Role of Heparan Sulfate Proteoglycans in the Uptake and Degradation of Tissue Factor Pathway Inhibitor-Coagulation Factor Xa Complexes*

(Received for publication, October 7, 1996, and in revised form, February 19, 1997)

Guyu Ho Dagger §, George J. Broze Jr. and Alan L. Schwartz Dagger

From the Dagger  Departments of Pediatrics, Molecular Biology and Pharmacology, and  Medicine, Washington University School of Medicine and Division of Hematology, Jewish Hospital of St. Louis, St. Louis, Missouri 63110

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
REFERENCES


ABSTRACT

Tissue factor pathway inhibitor (TFPI) is a potent inhibitor of blood coagulation factor Xa (fXa) and factor VIIa. We have recently shown that fXa binding stimulates the uptake and degradation of cell surface-bound 125I-TFPI (Ho, G., Toomey, J. R., Broze, G. J., Jr., and Schwartz, A. L. (1996) J. Biol. Chem. 271, 9497-9502). In the present study we examined the role of cell surface glycosaminoglycans (GAGs) in this process. Removal of cell surface GAG chains by treatment of cells with heparinase or heparitinase but not chondroitinase markedly reduced fXa-stimulated 125I-TFPI uptake and degradation. Inhibition of GAG sulfation by growth of cells in chlorate-containing medium similarly decreased fXa-stimulated 125I-TFPI degradation. These results suggest that heparan sulfate proteoglycans (HSPGs) are required for the uptake and degradation of 125I-TFPI·fXa complexes. Chemical cross-linking/immunoprecipitation analyses revealed that 125I-TFPI was directly associated with HSPGs on the cell surface and that fXa binding increased the amount of 125I-TFPI bound. Of the several cell lines evaluated, bend endothelial cells demonstrated the greatest fXa stimulation of 125I-TFPI uptake and degradation. Cross-linking/immunoprecipitation analyses on bend cells also revealed that HSPGs were specifically associated with TFPI and fXa. These data suggest that HSPGs may directly act as the uptake and degradation receptor for TFPI·fXa complexes.


INTRODUCTION

Coagulation factor Xa (fXa),1 a serine protease positioned at the convergence of the intrinsic and extrinsic pathways of the coagulation cascade, plays a central role in hemostasis. Activated fXa along with phospholipids and factor Va converts prothrombin to thrombin, which in turn generates the fibrin clot.

The nonthrombogenic properties of the endothelial cell surface are maintained in part by protease inhibitors of the coagulation cascade. Tissue factor pathway inhibitor (TFPI) is a potent direct inhibitor of fXa and, in a fXa-dependent fashion, produces feedback inhibition of the factor VIIa-tissue factor catalytic complex (1). The TFPI molecule consists of three tandem Kunitz-type protease inhibitor domains and a basic C-terminal region (2). Its second Kunitz domain is required for binding to fXa, and its first Kunitz domain appears to bind factor VIIa in the factor VIIa-tissue factor complex (1, 3). Unlike other coagulation protease inhibitors, whose binding is irreversible, TFPI binds to the proteases in a reversible manner (4, 5). Instead of transferring fXa to other protease inhibitors for final inactivation, we have shown that cell surface-bound TFPI mediates 125I-fXa uptake and degradation and, reciprocally, fXa binding stimulates the uptake and degradation of cell surface-bound 125I-TFPI (6). Unlike the uptake and degradation of uncomplexed 125I-TFPI, which is mediated via the endocytic receptor low density lipoprotein-related protein (LRP) (7), the uptake and degradation of the fXa·TFPI complex is independent of LRP (6).

Circumstantial evidence suggests that TFPI may be bound to glycosaminoglycans (GAGs) on the endothelial cell surface. This notion derives from the following facts. (a) TFPI binds to heparin agarose (8); (b) heparin and sulfated polysaccharides enhance the anticoagulant activity of TFPI (9); (c) after intravenous administration of heparin, plasma levels of TFPI increase severalfold (10, 11); and (d) heparin competes for abundant TFPI binding sites on the cell surface (7, 12, 13). In the present investigation, we attempted to address whether GAGs play a role in the uptake and degradation of the fXa·TFPI complex. Using PEA 13 fibroblasts, a cell line deficient in LRP (14), we show that enzymatic removal of cell surface heparin sulfate or growth of cells in chlorate reduced fXa-stimulated 125I-TFPI degradation. Chemical cross-linking of 125I-TFPI to cells coupled with immunoprecipitation analyses showed an association of high molecular weight species with 125I-TFPI and enhancement following fXa binding. These data suggest that heparan sulfate proteoglycans (HSPGs) may directly serve as the receptor for the uptake and degradation of TFPI·fXa complexes.


