The FASEB Journal express article 10.1096/fj.02-0687fje. Published online May 8, 2003. Pericellular plasmin induces smooth muscle cell anoikis Olivier Meilhac,* Benoît Ho-Tin-Noé,* Xavier Houard,� Monique Philippe,* Jean-Baptiste Michel,* and Eduardo Anglés-Cano*

*INSERM U460, bât.13, CHU Bichat-Claude Bernard, 46, rue Henri Huchard 75877- Paris Cedex 18, France; and �INSERM U36, Collège de France 11, Place Marcelin Bertelot, 75231 Paris Cedex 05, France

Corresponding author: Olivier Meilhac, INSERM U460, bât.13, CHU Bichat-Claude Bernard, 46, rue Henri Huchard 75877- Paris Cedex 18, France. E-mail: u460@bichat.inserm.fr

ABSTRACT

Smooth muscle cell (SMC) rarefaction is involved in the development of several vascular pathologies. We suggest that the plasminogen activation system is a potential extracellular signal that can induce pericellular proteolysis and apoptosis of vascular SMCs. Using primary cultures of arterial SMCs, we show that plasmin generated from plasminogen on the cell surface induces cell retraction and fibronectin fragmentation, leading to detachment and morphological/biochemical changes characteristic of apoptosis (also called anoikis). The generation of cell-bound plasmin mediated by tissue-type plasminogen activator (t-PA), constitutively expressed by VSMCs, requires binding of plasminogen to the cell surface and is inhibited by ε-aminocaproic acid (IC50=0.9±0.2 mM), a competitor of plasminogen binding to membrane glycoproteins. Conversely, addition of α2-antiplasmin, which blocks free plasmin in the cell supernatant, could not fully prevent anoikis. Finally, an MMP inhibitor failed to prevent VSMC anoikis, arguing for a direct involvement of plasmin in this phenomenon. Indeed, similar changes are induced by plasmin directly added to VSMCs or to arterial rings, ex-vivo. We show for the first time that pathological anoikis can be triggered by a process that requires functional assembly of the plasminogen activation system on the surface of VSMCs.

Key words: plasminogen activators • extracellular matrix • metalloproteinases • fibronectin

A noikis is a subtype of apoptosis induced by detachment of adherent cells from the ECM, which was first described for epithelial cells as a physiological phenomenon (1). Hitherto, most research has been focused on the intracellular signaling triggered by

rupture of appropriate cell/matrix communications, leading to physiological anoikis (2). In contrast, the extracellular communications able to induce cell anoikis in pathology have been less explored. Apoptosis of vascular smooth muscle cells (VSMCs) is involved in the development of atherothrombosis and plays a major role in weakening the fibrous cap of the atheroma (plaque rupture) and the entire media in aneurysm formation (3�5). During vascular remodeling, VSMCs can modulate their synthesis of matrix metalloproteinases (MMPs) or of plasminogen activators, in order to migrate or proliferate. Both MMPs and plasmin generated locally can degrade adhesive glycoproteins synthesized by VSMCs such as fibronectin (6). VSMCs represent the main source of protease inhibitors in the vascular wall, including TIMP-1 (7, 8) and Plasminogen Activator Inhibitor 1 (PAI-1) (9). The balance between proteases and their inhibitors is tightly

regulated in the vessel wall. The rarefaction of VSMCs and the associated decrease in protease inhibitors could play a central role in vascular pathologies involving proteolysis of the ECM (8, 10). Expression of plasminogen activators (tissue- or urokinase-type plasminogen activator, respectively, t-PA or u-PA) is modulated at various stages of vascular pathologies (11). Moreover, SMCs can bind exogenous t-PA and u-PA and potentiate their activity (12). Several studies point to involvement of the plasminogen activation system in pathologies, such as abdominal aortic aneurysms in which disappearance of VSMCs has been observed (10, 13). Disruption of the balance between plasminogen activators and their inhibitors could therefore be a trigger for the pericellular generation of plasmin and subsequent associated cellular events. We suggest that under pathological conditions, plasmin could induce detachment of VSMCs from their surrounding ECM, causing them to undergo anoikis. In the present study, using primary cultures, we investigated whether plasmin can induce anoikis in vitro and ex vivo in media isolated from rat aorta. We then tested the ability of vascular SMCs to convert plasminogen into plasmin by an endogenous activator system and subsequently to induce anoikis.

