Extracellular matrix regulation of growth factor and protease activity.

Extracellular matrices bind many growth factors, proteases, and protease inhibitors. These interactions not only localize these molecules to the pericellular environment, but also modulate their biological activities. Recent evidence suggests that some growth factors may be active in vivo primarily in complexes with extracellular matrix molecules and that this interaction may be essential to their activity.

Recombinant basic fibroblast growth factor stimulates wound healing in healing-impaired db/db mice.

The stimulatory effect of recombinant basic fibroblast growth factor (bFGF) on wound healing was assessed using healing-impaired (db/db) mice. Full-thickness wounds were made in female diabetic C57BL/KsJ db/db mice, and their normal (db/+) littermates with a punch biopsy instrument. Recombinant bFGF was applied locally to the open wound once a day. The mice were later killed and histological sections of the wounds were prepared. The degree of wound healing was evaluated using several histological parameters such as degree of reepithelialization, granulation tissue thickness, matrix density, number of infiltrated cells, and number of capillaries. Wounds from normal mice displayed good reepithelialization rates and granulation tissue formation, while wounds from db/db mice had poor responses, especially in the dermal parameters. Although the application of bFGF to wounds in the normal (db/+) mice had little effect, application of bFGF to wounds in db/db mice induced significant responses in all of the dermal parameters compared with nontreated db/db mice (p less than 0.001). In the presence of bFGF, these parameters approximated those observed in nontreated littermates. A minimum of 0.5 microgram bFGF in either single or multiple applications was required for a significant effect. bFGF that was either boiled or pretreated with neutralizing antibody had little stimulatory effect. Time-course experiments indicated that the granulation response in bFGF-treated mice peaked between 8 and 12 d, and decreased after 12 d, while matrix density continued to increase until the 18th day (p less than 0.05). The breaking strength of healed linear wounds in db/db mice was also decreased when compared with heterozygous littermates. This parameter was also improved by the administration of bFGF to the wounds (p less than 0.05).

Properties of plasminogen activators formed by neoplastic human cell cultures.

A series of human cell lines has been examined for fibrinolysis in culture. The sera that are activating for fibrinolysis by human cells are mouse, monkey, human, horse, and bovine. Individual human sera show considerable variation in the ability to activate fibrinolysis. In common with other neoplastic or transformed mammalian and avian cell cultures, human cell lines of neoplastic origin produce substantial amounts of plasminogen activator. Several cultures of nonmalignant origin also produce plasminogen activator, whereas cultures obtained from trypsinized human embryos, or from human embryonic skin do not. The human melanoma plasminogen activators are of two kinds: a major component with a mol wt of 50,000, and a minor species with a mol wt of approximately 60,000. Both are DFP sensitive, serine proteases.

An enzymatic function associated with transformation of fibroblasts by oncogenic viruses. II. Mammalian fibroblast cultures transformed by DNA and RNA tumor viruses.

Chick, hamster, mouse, and rat embryo fibroblast cultures, transformed by either DNA or RNA viruses, show fibrinolytic activity under suitable conditions of growth and in appropriate media; normal counterpart cultures do not. The fibrinolysin is produced by the interaction of two protein factors: one of these, a cell factor, is released by transformed cells and accumulates in the medium when cultures are incubated in the absence of scrum. The second factor, the serum factor, is a specific protein that is present in sera of many avian and mammalian species, including man. Not all sera yield fibrinolysin on interaction with any given transformed cell factor, and the spectrum of activating sera is distinctive for each cell factor. This pattern appears to be determined by the cell type, rather than by the transforming virus. An important role for the fibrinolysin in oncogenic transformation is suggested by the following correlations. (a) The initial appearance of fibrinolysin precedes the morphological change after the transfer to permissive temperatures of chick fibroblast cultures infected with a temperature-sensitive mutant of RSV. (b) The initiation of fibrinolysis and of morphological change both require the synthesis of new protein, but not the synthesis of either DNA or rRNA. (c) The activity of the fibrinolysin is correlated with the retention of abnormal morphology in hamster cells transformed by SV-40. (d) The sera of normal chicks effectively activate fibrinolysis with the cell factor from transformed chick cells. In contrast the sera of chicks with RSV tumors do not; these contain an inhibitor of the fibrinolytic activity.

