tissue culture media, sera, and cell culture supplements were from Beth-Haemek Biological Industries (Israel) or from Life Technologies, Inc. Disuccinimidyl suberate was obtained from Pierce. All other chemicals were purchased from Sigma.
Purification of Glypican, Enzymatic Deglycosylation, and Isolation of Glypican-1-associated GAG Side Chains-Glypican-1 was purified from salt extracts of subconfluent cultures of the rat myoblast cell line L6E9 by anion exchange chromatography with DEAE-Sephacel followed by FGF affinity purification as described previously (11). Enzymatic deglycosylation was carried out in PBS containing 125 I-glypican-1 and 0.5 units/ml of heparinase IϩIII. Incubation was for 2 h at 37°C. Degradation of the core protein of glypican-1 was carried out overnight with proteinase K (0.5 mg/ml). To ensure that digestion was complete, a parallel incubation was carried out in the presence of radioiodinated glypican-1, and digestion was monitored by SDS-polyacrylamide gel electrophoresis and autoradiography. The GAG side chains were then separated from the protease and degradation products by DEAE-Sephacel chromatography as described (11).
Quantitation of GAG and HSPG-The Safranin O dye that reacts with carboxyl and sulfate groups and can detect ng amounts of sulfated GAGs (32) was used to estimate GAG content.
The concentrations of glypican-1 are relative to its HS concentration, which accounts to about two-thirds of the total PG concentration (11). Prior to performing the study we examined the biological activity of glypican-1 by testing its ability to enhance binding of FGF-1 and FGF-2 to FGF receptor 1 and to modulate the biological activities of FGF-7 and FGF-1 as described previously (11).
Radioiodination of Glypican-1 and VEGFs-Radioiodination was carried out utilizing the chloramine T method as described previously (11,33,34). 125 I-VEGF 165 was separated from free iodine using a heparin-Sepharose column, and 125 I-VEGF 121 was separated from free iodine using size exclusion chromatography on Sephadex-G25 as described (34). The specific activities of the 125 I-VEGF 165 and the 125 I-VEGF 121 were about 10 5 cpm/ng. Radiolabeled glypican-1 was separated from free iodine by chromatography on DEAE-Sephacel (11). Specific activities of iodinated glypican-1 were in range of 6 -12 ϫ 10 6 cpm/g.
Binding of 125 I-Glypican-1 to VEGF-coated Wells-Binding of radioiodinated glypican-1 to wells coated with VEGF 165 and various other proteins was done essentially as described (35). Briefly, 0.2 g of each protein in 100 l of coating buffer were adsorbed to 96-well ELISA plates for 2 h at room temperature, and binding was performed with 125 I-glypican-1 for 2 h at room temperature. Free 125 I-glypican-1 was removed by three washes with wash buffer (35). Estimation of bound 125 I-glypican-1 was done following solubilization with 0.2 M NaOH. All the experiments were done in triplicate and were repeated at least three times. Nonspecific binding was estimated as described under "Results," and it was less than 10% of the total binding. Standard deviation between replicates in all the experiments was less than 10%.
Reverse Immunodot Assay on Nitrocellulose Filters- 35 S-HSPGs were partially purified from conditioned medium of metabolically labeled HUVEC cells as described (11). Various proteins were spotted onto nitrocellulose filters at 2 g/spot. The filters were blocked with coating buffer (35) containing 1% BSA, washed three times with coating buffer, and incubated with binding buffer (35) containing HUVECderived 35 S-HSPGs (15,000 cpm/cm 2 ) for 2 h at room temperature. Following extensive washes, one filter was air dried and taken for autoradiography to confirm binding of radiolabeled HSPGs, and the remaining filters were incubated with a monoclonal antibody directed against human glypican-1 (monoclonal antibody S1) or with a antiphosphotyrosine monoclonal antibody (Upstate Biotechnology Inc.) as described (36). Detection of bound antibodies was performed with enhanced chemiluminescence reagents.
