Caveolin-1 Enhances Tissue Factor Pathway InhibitorExposure and Function on the Cell Surface*□S

Received for publication, March 25, 2005Published, JBC Papers in Press, April 6, 2005, DOI 10.1074/jbc.M503333200

Cristina Lupu‡§, Xiaohong Hu‡, and Florea Lupu‡¶

From the ‡Cardiovascular Biology Research Program, Oklahoma Medical Research Foundation, and the¶Department of Pathology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma 73104

Tissue factor pathway inhibitor (TFPI) blocks tissuefactor-factor VIIa (TF-FVIIa) activation of factors X andIX through the formation of the TF-FVIIa-FXa-TFPIcomplex. Most TFPI in vivo associates with caveolae inendothelial cells (EC). The mechanism of this associa-tion and the anticoagulant role of caveolar TFPI are notyet known. Here we show that expression of caveolin-1(Cav-1) in 293 cells keeps TFPI exposed on the plasma-lemma surface, decreases the membrane lateral mobil-ity of TFPI, and increases the TFPI-dependent inhibi-tion of TF-FVIIa. Caveolae-associated TFPI supports theco-localization of the quaternary complex with caveo-lae. To investigate the significance of these observationsfor EC we used RNA interference to deplete the cells ofCav-1. Functional assays and fluorescence microscopyrevealed that the inhibitory properties of TFPI werediminished in EC lacking Cav-1, apparently through de-ficient assembly of the quaternary complex. These find-ings demonstrate that caveolae regulate the inhibitionby cell-bound TFPI of the active protease production bythe extrinsic pathway of coagulation.

Tissue factor (TF)1 is a transmembrane protein that triggersblood coagulation in vivo. Assembly of TF with factor VIIa(FVIIa) on cell surfaces initiates limited proteolysis of factorsIX and X (FX), leading to thrombin generation. TF elicitsthrombogenic responses in septicemia, cancer, and atheroscle-rosis (1–4), promotes metastasis, angiogenesis, and intima hy-perplasia after arterial injury (5, 6), and acts as signalingreceptor upon binding of FVIIa (7).

Tissue factor pathway inhibitor (TFPI) is the endogenousregulator of TF-FVIIa activity. TFPI is a Kunitz-type protease

inhibitor, which binds FVIIa via Kunitz-1 and FXa viaKunitz-2. Formation of the TF-FVIIa-FXa-TFPI complex pro-vides sustained repression of the TF pathway (8). TFPI wasfirst described as a soluble plasma protein that binds throughits C terminus to yet uncharacterized anionic sites on the cellsurface. Nevertheless, the majority of TFPI circulating inplasma is C-terminal truncated and bound to plasma lipopro-teins, and as such, has significantly less inhibitory activitythan the full-length TFPI (9).

TFPI in vivo is mainly produced by endothelial cells (EC),has an intact C terminus, and associates with the plasmamembrane through mechanisms that are not fully identified (3,10–13). Heparin releases a portion of the full-length, function-ally active cell-associated TFPI, either from cell surface-bind-ing sites or from intracellular stores (14, 15). However, the bulkof heparin-resistant cellular TFPI is released by phosphatidy-linositol-phospholipase C in vitro, which indicates that mostcellular TFPI associates with the cell surface via a glyco-sylphosphatidylinositol link (10–12, 16, 17). Furthermore, en-dogenous TFPI in resting endothelium (10), monocytes (18),and the ECV304 cell line (11, 16) partitions in low-densityfractions insoluble in cold detergent (lipid rafts). In monocytesand ECV304 endogenous TFPI inhibits efficiently FVIIa-TFactivity through translocation of the TF-FVIIa-FXa complex inlipid rafts (11). These findings suggest that cell-bound TFPI,particularly the lipid raft-associated pool, plays a critical rolein regulating cell surface FVIIa-TF and FXa activity.

Lipid rafts are domains rich in cholesterol and sphingolipidsthat can exist by themselves or as caveolae. Caveolae are small(50–80 nm) plasma membrane invaginations that have a pro-tein “coat” composed of caveolin family members. Caveolin-1(Cav-1) is an integral membrane protein and the principalcomponent of caveolae (19). The role of Cav-1 in caveolae for-mation was confirmed in Cav-1-deficient cells, which lack mor-phologically identifiable caveolae (20). Expression of Cav-1 inthese cells induces caveolae formation (21). Caveolae are mul-tifunctional organelles in which Cav-1 plays a direct role invarious events, such as membrane trafficking and cellular sig-nal transduction (reviewed in Ref. 22).

We reported previously that TFPI is localized in caveolae inEC both in vitro and in vivo (3, 10). The mechanism of TFPIassociation with caveolae and the anticoagulant role of caveolarTFPI are not yet known. We report here that Cav-1 regulatesthe distribution and function of TFPI in HEK293, a cell systemwhere we controlled the expression of Cav-1 and TFPI bytransfection. We show for the first time that caveolae keepTFPI associated with the cell surface and enhance the antico-agulant activity of the inhibitor. Furthermore, using RNA in-terference (RNAi) to deplete HUVEC and EA.hy926 cells ofCav-1, we show that EC which lacks Cav-1 displayed several-fold enhanced procoagulant activity. In conclusion, we identi-fied Cav-1 as an active regulator of TFPI-dependent inhibition

* This work was supported by American Heart Association BeginningGrant-in-aid 0265213Z and National Institutes of Health COBRE Pro-gram Grant 5P20RR018758-02. The costs of publication of this articlewere defrayed in part by the payment of page charges. This article musttherefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.

□S The on-line version of this article (available at http://www.jbc.org)contains data and Figs. S1—S5.

§ To whom correspondence should be addressed: Oklahoma MedicalResearch Foundation, 825 NE 13th St., Oklahoma City, OK 73104. Tel.:405-271-7206; Fax: 405-271-7417; E-mail: cristina-lupu@omrf.ouhsc.edu.

1 The abbreviations used are: TF, tissue factor; Cav-1, caveolin-1;(E)GFP, (enhanced) green fluorescent protein; FCS, fetal calf serum;FRAP, fluorescence recovery after photobleaching; EC, endothelial cells;FVIIa, coagulation factor VIIa; HEK293, human embryonic kidney epi-thelial cells; HUVEC, human umbilical vein endothelial cells; mAb, mono-clonal antibody; PBS, phosphate-buffered saline; RNAi, RNA interfer-ence; siRNA, silencing RNA; TBS, Tris-buffered saline; TFPI, tissue factorpathway inhibitor; wt, wild type; FX, factor X; FITC, fluorescein isothio-cyanate; FCS, fetal calf serum; MES, 4-morpholineethanesulfonic acid;FRAP, fluorescence recovery after photobleaching; ELISA, enzyme-linkedimmunosorbent assay.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 23, Issue of June 10, pp. 22308–22317, 2005© 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org22308

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

of TF-FVIIa activity, which adds the hemostatic function as anovel dimension to the biological significance of caveolae.

EXPERIMENTAL PROCEDURES

Reagents and Cells—Antibodies and suppliers used were: monoclonalantibodies (mAb) TFPIK-9 against human r-TFPI (gift from Dr. T.Hamuro, KAKETSUKEN, Kumamoto, Japan); rabbit anti-humanr-TFPI-(1–249) IgG (23); mAb JL-8 anti-green fluorescent protein (GFP,BD Biosciences Clontech); goat anti-human TF IgG (American Diagnos-tica); mAbs 10H10 and 9-5B7 against human r-TF (kind gifts from Dr.J. Morrissey, University of Illinois, Urbana-Champaign, IL); rabbitanti-Cav-1 IgG (sc-894, Santa Cruz Biotechnology); mAb anti-Cav-1(BD Transduction); mAb HPC4, recognizes a calcium-dependentepitope in activated protein C (24) (gift from Dr. C. Esmon in ourinstitute); mAb anti-human lamin A/C (Serotec); horse anti-mouse IgG/FITC (Vector Laboratories); donkey anti-goat IgG-Cy5, donkey anti-mouse IgG-Cy3, donkey anti-rabbit IgG-Cy5, and donkey anti-rabbitIgG-Cy3 (Biomeda). Human coagulation factors VIIa, X, and Xa werefrom Enzyme Research Laboratories. Chromogenic substrate S-2765specific for FXa was from DiaPharma. Human r-tumor necrosis factor-�was from R&D Systems, ionomycin was from EMD Biosciences, andhuman r-TF InnovinTM (0.22 �g of TF/vial) was from Dade Behring.Restriction enzymes were from Promega. Cell culture media and sup-plements were from Invitrogen and HyClone. Rabbit brain thrombo-plastin and all other reagents were from Sigma unless otherwise stated.Human embryonic kidney epithelial cells HEK293 were grown in Dul-becco’s modified Eagle’s medium containing 100 units/ml penicillin, 0.1mg/ml streptomycin, and 10% fetal calf serum (FCS) and incubated in5% CO2, 95% humidified air at 37 °C. The immortalized hybrid EC lineEA.hy926 (gift from Dr. Cora-Jean Edgell, University of North Caro-lina, Chapel Hill, NC) was maintained in Dulbecco’s modified Eagle’smedium supplemented with 4 mM glutamine, 15 mM HEPES, 10% FCS,and antibiotics, in 10% CO2, 90% humidified air at 37 °C. Humanumbilical vein EC (HUVEC) were used at passages 1–5, and weremaintained in antibiotic-free Medium 199 with Earles salts, supple-mented with 0.05 mg/ml heparin, 10% FCS, and 1:100 diluted crudebovine brain extract (as source of growth factors; gift from Dr. C.Esmon) in 5% CO2, 95% humidified air at 37 °C.