EXPERIMENTAL PROCEDURES

Materials

IODOGEN and dithiobis(sulfosuccinimidyl propionate) (DTSSP) were purchased from Pierce. [125I]Iodide was from Amersham Corp. Heparitinase (EC 4.2.2.8) and chondroitinase ABC (EC 4.2.2.4) were from ICN. Heparinase (EC 4.2.2.7, heparinase I), bovine serum albumin, and normal rabbit (nonimmune) IgG were from Sigma. Sodium chlorate and sodium sulfate were purchased from Aldrich. Immobilized rProtein A beads were from RepliGen (Cambridge, MA). Human factor Xa, goat-anti-human factor X polyclonal antibody, and normal goat IgG were obtained from American Diagnostica (Greenwich, CT). Anti-TFPI polyclonal antibodies were described previously (15). Tissue culture media and plasticware were obtained from Life Technologies, Inc.

Protein Iodination

Proteins (10-50 µg) were iodinated using the IODOGEN methods (16). Specific radioactivities were typically 0.5-3 × 106 cpm/pmol of protein.

Cell Culture

HepG2 cells (17), PEA 10 (LRP heterozygous), PEA 13 (LRP homozygous deleted) cells (14), and mouse brain endothelial cells (bend-3, gift of Dr. W. Frazier, Washington University at St. Louis) were cultured in Dulbecco's modified Eagle's medium (with glutamine) supplemented with 10% fetal calf serum, 100 units/ml penicillin, and 100 µg/ml streptomycin. Cells were incubated at 37 °C in humidified air containing 5% CO2.

Binding and Degradation Assays

Cells were seeded into 12-well dishes 1 day prior to assays. Cell monolayers were generally used at 70-80% confluence. Assay buffers were Dulbecco's modified Eagle's medium containing 3% bovine serum albumin. In general, binding was carried out by washing cell monolayers with prechilled assay buffer twice and binding was initiated by adding 0.5 ml of 4 °C assay buffer containing 125I-labeled proteins or unlabeled proteins. After incubation at 4 °C for the time duration specified in the figure legends, overlying buffer containing unbound ligand was removed, and the cells were washed three times with 4 °C assay buffer. The cells were then incubated for a second round of binding as indicated in the figure legends.

Degradation assays were carried out at 37 °C for 3 h in 0.5 ml of assay buffer. Radioligands were either prebound to the cell surface or contained in the assay buffer as indicated in the figure legends. Thereafter, the medium overlying the cell monolayers was removed and proteins were precipitated by addition of bovine serum albumin to 5 mg/ml and trichloroacetic acid to 10%. Degradation of radioligand was defined as the appearance of radioactive fragments in the overlying medium that were soluble in 10% trichloroacetic acid. Degradation of 125I-ligand in parallel dishes that did not contain cells was subtracted from the total degradation (17).

Digestion of Cell Surface GAGs

Cells grown in 12-well dishes (~70% confluence) were digested at 37 °C for 1.5 h in assay buffer containing the appropriate enzymes at concentrations indicated in the figure legends.

Chemical Cross-linking

Chemical cross-linking was performed as described (18). Briefly, cells (70-80% confluence) were incubated at 4 °C with assay buffer containing the desired concentrations of TFPI. After removing unbound ligand by washing twice with assay buffer, the cell monolayers were incubated at 4 °C with assay buffer alone or containing the desired concentrations of fXa. The cells were then washed with PBS supplemented with 1 mM CaCl2 and 0.5 mM MgCl2 (PBSc). Cross-linking of ligands to the cells was performed by incubating cell monolayers with PBSc containing 0.5 mM DTSSP. After 30 min at 4 °C, the reaction was quenched by washing cells with Tris-buffered saline. Cells were then lysed in PBSc containing 1% Triton X-100 and 1 mM phenylmethylsulfonyl fluoride for 30 min at 4 °C with intermittent votexing, followed by brief sonication at low power output. The cell lysates were subsequently used for immunoprecipitation.

Immunoprecipitation

Cell lysates from chemical cross-linking experiments were mixed with equal volumes of PBSc containing 1% Triton X-100, 0.5% sodium deoxycholate, 1% SDS, and 1 mM phenylmethylsulfonyl fluoride (immunomix). The mixture was aliquoted, and appropriate antibodies were added (10 µg of IgG in a volume of ~0.4 ml). After rocking overnight at 4 °C, the immunomixture was incubated with 30 µl of Protein A bead slurry at room temperature for 1 h. The Protein A beads were collected by spinning for 30 s at top speed in a Microfuge, washed twice with immunomix and twice with PBSc. The immunoprecipitates bound to Protein A beads were released by boiling for 5 min in 62.5 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol (Laemmli SDS sample buffer) (19) without 2-beta -mercaptoethanol.

SDS-Polyacrylamide Gel Electrophoresis and Autoradiography

SDS samples were analyzed by SDS-polyacrylamide gel electrophoresis (19) under nonreducing conditions. A set of prestained molecular weight standards (myosin, beta -galactosidase, bovine serum albumin, and ovalbumin) (Bio-Rad) was included for molecular weight estimation. Gels were dried and exposed to x-ray films for visualization.


