Characterization of endostatin binding to heparin and heparan sulfate by surface plasmon resonance and molecular modeling: role of divalent cations

Endostatin (20 kDa) is a C-terminal proteolytic fragment of collagen XVIII that is localized in vascular basement membrane zones in various organs. It binds zinc, heparin/heparan sulfate, laminin, and sulfatides and inhibits angiogenesis and tumor growth. Here we determined the kinetics and affinity of the interaction of endostatin with heparin/heparan sulfate and investigated the effects of divalent cations on these interactions and on the biological activities of endostatin. The binding of human recombinant endostatin to heparin and heparan sulfate was studied by surface plasmon resonance using BIAcore technology and further characterized by docking and molecular dynamics simulations. Kinetic data, evaluated using a 1:1 interaction model, showed that heparan sulfate bound to and dissociated from endostatin faster than heparin and that endostatin bound to heparin and heparan sulfate with a moderate affinity (K(D) approximately 2 microm). Molecular modeling of the complex between endostatin and heparin oligosaccharides predicted that, compared with mutagenesis studies, two further arginine residues, Arg(47) and Arg(66), participated in the binding. The binding of endostatin to heparin and heparan sulfate required the presence of divalent cations. The addition of ZnCl(2) to endostatin enhanced its binding to heparan sulfate by approximately 40% as well as its antiproliferative effect on endothelial cells stimulated by fibroblast growth factor-2, suggesting that this activity is mediated by the binding of endostatin to heparan sulfate. In contrast, no increase in the antiangiogenic and anti-proliferative activities of endostatin promoted by vascular endothelial growth factor was observed upon the addition of zinc.     

The morphogenic properties of oligomeric endostatin are dependent on cell surface heparan sulfate

Endostatin has attracted considerable attention because of its ability to inhibit angiogenesis. This property of monomeric endostatin contrasts with that of the trimeric endostatin moiety generated from the intact C-terminal domain of collagen XVIII that induces a promigratory phenotype in endothelial cells. This activity is inhibited by monomeric endostatin. In this study we demonstrate that the effect of oligomeric endostatin can also be inhibited by exogenous glycosaminoglycans in a size-dependent manner, with heparin oligosaccharides containing more than 20 monosaccharide residues having optimal inhibitory activity. Oligomeric endostatin was also found to induce morphological changes in Chinese hamster ovary cells, an epithelial cell line. This novel observation allowed the utilization of a panel of Chinese hamster ovary cell mutants with defined glycosaminoglycan biosynthetic defects. The action of oligomeric endostatin on these cells was shown to be dependent on cell surface glycosaminoglycans, principally heparan sulfate with N- and 6-O-sulfation of glucosamine residues rather than iduronate 2-O-sulfation being important for bioactivity. The responsiveness of a cell line (pgsE-606) with globally reduced heparan sulfate sulfation and shortened S domains, however, indicates that overall heparan sulfate domain patterning is the key determinant of the bioactivity of oligomeric endostatin. Purified heparin-monomeric endostatin constructs generated by zero-length cross-linking techniques were found to be unable to inhibit the action of oligomeric endostatin. This indicates a mechanism for the perturbation of oligomeric endostatin action by its monomeric counterpart via competition for glycosaminoglycan attachment sites at the cell surface.     

Multiple heparan sulfate proteoglycans synthesized by a basement membrane producing murine embryonal carcinoma cell line

The murine embryonal carcinoma derived cell line M1536-B3 secretes the basement membrane components laminin and entactin and, when grown in bacteriological dishes, produces and adheres to sacs of basement membrane components. Heparan sulfate proteoglycans have been isolated from these sacs, the cells, and the medium. At least three different heparan sulfate proteoglycans are produced by these cells as determined by proteoglycan size, glycosaminoglycan chain length, and charge density. The positions of the N- and O-sulfate groups in the glycosaminoglycan chains from each proteoglycan appear to be essentially the same despite differences in the size and culture compartment locations of the heparan sulfate proteoglycan. Additionally, small quantities of chondroitin sulfate proteoglycans are found in each fraction and copurify with each heparan sulfate proteoglycan. Because this cell line appears to synthesize at least three different heparan sulfate proteoglycans which are targeted to different final locations (basement membrane, cell surface, and medium), this will be a useful system in which to study the factors which determine final heparan sulfate proteoglycan structures and culture compartment targeting and the possible effects of the protein core(s) on heparan sulfate carbohydrate chain synthesis and secretion.     