Generation of 293 Cell Lines Stably Expressing EGFP-TFPI andCav1-HPC4ep—HEK293 express low levels of both Cav-1 (25, 26) andTFPI (27). We transfected 293 cells with pEGFP-C2/TFPI(C), pSVzeo-Cav1-HPC4ep, or co-transfected with both vectors using PolyFect (Qia-gen; efficiency �50–70%). Parallel assays were run on non-transfectedor sham-transfected cells. We selected 293 cell clones that stably ex-pressed EGFP-TFPI, Cav1-HPC4ep, or both chimeras (characterizationand selection of the clones detailed under supplemental materials).

Freeze-fracture EM—Freeze-fracture EM was performed as de-scribed (28) on wild-type and Cav-1 expressing cells.

Proteolytic Activity of Cell Surface TF-FVIIa—Confluent monolayersin 24-well plates were tested with a two-stage chromogenic assay (12,16). 50 �g/ml cycloheximide was added to prevent de novo proteinsynthesis. To de-encrypt TF and expose anionic phospholipids the cellswere briefly incubated with 1 �M ionomycin. Several concentrations ofFVIIa (0.2, 10, or 50 nM) were added for 30 min at room temperature,and then excess FX was added (5, 200, and 500 nM, respectively).Samples were taken during a 30-min incubation time at 37 °C, andquenched in ice-cold 50 mM Tris-buffered saline (TBS), pH 8.8, contain-ing 25 mM EDTA and 0.1% bovine serum albumin. FXa generated wasdetermined from the hydrolysis of the chromogenic substrate S-2765. Inantibody-blocking experiments, 50 �g/ml anti-TFPI IgG or 100 �g/ml ofthe anti-TF mAb 5B7 were included during the incubation with FVIIa.

Determination of TFPI Activity by Functional Assay—The activity ofcell surface TFPI was measured with a modification of a two-stagechromogenic assay in which we estimated the residual capability ofTF-FVIIa pre-formed mixtures to activate FX after incubation withTFPI-exposed cells. The original assay (10, 12) made use of 0.4 nM

FVIIa and 1:320 dilution of thromboplastin made from one vial recon-stituted with 1 ml of 50 mM TBS, pH 7.35. Here, TF-FVIIa complexeswere pre-formed by mixing up several combinations of thromboplastindilutions and FVIIa concentrations for 30 min at 37 °C in the presenceof 15 mM CaCl2. Each mixture was tested for the linearity and the ratesof FXa generation, first in the absence of any TFPI, and then in thepresence of serially diluted normal human plasma. Three mixturescontaining (nM FVIIa � thromboplastin dilution) 0.2 � 1:700, 10 � 1:15,and 50 � 1:3 were chosen. TFPI-bearing monolayers were then incu-bated for 30 min at room temperature with each pre-formed TF-FVIIacomplex, either alone or in the presence of 50 �g/ml anti-TFPI IgG (to

block the available TFPI). The supernatants were removed and anyunbound TF-FVIIa present was tested for activation of FX (added as 10times excess over FVIIa) by using S-2765. The activity of TFPI wasextrapolated from standard curves constructed with serial dilutions ofnormal human plasma, which was assigned a TFPI functional potencyof 1 unit/ml.

SDS-PAGE and Western Blotting—Cell monolayers were scraped inice-cold 0.1 M TBS, pH 7.8, containing 1 mM phenylmethylsulfonyl fluo-ride, 1% aprotinin, 1 mM sodium orthovanadate, 5 mM EDTA, and 0.02%NaN3. The cells were lysed with 1% Triton X-100 and 60 mM n-octyl-�-D-glucopyrannoside in TBS plus inhibitors for 30 min at 37 °C. After briefcentrifugation, aliquots of lysates containing 50 �g of total protein wereprecipitated with 40% trichloroacetic acid on ice and resolved by non-reducing SDS-PAGE (NuPAGE® MES 4–12% polyacrylamide gradientgel, Invitrogen). After electrotransfer, the polyvinylidene difluoride mem-branes were blocked for 1 h with 10% FCS and 0.5% Tween 20 in phos-phate-buffered saline (PBS), incubated for 1 h with primary antibodydiluted in PBS containing 1% bovine serum albumin and 0.1% Tween 20,washed with 0.3% bovine serum albumin and 3% Tween 20 in PBS, andthen incubated for 1 h with appropriate secondary antibodies conjugatedwith alkaline phosphatase or horseradish peroxidase. The blots weredeveloped with the DAB substrate kit for horseradish peroxidase or Vec-tor Red AP substrate kit I (Vector Labs).

Triton X-114 Cell Extraction—Cell monolayers on ice were scraped inice-cold 0.1 M TBS, pH 8.1, plus inhibitors, and lysed in pre-condensed1% Triton X-114 in the same buffer by incubating for 1 h on ice withrepeated mixing. Debris was removed by centrifugation and phaseseparation was induced for 5 min at 37 °C, followed by brief centrifu-gation to separate detergent and aqueous phases. The same bufferlacking detergent was added and the extraction repeated once beforethe separated phases were assayed or frozen down. In some experi-ments cells were treated with 1 unit/ml phosphatidylinositol-phospho-lipase C (Glyko) for 1 h at room temperature before extraction withTriton X-114.

Immunofluorescence—Immunofluorescence was carried out as de-scribed (10). The effect produced by FVIIa/FX on the distribution of TF,TFPI, and Cav-1 was studied in cells fixed with 3% paraformaldehydein PBS (1 h at room temperature) at the end of the FX activation assay.Intracellular immunostaining was achieved on cells permeabilized with0.1% Triton X-100.

Images were collected using a Nikon Eclipse inverted microscopeequipped with a confocal Nikon C1 system, using a computer-controlled488-nm argon laser to excite EGFP, and a helium-neon laser whosegreen line (543-nm) excites Cy3 and red line (633-nm) excites Cy5 orTO-PRO (nuclear stain). Samples mounted with Vectashield (Vector)were observed with a PlanApochromat oil-immersion objective (�60,NA: 1.4). Images were processed using MetaMorph (Universal Imag-ing). The overlap was quantified using Adobe Photoshop (Adobe Sys-tems), by determining the % of protein A that co-localizes with proteinB (29). A-channel minus B-channel gave the non-overlapping A ele-ments. Their intensity was subtracted from the total A-channel, and theresulting overlap signal was expressed as mean % of the total signal foreach channel. Triple co-localization was analyzed similarly. Fluores-cence intensity measurements were performed on a minimum of 50cells/group, randomly chosen from at least five pictures for each exper-imental condition. Experiments were repeated three to four times.

Statistical significance of the differences between groups was deter-mined by t test (Microsoft Excel), and the differences were consideredsignificant when p � 0.05. Descriptive statistics (median values andrange) and correlation analysis were performed with InStat (Macintosh).

Fluorescence Recovery after Photobleaching (FRAP)—For live cellrecordings, 293 cells stably expressing EGFP-TFPI, Cav1-HPC4ep, orboth, were grown on coverslips and mounted in a perfusion chamber at37 °C (Bioptech FCS2). The chamber was installed on the stage of theinverted microscope equipped with a temperature controlled �63 oilimmersion objective. The chamber was perfused with phenol red-freeculture medium at 37 °C. Bleaching in 3–5 spots in regions of interestwas performed at 37 °C with the 488-nm laser line at full power and fulltransmission for 2 s. Observation of fluorescence recovery over time wasdone at full laser power and 1% transmission, to avoid significantphotobleaching. Image J software (public domain) was used to measurepixel intensities in regions of interest, and to correct for overall bleach-ing by comparison with areas not exposed to the full power of the laser.The % recovery (recovery values � mean of pre-bleach values � 100)shows how much fluorescence returns to the area out of the amount offluorescence before photobleaching.