Heparinase Digestion and Receptor Binding Assays-Binding and cross-linking of VEGFs to a soluble extracellular domain of flk-1 fused to secreted human alkaline phosphatase (designated flk/SEAP) was done as described previously (18,26). Binding and cross-linking to HUVEC cells was essentially performed as described previously, except that instead of using adherent cells, the binding was performed using cells in suspension. The cells were first treated with heparinase (0.5 units/ml) for 1 h at 37°C and then detached from the tissue culture dishes using PBS solution containing 2 mM EDTA. The cells were collected, washed twice with PBS and once with binding buffer, and then subjected to the binding assay. Following binding the cells were washed extensively with PBS, and cross-linking was performed in solution. Equal amounts of total cell lysates were loaded onto each lane.
Inactivation of VEGF 165 by Oxidation-Oxidation of VEGF 165 was performed essentially as described (18). Briefly, 2 g of growth factor in 100 mM sodium phosphate buffer (pH 7.2) were incubated for 1 min with 0.1% of H 2 O 2 in a final reaction volume of 40 l. The reaction was terminated by sodium metabisulfite, and the growth factor was separated from the H 2 O 2 and sodium metabisulfite using heparin-Sepharose affinity chromatography (recovery was over 90%).

RESULTS
Glypican-1 Binds to VEGF 165 -A solid phase assay was utilized to investigate the ability of glypican-1 to bind VEGF 165 and VEGF 121 . Increasing concentrations of radioiodinated glypican-1 were bound to microtiter plates precoated with 0.2 g/well of VEGF 165 or VEGF 121 as well as FGF-7, which was previously reported to bind glypican (11). As shown in Fig. 1, only VEGF 165 and FGF-7 bind similarly to 125 I-glypican, whereas VEGF 121 or BSA did not exhibit any binding capacity. Binding was saturable and Scatchard analysis yielded an apparent dissociation constant of 1.2 ϫ 10 Ϫ10 M for VEGF 165 and 3 ϫ 10 Ϫ10 M for FGF-7 (Fig. 1B).
To test the specificity of this interaction we measured the ability of a panel of proteins to compete with 125 I-glypican for binding to VEGF 165 . As shown in Fig. 2, all three FGF mem-bers inhibited binding. This finding is with accordance with the previously reported capacity of these growth factors to bind glypican (11). By contrast, VEGF 121 , insulin, transferrin, and BSA could not compete with 125 I-glypican for binding to VEGF 165 .
Binding of Glypican-1 to VEGF 165 Is Mediated by Its HS Chain-To determine whether glypican-1-derived GAGs are involved in VEGF 165 binding, we prepared heparinase-digested or protease-digested glypican. Heparinase treatment abolished the ability of glypican-1 to compete with intact 125 I-glypican-1 for binding to VEGF 165 , whereas the free glypican-1-derived GAGs competed with 125 I-glypican-1 as efficiently as intact glypican-1 (Fig. 3A). These findings suggest that the interaction of glypican-1 with VEGF is mediated by its GAG side chains. To further verify these findings we compared the ability of intact glypican, glypican-1-derived GAGs, heparin, commercial HS, and non-HS GAGs to inhibit the binding of glypican-1 to VEGF 165. Glypican-1-derived HS inhibited the binding in a concentration-dependent manner (Fig. 3B). Half-maximal inhibition was obtained at a concentration of about 180 ng/ml glypican-1-derived HS. This concentration was 8-fold higher as compared with the heparin concentration required for halfmaximal inhibition but about 10-fold lower than that of commercial HS. By contrast neither chondroitin sulfate nor hyaluronic acid were able to compete with glypican-1 for binding to VEGF 165 . These results establish that the HS side chains of glypican-1 mediate binding to VEGF 165 .
Glypican-1 Enhances the Binding of VEGF 165 to VEGF Receptors-VEGF 165 binds to three VEGF receptors on HUVECderived endothelial cells (28). The larger of these receptors is KDR/flk-1, whereas the identity of the other two is not yet known (18). Heparin enhances the binding of VEGF 165 to all three receptors (28). Heparin is not a normal constituent of endothelial cells, whereas glypican-1 is known to be expressed in these cells (24). 2 We have therefore examined whether glypican-1 can modulate the binding of VEGF 165 to cell surface receptors in untreated and in heparinase-digested endothelial cells. Digestion of the endothelial cells by heparinase inhibited almost completely the binding of 125 I-VEGF 165 to the two smaller receptors, whereas the binding to KDR/flk-1 (18) was inhibited by about 80% (Fig. 4A, lane 3). Exogenously added glypican-1 restored the binding to all three receptors (Fig. 4A,  lane 4). Glypican-1 also potentiated the binding of 125 I-VEGF 165 to the two smaller receptors in untreated cells (Fig.  4A, lane 2).