Localization of binding sites for laminin, heparan sulfate proteoglycan and fibronectin on basement membrane (type IV) collagen

Rotary shadowing electron microscopy was used to examine complexes formed by incubating combinations of the basement membrane components: type IV collagen, laminin, large heparan sulfate proteoglycan and fibronectin. Complexes were analyzed by length measurement from the globular (COOH) domain of type IV collagen, and by examination of the four arms of laminin and the two arms of fibronectin. Type IV collagen was found to contain binding sites for laminin, heparan sulfate proteoglycan and fibronectin. With laminin the most frequent site was centered approximately 81 nm from the carboxy end of type IV collagen. Less frequent sites appeared to be present at approximately 216 nm and approximately 291 nm, although this was not apparent when the sites were expressed as a fraction of the length of type IV collagen to which they were bound. For heparan sulfate proteoglycan the most frequent site occurred at approximately 206 nm with a less frequent site at approximately 82 nm. For fibronectin, a single site was present at approximately 205 nm. Laminin bound to type IV collagen through its short arms, particularly through the end of the lateral short arms and to heparan sulfate proteoglycan mainly through the end of its long arm. Fibronectin bound to type IV collagen through the free end region of its arms. Using a computer graphics program, the primary laminin binding sites of two adjacent type IV collagen molecules were found to align in the "polygonal" model of type IV collagen, whereas with the "open network" model, a wide meshed matrix is predicted. It is proposed that basement membrane may consist of a lattice of type IV collagen coated with laminin, heparan sulfate proteoglycan and fibronectin.     

Structure and biological activity of basement membrane proteins

Collagen type IV, laminin, heparan sulfate proteoglycans, nidogen (entactin) and BM-40 (osteonectin, SPARC) represent major structural proteins of basement membranes. They are well-characterized in their domain structures, amino acid sequences and potentials for molecular interactions. Such interactions include self-assembly processes and heterotypic binding between individual constituents, as well as binding of calcium (laminin, BM-40) and are likely to be used for basement membrane assembly. Laminin, collagen IV and nidogen also possess several cell-binding sites which interact with distinct cellular receptors. Some evidence exists that those interactions are involved in the control of cell behaviour. These observations have provided a more defined understanding of basement membrane function and the definition of new research goals in the future.     

Human basement membrane heparan sulfate proteoglycan core protein: a 467-kD protein containing multiple domains resembling elements of the low density lipoprotein receptor, laminin, neural cell adhesion molecules, and epidermal growth factor

The primary structure of the large human basement membrane heparan sulfate proteoglycan (HSPG) core protein was determined from cDNA clones. The cDNA sequence codes for a 467-kD protein with a 21-residue signal peptide. Analysis of the amino acid sequence showed that the protein consists of five domains. The amino-terminal domain I contains three putative heparan sulfate attachment sites; domain II has four LDL receptor-like repeats; domain III contains repeats similar to those in the short arms of laminin; domain IV has lg-like repeats resembling those in neural cell adhesion molecules; and domain V contains sequences resembling repeats in the G domain of the laminin A chain and repeats in the EGF. The domain structure of the human basement membrane HSPG core protein suggests that this mosaic protein has evolved through shuffling of at least four different functional elements previously identified in other proteins and through duplication of these elements to form the functional domains. Comparison of the human amino acid sequence with a partial amino acid sequence from the corresponding mouse protein (Noonan, D. M., E. A. Horigan, S. R. Ledbetter, G. Vogeli, M. Sasaki, Y. Yamada, and J. R. Hassell. 1988. J. Biol. Chem. 263:16379-16387) shows a major difference between the species in domain IV, which contains the Ig repeats: seven additional repeats are found in the human protein inserted in the middle of the second repeat in the mouse sequence. This suggests either alternative splicing or a very recent duplication event in evolution. The multidomain structure of the basement membrane HSPG implies a versatile role for this protein. The heparan sulfate chains presumably participate in the selective permeability of basement membranes and, additionally, the core protein may be involved in a number of biological functions such as cell binding, LDL-metabolism, basement membrane assembly, calcium binding, and growth- and neurite-promoting activities.     