Depletion of Cav-1 in EA.hy926 and HUVEC by RNAi—Post-tran-scriptional silencing of the Cav-1 expression was achieved with two

Caveolin-1 Modulates TFPI Function 22309

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

duplex RNA oligonucleotides (Dharmacon Research), each composed of21 bases of sequence that is conserved in Cav-1 from several mamma-lian species (30). The sense strands of the duplexes were 5�-CCUgAUU-gAgAUUCAgUgCdTdT-3� and 5�-gAgAAgCAAgUgUACgAcgdTdT-3� (31).EC were seeded on coverslips to reach �60% confluence on the followingday. Silencing (si) RNA duplexes (25 nM final concentration) mixed withTransIT-TKO (Mirus) in Opti-MEM I were added to the cells in com-plete medium without antibiotics. Cells were incubated for 2–3 daysand analyzed by confocal microscopy after immunostaining. Controlswere as described under supplemental materials.

Supplemental Data—Details of the following protocols are avail-able: construction and expression of pEGFP-C2/TFPI(C) and pSVZeo-Cav1-HPC4ep, primers, and conditions for reverse transcriptase-PCR, characterization, and selection of 293 cell clones, and RNAiexperimental controls.

Fig. S1 demonstrates that EGFP-TFPI has distribution and functionsimilar to natural TFPI, therefore is a reliable marker for endogenousTFPI. Fig. S2 displays immunofluorescence characterization of the chi-meras expressed in 293 cells. Fig. S3 details the testing and selection of293 clones. Results of the RNAi controls are illustrated in Figs. S4 andS5 and demonstrate that the effects induced by Cav-1 siRNA transfec-tion of EC on TFPI distribution and activity are because of the silencingof Cav-1 expression.

RESULTS

EGFP-TFPI Is a Reliable Marker for Endogenous TFPI—Using EA.hy926, we compared the expression and activity ofEGFP-TFPI (a chimera with EGFP fused at the N terminus ofTFPI) with the properties of native TFPI. Western blottingshowed a protein with molecular mass �65,000 Da immunore-active with both anti-GFP and anti-TFPI IgGs, which wereabsent from native (N) and cells expressing EGFP only (GFP)(Fig. S1, A). The distribution and properties of native andchimerical TFPI were similar. Fluorescence imaging showedthat 98% of EGFP-TFPI overlapped the anti-TFPI IgG staining(Fig. S1, B, a, d, and e), indicating that the EGFP signaloriginated from the chimera and not from EGFP alone. BothEGFP-TFPI and native TFPI co-localized with endogenousCav-1 (b, c, and f).

Enzyme-linked immunosorbent assay (ELISA) for TFPI an-tigen confirmed that EC overexpressing EGFP-TFPI synthe-sized and secreted two times more TFPI than native cells (Fig.S1, C, a). TFPI-mediated inhibition of exogenously added TF-FVIIa was similarly increased (Fig. S1, C, b). EGFP-TFPImatched the Triton X-114 partition properties of native TFPI,a finding that suggests direct or indirect glycosylphosphatidy-linositol anchoring. Enzymatic cleavage of the glycosylphos-phatidylinositol anchor with phosphatidylinositol-phospho-lipase C (10, 11, 32) substantially reduced TFPI partitioninginto the detergent phase of Triton X-114 (Fig. S1, C, c).

Taken together, these data establish that EGFP-TFPI is avalid fluorescent indicator of native TFPI, and can be used tostudy the distribution and function of TFPI in live as well as infixed cells.

Analysis of the Expression of EGFP-TFPI and Cav1-HPC4epin 293 Cells—In HEK293 transiently expressing Cav1-HPC4ep(a chimera where the small HPC4 epitope was fused to the Cterminus of Cav-1), immunostaining with mAb HPC4 (24) re-vealed the typical distribution of native Cav-1 (Fig. S2, a–c). Incells expressing EGFP-TFPI 99% of the chimera overlapped theimmunolabeled TFPI, showing that no cleaved EGFP was pres-ent (Fig. S2, d–f). Co-localization of each protein with its tagexceeded 90% (Fig. S2).

Five stable cell clones were selected for each experimentalcondition. The cell clones were re-named: “TFPI” (expressEGFP-TFPI only); “Cav” (cells express Cav1-HPC4ep alone);and “TFPI � Cav” (express both proteins). As detailed in Fig.S3, the total level of TFPI was not significantly different be-tween TFPI cells and TFPI � Cav cells (panel A). Although theconstitutive secretion of TFPI was significantly decreased in

the presence of Cav-1. The capability of TFPI cells to preventendogenous TF-dependent FX activation was significantly di-minished as compared with TFPI � Cav cells (Fig. S3, B).

Native (wild type, wt) cells were also tested, either quiescent orafter transfection with pEGFP-C2. There was no significant var-iation in TFPI distribution or function between these two condi-tions; therefore we used wt cells in all subsequent experiments.

Wt and Cav cells expressed very low levels of TFPI mRNA,whereas TFPI cells and TFPI � Cav cells expressed mRNA forboth TFPI and EGFP-TFPI (Fig. 1A, semi-quantitative reversetranscriptase-PCR). Lysates of TFPI cells and TFPI � Cav cellsdisplayed a protein band of �65 kDa that was immunoreactivewith both anti-GFP and anti-TFPI IgGs, but was absent fromwt cells (Fig. 1B). Cav cells and TFPI � Cav cell lysates containa protein that was immunoreactive with anti-Cav-1 IgG (band

FIG. 1. Analysis of the expression of EGFP-TFPI and Cav1-HPC4ep in HEK293. A, reverse transcriptase-PCR of native TFPI andEGFP-TFPI in 293 cell clones. wt, wild type; C, Cav cells; T, TFPI cells;TC, TFPI � Cav cells. �-Actin, internal control. B, Western blot anal-ysis of protein expression. 293 cell lysates were resolved by SDS-PAGE,transferred to polyvinylidene difluoride, and the blots probed withanti-GFP, anti-Cav-1, and anti-TFPI IgGs. EA, EA.hy926 lysate (posi-tive control for Cav-1). C, immunostaining for TFPI and Cav-1 in thecell clones. Permeabilized cells were double labeled with rabbit anti-TFPI IgG/donkey anti-rabbit IgG-Cy3 (a, d, g, and j) and mouse anti-Cav-1 IgG/donkey anti-mouse IgG-Cy5 (b, e, h, and k). Double co-localization appears in purple (c, f, i, and l). Bars, 20 �m. D, TFPIantigen determination by enzyme-linked immunosorbent assay in Tri-ton X-114 extracts of 293 cell clones after temperature-induced phaseseparation. Values are mean � S.D. from triplicate assays.

Caveolin-1 Modulates TFPI Function22310

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

at �20 kDa), present only as trace in lanes “wt” and “T.” Theequivalent band appeared in EA.hy926 lysates used as positivecontrol (lane EA). Immunofluorescence confirmed that wt cellscontained very low levels of TFPI and Cav-1 (Fig. 1C, a and b),and that EGFP-TFPI largely co-localized with Cav-1 (j-l).

Similar to TFPI in EC, EGFP-TFPI in 293 cells partitionedpredominantly in the detergent phase of Triton X-114 (Fig. 1D).The equivalent of the 1:3 ratio between the water-soluble anddetergent-soluble TFPI normally found in EC was achieved in293 cells only in the presence of Cav-1 (Fig. 1D). The totalamount of TFPI was not significantly different among the TFPIcells and TFPI � Cav cells.

Cav-1 Expression in 293 Cells Induces Formation of Caveo-lae—We used freeze-fracture EM to analyze the formation ofinvaginations/caveolae in wt and Cav cells. In wt cells (Fig. 2a)the external leaflet of the membrane displayed a smooth as-pect, hence lack of caveolae. Cav cells exhibited pits with di-mensions and aspect typical of caveolae (Fig. 2b), thus confirm-ing that expression of Cav-1 in deficient cells inducedinvagination of caveolae.

