| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Departments of Pediatrics [C. I., W. E. L., M. F. S., D. L. D., A. E-E., Y. A. D.], Pathology [H. S.], and Biochemistry and Molecular Biology [Y. A. D.], Childrens Hospital Los Angeles and Keck School of Medicine, University of Southern California, Los Angeles, California 90027, and Laboratory for Pharmaceutical Biology and Phytopharmacology, Faculty of Pharmaceutical Sciences, Katholieke Universiteit Leuven, 3000 Leuven, Belgium [P. J. D.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
vß3 integrin in blood vessels of primary neuroblastoma tumors, suggesting that PAI-1 plays a role in angiogenesis. Using human brain microvascular endothelial cells (HBMECs), we found that PAI-1 inhibits vß3 integrin-mediated cell adhesion to vitronectin but promotes 5ß1-mediated migration from vitronectin toward fibronectin. Inhibition of vitronectin adhesion by PAI-1 did not induce HBMEC apoptosis. PAI-1 also inhibited endothelial tube formation on Matrigel in the presence of vitronectin but had a stimulatory effect in the presence of fibronectin. This effect of PAI-1 on microvascular endothelial cells is primarily related to the ability of PAI-1 to bind to vitronectin via its NH2-terminal domain and to interfere with cell adhesion to vitronectin. We propose that PAI-1 acts as a positive switch for angiogenesis by promoting endothelial cell migration away from their vitronectin-containing perivascular space toward fibronectin-rich tumor tissue. These observations provide a novel explanation for the enhancing effect of PAI-1 in cancer progression. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
vß3 integrin (6)
. By interfering with cell adhesion to vitronectin, PAI-1 represses integrin-mediated cell migration on vitronectin (7, 8, 9)
, suggesting an inhibitory rather than a stimulatory role in angiogenesis. This seemingly contradictory role in tumor progression is presently not fully understood. Elevated levels of PAI-1 have been reported by us in primary tumors of progressive neuroblastoma, and immunohistochemical examination showed that PAI-1 was strictly localized to the vascular endothelial cells, suggesting a possible involvement in angiogenesis (10) . Here we have tested this hypothesis and provide evidence for a mechanism supporting a positive function for PAI-1 in angiogenesis and a key role of fibronectin in this process.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
v antagonist EMD 121974 cyclo (Arg-Gly-Asp-D-Phe-[N-Met]-Val) was provided by A. Jon
zyk and S. L. Goodman (Merck KGaA, Darmstadt, Germany).
Tumor Specimens and Immunohistochemistry.
Primary tumor specimens of human neuroblastoma were obtained from institutions of the Children’s Cancer Group (Arcadia, CA) between 1985 and 1992 as described previously (12)
. Immunohistochemistry was performed on 6-µm-thick serial cryostat sections prepared from frozen tumor tissues preserved in OCT or on paraffin-embedded tissue sections. Frozen sections were immediately fixed in acetone for 5 min and air dried. Tissue for paraffin embedding was fixed in 4% paraformaldehyde, dehydrated through a graded series of ethanol washes, cleared with xylene, and embedded. Sections were then dewaxed in xylene and rehydrated through an ethanol series of decreased concentrations for 2 min each and rinsed with PBS. The tissue sections were exposed for 10 min to 2% goat serum in Tris-buffered saline (10 mM Tris-HCl, pH 7.6, 150 mM NaCl), washed, and incubated with the primary antibodies overnight at 4°C. The following primary antibodies were used: a rabbit antihuman PAI-1 polyclonal antibody (Molecular Innovations, Inc., Royal Oak, MI) at a 1:50 dilution, a mouse antihuman vß3 monoclonal antibody (LM609; Chemicon Int., Temecula, CA) at a 1:50 dilution, a mouse antihuman CD31 monoclonal antibody (DAKO, Carpinteria, CA), a mouse antihuman vitronectin monoclonal antibody (Biogenesis, Kingston, New Hampshire) at a 1:1000 dilution, and a rabbit antihuman fibronectin polyclonal antibody (Sigma Chemical Co., St. Louis, MI) at a 1:400 dilution. Slides were then washed three times in Tris-buffered saline prior to incubation with the secondary antibody [multi-link swine antigoat, mouse, rabbit immunoglobulins (DAKO) at a 1:50 dilution] for 30 min at room temperature. The sections were further processed for avidin-biotin horseradish peroxidase reaction using the Vectastain ABC kit (PK 400; Vector Laboratories Inc., Burlingame, CA) for 30 min. After washing three times, the reaction was developed in the presence of 3,3'-diaminobenzidine tetrahydrochloride (0.4 mg/ml). The slides were counterstained with Mayer’s hematoxylin, dehydrated, and mounted. A mouse IgG1 monoclonal antibody against Aspergillus niger glucose oxidase (DAKO) was used as a negative control.