High resolution immunoelectron microscopic localization of functional domains of laminin, nidogen, and heparan sulfate proteoglycan in epithelial basement membrane of mouse cornea reveals different topological orientations

Thin and ultrathin cryosections of mouse cornea were labeled with affinity-purified antibodies directed against either laminin, its central segments (domain 1), the end of its long arm (domain 3), the end of one of its short arms (domain 4), nidogen, or low density heparan sulfate proteoglycan. All basement membrane proteins are detected by indirect immunofluorescence exclusively in the epithelial basement membrane, in Descemet's membrane, and in small amorphous plaques located in the stroma. Immunoelectron microscopy using the protein A-gold technique demonstrated laminin domain 1 and nidogen in a narrow segment of the lamina densa at the junction to the lamina lucida within the epithelial basement membrane. Domain 3 shows three preferred locations at both the cellular and stromal boundaries of the epithelial basement membrane and in its center. Domain 4 is located predominantly in the lamina lucida and the adjacent half of the lamina densa. The low density heparan sulfate proteoglycan is found all across the basement membrane showing a similar uniform distribution as with antibodies against the whole laminin molecule. In Descemet's membrane an even distribution was found with all these antibodies. It is concluded that within the epithelial basement membrane the center of the laminin molecule is located near the lamina densa/lamina lucida junction and that its long arm favors three major orientations. One is close to the cell surface indicating binding to a cell receptor, while the other two are directed to internal matrix structures. The apparent codistribution of laminin domain 1 and nidogen agrees with biochemical evidence that nidogen binds to this domain.     

Interactions of basement membrane components

The binding of laminin, type IV collagen, and heparan sulfate proteoglycan to each other was assessed. Laminin binds preferentially to native type IV (basement membrane) collagen over other collagens. A fragment of laminin (Mr 600 000) containing the three short chains (Mr 200 000) but lacking the long chain (Mr 400 000) showed the same affinity for type IV collagen as the intact protein. The heparan sulfate proteoglycan binds well to laminin and to type IV collagen. These studies show that laminin, type IV collagen and heparan sulfate proteoglycan interact with each other. Such interactions in situ may determine the structure of basement membranes.     

Heparin modulation of laminin polymerization

Previously, it has been shown that laminin will self-assemble by a two-step calcium-dependent process using end-domain interactions (Yurchenco, P. D., Tsi-library, E. C., Charonis, A. S., and Furthmayr, H. (1985) J. Biol. Chem. 260, 7636-7644). We now find that heparin, at low concentrations, modifies this polymerization by driving the equilibrium further toward aggregation, by producing a denser polymer, and by inducing aggregation in the absence of calcium. This effect on self-assembly is specific in that it is observed with heparin but not with several heparan sulfates or other glycosaminoglycans: it correlates with affinity and depends on the degree of polysaccharide sulfation. Heparin binds to laminin in a calcium-dependent manner with a single class of interaction (KD = 118 +/- 18 nM) and with a binding capacity of one heparin for two laminins. We find the long arm globule (E3) is the only laminin domain which exhibits substantial heparin binding: heparin binds E3 with an affinity (KD = 94 +/- 12 nM) and calcium dependence similar to that for intact laminin. These data strongly suggest that heparin modifies laminin assembly by binding to pairs of long arm globular domains. As a result the polymer may be stabilized at domain E3 and laminin interdomain interactions induced or modified. We further postulate that heparins may act in vivo as specific regulators of the structure and functions of basement membranes by both altering the laminin matrix and by displacing weakly binding heparan sulfates.     

Endostatin's heparan sulfate-binding site is essential for inhibition of angiogenesis and enhances in situ binding to capillary-like structures in bone explants

The functional role of endostatin's affinity for heparan sulfates was addressed using an ex vivo bone angiogenesis model. Capillary-like sprouts showed prominent expression of collagen XVIII/endostatin. Outgrowth of endothelial cells was not altered in the absence of collagen XVIII but inhibited by the addition of recombinant endostatin. Mutant non-heparan sulfate binding endostatin and the collagen XV endostatin homologue were ineffective. The ability of mutant endostatin to bind to capillary structures was reduced when compared to endostatin. Endostatin-XV completely failed to bind to endothelial cells. Our data indicate that endostatin's angiostatic function is heparan sulfate-dependent, and that in situ-binding of endostatin to endothelial cells is increased by heparan sulfates.     