Kinetics of Fluorescent TFPI Were Different in the Presenceand Absence of Cav-1—We used FRAP to quantify the kineticproperties of EGFP-TFPI in the plasma membrane of livingcells. The fluorophore was bleached in “regions of interest” atthe peripheral EGFP-TFPI-labeled rim of each cell and thefluorescence recovery was followed for 10 min. Non-bleachedmolecules diffused into the bleached areas according to theirmembrane mobility.

In the presence of Cav-1, clusters next to regions of interestmaintained their fluorescence after bleaching and did not moveinto the bleach area, suggesting that only a little lateral diffu-sion of fluorescent TFPI took place (not shown). Fig. 3 showsthe FRAP curves after curve fitting of the means of 10 inde-pendent recordings for each cell clone. The recovery of EGFP-TFPI in TFPI � Cav cells was much slower than in TFPI cells.The mobile fraction of EGFP-TFPI was �35% in the presence ofCav-1 and �85% in its absence (Fig. 3, p � 0.01).

Expression of Cav-1 in 293 Cells Increases Both Cell SurfaceTFPI and the Inhibition of FVIIa-TF by TFPI—From the meas-urement of TFPI and Cav-1 fluorescence intensity after immu-nostaining (Fig. 4A), it results that expression of Cav-1 in TFPI �Cav cells doubled or even tripled the amount of TFPI exposedon the cell surface (Fig. 4A, compare TFPI � Cav cell clones(TC) 1, 2, 4, and 5, with TFPI cell clones). Only TFPI � Cavcells displayed a significantly positive correlation (r2 � 0.92)between Cav-1 expression and the level of cell surface TFPI. Todetermine the functional activity of the surface exposed TFPI,we assessed its inhibitory potency, both against exogenouslyadded TF, and toward the endogenous TF, which is normallyexpressed by HEK293.

The functional activity of cell surface TFPI tested againstexogenous TF-FVIIa was significantly different between theclones. We measured the capability of the TFPI exposed on thecell surface to prevent the activation of FX by pre-formedcomplexes consisting of 10 nM FVIIa and 1:15 diluted throm-boplastin added in the overlying cells medium. The activity ofTFPI, expressed as arbitrary units, was extrapolated fromstandard curves made with serial dilutions of normal humanplasma, run in parallel in identical conditions. In Fig. 4B weillustrated for each cell clone the activity of cell surface TFPIexpressed as milliunits, together with the corresponding TFPIantigen (ng) measured by enzyme-linked immunosorbent assayon cell monolayers (all values were normalized to 106 cells).

TFPI cells displayed a significant increase of both antigenand activity of cell surface TFPI as compared with wt cells (p �0.01 for all the clones). Nevertheless, there was no positivecorrelation between the activity and either TFPI or Cav-1 an-tigen levels (Fig. 4C, r2 � 0.31 and 0.35, respectively). In TFPI �Cav cells, increases of up to five times of cell surface antigen ledto an enhancement of TFPI activity of as much as 40 times overthe wt cell levels (Fig. 4B, right panel; note the different scaleof the y axis). In comparison with TFPI cells, the average2.5-fold increase of TFPI antigen on the cell surface of TFPI �Cav cells could be translated into as much as 10 times higherinhibitory potency (Fig. 4B). After normalization for TFPI an-tigen levels it results that TFPI on TFPI � Cav cells were, onaverage, 3.5 times more potent than TFPI on TFPI cells in

FIG. 2. Expression of Cav-1 in deficient cells induces forma-tion of caveolae. Wt and Cav1-HPC4ep-expressing 293 cells wereprocessed by freeze-fracture EM. During freeze-fracture, the plasmamembrane is split horizontally through the hydrophobic region of thebilayer, exposing the inner fracture faces of the two leaflets. View ofthe external leaflet of the plasma membrane shows a smooth aspect ofthe membrane in wt cells (a), and the presence of pits typical of caveolaein Cav cells (b, arrowheads). Bar, 500 nm.

FIG. 3. Cav-1 affects the lateral mobility of EGFP-TFPI in theplasma membrane. Comparative FRAP of EGFP-TFPI was performedin TFPI � Cav and TFPI cells. Similar bleach regions were selected atthe plasmalemma level in the two cell clones and all the parameterswere kept constant. FRAP was repeated three times for each cell clone.The fluorescence recovery curves illustrate that EGFP-TFPI in TFPI �Cav cells (bottom curve) recovered much slower than in TFPI cells(top curve).

Caveolin-1 Modulates TFPI Function 22311

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

inhibiting exogenous TF-FVIIa activity. Specificity controls in-cluded parallel assays run in the presence of 50 �g/ml anti-TFPI IgG, which abrogated almost entirely the activity of TFPI(not shown).

Only in Cav-1-expressing cells did the activity of cell surfaceTFPI correlate positively with the TFPI antigen, and implicitlywith Cav-1 levels (Fig. 4C, r2 � 0.95 and 0.93, respectively).The inhibitory activity of cell surface TFPI was severely im-paired in the absence of Cav-1, even for cells that displayedequal or higher amounts of TFPI antigen (Fig. 4C, compareTFPI cell clones 4, 3, 5, and 1, with TFPI � Cav cell clone 3).Conversely, the presence of Cav-1 alone was enough to bringthe activity of even very small amounts of TFPI antigen, suchas the one present in Cav cells, up to levels that were equivalentwith the activities measured on TFPI cells (Fig. 4B, compare Cavcell clones 3, 4, and 5 with TFPI cell clones 2, 4, and 5).

Next, we tested the inhibitory potency of TFPI against en-dogenous TF on TFPI cell clone 1 and TFPI � Cav cell clone 4.TF mRNA determined by reverse transcriptase-PCR and totalTF antigen in cell lysates were comparable (not shown), indi-cating that the expression of TF was similar among the 293 cellclones. TFPI cells and TFPI � Cav cells displayed similaramounts of cell surface TF antigen (11 and 12.2 pM, respec-tively). FX was efficiently activated after the addition ofFVIIa to 293 cell monolayers, following a brief incubationwith ionomycin to expose the cryptic cell surface TF and/orphosphatidylserine. Inclusion of cycloheximide precludedchanges because of de novo protein synthesis. Preincubationof cells with inhibitory anti-TF mAb 9-5B7 blocked the gen-eration of FXa, confirming that FX activation was dependenton TF-FVIIa.

The overall capability of the cells to activate FX, assessedafter inhibiting the available TFPI with anti-TFPI IgG, wasdefined as “total FXa.” FXa generation curves for both TFPIcells and TFPI � Cav cells were linear over time for eachconcentration of FVIIa tested (Fig. 5A). The specific potential ofthe cells to generate FXa in the presence of TFPI, assessedusing non-inhibited assays, was significantly different betweenthe two cell clones (Fig. 5A). Whereas FXa generation increasedcontinuously on TFPI cells, the activation of FX on TFPI � Cavcells stayed at very low levels, and, if it increased slightly, itwas with a significant delay and slower rate. The differences ofFXa generation were larger than what would be accounted foronly by the differences in cell surface TFPI antigen: 45 pM inTFPI cells and 100 pM in TFPI � Cav cells.

The differences between total FXa (open symbols in Fig. 5)and the FXa generated in 15 min in the absence of anti-TFPIIgG (closed squares and circles) represent the amount of FXwhose activation was prevented by TFPI, therefore giving themeasure of the functional activity of cell surface TFPI (closedtriangles). As shown in Fig. 5B, TFPI � Cav cells displayedsignificantly higher capabilities to prevent FXa generationthan TFPI cells for all three concentrations of FVIIa tested (p �0.01). Even after normalization of the values to account for thedifferences in the cell surface antigen, TFPI on TFPI � Cav

FIG. 4. Cav-1 supports TFPI-dependent inhibition of the pro-coagulant activity of exogenous TF-FVIIa. A, measurement ofTFPI and Cav-1 fluorescence intensity on each of the five clones wasperformed after immunostaining for cell surface TFPI (non-permeabi-lized cells, rabbit anti-TFPI IgG/donkey anti-rabbit IgG-Cy3) and in-tracellular Cav-1 (permeabilized cells, mouse anti-Cav-1 IgG/donkeyanti-mouse IgG-Cy5). The positive correlation between fluorescenceintensities of Cav-1 and cell surface TFPI was highly significant (r2 �0.92) only for TFPI � Cav cells. B, cell surface TFPI inhibits FXactivation by pre-formed TF-FVIIa complex added in the medium. Theactivity values of cell surface TFPI (black bars) for each cell clone wereextrapolated from the inhibitory potency measured for TFPI in seriallydiluted normal human plasma, which is arbitrarily assigned 1 unit/ml.The amount of TFPI antigen on the cell surface determined by enzyme-linked immunosorbent assay (ng, gray bars) is given for comparison.Values are mean � S.D. of triplicate assays. Both TFPI activity andantigen were normalized for cell numbers (106 cells). C, correlationanalysis was performed for each cell clone to analyze the relationshipbetween cell surface TFPI activity and either TFPI antigen or Cav-1levels. Only Cav-1-expressing cells displayed significant positive corre-lation for both conditions (r2 � 0.95 and 0.93, respectively). n � 100 forall the conditions. For simplicity, we represented median values offluorescence intensities for each clone.