Immunofluorescence Staining.
Cryostat sections were prepared as described above and incubated with a mouse monoclonal antibody against vß3 (LM609) overnight at 4°C. After washing with Tris-buffered saline, the sections were incubated with a rabbit antihuman PAI-1 antibody for 2 h at room temperature. The sections were washed and incubated with a mixture of Rhodamine-labeled antimouse IgG and FITC-labeled antirabbit IgG (Jackson Immuno Research Laboratories, Inc., West Grove, PA) for 60 min at room temperature. The stained slides were photographed under a fluorescence microscope (Vanox-S, AH-2, Olympus).
Cell Culture, Adhesion, and Migration Assays.
HBMECs were isolated from surgically resected brain tissue of children with seizure disorders after appropriate informed consent was obtained from the family as described previously (13)
. Cells used were between passages 8 and 10 and maintained in RPMI 1640 containing 10% fetal bovine serum supplemented with epidermal and endothelial growth factors, insulin, and transferrin (NuSerum IV; Becton Dickinson, Bedford, MA) and containing 2 mM
L-glutamine, 1 mM sodium pyruvate, 1% nonessential amino acids and penicillin (50 units/ml), and streptomycin (50 µg/ml). Cells were always split and maintained in subconfluent culture for experiments. Forty-eight well, non-tissue culture treated cluster plates were coated with 150 µl/well of native vitronectin (Promega Corp., Madison, WI; 5 µg/ml) or fibronectin (Promega; 10 µg/ml) at 37°C for 1 h. Nonspecific binding was blocked by the addition of 1% heat-treated BSA for 30 min at room temperature. When indicated, wells were incubated with stable recombinant PAI-1 for 30 min prior to the addition of cells. Cells (5 x 104/well) harvested by brief (1-min) treatment with trypsin, followed by neutralization with serum-containing medium, were washed with PBS (137 mM NaCl, 2.6 mM KCl, 10 mM Na2HPO4, and 1.7 mM KH2PO4, pH 7.4) and resuspended in 200 µl of adhesion buffer (serum-free culture medium containing 0.5% BSA, 1 mM CaCl2, 1 mM MgCl2, and 0.2 mM MnCl2). The cells were plated in the wells and allowed to attach for 1 h. Cells were then washed gently three times with adhesion buffer to remove nonadherent cells, and the adherent cells were stained with a 0.1% crystal violet solution at room temperature for 2 min. After washing three times, the insoluble dye taken up by adherent cells was dissolved in 200 µl of methanol, and the absorbance of the solution was read at 595 nm in an ELISA plate reader. Cell migration assays were performed in Transwell plates equipped with a 12-µm pore size polycarbonate filter (Costar, Cambridge, MA). The upper side of the filter was coated with vitronectin (150 ng/filter), and the lower side with vitronectin (750 ng/filter) or fibronectin (1.5 µg/filter) diluted in PBS. After incubation for 1 h at 37°C in the presence of each protein, the filters were treated with 1% heat-treated BSA in PBS to prevent nonspecific binding and washed twice with PBS. Cells suspended in serum-free medium were then placed in the upper chamber (5 x 104 cells/chamber) and incubated at 37°C in a humidified atmosphere of 5% CO2 in air. After 20 h, the cells on the upper side of the filter were removed with a cotton swab, and cells on the lower side of the filter were fixed and stained with Diff-Quik Stain Set (Dade Behring, Newark, DE). The cells on the lower side of the filter were counted in nine randomly selected fields using a x20 objective.
FACS Analysis.
Exponentially growing HBMECs were harvested in dissociation buffer (enzyme free/Hanks based; Life Technologies, Inc.), centrifuged, and resuspended on ice at a concentration of 106 cells/ml in PBS supplemented with 0.1% BSA. Cells were then incubated with a primary antibody or a nonspecific IgG at a final concentration of 1 µg/100 ml for 30 min at 4°C. The cells were then washed three times with PBS containing 0.1% BSA, resuspended in 50 µl of PBS/0.1% BSA, and incubated with a FITC-conjugated antimouse IgG antibody in the dark at 4°C for 20 min. After washing three times with PBS/0.1% BSA, the cells were fixed in 4% paraformaldehyde and kept at 4°C in the dark before analysis. FACS analysis was done using a Coulter Epics Elite ESP instrument. For data analysis, we used the Coulter Expo2 software provided by the manufacturer. The primary antibodies used were: a mouse antihuman vß3 integrin monoclonal antibody (LM609); a mouse antihuman vß5 integrin monoclonal antibody (P1F6; Chemicon Int.); a mouse antihuman 5ß1 integrin monoclonal antibody (HA5 and JBS5; Chemicon Int.); a mouse antihuman ß1 (activated conformation) monoclonal antibody (HUTS-4; Chemicon Int.); a mouse antihuman uPAR monoclonal antibody (American Diagnostica); a mouse antihuman von Willebrand factor (A0082; DAKO) monoclonal antibody; a mouse antihuman fibroblast growth factor receptor monoclonal antibody (VBS1; Chemicon Int.); a mouse anti-Flk-1 monoclonal antibody (CH-11; Chemicon Int.); and a mouse antihuman Flt-1 monoclonal antibody (C-17; Santa Cruz Biotechnology, Santa Cruz, CA).