RESULTS

Reduction of Cell Surface Heparan Sulfate Decreases the fXa-stimulated 125I-TFPI Degradation by PEA 13 Cells

We have previously shown that cell surface-bound 125I-TFPI was degraded at very low rates by PEA 13 cells but that this basal degradation rate is enhanced 5-fold upon binding of 125I-TFPI to fXa (6). Since PEA 13 cells are devoid of LRP (14), this observation suggests that the fXa-stimulated 125I-TFPI degradation is mediated via a non-LRP endocytic pathway. TFPI has been proposed to bind GAGs on the cell surface (10-12). To investigate the role of GAGs in TFPI·fXa degradation, PEA 13 cells were incubated with heparinase or heparitinase (1 unit/ml), which cleave GAG heparan sulfate moieties at distinct sites (20). The enzyme-treated cells were then incubated at 4 °C with 125I-TFPI to allow for binding, followed by incubation at 4 °C in the presence or absence of fXa. Thereafter, degradation of 125I-TFPI was assessed at 37 °C. As shown in Table I, degradation of fXa-stimulated 125I-TFPI, measured as the difference in 125I-TFPI degraded in the presence and absence of fXa, was reduced to ~48% of control (Table I). Removal of GAG chondroitin sulfate and dermatan sulfate by treating cells with chondroitinase ABC (21), however, did not affect fXa-stimulated 125I-TFPI degradation (Table I). Since heparan sulfate and chondroitin sulfate represent the two major types of GAGs (22), these data suggest that HSPGs participate in the uptake and degradation of TFPI·fXa complexes. The maximal effect of heparinase on fXa-stimulated 125I-TFPI degradation was determined, as shown in Fig. 1. The amount of fXa-stimulated 125I-TFPI degradation decreased with increasing concentrations of heparinase, reaching a maximal effect that was 20% of the control, at 4 units/ml heparinase. These data thus suggest that HSPGs play a major role in the uptake and degradation of fXa·TFPI complexes.

Table I. Effect of GAG-degrading enzymes on fXa-stimulated 125I-TFPI degradation

PEA 13 cells were incubated in the presence or absence of 1 unit/ml of the indicated enzymes (in culturing medium DMEM) for 1.5 h at 37 °C, after which the cells were incubated with 4 nM 125I-TFPI at 4 °C for 30 min to allow for binding. After washing to remove unbound radioligand, the cells were incubated with 4 nM fXa at 4 °C for 30 min before incubation at 37 °C for 3 h for assessment of 125I-TFPI degradation. The amount of fXa-stimulated 125I-TFPI was calculated as 125I-TFPI degraded in the presence of fXa minus that degraded in the absence of fXa. 125I-TFPI degradation in the absence of enzyme treatments were arbitrarily set at 100%. The mean ± S.D. are derived from at least three independent experiments.

Enzyme fXa-stimulated 125I-TFPI degradation

1 unit/ml %
0 100
Heparinase 47  ± 5
Heparitinase 48  ± 2
Chondroitinase 101  ± 4


Fig. 1. fXa-stimulated 125I-TFPI degradation following heparinase digestion of PEA 13 cells. PEA 13 cells were in incubated with various concentrations of heparinase under the conditions specified in Table I. Subsequent assays for fXa-stimulated 125I-TFPI degradation also followed those described in Table I. Each symbol represents the mean ± S.D. of three independent experiments.
[View Larger Version of this Image (15K GIF file)]

The C-terminal heparin-binding domain of TFPI consists of a cluster of positively charged residues, truncation of which abolishes the binding of TFPI to the cell surface (23). To determine whether the negatively charged sulfate groups on HSPGs were critical for TFPI·fXa binding and degradation, PEA 13 cells were cultured in medium containing chlorate, a competitive inhibitor of ATP sulfurylase (24). Cells grown in media containing various concentrations of chlorate were then analyzed for degradation of 125I-TFPI in the presence or absence of fXa as described above. As shown in Fig. 2, growth of cells for more than four doublings in the presence of chlorate reduced the level of fXa-stimulated 125I-TFPI degradation in a dose-dependent fashion. At 30 mM chlorate, the level of degradation was reduced to 9% of the control value. To assure that the inhibition by chlorate was not due to cytotoxicity but was a result of competitive inhibition of sulfation, cells were cultured in medium containing chlorate (20 mM or 30 mM) as well as sulfate (10 mM). As shown in Fig. 2, degradation of 125I-TFPI was fully rescued by simultaneous addition of sulfate in the medium. These data indicate that degradation of fXa·TFPI complexes requires sulfate groups on HSPGs, and support the observation from Fig. 1 that HSPGs play a key role in the uptake and degradation of the TFPI·fXa complex by cells.