An enzymatic function associated with transformation of fibroblasts by oncogenic viruses. I. Chick embryo fibroblast cultures transformed by avian RNA tumor viruses.

Chick embryo fibroblast cultures develop fibrinolytic activity after transformation by Rous sarcoma virus (RSV). This fibrinolytic activity is not present in normal cultures, and it does not appear after infection with either nontransforming strains of avian leukosis viruses or cytocidal RNA and DNA viruses. In cultures infected with a temperature sensitive mutant of RSV the onset of fibrinolysis appears after exposure to permissive temperatures and precedes by a short interval the appearance of morphological evidence of transformation. See PDF for Structure The rate of fibrinolysis in transformed cultures depends on the nature of the serum that is present in the growth medium: some sera (e.g., monkey or chicken serum) promote high enzymatic activity, while others (calf, fetal bovine) do not. Some sera contain inhibitors of the fibrinolysin. Based on the effect of a small number of known inhibitors, at least one step of the fibrinolytic process shows specificity resembling that of trypsin. The sera of sarcoma-bearing chickens contain an inhibitor of the fibrinolysin, whereas normal chicken sera do not. For general discussion, conclusions, and summary see the accompanying paper, part II, (J. Exp. Med. 137:112).

Inhibition of endothelial cell movement by pericytes and smooth muscle cells: activation of a latent transforming growth factor-beta 1-like molecule by plasmin during co-culture.

When a confluent monolayer of bovine aortic endothelial (BAE) cells is wounded with a razor blade, endothelial cells (ECs) spontaneously move into the denuded area. If bovine pericytes or smooth muscle cells (SMCs) are plated into the denuded area at low density, they block the movement of the ECs. This effect is dependent upon the number of cells plated into the wound area and contact between ECs and the plated cells. Antibodies to transforming growth factor-beta 1 (TGF-beta 1) abrogate the inhibition of BAE cell movement by pericytes or SMCs. TGF-beta 1, if added to wounded BAE cell monolayers, also inhibits cell movement. When cultured separately, BAE cells, pericytes, and SMCs each produce an inactive TGF-beta 1-like molecule which is activated in BAE cell-pericyte or BAE cell-SMC co-cultures. The activation appears to be mediated by plasmin as the inhibitory effect on cell movement in co-cultures of BAE cells and pericytes is blocked by the inclusion of inhibitors of plasmin in the culture medium.

Autocrine activities of basic fibroblast growth factor: regulation of endothelial cell movement, plasminogen activator synthesis, and DNA synthesis.

We have found that the spontaneous migration of bovine aortic endothelial cells from the edge of a denuded area in a confluent monolayer is dependent upon the release of endogenous basic fibroblast growth factor (bFGF). Cell movement is blocked by purified polyclonal rabbit IgG to bFGF as well as affinity purified anti-bFGF IgG and anti-bFGF F(ab')2 fragments. The inhibitory effect of the immunoglobulins is dependent upon antibody concentration, is reversible, is overcome by the addition of recombinant bFGF, and is removed by affinity chromatography of the antiserum through a column of bFGF-Sepharose. Cell movement is also reversibly inhibited by the addition of protamine sulfate and suramin; two agents reported to block bFGF binding to its receptor. The addition of recombinant bFGF to wounded monolayers accelerates the movement of cells into the denuded area. Transforming growth factor beta which has been shown to antagonize several other effects of bFGF also inhibits cell movement. The anti-bFGF IgG prevents the movement of bovine capillary endothelial cells, BHK-21, NIH 3T3, and human skin fibroblasts into a denuded area. Antibodies to bFGF, as well as suramin and protamine sulfate also suppress the basal levels of plasminogen activator and DNA synthesis in bovine aortic endothelial cells.

Production of plasminogen activator by established cell lines of mouse origin.

The correlation between malignant transformation and increased plasminogen activator synthesis has been studied in a variety of established cell lines. In contrast to the behavior of secondary mouse embryo cultures, which always show increased fibrinolytic activity when transformed, no such correlation was found within the BALB/c 3T3 line and its transformed derivatives. Cell lines were established from tumors initiated in BALB/c mice by several transformed cell lines. These lines were generally found to contain no more plasminogen activator than the cells used for inoculation. A correlation was found between transformation and plasminogen activator synthesis within Swiss 3T3 cell lines. However, the correlation was not maintained by serum revertants of transformed Swiss 3T3 cells.