A Mr 80K hepatocyte surface protein(s) interacts with basement membrane components

We have identified a Mr 80K cell surface protein(s) from adult rat hepatocytes that binds basement membrane components, including collagen IV, heparan sulfate proteoglycan, and laminin. Freshly isolated hepatocytes were cell surface-labeled with 125I using the lactoperoxidase-catalyzed method, and detergent-solubilized membrane proteins were chromatographed on affinity columns prepared with purified basement membrane components. A Mr 80K protein was eluted with 0.15-1 M NaCl from a collagen IV column. Two proteins (Mr 80K and 38K) were eluted from a heparan sulfate proteoglycan column. The larger protein was also eluted from a column made with heparan sulfate side chains. Several proteins (Mr 80K, 67K, 45K, and 32K) bound to an affinity chromatography column made with the laminin A chain-derived synthetic peptide PA22-2, which is active for promoting cell attachment. When fractions eluted from these columns were analyzed by two-dimensional gel electrophoresis, the Mr 80K proteins showed similar patterns with a pI ranging from 8 to 9. The Mr 80K protein(s) may have an important role in the interaction of hepatocytes with basement membrane.     

Basic fibroblast growth factor (bFGF) dissociates rapidly from heparan sulfates but slowly from receptors. Implications for mechanisms of bFGF release from pericellular matrix.

The effect of heparin on the rate of binding of basic fibroblast growth factor (bFGF) to high affinity (receptor) and low affinity (heparan sulfate) binding sites on endothelial cells and CHO cells transfected with FGF receptor-1 or FGF receptor-2 was investigated. Radiolabeled bFGF bound rapidly to both high and low affinity sites on all three types of cells. Addition of 10 micrograms/ml heparin eliminated binding to low affinity sites and decreased the rate of binding to high affinity sites to about 30% of the rate observed in the absence of heparin. However, the same amount of 125I-bFGF bound to high affinity sites at equilibrium in the presence and absence of heparin. The effect of heparin on the initial rate of binding to high affinity sites was related to the log of the heparin concentration. Depletion of the cells of heparan sulfates by treatment with heparinase also decreased the initial rate of binding to high affinity receptors. These results suggest that cell-surface heparan sulfates facilitate the interaction of bFGF with its receptor by concentrating bFGF at the cell surface. Dissociation rates for receptor-bound and heparan sulfate-bound bFGF were also measured. Dissociation from low affinity sites was rapid, with a half-time of 6 min for endothelial cell heparan sulfates and 0.5 min for Chinese hamster ovary heparan sulfates. In contrast, dissociation from receptors was slow, with a half-time of 46 min for endothelial cell receptors, 2.5 h for FGF receptor-1, and 1.4 h for FGF receptor-2. These results suggest that degradative enzymes may not be needed to release bFGF from the heparan sulfates in instances where receptors and heparan sulfate-bound bFGF are in close proximity because dissociation from heparan sulfates occurs rapidly enough to allow bFGF to bind to unoccupied receptors by laws of mass action.     

The beta-amyloid peptide of Alzheimer's disease decreases adhesion of vascular smooth muscle cells to the basement membrane

Cerebral amyloid angiopathy (CAA) is a major feature of Alzheimer's disease pathology. In CAA, degeneration of vascular smooth muscle cells (VSMCs) occurs close to regions of the basement membrane where the amyloid protein (Abeta) builds up. In this study, the possibility that Abeta disrupts adhesive interactions between VSMCs and the basement membrane was examined. VSMCs were cultured on a commercial basement membrane substrate (Matrigel). The presence of Abeta in the Matrigel decreased cell-substrate adhesion and cell viability. Full-length oligomeric Abeta was required for the effect, as N- and C-terminally truncated peptide analogues did not inhibit adhesion. Abeta that was fluorescently labelled at the N-terminus (fluo-Abeta) bound to Matrigel as well as to the basement membrane heparan sulfate proteoglycan (HSPG) perlecan and laminin. Adhesion of VSMCs to perlecan or laminin was decreased by Abeta. As perlecan influences VSMC viability through the extracellular signal-regulated kinase (ERK)1/2 signalling pathway, the effect of Abeta1-40 on ERK1/2 phosphorylation was examined. The level of phospho-ERK1/2 was decreased in cells following Abeta treatment. An inhibitor of ERK1/2 phosphorylation enhanced the effect of Abeta on cell adhesion. The studies suggest that Abeta can decrease VSMC viability by disrupting VSMC-extracellular matrix (ECM) adhesion.     