Fig. 2. Chlorate inhibits fXa-stimulated 125I-TFPI degradation. PEA 13 cells were cultured for 2 days in medium containing various concentrations of sodium chlorate with or without additional 10 mM sodium sulfate. Following ~4 doublings the cells reached a density of ~70% on the day of assay. Degradation studies were conducted as described in Table I. The data were generated from at least three independent experiments.
[View Larger Version of this Image (21K GIF file)]

In addition to serving as a binding site for fXa on the cell surface (25), we have shown that TFPI also mediates the cellular uptake and degradation of fXa (6). To determine whether that fXa degradation was similarly affected by HSPGs, fXa was radiolabeled and its degradation was analyzed in PEA 13 cells that were either treated with 1 unit/ml heparinase or cultured in 10 mM chlorate. The respective cells were then incubated at 4 °C with TFPI to allow for binding. Following washing to remove unbound TFPI, the cell monolayers were then incubated with various concentrations of 125I-fXa and the degradation was evaluated upon incubation at 37 °C as above. As shown in Fig. 3A, heparinase treatment reduced TFPI-mediated 125I-fXa degradation by ~50% at each concentration used. Growing cells in chlorate also markedly decreased 125I-fXa degradation (Fig. 3B), and this inhibitory effect was abrogated in cells cultured with an additional 5 mM sulfate (Fig. 3B).


Fig. 3. Heparinase and chlorate reduce the cellular uptake and degradation of 125I-fXa mediated by TFPI. A, PEA 13 cells were treated with or without 1 unit/ml heparinase in a manner identical to that described in Table I. The cells were then incubated at 4 °C for 30 min in the presence of 50 nM TFPI, following which unbound TFPI was removed. After incubation at 4 °C for 30 min in the presence of increasing concentrations of 125I-fXa, the cells were placed at 37 °C for 3 h to assay degradation. B, PEA 13 cells were grown in medium containing 10 mM chlorate, or 10 mM chlorate plus 5 mM sulfate, or no additional chemicals in a manner similar to that described in Fig. 2. TFPI prebinding to and subsequent 125I-fXa degradation by the cells were carried out as described in A. Numbers represent duplicate determinations.
[View Larger Version of this Image (19K GIF file)]

Degradation of Uncomplexed 125I-TFPI Does Not Require Cell Surface HSPGs

We have shown previously that the cellular degradation of uncomplexed 125I-TFPI is inhibited >80% both by antibodies directed against LRP and by the 39-kDa protein, an LRP-associated protein that inhibits all ligand interactions with LRP (7). LRP, however, is not the major cell surface receptor, as the vast majority of 125I-TFPI binding is not inhibitable by the 39-kDa protein (7, 12). HSPGs have been proposed as the initial binding site for several LRP ligands in addition to TFPI (7, 12) including apoE-enriched remnant lipoproteins (26), hepatic lipase (27), and thrombospondin (28), whose uptake and degradation by LRP requires HSPGs. To further determine the role of HSPGs in the LRP-mediated uptake and degradation of TFPI, degradation of 125I-TFPI was analyzed in PEA 10 cells, which express LRP (14). The cells were incubated with 2 units/ml heparinase or grown in medium containing 10 mM chlorate as described in Fig. 3. The cells were then incubated with 125I-TFPI at 37 °C to assess degradation. As shown in Fig. 4, neither treatment of PEA 10 cells with heparinase nor culture of cells in chlorate affected 125I-TFPI degradation.


Fig. 4. Heparinase and chlorate have no effect on the cellular degradation of uncomplexed 125I-TFPI. PEA 10 cells were incubated without or with 2 units/ml heparinase in conditions described in Table I (A) or grown in the absence or presence of 10 mM sodium chlorate for 2 days (B). Cells from A and B were then incubated with increasing concentrations (2-12 nM) of 125I-TFPI at 37 °C for 3 h to assess degradation. Numbers represent duplicate determinations.
[View Larger Version of this Image (18K GIF file)]

Binding of fXa Increases the Interaction of 125I-TFPI for HSPGs

We showed previously that fXa-stimulated 125I-TFPI degradation results from an enhanced uptake of 125I-TFPI by cells (6). One potential explanation for this notion is that the interaction of TFPI for its endocytic receptor is increased upon association with fXa, resulting in the enhanced cellular uptake. To address this question as well as to define candidate receptors for the fXa·TFPI complex, chemical cross-linking coupled with immunoprecipitation analyses was performed. Initially fXa was radioiodinated and cross-linking was performed on PEA 13 cells. Cells were incubated with or without TFPI at 4 °C, followed by incubation with 125I-fXa to allow for binding. Chemical cross-linking of the bound ligands to the cells was performed with DTSSP, a thio-cleavable, water-soluble, and membrane-impermeable reagent (29). After cross-linking, the cells were lysed and immunoprecipitated with anti-TFPI or anti-fX polyclonal antibodies. To characterize membrane proteins cross-linked to 125I-fXa, the immunoprecipitates were resolved on 7.5% SDS gels under nonreducing conditions. As shown in Fig. 5A, in the presence of TFPI prebinding and DTSSP, a high molecular weight (high Mr) band at the top of the gel was immunoprecipitated by both anti-fX and anti-TFPI antibodies (lanes 3 and 6), suggesting its association with 125I-fXa as well as TFPI. The specificity of association is affirmed by showing that this high Mr band is absent without the cross-linking reagent (lanes 4 and 7) and is not precipitated by normal rabbit IgG (lane 1). The high Mr band is undetectable when cells were not prebound with TFPI (lane 2), suggesting that binding of 125I-fXa to the high Mr band is mediated by TFPI. Species at size ~90 kDa are free cross-linked 125I-fXa·TFPI complexes (lanes 3 and 6). Intense bands at 50 kDa are free 125I-fXa (lanes 3 and 4).