The ability of glypican-1 to potentiate the binding of VEGF 165 to KDR/flk-1 was also confirmed utilizing a soluble extracellular domain of this receptor (flk-1/SEAP) in a cell-free assay (26). The potentiation of 125 I-VEGF 165 binding to the soluble receptor was concentration-dependent and was maximal at about 1 g/ml of glypican-1 (Fig. 4B). These findings were also confirmed in cross-linking experiments (data not shown).
The ability of myoblast-derived glypican-1 to restore binding of VEGF 165 to heparinase-treated HUVEC to a level that is higher than the binding observed in untreated cells strongly suggests that glypican-1 from endothelial cells also fulfills a similar function. We have therefore partially purified glypican-1 from HUVEC cells and have tested its ability to interact with VEGF 165 . VEGF 165 was bound to nitrocellulose membranes and incubated with partially purified 35 S-HSPGs. As shown in Fig. 5A, VEGF 165 and FGF-2 bound 35 S-HSPGs, whereas VEGF 121 , BSA, transferrin, and insulin did not exhibit any HSPG binding ability. Moreover, a monoclonal antibody directed against human glypican-1 interacted specifically with the VEGF 165 and FGF-2-bound HSPGs (Fig. 5B). These observations strongly suggest that HUVEC-associated glypican-1 is a candidate modulator of VEGF 165 activity in these cells.
Glypican  1. Binding of glypican-1 to VEGF. A, binding saturation of 125 I-glypican-1 to VEGF. 125 I-Labeled glypican-1 (specific activity of 12 ϫ 10 6 cpm/g) was bound to VEGF 165 -coated (q), VEGF 121coated (), and FGF-7-coated (E) wells as described under "Experimental Procedures." After 2 h of incubation at room temperature, the wells were extensively washed, and quantification of bound 125 Iglypican-1 was done as described under "Experimental Procedures." B, Scatchard analysis of glypican-1 binding to VEGF 165 (q) and FGF-7 (E). The Ligand program was used for the analysis (55). receptor and that heparin restores the receptor binding capacity of VEGF 165 (18). Because glypican-1 binds efficiently to VEGF 165 , we examined whether it displays a similar restorative capacity. As shown in Fig. 6, H 2 O 2 -treated VEGF 165 lost the ability to efficiently compete with VEGF 121 for binding to flk-1/SEAP, and addition of glypican-1 restored the receptor binding capacity of the oxidized VEGF 165 .

PF4 Inhibits the Binding of Glypican-1 to 125 I-VEGF 165 and Abrogates the Stimulatory Effect of Glypican-1 on 125 I-VEGF 165
Receptor Binding-PF4 is a heparin-binding protein that is synthesized by megakaryocytes, sequestered in platelets, and released from ␣-granules as a complex with chondroitin 4-sulfate proteoglycan (37,38). PF4 displays an anti-angiogenic activity in vivo that is attributed in part to its heparin binding capacity. Thus, a peptide derived from the heparin-binding carboxyl-terminal domain of PF4 possesses anti-angiogenic properties (29). The receptor binding ability of VEGF 165 and its mitogenic activity are inhibited by PF4 (35). Our results indicate that glypican may be one of the endothelial cell-associated heparan sulfate proteoglycans that bind VEGF 165 and mediate its biological activity. We reasoned that PF4 may inhibit an-giogenesis by preventing the interaction between VEGF and glypican. We have therefore examined whether PF4 can inhibit the binding of VEGF 165 to glypican-1 and abrogate the stimulatory effect of glypican-1 on VEGF receptor binding. As shown in Fig. 7A, PF4 efficiently inhibited the binding of glypican-1 to VEGF 165 -coated wells. Half-maximal inhibition was obtained at a PF4 concentration of 100 ng/ml. Moreover, PF4 nullifies the stimulatory effect of glypican-1 on the binding of 125 I-VEGF 165 to flk-1/SEAP (Fig. 7B). DISCUSSION Heparin had been previously shown to bind to VEGF 165 and to act as an accessory receptor that enhances the interaction of VEGF 165 with its signaling receptors. Furthermore, heparin restores the bioactivity of damaged VEGF 165 (18). However, because cells express several types of HSPGs on their surfaces but not heparin, it was important to identify native HSPGs that bind VEGF and modulate its activity. In the present study we have demonstrated for the first time an interaction of VEGF with the lipid anchored cell surface heparan sulfate proteoglycan glypican-1. Glypican-1 was found to interact with VEGF 165 FIG. 3. Binding of glypican-1 to VEGF 165 is mediated by its HS side chains. A, VEGF does not interact with glypican-1 core protein. 