Laminin promotes formation of cord-like structures by Sertoli cells in vitro

Basement membranes are thin extracellular matrices which contact epithelial cells and promote their adhesion, migration, differentiation, and morphogenesis. These matrices are composed of collagen IV, heparan sulfate proteoglycan, laminin, and entactin as well as other minor components. Sertoli cells, like most epithelial cells, are in contact at their basal surface with a basement membrane. When cultured within three-dimensional basement membrane gels (Matrigel), Sertoli cells reorganize into cords that resemble testicular seminiferous cords found in the in vivo differentiating testis. Anti-laminin and anti-entactin antisera inhibit this cord morphogenesis by Sertoli cells whereas antisera against type IV and type I collagen, heparan sulfate proteoglycan, fibronectin, and preimmune sera had no effect. The RGD (RGDS-NH2) sequence, found in the cell binding domain of the integrin family of cell adhesion molecules as well as in the A chain of laminin and in entactin, effectively inhibited Sertoli cell cord formation at a concentration of 1.0 mg/ml but was unable to prevent Sertoli cell attachment at concentrations as high as 2.0 mg/ml. A synthetic pentapeptide from a cell-binding domain of the B1 chain of laminin. YIGSR-NH2, inhibited cord formation at a concentration of 0.25 mg/ml, but Sertoli cells were still adherent to the basement membrane matrix. At concentrations greater than 0.50 mg/ml, Sertoli cells detached. Antiserum against the YIGSR-NH2-containing sequence was also effective in inhibiting cord formation by Sertoli cells. Ligand (YIGSR-NH2 peptide) blot analysis of Sertoli cell lysates revealed an interaction with a major band at 60 kDa and with minor bands at 39 and 127 kDa. Furthermore, in Western blot analysis the anti-67-kDa laminin-binding protein antibody recognized a 59- to 60-kDa protein in Sertoli cells. The data indicate that laminin is involved in both Sertoli cell attachment and migration during formation of histotypic cord structures by these cells in culture. Two separate laminin cell-binding domains appear to be involved in Sertoli cell cord morphogenesis in vitro and are likely to participate in the formation of seminiferous cords in vivo.     

Endothelial cells interact with the core protein of basement membrane perlecan through beta 1 and beta 3 integrins: an adhesion modulated by glycosaminoglycan

Aortic endothelial cells adhere to the core protein of murine perlecan, a heparan sulfate proteoglycan present in endothelial basement membrane. We found that cell adhesion was partially inhibited by beta 1 integrin-specific mAb and almost completely blocked by a mixture of beta 1 and alpha v beta 3 antibodies. Furthermore, adhesion was partially inhibited by a synthetic peptide containing the perlecan domain III sequence LPASFRGDKVTSY (c-RGD) as well as by GRGDSP, but not by GRGESP. Both antibodies contributed to the inhibition of cell adhesion to immobilized c-RGD whereas only beta 1-specific antibody blocked residual cell adhesion to proteoglycan core in the presence of maximally inhibiting concentrations of soluble RGD peptide. A fraction of endothelial surface-labeled detergent lysate bound to a core affinity column and 147-, 116-, and 85-kD proteins were eluted with NaCl and EDTA. Polyclonal anti-beta 1 and anti-beta 3 integrin antibodies immunoprecipitated 116/147 and 85/147 kD surface-labeled complexes, respectively. Cell adhesion to perlecan was low compared to perlecan core, and cell adhesion to core, but not to immobilized c-RGD, was selectively inhibited by soluble heparin and heparan sulfates. This inhibition by heparin was also observed with laminin and fibronectin and, in the case of perlecan, was found to be independent of heparin binding to substrate. These data support the hypothesis that endothelial cells interact with the core protein of perlecan through beta 1 and beta 3 integrins, that this binding is partially RGD- independent, and that this interaction is selectively sensitive to a cell-mediated effect of heparin/heparan sulfates which may act as regulatory ligands.     

Interactions of basement membrane components

The binding of laminin, type IV collagen, and heparan sulfate proteoglycan to each other was assessed. Laminin binds preferentially to native type IV (basement membrane) collagen over other collagens. A fragment of laminin (M_r 600 000) containing the three short chains (M_r 200 000) but lacking the long chain M_r 400 000) showed the same affinity for type IV collagen as the intact protein. The heparan sulfate proteoglycan binds well to laminin and to type IV collagen. These studies show that laminin, type IV collagen and heparan sulfate proteoglycan interact with each other. Such interactions in situ may determine the structure of basement membranes.     

Isolation of a tumor cell laminin receptor

BL6 murine melanoma cells contain approximately 110,000 cell surface binding sites for the basement membrane glycoprotein laminin. Treatment of isolated melanoma cell plasma membranes with detergent yields a single class of laminin receptor. The receptor was purified 900 fold by laminin affinity chromatography. The isolated receptor has a Mr of 67,000 and binds laminin with high affinity: kd = 2 nm. The binding affinity of the isolated receptor was similar to that of the plasma membranes or the whole cells. Such a laminin receptor, isolated here for the first time, could facilitate the interaction of metastasizing tumor cells with the basement membrane.     