Fig. 5. Binding of fXa increases the interaction of 125I-TFPI with HSPGs on the PEA 13 cell surface. A, PEA 13 cells (grown in 100-mm dishes) were incubated at 4 °C with or without 10 nM TFPI for 1.5 h. After washing to remove unbound TFPI, the cells were incubated at 4 °C with 10 nM 125I-fXa for another 1.5 h. Following washing to remove unbound radioligand, the cells were incubated with or without 0.5 mM DTSSP at 4 °C for 30 min. Following cell lysis, immunoprecipitations were performed with anti-TFPI, anti-fX, or normal rabbit IgG (N.R.) as described under "Experimental Procedures." The immunoprecipitates were run on 7.5% SDS gels under nonreducing conditions. Equal amounts of total cellular proteins were loaded in each lane. B, PEA 13 cells were incubated at 4 °C with 10 nM 125I-TFPI for 1.5 h. After washing to remove unbound radioligand, the cells were incubated at 4 °C with or without 10 nM fXa for an additional 1.5 h. Subsequent cross-linking, immunoprecipitation, and gel analysis were identical to those specified in A. C, PEA 13 cells were incubated without (lane 1) or with (lane 2) 2 units/ml heparinase at 37 °C for 1.5 h. The enzyme-treated cells were then incubated sequentially with 125I-TFPI and fXa as in B. After incubation at 4 °C for 30 min with DTSSP, the cell lysates were immunoprecipitated with anti-TFPI antibodies. Gel analysis was identical to that described in A.
[View Larger Version of this Image (25K GIF file)]

Since 125I-fXa binding to the high Mr band is mediated through TFPI, we next examined the effect of fXa on 125I-TFPI binding to the receptor by cross-linking/immunoprecipitation analyses. PEA 13 cells were incubated at 4 °C with 125I-TFPI and thereafter with or without fXa at 4 °C. Subsequent cross-linking and immunoprecipitation were performed as above. As shown in Fig. 5B, in the absence of fXa addition but presence of DTSSP, the same high Mr band was immunoprecipitated with anti-TFPI antibodies (lane 4). The band intensity, however, was enhanced when cells were incubated with fXa prior to cross-linking (lane 5). Densitometric analysis showed a 2.7-fold enhancement. These data indicate that fXa binding increases the interaction of TFPI with the high Mr band species. The high Mr band was precipitated by anti-fX antibodies in the presence of fXa and DTSSP (lane 2) but not in the absence of either cross-linking (lane 3) or prior fXa binding (lane 1), reaffirming the specific association of the high Mr band with the TFPI·fXa complex. In the absence of cross-linking (lane 6), there appears to be a high Mr band at the top of the gel. This is likely the result of high background during sample preparation. To confirm the identity of the high Mr species as HSPGs, PEA 13 cells were treated with heparinase prior to binding to 125I-TFPI and fXa. The cross-linking and immunoprecipitation with anti-TFPI antibodies were conducted in an identical manner to that in Fig. 5B. As shown in Fig. 5C, heparinase pretreatment diminished the intensity of the high Mr band (lane 2) (~ 2.9-fold reduction) relative to the control (lane 1). These results demonstrate that TFPI and fXa are indeed associated with cell surface HSPGs and that the reduced uptake and degradation of fXa·TFPI complexes observed in Table I and Figs. 1 and 3 is likely secondary to the reduction of TFPI binding to the cells following heparinase treatment.

Degradation of 125I-TFPI Is Markedly Potentiated by fXa in Vascular Endothelial Cells

Under physiological conditions, fXa·TFPI complexes form within the vasculature. Thus, we compared the relative ability of the bend-3 microvasculature endothelial cells, hepatoma HepG2, and fibroblast PEA 10 cells to undergo the fXa-stimulated TFPI degradation. All of these cell lines express functional LRP. The cells were incubated at 4 °C with radiolabeled TFPI to allow for cell surface binding. After washing to remove unbound radioligand, 125I-TFPI degradation was assessed in the absence or presence of varying concentrations of fXa. As shown in Fig. 6, in the absence of fXa the degradation of 125I-TFPI was ~6, 40, and 90 fmol/106 cells for bend, PEA 13, and HepG2, respectively. These rates essentially reflect the amount of 125I-TFPI degraded via LRP. With increasing concentrations of fXa added, bend cells exhibited >10-fold enhancement in 125I-TFPI degradation, whereas PEA 10 and HepG2 cells displayed only ~1.5- and ~1.1-fold enhancement, respectively. These data thus suggest that bend cells express low levels of LRP but high levels of the endocytic receptor for the TFPI·fXa complex. The fXa-stimulated 125I-TFPI degradation in bend cells was also markedly reduced by treatments of cells with heparinase, heparitinase, and chlorate (data not shown).