FIG. 5. 293 cells promote endogenous TF-dependent activationof FX in the presence of different concentrations of FVIIa. A,FXa generation is shown over time on TFPI cells (circles) and TFPI �Cav cells (squares) in the presence (open symbols) or absence (closedsymbols) of 50 �g/ml inhibitory anti-TFPI IgG, for each of the threeconcentrations of FVIIa tested (a–c). The difference between the twoexperimental conditions, which gives the measure of the inhibitorycapability of TFPI, is represented in panel B as a function of the addedFVIIa concentration (triangles; a, TFPI � Cav cells; b, TFPI cells). Allvalues are mean � S.D. of triplicate determinations.

Caveolin-1 Modulates TFPI Function22312

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

cells was still �2 times more active in inhibiting FX activationthan on TFPI cells.

Cav-1 Co-localizes with TF-FVIIa-FXa-TFPI in 293 Cells—Using fluorescence microscopy we studied the spatial relation-ship between Cav-1, TF, and TFPI in the presence of FVIIa/FX.Cells kept in assay buffer (control) or incubated with FVIIa andFX were fixed with 3% paraformaldehyde in PBS. Cell surfaceTF was immunostained on non-permeabilized cells with mAb10H10 followed by donkey anti-mouse IgG-Cy3. EGFP-TFPIwas visualized through its intrinsic fluorescence. To detectCav-1, and thus to observe the relationship between the cellsurface TF and TFPI, and the submembrane Cav-1, the cellswere permeabilized with 0.1% Triton X-100 and immuno-stained with rabbit anti-Cav-1 IgG followed by donkey anti-rabbit IgG-Cy5.

In the absence of FVIIa/FX, TFPI cells displayed little to noco-localization between TF and TFPI, and the triple co-local-ization was scarce (Fig. 6A, a, yellow, and c, white). Quantita-tively, 20% of the total TF co-localized with TFPI, and �15% ofthe total TFPI co-localized with TF.

After adding FVIIa/FX to TFPI cells the overlap betweenTFPI and TF increased to 30% of both total TFPI and total TF.TFPI patched with TF over the traces of Cav-1 present (Fig. 6A,d–f) resulting in the odd “rosette” formation (f, arrow).

50–60% of TFPI co-localized with Cav-1, and �40% of TFPIoverlapped both TF and Cav-1 in normal TFPI � Cav cells (Fig.6B, Control, a–c). After incubation with FVIIa-FX (Fig. 6B, d–f)

the patching of TFPI with TF increased to 70–80% co-localiza-tion (p � 0.01), and both overlapped Cav-1 (d, yellow, and f,white). Formation of rosettes frequently occurred on the apicalsurface of TFPI � Cav cells (Fig. 6B, inset in f) where TF, TFPI,and Cav-1 overlapped almost perfectly (white).

Such co-localization was no longer observed if TFPI � Cavcells were incubated with inhibitory anti-TFPI IgG before ad-dition of FVIIa. The activity assays indicated that this treat-ment blocked the capability of TFPI to inhibit FX activation.Fluorescence microscopy now showed that TF largely failed toco-localize with TFPI and/or Cav-1 (not shown). The percentageof TF overlapping TFPI, Cav-1, or both was brought down tothe control levels for all the conditions.

Silencing of Cav-1 Expression in EC Decreases the SurfaceExposure and Activity of TFPI—The role of Cav-1 in the distri-bution and function of TFPI in EC was studied by siRNAdepression of Cav-1. Incubation of EA.hy926 and HUVEC withtwo siRNA duplexes homologous to highly conserved sequencesin cav1, from several species, led to a significant decrease ofCav-1 expression, as assessed by immunofluorescence (Fig. 7,EA.hy926, compare A, a with B, a). For quantification, imagesof mock and siRNA cells were collected keeping the confocalparameters unchanged. The intensity of fluorescence was

FIG. 6. TFPI and TF co-localize with Cav-1 after formation ofthe quaternary complex. Green, EGFP-TFPI chimera visualizedthrough its intrinsic fluorescence. Red, mAb anti-TF 10H10/donkeyanti-mouse IgG-Cy3. Blue, rabbit anti-Cav-1 IgG/donkey anti-rabbitIgG-Cy5. Non-permeabilized cells were immunostained for cell surfaceTF, then permeabilized and immunolabeled for Cav-1. The distributionof the three proteins is shown for cells stably expressing EGFP-TFPI (A)or both EGFP-TFPI and Cav1-HPC4ep (B), in the absence (Control) orpresence of FVIIa � FX. For all the images: yellow, co-localizationbetween EGFP-TFPI and TF; aqua, EGFP-TFPI and Cav-1; and white,triple co-localization EGFP-TFPI/TF/Cav-1. Arrow in A (f) and inset inB (f) show rosettes, regions of the apical plasma membrane where Cav-1accumulates together with TFPI and TF. Bars, 20 �m. Bar for inset,1 �m.

FIG. 7. Silencing of Cav-1 expression decreases cell surfaceTFPI. EA.hy926 cells and HUVEC transfected with Cav-1 siRNA du-plexes were studied by immunostaining and confocal microscopy. A,distribution of Cav-1 (a, rabbit anti-Cav-1 IgG/goat anti-rabbit IgG-FITC) and cell surface TFPI (b, rabbit anti-TFPI IgG/goat anti-rabbitIgG-FITC) in mock-transfected EA.hy926. B, double immunostainingfor Cav-1 (a, rabbit anti-Cav-1 IgG/goat anti-rabbit IgG-FITC) and cellsurface TFPI (b, mAb TFPIK-9/donkey anti-mouse IgG-Cy3) onEA.hy926 transfected with Cav-1 siRNA. Blue, TO-PRO (nuclear stain-ing). Bars, 20 �m. C, median values and range of fluorescence intensity(arbitrary units) for Cav-1 and cell surface TFPI in HUVEC. Mock, ECincubated with transfection medium only; Cav-1 siRNA, cells trans-fected with Cav-1 siRNA. D, the positive correlation between fluores-cence intensity of Cav-1 and cell surface TFPI in both control (mock-transfected) and Cav-1 siRNA-treated HUVEC was highly significant(r2 � 0.86 and 0.95, respectively). Number of cells: n � 201 for mockcells and n � 174 for siRNA cells.

Caveolin-1 Modulates TFPI Function 22313

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

measured on three groups of images taken during three differ-ent experiments (total evaluated: 170–200 cells for each con-dition). Results were similar between HUVEC and EA.hy926.Median and range values of the fluorescence intensity meas-ured on HUVEC are represented in Fig. 7C. Cav-1 fluorescencein siRNA-treated HUVEC was �6-fold diminished as comparedwith mock cells (p � 0.001). The fluorescence intensity of Cav-1was �10% of mock cells levels in 50% of the siRNA cells. 18%of the siRNA-treated cells exhibited fluorescence intensityequivalent to mock cells. The mean fluorescence intensity ofthe remaining cells was 20% (�5) of the mock cell levels. TFPIexposed on the surface of Cav-1 siRNA cells was also decreased,for both EA.hy926 (Fig. 7, compare A, b with B, b, and Fig. S4,compare A with C, b), and for HUVEC. TFPI fluorescence was�3-fold diminished versus mock cells (Fig. 7C, p � 0.01). TFPIfluorescence intensity was �20% of the mock cells level in 40%of the siRNA cells. 16% of the cells displayed fluorescenceintensity for TFPI similar with mock cells. The remainderdisplayed fluorescence intensity of 30% (�7) of the mock cellslevels.

The surface exposure of TFPI correlated positively andhighly significantly with the expression of Cav-1, for bothEA.hy926 (not shown) and HUVEC, as indicated by the corre-lation analysis of the fluorescence intensity performed on 201cells (mock) and 174 cells (Cav-1 siRNA) following double im-munostaining for TFPI and Cav-1 (Fig. 7D, r2 � 0.86 for mockcells, and r2 � 0.94 for siRNA cells).