125 I-Glypican-1 (5 ng/well) was bound to VEGF 165 -coated wells in the absence or presence of 1 g/ml of unlabeled intact glypican-1 (Gly), glypican-1-free GAGs (Gly-HS) obtained by proteinase K digestion, or glypican-1 core protein (Gly-core) obtained by heparinase digestion. B, inhibition of the binding of 125 I-glypican-1 to VEGF 165 by glycosaminoglycans. Binding was performed in the presence of the indicated concentrations of heparin (f), glypican-1-derived HS (ƒ), commercial HS (), chondroitin sulfate (E), and hyaluronic-acid (q). The wells were washed, and the amount of bound radioactivity was quantified.
FIG. 4. Effect of glypican-1 on the binding of 125 I-VEGF 165 to its receptors. A, glypican-1 restores binding of VEGF 165 to heparinasetreated HUVEC cells. Cells were treated with heparinase and detached from the plate using PBS solution containing EDTA as described under "Experimental Procedures." Cells were allowed to bind 125 I-VEGF 165 (10 ng/ml) for 2 h at 4°C, the cells were then washed, and bound 125 I-VEGF 165 was cross-linked to cell surface receptors. Cross-linked complexes were visualized following SDS-polyacrylamide gel electrophoresis and autoradiography. Lane 1, untreated cells; lanes 2, untreated cells bound to VEGF in the presence of 3 g/ml of glypican; lane 3, heparinase-treated cells; lane 4, heparinase-treated cells bound to VEGF in the presence of 3 g/ml of glypican. B, glypican-1 potentiates the binding of VEGF 165 to a soluble form of KDR/flk-1. 125 I-VEGF 165 (20 ng/ml) was bound to flk-1/SEAP-coated wells in the presence of increasing concentrations of glypican-1 (E) or heparin (q), with or without 1 g/ml unlabeled VEGF 165 , as described under "Experimental Procedures." Specific binding was determined by subtracting normalized cpm of samples incubated with 1 g/ml unlabeled VEGF 165 from the normalized cpm bound in the absence of unlabeled ligand. and to modulate its receptor binding properties. We show that glypican-1 binds to VEGF 165 via its heparan sulfate chains and that the binding is specific and saturable. The affinity of glypican-1 to VEGF 165 is high and has an apparent dissociation constant of 1.2 ϫ 10 Ϫ10 M. Glypican-1 not only binds to VEGF 165 but is also capable of enhancing the interaction of VEGF 165 with its signaling receptors both in cell-free binding assays and in heparinase-treated endothelial cells.
VEGF 165 is a major angiogenic factor that is active in processes such as wound repair, hypoxia-induced angiogenesis, and inflammation (15), processes associated with the generation of oxidizing agents and free radicals. VEGF 165 is inactivated by these agents (18), and therefore their presence may inhibit VEGF-induced angiogenesis. Therefore, repair mecha-nisms allowing restoration of VEGF activity could fulfill an important biological role. Glypican-1 restores the biological activity of oxidized VEGF and is expressed on the cell surface of the endothelial cells that are the natural target for VEGF (24). The presence of glypican on the surface of endothelial cells could therefore be a fail-safe mechanism ensuring that every VEGF 165 molecule that reaches the endothelial cells will be eventually able to interact and activate signaling VEGF receptors such as the KDR receptor. This function of glypican-1 may be critical under conditions in which the concentrations of VEGF are limited or under conditions in which the activity of VEGF is compromised.