Fig. 6. fXa-stimulated 125I-TFPI degradation by various cell lines. HepG2, bend, and PEA 10 cells were incubated with 12 nM 125I-TFPI at 4 °C for 30 min to allow for binding. After washing to remove unbound radioligand, the cells were incubated at 4 °C for 30 min with increasing concentrations (0-12 nM) of fXa and subsequently incubated at 37 °C for 3 h to assess 125I-TFPI degradation. Means ± S.D. are derived from two or three independent experiments.
[View Larger Version of this Image (22K GIF file)]

The possibility that bend cells may express high levels of TFPI·fXa specific endocytic receptors prompted us to examine the receptor(s) in this cell line. Under the conditions described in Fig. 5, TFPI was radiolabeled, cross-linked to bend cells, and immunoprecipitated with anti-fXa and anti-TFPI antibodies. The immunoprecipitates were run on SDS gels under non-reducing conditions. As shown in Fig. 7A, a high Mr band was cross-linked to 125I-TFPI (lane 4), and the band intensity was greatly potentiated in the presence of fXa (lane 5). Densitometric analysis showed a 8.7-fold enhancement. Similar to that seen in Fig. 5B, this high Mr band-125I-TFPI complex was also specifically cross-linked to fXa (lane 2). To ascertain the identity of the high Mr band as HSPGs, heparinase treatment of bend cells was carried out as in Fig. 5C. As shown in Fig. 7B, heparinase-treated cells displayed a 2.7-fold reduced level of surface-bound 125I-TFPI·fXa complexes. These data demonstrate that HSPGs are also the binding species for fXa·TFPI complexes on the endothelial cell.


Fig. 7. TFPI and fXa associate with HSPGs on bend cells. A, binding of 125I-TFPI and fXa to bend cells, and subsequent cross-linking/immunoprecipitation were performed as described in Fig. 5B. B, heparinase treatment of bend cells and the subsequent experimental scheme followed that described in Fig. 5C.
[View Larger Version of this Image (24K GIF file)]


DISCUSSION

Using both enzymatic and chemical approaches, we have shown that fXa-stimulated 125I-TFPI uptake and degradation by PEA 13 cells requires cell surface HSPGs. Similarly, HSPGs are necessary for TFPI-mediated 125I-fXa uptake and degradation. Using chemical cross-linking in combination with immunoprecipitation analyses, we have demonstrated that 125I-TFPI is physically associated with HSPGs on the PEA 13 cells and that TFPI's interaction with HSPGs is potentiated upon association with fXa. Of the various cell lines examined for fXa-stimulated 125I-TFPI degradation, the bend endothelial cell displays the highest level of enhancement, suggesting high levels of the endocytic receptor(s) for the fXa·TFPI complex are present in these cells. Cross-linking/immunoprecipitation analysis performed with bend cells also revealed HSPGs as binding species for TFPI·fXa complexes. Thus the direct association of the TFPI·fXa complex with HSPGs on the cell surface of both PEA 13 and bend cells, as well as the diminution of 125I-TFPI·fXa degradation by cells following removal of HSPGs from the cell surface, strongly suggest that HSPGs as receptors for the uptake and degradation of the TFPI·fXa complex. Chlorate and heparinase treatments reduced the fXa-stimulated 125I-TFPI degradation to 10-20% of control (Figs. 1 and 2), and the reduced levels do not seem to have plateaued in the dose-response curves (Figs. 1 and 2). Thus it is likely that HSPGs are the major, if not the only, endocytic receptors for the fXa·TFPI complex.

HSPGs are a complex and heterogeneous family of macromolecules composed of linear sulfated polyaccharide chains, the heparan sulfate moieties, that are covalently attached to core proteins (30, 31). HSPGs are ubiquitously distributed on the plasma membrane. They are anchored to the membrane either via a linkage with membrane phospholipid (32) or via a hydrophobic transmembrane domain (33). Membrane HSPGs have been implicated in an array of cellular functions, among which are ligand binding and endocytosis (31, 34). Apart from serving as initial binding sites for ligands that are eventually transferred to classical endocytic receptors (e.g. LRP) for cellular uptake and degradation (26-28), several lines of evidence suggest that HSPGs may also act directly in the internalization of ligands. Lipoprotein lipase is anchored to the cell surface via HSPGs, through which it is internalized and recycled to intracellular compartments (35). Herpes simplex virus does not bind to nor is internalized by cells harboring mutations in HSPG synthesis (36). While it is possible that the virus may bind to HSPGs initially and thereafter be transferred to another receptor that mediates infection, the observation that cell surface proteoglycans, especially those containing heparan sulfate, are rapidly internalized and degraded (37) argues for a direct role of the proteoglycans.