Treatment with PI-PLC.
Treatment with PI-PLC (Sigma Chemical Co.) was used to remove cell surface-associated uPAR. For these experiments, cells were harvested in trypsin-EDTA, followed by neutralization with serum-containing medium. The cells were then washed twice in PBS and resuspended at 5 x 106 cells/ml in RPMI containing 0.1% BSA. PI-PLC was added at a final concentration of 10 units/ml, and cells were further incubated for 90 min at 4°C. Cells were then centrifuged and washed twice in PBS/0.1% BSA. An aliquot of the cell suspension was used to verify the expression of uPAR by FACS analysis as above described, whereas another aliquot was used for adhesion and migration assays. Cells similarly incubated without the addition of PI-PLC were used as control.
Apoptosis.
HBMECs were allowed to attach to protein-coated wells used for adhesion assay. After incubation, nonadherent cells were harvested, and adherent cells were collected by trypsinization and added to nonadherent cells. Cytospin preparations of cells were made and fixed in 4% paraformaldehyde for 1 h at room temperature. Apoptotic cells were evaluated using the TUNEL detection kit (Boehringer Mannheim, Indianapolis, IN). Cells were then lightly counterstained with hematoxylin, and the number of TUNEL-positive cells was counted by light microscopy in 10 random fields under a x40 objective.
Western Blot.
Cells were lysed in buffer containing 1% NP40, 0.1% SDS, and a freshly added mixture of protease inhibitors (Roche, Mannheim, Germany). After incubation for 30 min and centrifugation, the protein concentration in the supernatant was measured by a Protein Quantification kit (Bio-Rad). Forty µg of each sample were electrophoresed in 0.1% SDS-12.5% polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were blocked in Tris-buffered saline containing 0.05% Tween 20 and 5% nonfat dry milk at 37°C for 1 h and incubated overnight at 4°C with a rabbit polyclonal anti-Bcl-XL antibody (Santa Cruz Biotechnology) or a rabbit polyclonal anti-Bax antibody (PharMingen, San Diego, CA) using dilutions recommended by the manufacturer. To ensure that equal amounts of proteins were loaded and blotted, the membranes were stripped and reprobed with a mouse antihuman ß-tubulin monoclonal antibody (Sigma Chemical Co.).
Endothelial Tube Assay.
Matrigel (Becton Dickinson) was added (300 µl) to each well of a 24-well plate and allowed to polymerize. This Matrigel was found to be free of vitronectin and/or fibronectin by Western blot analysis (data not shown). When indicated, vitronectin (5 µg/ml) and/or fibronectin (10 µg/ml) were added to the Matrigel before polymerization. A suspension of 2 x 104 HBMECs in culture medium was then added into each coated well in the presence or absence of rPAI-1 (100 nM) or Q123K PAI-1 mutant (100 nM). Cells were incubated for 24 h at 37°C, viewed using a Nikon TMS microscope (Tokyo, Japan), and photographed using Matrox Inspector software. The number of tubes were counted in 10 fields and averaged.
Statistical Analysis.
The statistical analysis of the results was done using Student’s t test. P < 0.05 was considered significant.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
vß3 integrin, a specific marker for angiogenesis (14)
. We observed that PAI-1 and vß3 colocalized in blood vessels of primary neuroblastoma tumor specimens (Fig. 1, A-C)
. Furthermore, in a total of 18 primary tumors examined, there was a direct correlation between the proportion of blood vessels expressing PAI-1 and those expressing vß3 (Fig. 1D)
. We also examined these tumors for the presence of vitronectin and fibronectin by immunohistochemistry. Whereas vitronectin was specifically present in endothelial cells and around microvessels (Fig. 1E)
, fibronectin was abundantly present in the tumor tissue and in septae of connective tissues that separated nests of tumor cells (Fig. 1, F and G)
.