MATERIALS AND METHODS

Chemicals and reagents

D-valyl-L-phenylalanyl-L-lysine chloromethane (ValPheLysCH2Cl) and 1,5-Dansyl-Glu-Gly-Arg chloromethane (dGGACK) were from France Biochem (Meudon, France), the chromogenic substrate methylmalonylhydroxyprolylarginine p-nitroanilide (CBS0065) from Diagnostica Stago (Asnières, France), and the GM6001 (N-[2(R)-2-(hydroxamidocarbonylmethyl)-4-methylpentanoyl]�L-tryptophanemethylamide) from Calbiochem. α-2-antiplasmin was a generous gift from Dr. Lijnen, (Centrum voor transgene technologie en gen therapie, Belgium). Fibronectin was obtained from Invitrogen (Cergy Pontoise, France). Plasminogen was purified from fresh-frozen human plasma as described previously (14). Plasmin was prepared by activation of plasminogen with immobilized u-PA (15).

Cell culture and aorta isolation

Vascular smooth muscle cells (VSMCs) were isolated from Wistar rat abdominal aortas (16). VSMCs were cultured in DMEM containing 10% fetal calf serum and used for experiments before the 4th passage. Human VSMCs were isolated following the same protocol from radial arteries (surgical waste). Alternatively, adventitia-free aortas from the same source were cut into 2 mm rings, carefully washed with PBS, and placed in RPMI with or without various concentrations of plasmin.

Cell detachment assay

VSMCs were grown to confluence in 96-well plates and were serum-deprived for 24 h before stimulation. At the end of the experiment, the cells were washed with PBS. Remaining viable adherent cells were assessed by using the MTT test (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) (17). Trypan blue exclusion test performed on trypsinized cells confirmed that 95% of remaining adherent cells did not exhibit membrane permeabilization.

Determination of nuclear morphology

VSMCs were grown on four-well slide chambers until confluence and serum-deprived for 24 h. After stimulation, the cells were fixed with cold methanol for 5 min, and stained by standard hematoxylin-eosin procedure before microscopic observation.

TUNEL assay

Terminal transferase dUTP Nick End Labeling (TUNEL) technique was used in situ to visualize DNA fragmentation, according to the manufacturer�s instructions (Roche Molecular Biochemicals). Total nuclei number was determined after counterstaining with 100 ng/ml DAPI (4',6'-diamidino-2-phenylindole hydrochloride). Apoptotic index was calculated as the ratio of TUNEL-positive nuclei relative to total DAPI-stained nuclei (Mean count of three representative fields from two separate experiments containing ~200 cells for each condition).

DNA ladders

Genomic DNA was isolated from VSMCs by using standard DNA extraction methods (G-NOME kit, Bio101, Vista, CA). Adherent cells were scraped off into the culture medium and pooled with floating cells. DNA (10 µg) extracted according to the manufacturer�s instruction, was separated by electrophoresis in a 1.8% agarose gel containing 0.5 µg/mL ethidium bromide.

Western blot analysis

Cells were washed twice with PBS, the pellet was then lysed in a buffer containing 1% Triton X-100 in PBS (15 min at 4°C), sonicated, and centrifuged at 10,000 g for 10 min at 4°C. Equal amounts of protein extracts were separated by a 10% SDS-PAGE. The gels were transblotted onto a nitrocellulose membrane, blocked with 5% milk powder in TBS-T (tris buffer saline, pH 7.4, 0.1% Tween 20) for 1 h and then incubated with either a polyclonal serum anti-rat fibronectin (Biogenesis, Poole, UK) at a 1:1000 dilution as suggested by the manufacturer, an anti-α-2 antiplasmin IgG (Biopool, Uppsala, Sweden) at 10 µg/mL or a monoclonal IgG (CPL15) anti-plasmin(ogen) at 1 µg/mL as described (18). Membranes were then washed with TBS-T and incubated with appropriate secondary antibody conjugated to horseradish peroxidase (1:1500 for 1 h). After five washes, the signal was detected by using the chemiluminescence kit (Amersham, ECL kit, Arlington Heights, IL).

Determination of zymographic activities

Electrophoresis of samples prepared as described above was performed under non-reducing conditions in a 10% polyacrylamide gel containing or not 1 mg/mL gelatin. The gels were renatured as described previously (19) and incubated, respectively, in an appropriate buffer before staining with Coomassie blue, or carefully overlaid on a fibrin-agar gel (20). Fibrin autography was followed at 37°C for 4 to 24 h and pictures of lysis area were taken on a dark background at regular intervals. The plasminogen activator was further identified in fibrin-agar gels containing 100 µg/mL of immunoglobulins directed against t-PA.

Plasminogen activation on VSMCs

VSMCs were grown to confluence in 96-well plates and deprived of serum 24 h before stimulation. The cells were then incubated with various inhibitors in presence or absence of plasminogen in a total volume of 100 µl/well. Plasmin generation was assessed by addition of 0.75 mM CBS0065 and by measuring the release of p-nitroaniline at A405/nm vs. time with a thermostated plate reader at 37°C (MR5000, Dynatech) (21). Following activation, the plate was washed with PBS, and the amount of cell-bound plasmin was detected by adding 50 µl of chromogenic substrate as indicated above. Alternatively, cells incubated with plasminogen were centrifuged and plasmin was quantified in the cell-free conditioned medium by mixing an aliquot of 50 µl with an equal volume of CBS0065. This latter assay measures the plasmin generated and accumulated in the cell supernatant after 16 h of incubation. Plasmin concentration was calculated from rates of A405/minute according to Fleury et al. (22).