The RNAi experimental controls (see supplemental materi-als) verified that the above described effects were due indeed tospecific Cav-1 gene silencing. Neither TFPI nor Cav-1 wereaffected by transfection of cells with siGLO RISC-Free siRNA,a non-functional non-targeting siRNA (Fig. S4, A and B), butdecreased drastically when the cells were co-transfected withCav-1 siRNA (Fig. S4, C). Positive controls consisted of ECtransfected with siCONTROL lamin A/C siRNA (Fig. S5).Whereas the expression of lamin A/C was strongly decreased(compare A, b and c, with B, a and d), neither cell surface TFPI(B, b and c) nor Cav-1 (B, e and f) were affected by depressionof lamin A/C.

Silencing of Cav-1 Decreases the Activity of TFPI on the ECSurface—The activity of TFPI was assessed through the inhi-bition of TF-FVIIa-mediated activation of FX. To study endog-enous TF-dependent FXa generation, we induced TF expres-sion in EA.hy926 and HUVEC through incubation with 10ng/ml r-tumor necrosis factor-� for 5 h (16). Fig. 8A displaysFXa generation curves on mock-transfected and Cav-1 siRNA-treated HUVEC. The overall capability of the cells to activateFX (total FXa, open symbols), assessed in the presence of anti-TFPI IgG, was linear over time for all three concentrations ofFVIIa tested, as well as similar among mock- and Cav-1siRNA-transfected HUVEC, thus suggesting equivalent ex-pression of TF.

FXa generation measured in non-inhibited assays (Fig. 8A,closed squares and circles) varied with the FVIIa concentration,as well as between the mock- and siRNA-transfected cells.Whereas the cells in both groups proved efficient in inhibiting FXactivation at 0.2 nM FVIIa (Fig. 8A, a), only TFPI on mock cellswas able to keep FX activation at low levels for 10 and 50 nM ofadded FVIIa (Fig. 8A, b and c). Meanwhile, the generation of FXaincreased constantly on Cav-1 siRNA-transfected HUVEC.

The amount of FX whose activation was prevented by TFPIon HUVEC was calculated for each condition as described for293 cells. As shown in Fig. 8B (closed triangles), the inhibitorycapability of TFPI on Cav-1 depressed cells was significantlyimpaired (�6 times diminished, p � 0.005). After normalizingfor differences of the surface TFPI antigen, 150 pM for normal

HUVEC and 63 pM for Cav-1 siRNA-transfected cells, the twocell groups had similar inhibitory potency against the low levelof FXa generated by 0.2 nM FVIIa (�80% of the total FXa).When challenged with larger amounts of FVIIa and faster ratesof FX activation, TFPI on mock cells preserved its inhibitorypotency almost intact, being still capable of preventing thegeneration of 60–80% of the total FXa in the presence of 10 and50 nM FVIIa. Conversely, equimolar levels of TFPI were �3times less potent inhibitors of FXa generation (p � 0.0001) inthe absence of Cav-1, thus allowing the activation of FX toproceed almost uninhibited (Fig. 8B, b).

Similar differences were observed when the inhibitory effi-ciency of cell surface TFPI was tested against exogenous TF-FVIIa (Fig. 8C). Although TFPI in mock cells proved equallyefficient in inhibiting FX activation for all the concentrations ofTF-FVIIa tested, the inhibitory capabilities of TFPI could di-minish by as much as 7-fold in Cav-1 siRNA-transfected HU-VEC in the presence of 50 nM FVIIa (Fig. 8C).

The inhibitory activity of TFPI in EA.hy926 was also signif-icantly diminished following Cav-1 depression, albeit to a lowerextent than in HUVEC. After normalization of the values forcell surface antigen, we found that the capability of TFPI toinhibit TF-FVIIa-dependent generation of FXa was �1.5 timeslower in Cav-1 silenced cells than in mock cells (not shown).

Effect of Cav-1 Silencing on the Formation of the TF-FVIIa-FXa-TFPI Complex—At the end of the activity assay, bothHUVEC and EA.hy926 cells were fixed and double immuno-stained for cell surface TFPI (mAb TFPIK-9/horse anti-mouseIgG-FITC) and TF (goat anti-human TF IgG/donkey anti-goat

FIG. 8. Depletion of Cav-1 in HUVEC decreases TFPI activity.A and B, tumor necrosis factor-�-stimulated HUVEC promote endoge-nous TF-dependent activation of FX in the presence of different concen-trations of FVIIa. FXa generation over time is shown on mock cells(squares) and Cav-1 siRNA-treated cells (circles) in the presence (opensymbols) or absence (closed symbols) of 50 �g/ml inhibitory anti-TFPIIgG, for each of the three concentrations of FVIIa tested (panel A, a–c).As described for 293 cells, the capability of TFPI to prevent FX activa-tion is represented in panel B as a function of the added FVIIa concen-tration (triangles; a, mock cells; b, Cav-1 siRNA cells). C, TFPI exposedon the cell surface of non-stimulated HUVEC inhibits FX activation bypre-formed TF-FVIIa complexes (exogenous TF) added in the medium.The TFPI specific activity (milliunits/pmol) is represented for the threedifferent concentrations of FVIIa and thromboplastin tested. As op-posed to Cav-1-depleted HUVEC (circles), TFPI in normal cells (mockcells, squares) preserved almost unchanged its potency against FXactivation regardless of the concentration of FVIIa-TF added. All valuesare mean � S.D. of triplicate determinations.

Caveolin-1 Modulates TFPI Function22314

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

IgG-Cy5), then permeabilized and labeled with rabbit anti-Cav-1 IgG/donkey anti-rabbit IgG-Cy3. Fig. 9 illustrates thedistribution and overlap between the three proteins. Measure-ment of fluorescence intensity and calculation of co-localizationpercentages were done as for 293 cells. Mock cells exhibitedstrong patching of cell surface TF and TFPI over areas rich inCav-1 (Fig. 9, A, a–d, and B, a–d, �80% mean overlap), indi-cating that the quaternary complex did co-localize with caveo-lae (panels d, white indicates triple co-localization). The co-localization reached almost 100% on the apical surface and onthe cell edges (A and B, panels d, yellow arrows).

Such co-localization was no longer observed if EC were incu-bated with inhibitory anti-TFPI IgG before addition of FVIIa.Similar to the 293 cell clones (see above), the percentage ofTF overlapping TFPI, Cav-1, or both, decreased to �30% (notshown).

Although expressing lower intensity of fluorescence for cellsurface TFPI than mock cells (Fig. 9, A and B, panels e and f),the cells where Cav-1 was silenced displayed normal levels ofTF (panels g). Mean fluorescence intensity for TF (arbitraryunits) was not significantly different among HUVEC andEA.hy926 cells. The values determined for mock cells and siRNAcells with normal levels of Cav-1 were 114 � 11 and 109 � 10(HUVEC), respectively, 100 � 11 and 90 � 12 (EA.hy926). ForCav-1 siRNA-transfected cells with low levels of Cav-1 wemeasured 106 � 11 (HUVEC) and 95 � 7.2 (EA.hy926) arbi-trary units of TF fluorescence intensity. The response of Cav-1siRNA cells was heterogeneous, most likely because of therelatively significant proportion of cells in this population thatstill had normal levels of Cav-1 and TFPI (Fig. 9A, e–h; note the

very low magnification images of Cav-1 siRNA-transfectedHUVEC).

The triple co-localization TF/TFPI/Cav-1 was similar be-tween mock cells and the population of siRNA cells that hadnormal levels of Cav-1 (white; Fig. 9B, d and h, yellow arrows).In contrast, we observed very little co-localization between TFand TFPI and/or Cav-1 in Cav-1-depressed cells (�15% averageoverlap, Fig. 9, A and B, panels h).

DISCUSSION

We showed here that Cav-1 and/or caveolae actively regulatethe anticoagulant activity of TFPI, apparently through a dualmechanism: by stabilizing the exposure of the inhibitor on theplasmalemma surface and by creating a microenvironmentthat enhances the down-regulation of TF-FVIIa activity byTFPI.

Despite reports trying to identify the factors that regulatethe inhibition of TF-FVIIa activity by TFPI, the molecularmechanism of TF-FVIIa-FXa-TFPI complex formation is notyet fully understood. In ECV304 and monocytes, TFPI medi-ates the translocation of the quaternary complex to lipid rafts(11, 16, 18). In EC, where the majority of membrane TFPIresides in caveolae/lipid rafts, disturbing the rafts with choles-terol-extracting agents decreases the activity of TFPI (10). Be-cause membrane TFPI is now believed to play the major role indown-regulation of TF-FVIIa activity on cell surfaces, we de-cided to investigate whether Cav-1/caveolae regulate the TFPI-dependent inhibition of TF-FVIIa activity.