|
5ß1) and vitronectin (vß3 and vß5; Table 1
). We next verified that these integrins were involved in the adhesion of HBMECs to vitronectin and fibronectin by testing cell adhesion in the absence and presence of integrin-specific blocking monoclonal antibodies. The data (Fig. 2)
show that adhesion of HBMECs to vitronectin was specifically inhibited in the presence of an anti-vß3 blocking antibody (LM609) and to a lesser extent by an anti-vß5 (P1F6) blocking antibody and not in the presence of an anti-5ß1 (JBS5) blocking antibody. In contrast, adhesion to fibronectin was not affected by anti-vß3 and anti-vß5 antibodies but was inhibited by the anti-5ß1 antibody. The data therefore indicate that under the established culture conditions, these microvascular cells closely mimic angiogenic endothelial cells and adhere to vitronectin and fibronectin via specific integrins.
|
|
. The data show a 76 and 77% inhibition of adhesion to vitronectin in the presence of 40 and 100 nM of rPAI-1, respectively, and no effect on adhesion to fibronectin. When the amount of vitronectin in the well was lowered from 750 µg/well to 400 ng/well, lower concentrations of rPAI-1 (up to 1 nM) inhibited HBMEC adhesion to vitronectin (Fig. 3B)
. This is consistent with the requirement of rPAI-1 to occupy cell binding sites on vitronectin to inhibit adhesion. Cell binding to vitronectin is mediated by at least two types of cell surface receptors, integrins of the v family and uPAR (6)
. Because both receptors were present on the surface of HBMECs (Table 1)
, we asked which one of these two receptors was primarily affected by PAI-1 (Fig. 3C)
. We first tested whether removal of uPAR on the cell surface by treatment with PI-PLC would affect vitronectin binding in the absence or presence of rPAI-1. Treatment of HBMECs with PI-PLC reduced the level of uPAR expression from 82 to 17% (Fig. 3C
, inset). This treatment had no effect on basal adhesion of HBMECs to vitronectin and did not alter the inhibitory effect of PAI-1 on cell binding to vitronectin. Next, we observed that addition of a function-blocking monoclonal antibody against uPAR did not inhibit binding to vitronectin of cells treated or untreated with PI-PLC. Then, we found that an v antagonist cyclic penta-peptide (EMD 121974) had a strong inhibitory effect on HBMECs binding to vitronectin, independent of PI-PLC treatment. Altogether, the data indicate that the binding of HBMECs to vitronectin is primarily mediated by integrins and that PAI-1 specifically interferes with this binding.
|
vß3 integrin-mediated endothelial cell binding to vitronectin (16
, 17)
. Considering the inhibitory effect of PAI-1 on vß3-mediated HBMEC binding to vitronectin, a negative rather than a positive role for PAI-1 on angiogenesis would be predicted. To resolve this paradox, we designed a series of experiments to compare the effects of PAI-1 and the v antagonist EMD 121974 on HBMEC adhesion and apoptosis. We first asked whether these two inhibitors of vitronectin adhesion would have a similar inhibitory activity when added prior to or after cell contact with vitronectin, because differences in inhibition of Hep-2 human epidermoid carcinoma-derived cell adhesion to vitronectin by PAI-1 and RGD peptides have been reported previously (18)
. The data (Fig. 4A)
indicated that PAI-1 and EMD 121974 effectively inhibited the attachment of HBMECs to vitronectin when they were added to vitronectin-coated wells prior to plating the endothelial cells. However, when PAI-1 was added after the endothelial cells were plated on vitronectin, no inhibition was seen, and the cells did not detach from the substrate. In contrast, addition of the EMD 121974 cyclic peptide to HBMECs attached to vitronectin resulted in a rapid loss of adhesion, even when the pentapeptide was added 3 h after cells had adhered to vitronectin. In a second set of experiments, we compared the effect of PAI-1 and EMD 121974 on endothelial cell apoptosis (Fig. 4, B and C)
. Whether added before or after cell plating, PAI-1 had no effect on endothelial cell apoptosis up to 24 h after cells were plated. In contrast, the percentage of cells going into apoptosis significantly increased 16 and 24 h after plating the cells in the presence of the EMD 121974 cyclic peptide. This increase in apoptosis was observed independently, whether the v antagonist was added before or after the cells were in contact with vitronectin. Consistently, rPAI-1 had no effect on the expression of the antiapoptotic protein Bcl-XL. In contrast, in the presence of the EMD 121974 cyclic peptide, a decrease in the expression of Bcl-XL was observed after 24 h. Neither EMD 121974 nor PAI-1 had any effect on the expression of Bax. Bcl-2 was not detectable by Western blot in HBMECs and was not induced in response to PAI-1 or EMD 121974 (not shown). Thus, the data point to a significant difference between the two inhibitors of vitronectin binding, with regard to their ability to detach cells from vitronectin and to induce apoptosis.