Release of basic fibroblast growth factor-heparan sulfate complexes from endothelial cells by plasminogen activator-mediated proteolytic activity.

Cultured bovine capillary endothelial (BCE) cells synthesize heparan sulfate proteoglycans (HSPG), which are both secreted into the culture medium and deposited in the cell layer. The nonsoluble HSPGs can be isolated as two predominant species: a larger 800-kD HSPG, which is recovered from preparations of extracellular matrix, and a 250-kD HSPG, which is solubilized by nonionic detergent extraction of the cells. Both HSPG species bind bFGF. 125I-bFGF bound to BCE cell cultures is readily released by either heparinase or plasmin. When released by plasmin, the growth factor is recovered from the incubation medium as a complex with the partly degraded high molecular mass HSPG. Endogenous bFGF activity is released by a proteolytic treatment of cultured BCE cells. The bFGF-binding HSPGs are also released when cultures are incubated with the inactive proenzyme plasminogen. Under such experimental conditions, the release of the extracellular proteoglycans can be enhanced by treating the cells either with bFGF, which increases the plasminogen activating activity expressed by the cells, or decreased by treating the cells with transforming growth factor beta, which decreases the plasminogen activating activity of the cells. Specific immune antibodies raised against bovine urokinase also block the release of HSPG from BCE cell cultures. We propose that this plasminogen activator-mediated proteolysis provides a mechanism for the release of biologically active bFGF-HSPG complexes from the extracellular matrix and that bFGF release can be regulated by the balance between factors affecting the pericellular proteolytic activity.

Requirement for transglutaminase in the activation of latent transforming growth factor-beta in bovine endothelial cells.

A hitherto unknown function for transglutaminase (TGase; R-glutaminyl-peptide: amine gamma-glutamyltransferase, EC 2.3.2.13) was found in the conversion of latent transforming growth factor-beta (LTGF-beta) to active TGF-beta by bovine aortic endothelial cells (BAECs). The cell-associated, plasmin-mediated activation of LTGF-beta to TGF-beta induced either by treatment of BAECs with retinoids or by cocultures of BAECs and bovine smooth muscle cells (BSMCs) was blocked by seven different inhibitors of TGase as well as a neutralizing antibody to bovine endothelial cell type II TGase. Control experiments indicated that TGase inhibitors and/or a neutralizing antibody to TGase did not interfere with the direct action of TGF-beta, the release of LTGF-beta from cells, or the activation of LTGF-beta by plasmin or by transient acidification. After treatment with retinoids, BAECs expressed increased levels of TGase coordinate with the generation of TGF-beta, whereas BSMCs and bovine embryonic skin fibroblasts, which did not activate LTGF-beta after treatment with retinoids, did not. Furthermore, both TGase inhibitors and a neutralizing antibody to TGase potentiated the effect of retinol in enhancing plasminogen activator (PA) levels in cultures of BAECs by suppressing the TGF-beta-mediated enhancement of PA inhibitor-1 (PAI-1) expression. These results indicate that type II TGase is a component required for cell surface, plasmin-mediated LTGF-beta activation process and that increased expression of TGase accompanies retinoid-induced activation of LTGF-beta.

Lipoprotein (a) inhibits the generation of transforming growth factor beta: an endogenous inhibitor of smooth muscle cell migration.

Conditioned medium (CM) derived from co-cultures of bovine aortic endothelial cells (BAECs) and bovine smooth muscle cells (BSMCs) contains transforming growth factor-beta (TGF-beta) formed via a plasmin-dependent activation of latent TGF-beta (LTGF beta), which occurs in heterotypic but not in homotypic cultures (Sato, Y., and D. B. Rifkin. 1989. J. Cell Biol. 107: 1199-1205). The TGF-beta formed is able to block the migration of BSMCs or BAECs. We have found that the simultaneous addition to heterotypic culture medium of plasminogen and the atherogenic lipoprotein, lipoprotein (a) (Lp(a)), which contains plasminogen-like kringles, inhibits the activation of LTGF-beta in a dose-dependent manner. The inclusion of LDL in the culture medium did not show such an effect. Control experiments indicated that Lp(a) does not interfere with the basal level of cell migration, the activity of exogenous added TGF-beta, the release of LTGF-beta from cells, the activation of LTGF-beta either by plasmin or by transient acidification, or the activity of plasminogen activator. The addition of Lp(a) to the culture medium decreased the amount of plasmin found in BAECs/BSMCs cultures. Similar results were obtained using CM derived from cocultures of human umbilical vein endothelial cells and human foreskin fibroblasts. These results suggest that Lp(a) can inhibit the activation of LTGF-beta by competing with the binding of plasminogen to cell or matrix surfaces. Therefore, high plasma levels of Lp(a) might enhance smooth muscle cell migration by decreasing the levels of the migration inhibitor TGF-beta thus contributing to generation of the atheromatous lesions.