Studies in recent years have strongly indicated that HSPGs are important modulators of the activity of heparin-binding growth factors, an issue that is particularly well studied for FGFs (39,40). FGFs bind to cell surface and extracellular matrix-associated HSPGs. The mechanism by which these HSPGs regulate FGF activity is complex and can be manifested at several levels. These molecules protect FGFs from protease digestion or from heat/acid inactivation, and it was shown that they can restore the activity of heat inactivated FGF (41). HSPGs are also thought to provide a reservoir from which FGFs can be rapidly released in response to specific triggering events (42). In addition to these effects, cell-associated HSPGs can increase the affinity of FGFs to their signaling receptors by stabilizing ligand-receptor complexes (43)(44)(45). VEGFs interact with several receptor types, some of which may also be able to interact directly with heparan sulfates (46 -48). It is therefore possible that in addition to its chaperone-like role, glypican-1 may be able to directly modulate the interaction between VEGF and these receptors using similar mechanisms.
The defined structure of the HSPGs side chains is regulated in a cell type-dependent manner. For example, human lung fibroblast-derived glypican-1 cannot enhance the biological activity of FGF-2 (49). On the other hand, glypican-1 derived from rat myoblasts or expressed ectopically in K562 cells can stimulate the binding of FGF-2 to FGF-receptor 1 and potentiates cellular responses to FGF-2 (11,12). The ability of glypican-1 from rat myoblasts and from human endothelial cells to interact with VEGF 165, together with the finding that rat myoblast glypican-1 can restore VEGF receptor binding in heparinasetreated HUVEC cells, strongly suggests that the HS side chains attached to glypican-1 in both cell types contain similar VEGF 165 -binding domains. Glypican-1 exists on the surface of cells both as a lipid anchored form and as a peripheral membrane proteoglycan. The peripheral form is generated following cleavage of the lipid anchored form by a specific phospholipase and can be shed from cell surfaces as a soluble proteoglycan (50). The fact that soluble myoblast-derived glypican-1 can bind and modulate VEGF 165 activity suggests that HSPGs derived from one cell type can act in a paracrine manner to affect biological responses to heparin-binding growth factors.
Beside glypican-1, endothelial cells express other HSPGs such as fibroglycan, perlecan, and syndecan (24,51,52). Studies conducted with growth factors of the FGF family revealed that distinct core proteins can bear HS side chains with similar function (12,49). A similar situation might exist for growth factors of the VEGF family. The ability of glypican-1 to fully restore VEGF receptor binding to heparinase-treated endothelial cells indicates that even if endothelial cells do express more than one type of stimulatory HSPG species for VEGF, it is likely that they bear HS with a fine structure similar to that of glypican-1 side chains. In addition, in heparinase-digested endothelial cells, glypican-1 enhances VEGF165 binding to its receptors to a level that is higher than that observed in untreated cells incubated with glypican-1 (compare lanes 2 and 4,  35 S-HSPGs to nitrocellulose filters. B and C, immunodetection of bound glypican by monoclonal antibody S1 (B) or with a monoclonal anti-phosphotyrosine antibody (C). The filters were also autoradiographed at the end of the assay to confirm that bound 35 4A). These findings could also imply that endothelial cells express HSPG species that inhibit the binding of VEGF to its receptors.
Heparin-like molecules have long been implicated in the control of angiogenesis. The earliest indication that heparin may be involved in regulation of the angiogenic process was the finding that mast cells accumulate at the site of tumor angiogenesis before capillary ingrowth, that conditioned medium from mast cells induces angiogenesis in vitro, and that this activity can be abolished by protamine and heparan sulfate degrading enzymes (53,54). It was subsequently found that many heparin-binding growth factors, including members of the FGF and VEGF family, are highly angiogenic. It soon became apparent that heparin and HS play an important role in modulating the biological activity of the growth factors which bind to them. However, only recently were efforts taken to identify native HSPGs that bind and modulate the activity of these growth factors (11)(12)(13)49). Our previous studies have shown that glypican-1 modulates the biological activity of FGF-2, suggesting a role for glypican-1 in the regulation of angiogenesis (11). The present study lends support to this hypothesis and further suggests that glypican-1 may play a more general modulatory role in angiogenesis by regulating the stability and activity of VEGFs.