|
. The stimulatory effect of PAI-1 on migration toward fibronectin was specific because it was not observed when the lower side of the filter was coated with other ECM proteins including laminin and types I and IV collagens (data not shown). Basal migration from vitronectin toward fibronectin in the absence of rPAI-1 was also stimulated in the presence of an anti-vß3 blocking antibody (LM609), confirming that inhibition of integrin-mediated vitronectin binding is essential to promote migration toward fibronectin (Fig. 5B)
. Basal migration toward fibronectin was, however, inhibited in the presence of an anti-5ß1 blocking antibody (JBS5), indicating the involvement of the fibronectin binding integrin 5ß1 in fibronectin haptotaxis. The addition of an anti-vß3 antibody to rPAI-1 had no additional stimulatory effect on HBMEC migration from vitronectin toward fibronectin, whereas the addition of an anti-5ß1 antibody abolished the stimulatory effect of rPAI-1 on migration. To confirm that the primary role of PAI-1 in promoting HBMEC migration from vitronectin toward fibronectin involves its vitronectin binding inhibitory function and does not involve changes in 5ß1 expression or activation, we examined the effect of PAI-1 on the expression of 5ß1 and on the active confirmation of ß1 in HBMECs by FACS analysis. The data (Table 2)
indicated no difference in expression of 5ß1 or active ß1 by PAI-1.
|
|
. Removal of uPAR by treatment of HBMECs with PI-PLC or blocking uPAR with an antibody did not affect basal endothelial cell migration toward vitronectin or fibronectin, indicating therefore an absence of involvement of uPAR in HBMEC migration. Furthermore, treatment of HBMECs with PI-PLC did not modify the inhibitory activity of PAI-1 on migration on vitronectin nor its stimulatory activity on migration toward fibronectin. To confirm that the effect of PAI-1 on HBMEC migration involves its vitronectin binding inhibitory function and to test whether the antiproteolytic function of PAI-1 is required, we tested the effect of two rPAI-1 mutants on HBMEC migration from vitronectin to fibronectin. These experiments indicated that a mutant PAI-1 (Q123K) that lacks vitronectin binding activity had lost its effect on HBMEC migration. Conversely a mutant PAI-1 (R346A) that lacks antiproteolytic activity but retains vitronectin binding capacity had an activity similar to the wild-type protein. Altogether the data indicate that the stimulatory effect of PAI-1 on HBMEC migration is primarily related to its inhibitory function on integrin-mediated vitronectin binding and does not require its antiproteolytic activity.
The inability of PAI-1 to detach endothelial cells already bound to vitronectin (Fig. 4A)
suggests that it cannot compete with existing vß3-vitronectin binding and that it can only bind to unoccupied sites on vitronectin. We therefore asked whether PAI-1 would have any effect on endothelial cell migration when added after the cells were attached to vitronectin (Fig. 5D)
. Under these conditions, PAI-1 retained its ability to inhibit migration toward vitronectin and to stimulate migration toward fibronectin. In contrast, the delayed addition of the EMD 121974 cyclic-peptide inhibited migration under both conditions. Because endothelial cells need to form new contacts with the ECM at their leading edge to advance, these experiments suggest that occupation of cell-binding sites on vitronectin by PAI-1 prevents further engagement with vitronectin and favors engagement with fibronectin.
Effect of PAI-1 on Endothelial Cell Tube Formation.
To demonstrate that the effect of PAI-1 on HBMEC migration is responsible for its stimulatory role on angiogenesis, we examined the effect of PAI-1 on the formation of tubes by HBMECs plated on Matrigel supplemented with vitronectin and/or fibronectin. The data (Fig. 6)
reveal a remarkable difference in the effect of PAI-1 on tube formation between vitronectin and/or fibronectin. Whereas neither vitronectin nor fibronectin altered tube formation in the absence of PAI-1, the addition of PAI-1 inhibited tube formation in vitronectin-supplemented Matrigel but stimulated it in fibronectin-supplemented Matrigel. Consistent with our hypothesis, the inhibitory effect of PAI-1 on tube formation in vitronectin-supplemented Matrigel was converted to a stimulatory effect when fibronectin was simultaneously added to vitronectin in the Matrigel. This effect required the vitronectin-binding capacity of PAI-1 as demonstrated by an absence of effect with the stable Q123K PAI-1 mutant. This stimulatory effect on tube formation in fibronectin-supplemented Matrigel did not involve growth stimulation because PAI-1 had no effect on [3H]thymidine incorporation in HBMECs grown on plastic dishes or on fibronectin-coated dishes (data not shown). Thus altogether, these experiments support a positive role for PAI-1 in angiogenesis.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Here we have examined whether the proangiogenic action of PAI-1 involves control of specific proteolysis or a modulation of cell migration, and we show that PAI-1 acts as a stimulator of angiogenesis by promoting the migration of endothelial cells from vitronectin toward fibronectin without triggering apoptosis. Our findings resolve apparently conflicting data indicating that PAI-1 can inhibit as well as promote cell migration (8
, 9 , 23)
by demonstrating that the effect of PAI-1 on endothelial cell migration is dependent on the composition of the extracellular matrix. Although PAI-1 inhibits migration on vitronectin, it stimulates migration from vitronectin toward fibronectin. This effect is primarily related to the inhibitory function of PAI-1 on vitronectin binding and does not involve changes in 5ß1 expression or activation. Further supporting the importance of fibronectin is our observation that the addition of fibronectin reverts the inhibitory effect of PAI-1 on endothelial cell tube formation in vitronectin-supplemented Matrigel to a stimulatory effect. In view of the abundance of fibronectin within tumor tissues and the more selective presence of vitronectin in close proximity to endothelial cells, as illustrated in neuroblastoma, this modulatory effect of PAI-1 is likely critical. It allows endothelial cells to escape the vitronectin-rich environment of their perivascular space and to penetrate the unvascularized and fibronectin rich tumor stroma. Accordingly, it has been shown previously that fibronectin promotes the elongation of microvessels during angiogenesis (24)
. Fibronectin-deficient mice have also been reported to have a poor development of endothelial structures (25)
.