The opposing effects of basic fibroblast growth factor and transforming growth factor beta on the regulation of plasminogen activator activity in capillary endothelial cells.

Basic fibroblast growth factor (bFGF), a potent inducer of angiogenesis in vivo, stimulates the production of both urokinase- and tissue-type plasminogen activators (PAs) in cultured bovine capillary endothelial cells. The observed increase in proteolytic activity induced by bFGF was effectively diminished by picogram amounts of transforming growth factor beta (TGF beta), but could not be abolished by increasing the amount of TGF beta. However, the inhibition by TGF beta was greatly enhanced if the cells were pretreated with TGF beta before addition of bFGF. After prolonged incubation of cultures treated simultaneously with bFGF and TGF beta, the inhibitory effect of TGF beta diminished and the stimulatory effect of the added bFGF dominated as assayed by PA levels. TGF beta did not alter the receptor binding of labeled bFGF, nor did a 6-h pretreatment with TGF beta reduce the amount of bFGF bound. The major difference between the effects of bFGF and TGF beta was that while bFGF effectively enhanced PA activity expressed by the cells, TGF beta decreased the amounts of both cell-associated and secreted PA activity by decreasing enzyme production. Both bFGF and TGF beta increased the secretion of the endothelial-type plasminogen activator inhibitor.

Fibronectin from chicken embryo fibroblasts contains covalently bound phosphate.

Fibronectin isolated from cultures of chicken embryo fibroblasts (CEF) contains phosphorus linked to serine and threonine by monoester bonds. Normal and Rous sarcoma virus (RSV)-transformed cells were incubated with [32P]orthophosphate, and fibronectin was isolated from the cell surfaces and conditioned media. 32P was stably associated with fibronectin during immunoprecipitation, SDS-polyacrylamide gel electrophoresis, phospholipid solvent extraction, and hot acid but not alkaline treatment. After a limited acid hydrolysis of fibronectin, both phosphoserine and phosphothreonine were found. The specific radioactivity of the 32P-labeled fibronectin from the conditioned medium of normal CEF was higher than that from the cultures of transformed CEF.

Expression of the urokinase receptor in vascular endothelial cells is stimulated by basic fibroblast growth factor.

Basic fibroblast growth factor, a potent angiogenesis inducer, stimulates urokinase (uPA) production by vascular endothelial cells. In both basic fibroblast growth factor-stimulated and -nonstimulated bovine capillary endothelial and human umbilical vein endothelial cells single-chain uPA binding is mediated by a membrane protein with a Mr of 42,000. Exposure of bovine capillary or endothelial human umbilical vein endothelial cells to pmolar concentrations of basic fibroblast growth factor results in a dose-dependent, protein synthesis-dependent increase in the number of membrane receptors for uPA (19,500-187,000) and in a parallel decrease in their affinity (KD = 0.144-0.790 nM). With both cells, single-chain uPA binding is competed by synthetic peptides whose sequence corresponds to the receptor-binding sequence in the NH2-terminal domain of uPA. Exposure of bovine capillary endothelial cells to transforming growth factor beta 1, which inhibits uPA production and upregulates type 1 plasminogen activator inhibitor, the major endothelial cell plasminogen activator inhibitor, has no effect on uPA receptor levels. These results show that basic fibroblast growth factor, besides stimulating uPA production by vascular endothelial cells, also increases the production of receptors, which modulates their capacity to focalize this enzyme on the cell surface. This effect may be important in the degradative processes that occur during angiogenesis.

Correlation of cell migration, cell invasion, receptor number, proteinase production, and basic fibroblast growth factor levels in endothelial cells.