Heparan sulfate moieties bind proteins predominantly via electrostatic interactions between the highly charged anionic sulfate groups and clusters of basic amino acid residues arranged in a three-dimensional array on the protein (31). TFPI may bind to cell surface HSPGs in a similar manner, since TFPI contains a stretch of basic residues at its C terminus (2), deletion of this C-terminal region abolishes binding of TFPI to the cell surface (23), and a complementary reduction of HSPG sulfation in cells grown in chlorate-containing medium decreases the extent of TFPI binding (data not shown) and degradation (Fig. 2). Electrostatic interactions may render TFPI binding to a wide range of HSPGs expressed on cell surfaces. However, not all TFPI-bound HSPGs necessarily participate in the uptake and degradation of TFPI·fXa complexes. Whether the TFPI·fXa complex receptor is a specific core protein containing diverse GAG chains or a specific GAG on heterogeneous core proteins is a matter of future studies.

It is interesting that among the cell lines tested, the endothelial cells (bend) exhibited >10-fold augmentation of fXa·TFPI uptake and degradation over the basal rate. The hepatoma cells (HepG2) and the fibroblasts (PEA 10) displayed 1.5-fold or less enhancement over the basal rates. Interestingly, the basal rate of TFPI degradation, which represents essentially the amount of TFPI degraded via LRP, was 15-fold lower in bend cells than in HepG2 cells (Fig. 6). These data suggest that bend cells express low levels of LRP, the endocytic receptor for uncomplexed TFPI (7), but high levels of the endocytic receptor for TFPI·fXa complexes. Physiologically this may be important since a low rate of removal (uptake and degradation) of uncomplexed TFPI may be necessary to maintain high local concentrations of TFPI on vascular surfaces. Rapid turnover may be initiated as a means of clearance of fXa or the TFPI·fXa complex from the cell surface.

Heparin has long been used clinically as an anticoagulant. In addition to direct inhibition of coagulation proteases, heparin greatly enhances the inhibitory activities of antithrombin III and TFPI. Anticoagulant heparan sulfate moieties on the vascular endothelial cell surfaces thus may endow them with non-thrombogenic properties (38). Such anticoagulant HSPGs, which bind antithrombin III and greatly facilitate the inhibition to thrombin, have been isolated from rat fat pad microvascular endothelial cells (39). The present study provides another potential anticoagulant role of cell surface HSPGs, one which directly involves TFPI and the active uptake and degradation of fXa·TFPI complexes generated on the cell surface.


FOOTNOTES

*   This work was supported in part by the National Institutes of Health and Monsanto.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§   To whom correspondence should be addressed: Dept. of Pediatrics, Box 8116, Washington University School of Medicine, St. Louis, MO 63110. Tel.: 314-454-6286; Fax: 314-454-2685.
1   The abbreviations used are: fXa, factor Xa; TFPI, tissue factor pathway inhibitor; GAG, glycosaminoglycan; HSPG, heparan sulfate proteoglycan; LRP, low density lipoprotein-related protein; DTSSP, dithiobis(sulfosuccinimidyl propionate); PBS, phosphate-buffered saline.