Our findings also demonstrate that the modulatory effect of PAI-1 on endothelial cell migration involves an effect on cell attachment to vitronectin and not a control over plasminogen-mediated proteolysis. Mutation of the glutamine residue at position 123 to lysine in PAI-1 suppressed its capacity to bind to vitronectin and resulted in a loss of stimulatory effect on migration toward fibronectin and tube formation in fibronectin-supplemented Matrigel. In contrast, a mutation that suppressed the antiproteolytic activity of the inhibitor (R346A) maintained the stimulatory activity of PAI-1 on endothelial cell migration. Interference with vß3-mediated adhesion to vitronectin is therefore required to promote endothelial cell migration toward fibronectin, whereas control of plasminogen activator-mediated proteolysis is not. However, in vivo, modulation of proteolysis by PAI-1 may still play an additional regulatory function, and therefore our data do not necessarily rule out a possible contribution of the antiproteolytic function of PAI-1 in angiogenesis.
Our finding that PAI-1 affects the interaction between vß3 and vitronectin in endothelial cells without triggering apoptosis is also significant because inhibition of this interaction by anti-vß3 antibodies or RGD-based peptides has been shown to induce endothelial cell apoptosis and to inhibit angiogenesis (15, 16, 17)
. This absence of effect of PAI-1 on apoptosis may be attributable to an absence of anoikis because our data demonstrate that an important difference between PAI-1 and the v antagonist cyclic peptide EMD 121974 is that PAI-1 does detach endothelial cells from vitronectin. However, we show that the addition of PAI-1 to vitronectin before endothelial cells are in contact with vitronectin inhibits adhesion without inducing apoptosis. Thus, lack of adhesion alone does not explain the absence of apoptosis in the presence of PAI-1. Interference with vß3-mediated contact with vitronectin is achieved in the case of PAI-1 by binding to the ligand (vitronectin) and in the case of EMD 121974 by binding to the receptor (vß3). Thus, the fact that PAI-1 does not interfere with the receptor may be responsible for the lack of a proapoptotic signal, as suggested by the absence of changes in the level of Bcl-XL and Bax in endothelial cells exposed to PAI-1. An additional direct protective effect of PAI-1 on induced apoptosis as demonstrated recently in tumor cells and endothelial cells cannot be eliminated at present (26)
. In contrast, upon exposure to EMD 121974, HBMECs undergo apoptosis, and Bcl-XL levels decrease. The decrease, however, was only detected after 24 h and not at 16 h, when a significant increase in apoptosis was already noticed, suggesting that other mechanisms may be involved. We have, for example, shown previously that treatment of bovine brain endothelial cells with an RGD function-blocking cyclic peptide was associated with a 50% increase in endogenous ceramide, a lipid second messenger that can mediate apoptosis (12)
. Whether a similar pathway is involved in the effect of EMD 121974 on cell apoptosis is being investigated.5
As this report was in preparation, Stefansson et al. (27) reported recently that PAI-1 inhibits VEGF-induced angiogenesis in the chicken chorioallantoid membrane assay. This effect was attributed to inhibition of endothelial cell vitronectin binding. The data do not contradict our observation which also indicates that PAI-1 inhibits endothelial cell adhesion and migration on vitronectin. However, these investigators have not explored the presence and distribution of fibronectin in their model.