Correlation between cell substrate attachment in vitro and cell surface heparan sulfate affinity for fibronectin and collagen

Heparan sulfate glycosaminoglycan, isolated from the cell surface of nonadhering murine myeloma cells (P3X63-Ag8653), does not bind to plasma fibronectin, but binds partially to collagen type I, as assayed by affinity chromatography with proteins immobilized on cyanogen bromide-activated Sepharose 4B. Identical results were obtained when myeloma heparan sulfate was cochromatographed, on the same fibronectin and collagen columns, with cell surface heparan sulfates collagen columns, with cell surface heparan sulfates from adhering Swiss mouse 3T3 and SV3T3 cells. These latter heparan sulfates do, however, bind to both fibronectin and collagen, as reported earlier (Stamatoglou, S.C., and J.M. Keller, 1981, Biochim. Biophys. Acta., 719:90-97). Cell adhesion assays established that hydrated collagen substrata can support myeloma cell attachment, but fibronectin cannot. Saturation of the heparan sulfate binding sites on the collagen substrata with heparan sulfate or heparin, prior to cell inoculation, abolished the ability to support cell adhesion, whereas chondroitin 4 sulfate, chondroitin 6 sulfate, and hyaluronic acid had no effect.     

Binding of mouse nidogen-2 to basement membrane components and cells and its expression in embryonic and adult tissues suggest complementary functions of the two nidogens

Nidogen-1 binds several basement membrane components by well-defined, domain-specific interactions. Organ culture and gene targeting approaches suggest that a high-affinity nidogen-binding site of the laminin gamma1 chain (gamma1III4) is important for kidney development and for nerve guidance. Other proteins may also bind gamma1III4, although human nidogen-2 binds poorly to the mouse laminin gamma1 chain. We therefore characterized recombinant mouse nidogen-2 and its binding to basement membrane proteins and cells. Mouse nidogen-1 and -2 interacted at comparable levels with collagen IV, perlecan, and fibulin-2 and, most notably, also with laminin-1 fragments P1 and gamma1III3-5, which both contain the gamma1III4 module. In embryos, nidogen-2 mRNA was produced by mesenchyme at sites of epithelial-mesenchymal interactions, but the protein was deposited on epithelial basement membranes, as previously shown for nidogen-1. Hence, binding of both nidogens to the epithelial laminin gamma1 chain is dependent on epithelial-mesenchymal interactions. Epidermal growth factor stimulated expression of both nidogens in embryonic submandibular glands. Both nidogens were found in all studied embryonic and adult basement membranes. Nidogen-2 was more adhesive than nidogen-1 for some cell lines and was mainly mediated by alpha3beta1 and alpha6beta1 integrins as shown by antibody inhibition. These findings revealed extensive coregulation of nidogen-1 and -2 expression and much more complementary functions of the two nidogens than previously recognized.     

Multiple mechanisms for exogenous heparin modulation of vascular endothelial growth factor activity

Heparin and heparin‐like molecules are known to modulate the cellular responses to vascular endothelial growth factor‐A (VEGF‐A). In this study, we investigated the likely mechanisms for heparin's influence on the biological activity of VEGF‐A. Previous studies have shown that exogenous heparin's effects on the biological activity of VEGF‐A are many and varied, in part due to the endogenous cell‐surface heparan sulfates. To circumvent this problem, we used mutant endothelial cells lacking cell‐surface heparan sulfates. We showed that VEGF‐induced cellular responses are dependent in part on the presence of the heparan sulfates, and that exogenous heparin significantly augments VEGF's cellular effects especially when endogenous heparan sulfates are absent. Exogenous heparin was also found to play a cross‐bridging role between VEGF‐A₁₆₅ and putative heparin‐binding sites within its cognate receptor, VEGFR2 when they were examined in isolation. The cross‐bridging appears to be more dependent on molecular weight than on a specific heparin structure. This was confirmed by surface plasmon resonance binding studies using sugar chips immobilized with defined oligosaccharide structures, which showed that VEGF‐A₁₆₅ binds to a relatively broad range of sulfated glycosaminoglycan structures. Finally, studies of the far‐UV circular dichroism spectra of VEGF‐A₁₆₅ showed that heparin can also modulate the conformation and secondary structure of the protein. J. Cell. Biochem. 111: 461–468, 2010. © 2010 Wiley‐Liss, Inc.