REFERENCES

  1. Broze, G. J. J., Warren, L. A., Novotny, W. F., Higuchi, D. A., Girard, J. J., and Miletich, J. P. (1988) Blood 71, 335-343 [Abstract]
  2. Wun, T.-C., Kretzmer, K. K., Girard, T. J., Miletich, J. P., and Broze, G. J., Jr. (1988) J. Biol. Chem. 263, 6001-6004 [Abstract/Free Full Text]
  3. Girard, T. J., Warren, L. A., Novotny, W. F., Likert, K. M., Brown, S. G., Mitletich, J. P., and Broze, G. J., Jr. (1989) Nature 338, 518-520 [CrossRef][Medline] [Order article via Infotrieve]
  4. Broze, G. J. J., Girard, T. J., and Novotny, W. F. (1990) Biochemistry 29, 7539-7546 [Medline] [Order article via Infotrieve]
  5. Huang, Z.-F., Wun, T.-C., and Broze, G. J., Jr. (1993) J. Biol. Chem. 268, 26950-26955 [Abstract/Free Full Text]
  6. Ho, G., Toomey, J. R., Broze, G. J., Jr., and Schwartz, A. L. (1996) J. Biol. Chem. 271, 9497-9502 [Abstract/Free Full Text]
  7. Warshawsky, I., Broze, G. J., Jr., and Schwartz, A. L. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 6664-6668 [Abstract]
  8. Wesselschimidt, R., Likert, K., Girald, T., Wun, T.-C., and Broze, G. J., Jr. (1992) Blood 79, 2004-2010 [Abstract]
  9. Wun, T.-C. (1992) Blood 79, 430-438 [Abstract]
  10. Sanset, P. M., Abildgaard, U., and Larsen, M. L. (1988) Thromb. Res. 50, 803-813 [Medline] [Order article via Infotrieve]
  11. Novotny, W. F., Palmier, M., Wun, T.-C., Broze, G. J., Jr., and Miletich, J. P. (1991) Blood 78, 394-400 [Abstract]
  12. Warshawsky, I., Herz, J., Broze, G. J., Jr., and Schwartz, A. L. (1996) J. Biol. Chem. 271, 25873-25879 [Abstract/Free Full Text]
  13. Narita, M., Bu, G., Olins, G. M., Higuchi, D. A., Herz, J., Broze, G. J., Jr., and Schwartz, A. L. (1995) J. Biol. Chem. 270, 24800-24804 [Abstract/Free Full Text]
  14. Willnow, T. E., and Herz, J. (1994) J. Cell Sci. 107, 719-726 [Abstract/Free Full Text]
  15. Broze, G. J. J., Lange, G. W., Duffin, K. L., and MacPhail, L. (1994) Blood Coag. Fibrinol. 5, 551-559 [Medline] [Order article via Infotrieve]
  16. Bu, G., Morton, P. A., and Schwartz, A. L. (1992) J. Biol. Chem. 267, 15595-15602 [Abstract/Free Full Text]
  17. Owensby, D. A., Sobel, E. B., and Schwartz, A. L. (1988) J. Biol. Chem. 263, 10587-10594 [Abstract/Free Full Text]
  18. Bu, G., and Schwartz, A. L. (1994) Cell Biology: A Laboratory Handbook, pp. 199-204, Academic Press, Inc., Orlando, FL
  19. Laemmli, U. K. (1970) Nature 227, 680-685 [Medline] [Order article via Infotrieve]
  20. Ototani, N., Kikuchi, M., and Yoshizawa, Z. (1981) Carbohydr. Res. 88, 291-303 [CrossRef][Medline] [Order article via Infotrieve]
  21. Oike, Y., Kimata, K., Shinomura, T., Nakazawa, K., and Suzuki, S. (1980) Biochem. J. 191, 193-207 [Medline] [Order article via Infotrieve]
  22. Kjellén, L., and Lindahl, U. (1991) Annu. Rev. Biochem. 60, 443-475 [CrossRef][Medline] [Order article via Infotrieve]
  23. Warshawsky, I., Bu, G., Mast, A., Saffitz, J. E., Broze, G. J., Jr., and Schwartz, A. L. (1995) J. Clin. Invest. 95, 1773-1781 [Medline] [Order article via Infotrieve]
  24. Farley, J. R., Nakayama, G., Cryns, D., and Segel, I. H. (1978) Arch. Biochem. Biophys. 185, 376-390 [Medline] [Order article via Infotrieve]
  25. Kazama, Y., Komiyama, Y., and Kisiel, W. (1993) Blood 81, 676-682 [Abstract]
  26. Ji, Z.-S., Fazio, S., Lee, Y.-L., and Mahley, R. W. (1994) J. Biol. Chem. 269, 2764-2772 [Abstract/Free Full Text]
  27. Kounnas, M. Z., Chappell, D. A., Wong, H., Argraves, W. S., and Strickland, D. K. (1995) J. Bio. Chem. 270, 9307-9312 [Abstract/Free Full Text]
  28. Godyna, S., Liau, G., Popa, I., Stefansson, S., and Argraves, W. S. (1995) J. Cell Biol. 129, 1403-1410 [Abstract]
  29. Bu, G., Maksymovitch, E. A., and Schwartz, A. L. (1993) J. Biol. Chem. 268, 13002-13009 [Abstract/Free Full Text]
  30. Gallagher, J. T., Lyon, M., and Steward, W. P. (1986) Biochem. J. 236, 313-325 [Medline] [Order article via Infotrieve]
  31. Bernfield, M., Kokenyesi, R., Kato, M., Hinkes, M. T., Spring, J., Gallo, R. L., and Lose, E. (1992) Annu. Rev. Cell Biol. 8, 365-393 [CrossRef]
  32. David, G., Lories, V., Decock, B., Marynen, P., Cassiman, J.-J., and Van der Berghe, H. (1990) J. Cell Biol. 111, 3165-3176 [Abstract]
  33. Kjellén, L., Oldberg, A., and Höök, M. (1980) J. Biol. Chem. 255, 10407-10413 [Free Full Text]
  34. Esko, J. D. (1991) Cur. Opin. Cell Biol. 3, 805-816 [Medline] [Order article via Infotrieve]
  35. Saxena, U., Klein, M. G., and Goldberg, I. J. (1990) J. Biol. Chem. 265, 12880-12886 [Abstract/Free Full Text]
  36. Shieh, M.-T., WuDunn, D., Montgomery, R. I., Esko, J. D., and Spear, P. G. (1992) J. Cell Biol. 116, 1273-1281 [Abstract]
  37. Yanagishita, M., and Hascall, V. C. (1984) J. Biol. Chem. 259, 10270-10283 [Abstract/Free Full Text]
  38. Colburn, P., and Buonassisi, V. (1982) Biochem. Biophys. Res. Commun. 104, 220-227 [Medline] [Order article via Infotrieve]
  39. Kojima, T., Leone, C. W., Marchildon, G. A., Marcum, J. A., and Rosenberg, R. D. (1992) J. Biol. Chem. 267, 4859-4869 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.