Our strategy involved first the expression of tagged full-length TFPI and Cav-1 in HEK293, a cell line that is naturallydeficient of TFPI and Cav-1. We found that the properties andfunctional activity of EGFP-TFPI matched those of the endog-enous TFPI in EA.hy926 cells. We thus confirmed that EGFP-TFPI is a valid fluorescent indicator of native TFPI. Next weconfirmed by immunofluorescence and freeze-fracture EM thatthe expression of Cav-1 in HEK293 induced the appearance ofcaveolae proper, as it was described for other cell types (21, 33).

Analysis of HEK293 expressing TFPI, Cav-1, or both pro-teins, by live cell and (immuno)fluorescence microscopy, and byfunctional assays, revealed novel aspects of TFPI distributionand function. In the absence of Cav-1, TFPI had predominantintracellular localization and was largely secreted. Withoutaffecting the overall levels of TFPI, Cav-1 expression instru-mented a shift in the distribution of TFPI, with the inhibitorbecoming predominantly associated with the plasma mem-brane surface and less secreted into the medium.

FRAP results showed that cell surface TFPI was signifi-cantly less laterally mobile in the presence of Cav-1. Understandard culture conditions Cav-1 becomes highly immobileonce it reaches the plasma membrane, and shows very limitedlateral diffusion and exchange with the intracellular pool (34).The fluorescence recovery profiles of EGFP-TFPI in TFPI �Cav cells (present paper, Fig. 3) and caveolae-associated GFP-tagged Cav-1 (35) are strikingly similar. We suggest that Cav-1/caveolae retain and stabilize TFPI on the membrane surface,hence the increased surface exposure and decreased lateralmobility of EGFP-TFPI in Cav-1 expressing cells.

The functional impact of TFPI localization in caveolae wasanalyzed through the inhibition of TF-FVIIa activity, as re-flected by inhibition of FXa generation. Activation of FX by apre-formed mixture of 10 nM FVIIa, which is the equivalent ofthe plasma level of FVII (7), and thromboplastin (crude TF)was almost completely inhibited by the TFPI on TFPI � Cavcells, but not on TFPI cells. As observed on the 293 cell clonesthat express different levels of TFPI and Cav-1, the inhibitoryactivity of TFPI was dependent on the level of antigen onlywhen Cav-1 was also present. In the absence of Cav-1, varia-

FIG. 9. Triple immunolabeling for TFPI, TF, and Cav-1 in ECshows the formation of the TF-FVIIa-FXa-TFPI complex. A, HU-VEC; B, EA.hy926. Non-permeabilized cells, either mock-transfected(a–d) or Cav-1 siRNA-transfected (e–h) were incubated with mAb TF-PIK-9 and goat anti-human TF IgG, followed by anti-mouse IgG-FITC(green) and donkey anti-goat IgG-Cy5 (blue). The cells were then per-meabilized and labeled with rabbit anti-Cav-1 IgG/anti-rabbit IgG-Cy3(red). Triple co-localization appears white in both panels (merge, d andh, yellow arrows). Bars, 20 �m.

Caveolin-1 Modulates TFPI Function 22315

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

tions of cell surface TFPI antigen had no noticeable impactupon the functional capabilities of the inhibitor. Accordingly,equal or even larger amounts of TFPI were always far lessactive against exogenous TF in TFPI cells.

When assessing the inhibitory potency of TFPI against en-dogenous TF, we used concentrations of FVIIa meant to spanboth sides of 10 nM (the equivalent of the plasma level of FVII).The higher concentration of FVIIa (50 nM) was intended tomimic locally increased levels of FVIIa. These may occur innormal circumstances, because of FVIIa interactions withphospholipids and/or proteoglycans (7), or in pathological con-ditions where a hypercoagulable state arises. Regardless of theconcentration of FVIIa added, equimolar amounts of TFPI pre-vented twice as much FXa generation in the presence of Cav-1.The increase in TFPI activity was independent of the level ofTF, which was also not affected by Cav-1 expression. TFPI �Cav cells were able to keep FX activation at very low levels forboth 10 and 50 nM added FVIIa, preventing, in both cases,�85% of total FX activation. The fact that TFPI cells werecapable, in similar conditions, to inhibit only �20% of FXactivation suggests that the presence of Cav-1 was essential forthe preservation of the TFPI inhibitory potential, especiallywhen high levels of FVIIa, and consequently FXa, were pres-ent. Altogether, these findings suggest that, when expressed innaturally deficient cells such as HEK293, Cav-1 actively en-hanced the anticoagulant activity of TFPI, probably through acombination of increased TFPI antigen retention on the cellsurface and microenvironment modifications brought about bythe formation of caveolae.

Next we verified that Cav-1 played a similar role in EC. Tothe best of our knowledge, this is the first direct proof thatcaveolae (or Cav-1) regulate the activity of TFPI against TF-FVIIa in EC. Using two well characterized duplex RNA oligo-nucleotides (30, 31), we achieved both Cav-1 depletion andconsiderable reduction of cell surface TFPI antigen and activityin both HUVEC and EA.hy926. This indicates that Cav-1 onthe cytosolic face of the membrane might control the exposureof TFPI on the plasmalemma surface.

Cav-1 depletion decreased the capability of TFPI to down-regulate the activity of endogenous TF-FVIIa, an effect thatwas more visible for high FVIIa concentrations. Similar with293 cells, equimolar levels of TFPI were �3 times less potent inpreventing TF-FVIIa-dependent FX activation in the absenceof Cav-1. As a result of combined lower cell surface TFPIantigen and diminished functionality, the procoagulant activ-ity of Cav-1 depressed EC could increase by as much as �5times.

TF-bearing circulating microparticles derived from leuko-cytes and/or other blood cells probably represent a very signif-icant source of procoagulant species in several diseased states.We sought to find out whether cell surface TFPI on EC couldinhibit the procoagulant activity of preformed TF-FVIIa com-plexes, used as an in vitro equivalent of microparticle-associ-ated TF-FVIIa. Accordingly, we assayed in HUVEC the inhib-itory activity of TFPI against several mixtures of FVIIa(including the 10 nM equivalent of plasma FVII) and crude TF.Depletion of Cav-1 decreased by �6-fold the potency of cellsurface TFPI to block FX activation. Furthermore, high levelsof FVIIa-TF and, consequently large amounts of FXa, couldoverwhelm the inhibitory capability of TFPI in the absence ofCav-1.

The nature of the interaction(s) that keeps TFPI in caveolaeis under investigation in our group. TFPI associates through adirect or indirect glycosylphosphatidylinositol anchor with lipidrafts (10, 17), and both cholesterol and sphingolipids influence

the association of TFPI with lipid rafts in EC.2 Because caveo-lae are highly enriched in cholesterol and sphingolipids, it isconceivable that these lipids mediate the caveolar distributionand function of TFPI. Whether such a mechanism is instru-mental only in cells that express caveolins, as suggested by thelack of effect of cholesterol depletion on the inhibition of TF-FVIIa by TFPI in HEK293 (36), remains to be determined.

As indicated by the fluorescence microscopy, a high percent-age of the cell surface TF and TFPI becomes co-localized withthe submembrane Cav-1 during and/or after the formation ofthe quaternary complex. Blocking TFPI-FXa and TFPI-FVIIainteractions with anti-TFPI IgG prevents the redistribution ofTF and the overlap of the complex with Cav-1. We suggest thatthe complex either forms within caveolae or reaches themafterward. Regardless of the mechanism, the association ofTFPI with caveolae plays the determinant role in targeting theTF-FVIIa-FXa-TFPI complex to caveolae.

How does Cav-1, a protein located on the inner surface of themembrane, influence localization and function of outer surfaceproteins such as TFPI and TF? In normal conditions, lipid-associated proteins, TFPI included, reside very briefly withintransient rafts, which are extremely small in size and verydynamic (37). Clustering of the small rafts and their associatedproteins leads to the formation of larger, less mobile rafts thatcan be further stabilized by Cav-1. When cells are exposed toFVIIa/FX, TFPI forms complexes with TF-FVIIa/FXa regard-less of the presence of Cav-1. In Cav-1-deficient cells, whichlikely contain less cholesterol in the membrane (35), thestrength of the interaction between the complex and rafts maybe too low, or the residence time too short, to “lock” the qua-ternary complex in the condensed cholesterol-sphingolipid do-mains. If the complex dissociates, a dynamic equilibrium be-tween the formation and the dissociation of the complex wouldbe expected, with a significant fraction of TF-FVIIa being leftfree to generate FXa at any given moment. This would explainthe continuous generation of FXa in Cav-1-deficient cells. Thepresence of Cav-1, which poses a barrier into the lateral mo-bility of TFPI, would produce a longer lasting inhibitory effectonce the complex is stabilized in the cholesterol-sphingolipid-rich environment of caveolae/rafts.