The levels of endogenous basic fibroblast growth factor (bFGF) in seven clones of cultured bovine capillary endothelial (BCE) cells were assayed, and their relation to cell morphology, bFGF receptor number, cell migration, amniotic membrane invasivity, and proteinase levels were studied. Immunoblotting experiments with anti-bFGF IgG demonstrated that cells from these clones contained different amounts of bFGF. The cells containing high levels of bFGF had a spindle or elongated appearance at confluence and a low number of high affinity receptors for bFGF. The cells containing low levels of bFGF had a cobblestone-like appearance and a higher number of high affinity receptors. When exposed to 10 ng/ml bFGF, cells containing a low level of bFGF took on an elongated appearance with a crisscross pattern similar to that seen with the high producer bFGF cells. The endogenous bFGF levels of the BCE cell clones correlated with the extent of cell migration after wounding of a monolayer and the degree of invasion of the human amniotic membrane. Cells from the clone with the highest endogenous bFGF level migrated well, invaded the amnion membrane without the addition of exogenous bFGF, and were relatively unaffected by the addition of bFGF. Cells from the clone containing the lowest level of bFGF did not migrate or invade under normal conditions. However, the addition of bFGF to the culture medium strongly enhanced both of these processes. The inclusion of anti-bFGF IgG in the media suppressed cell migration and invasion. The plasminogen activator (PA) activities of cell lysates of the clones, assayed by the 125I-fibrin plate technique, indicated that the PA levels did not correlate with the bFGF levels. Metalloproteinase activities in the conditioned medium, assayed by gelatin zymography, correlated with the endogenous bFGF levels, suggesting that the degree of expression of metalloproteinases might be critical for cell migration and invasion. These data suggest that endogenous bFGF may have an important role for migration and invasion of BCE cells during neovascularization via the induction and/or activation of specific metalloproteinases.

Patterns of plasminogen activator production in cultured normal embryonic cells.

Cultured normal low-passage embryo fibroblasts, from a number of species, and two untransformed clones of a Balb/3T3 line elaborate increasing amounts of plasminogen activator (PA) as they approach confluence; the low-passage cells then lose this PA activity after reaching confluence, while the 3T3 cells retain it indefinitely. Even at their peaks, however, the PA activities of the low-passage cells remain well below those of the corresponding virally or spontaneously transformed cells. The PA increases in normal cells are probably a result of PA production rather than of adsorption of secreted PA to the cell surface, or of changes in cell-associated protease inhibitors. The elaboration of PA by normal cells is dependent upon their metabolic activity, such that the level of serum supplementation and the growth phase of the culture directly influence the level of cell-associated PA observed. In addition, there may be a component of serum which exerts a negative control on PA production and which is not an acid-labile protease inhibitor.

Heparin and heparan sulfate increase the radius of diffusion and action of basic fibroblast growth factor.

The radius of diffusion of basic FGF (bFGF) in the presence and in the absence of the glycosaminoglycans heparin and heparan sulfate was measured. Iodinated 125I-bFGF diffuses further in agarose, fibrin, and on a monolayer of bovine aortic endothelial (BAE) cells in the presence of heparin than in its absence. Heparan sulfates affected the diffusion of 125I-bFGF in a manner similar to, though less pronounced than, heparin. When applied at the center of a monolayer of BAE cells, bFGF plus heparin stimulated morphological changes at a 10-fold greater radius than bFGF alone. These results suggest that bFGF-heparin and/or heparan sulfate complexes may be more effective than bFGF alone in stimulating cells located away from the bFGF source because the bFGF-glycosaminoglycan complex partitions into the soluble phase rather than binding to insoluble glycosaminoglycans in the extracellular matrix. Thus, the complex of bFGF and glycosaminoglycan may represent one of the active forms of bFGF in vivo.

In vitro angiogenesis on the human amniotic membrane: requirement for basic fibroblast growth factor-induced proteinases.

The role of basic fibroblast growth factor-(bFGF) induced proteinases in basement membrane (BM) invasion by bovine capillary endothelial (BCE) cells was studied using a quantitative in vitro assay previously described (Mignatti et al., 1986). 125I-iododeoxyuridine-labeled BCE cells were grown for 72 h on the human amnion BM, and cell invasion was determined by measuring the radioactivity associated with the tissue after removal of the noninvasive cell layer. BCE cells were noninvasive under normal conditions. Addition of human bFGF to either the BM or to the stromal aspect of the amnion induced BCE cell invasion with a dose-dependent response. This effect was maximal in the presence of 70 ng/ml bFGF, and was inhibited by anti-FGF antibody. Transforming growth factor beta, as well as plasmin inhibitors and anti-tissue type plasminogen activator antibody inhibited BCE cell invasion. The tissue inhibitor of metalloproteinases, 1-10 phenanthroline, anti-type IV and anti-interstitial collagenase antibodies had the same effect. On the contrary, anti-stromelysin antibody and Eglin, an inhibitor of elastase, were ineffective. The results obtained show that both the plasminogen activator-plasmin system and specific collagenases are involved in the invasive process occurring during angiogenesis.