In summary, our data provide a mechanism to explain the essential role of stromal-derived PAI-1 in angiogenesis and tumor invasion reported by Bajou et al. (4) and to support the paradoxical clinical finding made by many investigators, including our group, that there is a strong relationship between the levels of PAI-1 in cancer and a poor clinical outcome. PAI-1 is increasingly expressed during tissue remodeling and wound healing, where it maintains the integrity of the fibrin matrix required for cellular activities such as migration (28) . Our observation suggests an additional role for PAI-1 as a stimulator of neovascularization in wounded tissues. Our studies also suggest that interfering with the capacity of PAI-1 to bind to vitronectin may suppress the stimulatory effect of PAI-1 on angiogenesis in vivo (11 , 29) , which might be of future therapeutic value.
|
ACKNOWLEDGMENTS |
|---|
zyk and S. L. Goodman for the gift of the cyclic peptide EMD 121974, J. Rosenberg for typing the manuscript, and Dr. Z. Werb (University of California-San Francisco, San Francisco, CA) for critical review of the manuscript. |
FOOTNOTES |
|---|
1 This work was supported by Grants CA84103 and CA82989 from the NIH and by a grant from the T. J. Martell Foundation. 
2 Present address: Herman B. Wells Center for Pediatric Research, Cancer Research Institute, Indiana University School of Medicine, Indianapolis, IN 46202.
3 To whom requests for reprints should be addressed, at Childrens Hospital Los Angeles, 4650 Sunset Boulevard, MS #54, Los Angeles, CA 90027. Phone: (323) 669-2150; Fax: (323) 664-9455; E-mail: declerck@hsc.usc.edu.
4 The abbreviations used are: PAI-1, plasminogen activator inhibitor-1; rPAI-1, recombinant PAI-1; HBMEC, human brain microvascular endothelial cell; uPAR, urokinase plasminogen activator receptor; FACS, fluorescence-activated cell sorter; ECM, extracellular matrix; PI-PLC, phosphatidylinositol-specific phospholipase C; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling; VEGF, vascular endothelial growth factor.
5 A. Suzuki, C. Isogai, I. Gonzalez-Gomez, F. H. Gilles, Y. A. DeClerck, and W. E. Laug. Apoptosis induced by a cyclic v antagonist pentapeptide, comparison between dermal and brain microvascular endothelial cells, manuscript in preparation.
Received 11/27/00. Accepted 5/16/01.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Vß3 binding to vitronectin. Nature (Lond.), 383: 441-443, 1996.[Medline]
vß3 and vß5 are expressed by endothelium of high-risk neuroblastoma and their inhibition is associated with increased endogenous ceramide. Cancer Res., 60: 712-721, 2000.vß3 for angiogenesis. Science (Wash. DC), 264: 569-571, 1994.vß3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. Cell, 79: 1157-1164, 1994.[Medline]
vß3 blocks human breast cancer growth and angiogenesis in human skin. J. Clin. Investig., 96: 1815-1822, 1995.
This article has been cited by other articles:
|
|
 |
|
  M.-T. Lin, I-H. Kuo, C.-C. Chang, C.-Y. Chu, H.-Y. Chen, B.-R. Lin, M. Sureshbabu, H.-J. Shih, and M.-L. Kuo Involvement of Hypoxia-inducing Factor-1{alpha}-dependent Plasminogen Activator Inhibitor-1 Up-regulation in Cyr61/CCN1-induced Gastric Cancer Cell Invasion J. Biol. Chem., June 6, 2008; 283(23): 15807 - 15815. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Vial and P. J. McKeown-Longo PAI1 stimulates assembly of the fibronectin matrix in osteosarcoma cells through crosstalk between the {alpha}v{beta}5 and {alpha}5{beta}1 integrins J. Cell Sci., May 15, 2008; 121(10): 1661 - 1670. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-C. Chen, D. O. Henry, P. R. Reczek, and M. K.K. Wong Plasminogen activator inhibitor-1 inhibits prostate tumor growth through endothelial apoptosis Mol. Cancer Ther., May 1, 2008; 7(5): 1227 - 1236. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H Kliem, H Welter, W D Kraetzl, M Steffl, H H D Meyer, D Schams, and B Berisha Expression and localisation of extracellular matrix degrading proteases and their inhibitors during the oestrous cycle and after induced luteolysis in the bovine corpus luteum Reproduction, September 1, 2007; 134(3): 535 - 547. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Sternlicht, A. M. Dunning, D. H. Moore, P. D.P. Pharoah, D. G. Ginzinger, K. Chin, J. W. Gray, F. M. Waldman, B. A.J. Ponder, and Z. Werb Prognostic Value of PAI1 in Invasive Breast Cancer: Evidence That Tumor-Specific Factors Are More Important Than Genetic Variation in Regulating PAI1 Expression. Cancer Epidemiol. Biomarkers Prev., November 1, 2006; 15(11): 2107 - 2114. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. M. De Taeye, T. Novitskaya, L. Gleaves, J. W. Covington, and D. E. Vaughan Bone Marrow Plasminogen Activator Inhibitor-1 Influences the Development of Obesity J. Biol. Chem., October 27, 2006; 281(43): 32796 - 32805. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Dekker, R. A. Boon, M. G. Rondaij, A. Kragt, O. L. Volger, Y. W. Elderkamp, J. C. M. Meijers, J. Voorberg, H. Pannekoek, and A. J. G. Horrevoets KLF2 provokes a gene expression pattern that establishes functional quiescent differentiation of the endothelium Blood, June 1, 2006; 107(11): 4354 - 4363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Kaur, F. W. Khwaja, E. A. Severson, S. L. Matheny, D. J. Brat, and E. G. Van Meir Hypoxia and the hypoxia-inducible-factor pathway in glioma growth and angiogenesis Neuro-oncol, April 1, 2005; 7(2): 134 - 153. [Abstract] [PDF]
|
|
|
|
|
|
M. D. Hjelmeland, A. B. Hjelmeland, S. Sathornsumetee, E. D. Reese, M. H. Herbstreith, N. J. Laping, H. S. Friedman, D. D. Bigner, X.-F. Wang, and J. N. Rich SB-431542, a small molecule transforming growth factor-{beta}-receptor antagonist, inhibits human glioma cell line proliferation and motility Mol. Cancer Ther., June 1, 2004; 3(6): 737 - 745. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. A. DeClerck, A. M. Mercurio, M. S. Stack, H. A. Chapman, M. M. Zutter, R. J. Muschel, A. Raz, L. M. Matrisian, B. F. Sloane, A. Noel, et al. Proteases, Extracellular Matrix, and Cancer: A Workshop of the Path B Study Section Am. J. Pathol., April 1, 2004; 164(4): 1131 - 1139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R L Bates, S J Payne, S L Drury, P N Nelson, D A Isenberg, J J Murphy, and G Frampton The prevalence and clinical significance of autoantibodies to plasminogen activator inhibitor 1 in systemic lupus erythematosus Lupus, August 1, 2003; 12(8): 617 - 622. [Abstract] [PDF]
|
|
|
|
 |
|
 J. D. Su, L. D. Mayo, D. B. Donner, and D. L. Durden PTEN and Phosphatidylinositol 3'-Kinase Inhibitors Up-Regulate p53 and Block Tumor-induced Angiogenesis: Evidence for an Effect on the Tumor and Endothelial Compartment Cancer Res., July 1, 2003; 63(13): 3585 - 3592. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Stefansson and D. A. Lawrence Old Dogs and New Tricks, Proteases, Inhibitors, and Cell Migration Sci. Signal., July 1, 2003; 2003(189): pe24 - pe24. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Vaughan PAI-1 and Cellular Migration: Dabbling in Paradox Arterioscler Thromb Vasc Biol, October 1, 2002; 22(10): 1522 - 1523. [Full Text] [PDF]
|
|
|
|
|
|
Y. Guo, A. P. Mazar, J.-J. Lebrun, and S. A. Rabbani An Antiangiogenic Urokinase-derived Peptide Combined with Tamoxifen Decreases Tumor Growth and Metastasis in a Syngeneic Model of Breast Cancer Cancer Res., August 15, 2002; 62(16): 4678 - 4684. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Eddy Plasminogen activator inhibitor-1 and the kidney Am J Physiol Renal Physiol, August 1, 2002; 283(2): F209 - F220. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
) and fibronectin-coated (1.5 µg; 
) wells were incubated in the presence or absence of rPAI-1 for 30 min at 37°C. HBMECs (5 x 104 cells/well) suspended in adhesion buffer were then added to the wells and incubated for 1 h at 37°C. The amount of adherent cells was then determined as described in "Materials and Methods." The data represent the mean absorbance of triplicate wells; bars, SE. The data represent one experiment from a total of three similar experiments. *, P < 0.01 in comparison with control (0 nM). B, rPAI-1 was preincubated at indicated concentrations in vitronectin-precoated tissue culture wells (400 ng of vitronectin/well) for 30 min at 37°C. HBMECs (5 x 104 cells/well) were then added and incubated for 1 h at 37°C. The amount of adherent cells was determined. The data represent the mean absorbance of triplicate wells; bars, SE. C, HBMECs were incubated in the absence (
, rPAI-1 added 1 h before cells; 
, EMD 121974 added 1 h before cells; 
, rPAI-1 added 1 h after cells; •, EMD 121974 added 1 h after cells. C, cell lysates (40 µg) from HBMECs were obtained at the indicated times after plating and analyzed by Western blot for the expression of Bax, Bcl-XL, and ß-tubulin. rPAI-1 and EMD 121974 (RGD) were added 1 h before plating the cells.
, Matrigel supplemented with fibronectin;
, Matrigel supplemented with vitronectin and fibronectin. *, P < 0.02; **, P < 0.01.