Our results suggest that caveolae could concentrate the in-teracting molecules in particular regions of the cell surface andfacilitate the formation of the quaternary complex. The processmay have high biological relevance because it identifies Cav-1and/or caveolae as a key factor in the regulation of TFPI-de-pendent inhibition of TF-driven coagulation pathway in EC. Assuch, our findings convey a novel function to caveolae, namelyactive regulation of hemostasis.

Acknowledgments—We thank Dr. D. Qu and Dr. C. Esmon in ourinstitute for the pSVZeo-Cav1-HPC4ep rDNA and mAb HPC4,Dr. J. Morrissey for anti-TF antibodies, Dr. T. Hamuro for mAb TF-PIK-9, Dr. J. Crawley (Imperial College, London, United Kingdom) forinitial work on EGFP-TFPI chimeras, and Dr. C.-J. Edgell for thehybrid cell line EA.hy926. We are grateful to Dr. C. Esmon for contin-ued support and suggestions. We thank Drs. C. Esmon, R. McEver, andJ. Morrissey for critical reading of the manuscript.

REFERENCES

1. Drake, T. A., Cheng, J., Chang, A., and Taylor, F. B., Jr. (1993) Am. J. Pathol.142, 1458–1470

2. Contrino, J., Hair, G. A., Schmeizl, M. A., Rickles, F. R., and Kreutzer, D. L.(1994) Am. J. Pathol. 145, 1315–1322

3. Crawley, J., Lupu, F., Westmuckett, A. D., Severs, N. J., Kakkar, V. V., andLupu, C. (2000) Arterioscler. Thromb. Vasc. Biol. 20, 1362–1373

4. Westmuckett, A. D., Lupu, C., Roquefeuil, S., Krausz, T., Kakkar, V. V., andLupu, F. (2000) Arterioscler. Thromb. Vasc. Biol. 20, 2474–2482

5. Roque, M., Reis, E. D., Fuster, V., Padurean, A., Fallon, J. T., Taubman, M. B.,Chesebro, J. H., and Badimon, J. J. (2000) J. Am. Coll. Cardiol. 36,

2 C. Lupu, X. Hu, and F. Lupu, unpublished results.

Caveolin-1 Modulates TFPI Function22316

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2303–23106. Ruf, W., and Mueller, B. M. (1996) Curr. Opin. Hematol. 3, 379–3847. Rottingen, J. A., Enden, T., Camerer, E., Iversen, J. G., and Prydz, H. (1995)

J. Biol. Chem. 270, 4650–46608. Broze, G. J., Jr., Warren, L. A., Novotny, W. F., Higuchi, D. A., Girard, J. J.,

and Miletich, J. P. (1988) Blood 71, 335–3439. Nordfang, O., Bjorn, S. E., Valentin, S., Nielsen, L. S., Wildgoose, P., Beck,

T. C., and Hedner, U. (1991) Biochemistry 30, 10371–1037610. Lupu, C., Goodwin, C. A., Westmuckett, A. D., Emeis, J. J., Scully, M. F.,

Kakkar, V. V., and Lupu, F. (1997) Arterioscler. Thromb. Vasc. Biol. 17,2964–2974

11. Ott, I., Miyagi, Y., Miyazaki, K., Heeb, M. J., Mueller, B. M., Rao, L. V., andRuf, W. (2000) Arterioscler. Thromb. Vasc. Biol. 20, 874–882

12. Lupu, C., Poulsen, E., Roquefeuil, S., Westmuckett, A. D., Kakkar, V. V., andLupu, F. (1999) Arterioscler. Thromb. Vasc. Biol. 19, 2251–2262

13. Bajaj, M. S., Kuppuswamy, M. N., Saito, H., Spitzer, S. G., and Bajaj, S. P.(1990) Proc. Natl. Acad. Sci. U. S. A. 87, 8869–8873

14. Lupu, C., Kruithof, E. K., Kakkar, V. V., and Lupu, F. (1999) Thromb. Hae-mostasis 82, 1652–1658

15. Hansen, J. B., Naalsund, T., Sandset, P. M., and Svensson, B. (2000) Thromb.Res. 100, 413–417

16. Sevinsky, J. R., Rao, L. V., and Ruf, W. (1996) J. Cell Biol. 133, 293–30417. Mast, A. E., Acharya, N., Malecha, M. J., Hall, C. L., and Dietzen, D. J. (2002)

Arterioscler. Thromb. Vasc. Biol. 22, 2099–210418. Ott, I., Andrassy, M., Zieglgansberger, D., Geith, S., Schomig, A., and Neu-

mann, F. J. (2001) Blood 97, 3721–372619. Rothberg, K. G., Heuser, J. E., Donzell, W. C., Ying, Y. S., Glenney, J. R., and

Anderson, R. G. (1992) Cell 68, 673–68220. Fra, A. M., Williamson, E., Simons, K., and Parton, R. G. (1994) J. Biol. Chem.

269, 30745–30748

21. Fra, A. M., Williamson, E., Simons, K., and Parton, R. G. (1995) Proc. Natl.Acad. Sci. U. S. A. 92, 8655–8659

22. Krajewska, W. M., and Maslowska, I. (2004) Cell Mol. Biol. Lett. 9, 195–22023. Westmuckett, A. D., Kakkar, V. V., Hamuro, T., Lupu, F., and Lupu, C. (2001)

Thromb. Haemostasis 86, 1547–155424. Stearns, D. J., Kurosawa, S., Sims, P. J., Esmon, N. L., and Esmon, C. T. (1988)

J. Biol. Chem. 263, 826–83225. Feng, X., Gaeta, M. L., Madge, L. A., Yang, J. H., Bradley, J. R., and Pober,

J. S. (2001) J. Biol. Chem. 276, 8341–834926. Xu, J., Qu, D., Esmon, N. L., and Esmon, C. T. (2000) J. Biol. Chem. 275,

6038–604427. Caplice, N. M., Mueske, C. S., Kleppe, L. S., Peterson, T. E., Broze, G. J., Jr.,

and Simari, R. D. (1998) Circ. Res. 83, 1264–127028. Simionescu, N., Lupu, F., and Simionescu, M. (1983) J. Cell Biol. 97,

1592–160029. Pelkmans, L., Kartenbeck, J., and Helenius, A. (2001) Nat. Cell Biol. 3,

473–48330. Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., and Tuschl,

T. (2001) Nature 411, 494–49831. Nichols, B. J. (2002) Nat. Cell Biol. 4, 374–37832. Zhang, J., Piro, O., Lu, L., and Broze, G. J., Jr. (2003) Circulation 108,

623–62733. Vogel, U., Sandvig, K., and van Deurs, B. (1998) J. Cell Sci. 111, 825–83234. Lippincott-Schwartz, J., Snapp, E., and Kenworthy, A. (2001) Nat. Rev. Mol.

Cell. Biol. 2, 444–45635. van Deurs, B., Roepstorff, K., Hommelgaard, A. M., and Sandvig, K. (2003)

Trends Cell Biol. 13, 92–10036. Dietzen, D. J., Jack, G. G., Page, K. L., Tetzloff, T. A., Hall, C. L., and Mast,

A. E. (2003) Thromb. Haemostasis 89, 65–7337. Jacobson, K., and Dietrich, C. (1999) Trends Cell Biol. 9, 87–91

Caveolin-1 Modulates TFPI Function 22317

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Cristina Lupu, Xiaohong Hu and Florea Luputhe Cell Surface

Caveolin-1 Enhances Tissue Factor Pathway Inhibitor Exposure and Function on

doi: 10.1074/jbc.M503333200 originally published online April 6, 20052005, 280:22308-22317.J. Biol. Chem. 

  10.1074/jbc.M503333200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

  http://www.jbc.org/content/suppl/2005/04/15/M503333200.DC1

  http://www.jbc.org/content/280/23/22308.full.html#ref-list-1

This article cites 37 references, 20 of which can be accessed free at

by guest on March 26, 2018

http://ww

w.jbc.org/

Dow

nloaded from