Basic fibroblast growth factor-induced activation of latent transforming growth factor beta in endothelial cells: regulation of plasminogen activator activity.

Exposure of bovine aortic or capillary endothelial cells to basic FGF (bFGF) for 1 h resulted in an approximately sixfold increase in plasminogen activator (PA) activity by 18 h that returned nearly to basal levels by 36 h. We hypothesized that the decrease in PA activity following bFGF stimulation was mediated by transforming growth factor beta (TGF-beta) formed from its inactive precursor. Conditioned medium collected from endothelial cells 36 h after a 1-h exposure to bFGF, but not control medium, inhibited basal levels of PA activity when transferred to confluent monolayers of bovine aortic endothelial cells. Antibody to TGF-beta neutralized the inhibitory activity of this conditioned medium, indicating that the medium contained active TGF-beta. Northern blot analysis and quantitation of acid activatable latent TGF-beta in conditioned medium demonstrated that bFGF exposure did not increase the amount of transcription or secretion of latent TGF-beta by the endothelial cells. Both aprotinin, an inhibitor of plasmin, and anti-urokinase type PA IgG blocked the generation of active TGF-beta in cultures exposed to bFGF. These results demonstrated that plasmin generated by uPA activity is required for the activation of latent TGF-beta in endothelial cell cultures treated with bFGF. Activation of TGF-beta by endothelial cells exposed to bFGF appears to limit both the degree and duration of PA stimulation. Thus, in bFGF-stimulated endothelial cell cultures, PA levels are controlled by a negative feedback loop: PA, whose expression is stimulated by bFGF, contributes to the formation of TGF-beta, which in turn opposes the effects of bFGF by limiting PA synthesis and activity. These studies suggest a role for TGF-beta in reversing the invasive stage of angiogenesis and contributing to the formation of quiescent capillaries.

Endothelial cell-derived heparan sulfate binds basic fibroblast growth factor and protects it from proteolytic degradation.

Cultured bovine capillary endothelial (BCE) cells were found to synthesize and secrete high molecular mass heparan sulfate proteoglycans and glycosaminoglycans, which bound basic fibroblast growth factor (bFGF). The secreted heparan sulfate molecules were purified by DEAE cellulose chromatography, followed by Sepharose 4B chromatography and affinity chromatography on immobilized bFGF. Most of the heparinase-sensitive sulfated molecules secreted into the medium by BCE cells bound to immobilized bFGF at low salt concentrations. However, elution from bFGF with increasing salt concentrations demonstrated varying affinities for bFGF among the secreted heparan sulfate molecules, with part of the heparan sulfate requiring NaCl concentrations between 1.0 and 1.5 M for elution. Cell extracts prepared from BCE cells also contained a bFGF-binding heparan sulfate proteoglycan, which could be released from the intact cells by a short proteinase treatment. The purified bFGF-binding heparan sulfate competed with 125I-bFGF for binding to low-affinity binding sites but not to high-affinity sites on the cells. Heparan sulfate did not interfere with bFGF stimulation of plasminogen activator activity in BCE cells in agreement with its lack of effect on binding of 125I-bFGF to high-affinity sites. Soluble bFGF was readily degraded by plasmin, whereas bFGF bound to heparan sulfate was protected from proteolytic degradation. Treatment of the heparan sulfate with heparinase before addition of plasmin abolished the protection and resulted in degradation of bFGF by the added proteinase. The results suggest that heparan sulfate released either directly by cells or through proteolytic degradation of their extracellular milieu may act as carrier for bFGF and facilitate the diffusion of locally produced growth factor by competing with its binding to surrounding matrix structures. Simultaneously, the secreted heparan sulfate glycosaminoglycans protect the growth factor from proteolytic degradation by extracellular proteinases, which are abundant at sites of neovascularization or cell invasion.