Statistical analysis

Results are expressed as means ± SD unless otherwise stated. A one-factor ANOVA was performed followed by a Fisher test in order to evaluate the difference between each condition and the control (Statview software). Statistical significance was accepted for P<0.05.

RESULTS

Plasmin induces anoikis of VSMCs

Anoikis is characterized by apoptotic features in cells following their detachment from the surrounding ECM (23). In situ TUNEL reaction provided evidence of DNA fragmentation in both adherent and floating plasmin-treated cells (Fig. 1A2, A3, respectively) as compared with untreated control cells (Fig. 1A1). After cyto-spinning the floating cells, hematoxylin/eosin (H/E) staining showed morphological changes such as nuclear condensation (black arrows) and fragmentation (white arrow) as well as membrane blebbing (circled, Fig. 1A4). VSMCs treated by 0.25 µM of plasmin for 16 h exhibited a DNA-ladder profile characteristic of apoptosis (Fig. 1B) as well as ~40% of cell detachment assessed using the MTT test (which quantifies the viable adherent cells remaining after plasmin stimulation) (Fig. 1C).

TUNEL reaction performed on aortic segments treated ex vivo by 0.25 µM of plasmin for 16 h showed 40.5 ± 3% of positive nuclei versus 11 ± 2% in control, untreated aortas (Fig. 1D4 vs. D2). The total number of nuclei was determined after DAPI counterstaining (Fig. 1D1 and D3). Fibronectin fragmentation was assessed by Western blot performed on conditioned medium (Fig. 1E). Plasmin induces fibronectin proteolysis both in the aorta (lane 2) and in vitro (lane 4) as shown by Western blot analysis of the corresponding conditioned medium, which displays bands representing cellular (240 kDa) and remnant plasma fibronectin (70 kDa) (24) fragments.

VSMCs convert plasminogen to plasmin

Cultured VSMCs can convert plasminogen into plasmin in a time- and dose-dependent manner (Fig. 2A). The generation of plasmin at the cell surface proceeds at a very low rate (Vmax=2.93 nmol/min.) due to the limited amount of t-PA expressed by SMCs in our cell culture system. However, the value for apparent Km (20.2 ± 1.3 nM) is within the range determined for t-PA bound to human VSMCs (12) and to a variety of other cells (25). This is in contrast with previously reported values for t-PA in solution (Km = 65 µM), thus indicating that VSMCs stimulate plasminogen activation. Indeed, the cell extract but not the conditioned medium of VSMCs displays fibrinolytic activity corresponding to a free form of t-PA, as indicated by its electrophoretic migration (Fig. 2B top) and its inhibition in gels supplemented with 100 µg/mL of purified anti-t-PA IgG (Fig. 2B bottom). Furthermore, the conversion of plasminogen into plasmin by VSMCs was selectively inhibited by a t-PA inhibitor (dGGACK) but not by amiloride, a u-PA inhibitor (Fig. 2C). Both cell extract and supernatant exhibited activity corresponding most probably to t-PA/PAI-1 complexes suggesting that PAI-1 is present in excess in the supernatant. Altogether, these data indicate that in the cell system used in our experiments, constitutively expressed t-PA appears to be responsible for pericellular plasminogen activation and the subsequent generation of plasmin. Similar results were obtained by using human VSMCs in which conversion of plasminogen into plasmin was observed (data not shown).

Pericellular proteolysis of ECM

Plasmin generated from plasminogen by VSMCs induced cell retraction and detachment (Fig. 3 A), and eventually cell death by anoikis (Fig. 3 A �inset). Cell detachment was correlated with the accumulated amount of cell-bound plasmin (r=0.98). Plasmin was able to trigger cell retraction a few hours after stimulation whereas incubation with 0.5 µM plasminogen required at least 16 h before such changes could be observed. Similar results were obtained using human VSMCs (57±6% of detached cells after treatment by 1 µM of plasminogen for 24 h �data not shown). The pericellular generation of plasmin induced the degradation of ECM as shown in Fig. 3B, in which an accumulation of fibronectin fragments was detectable in the conditioned medium. This degradation is due to plasmin activity as it was prevented by ValPheLysCH2Cl, a peptide that blocks the catalytic site of plasmin (data not shown). We also observed the 66 kDa active form of matrix metalloproteinase 2 (MMP2) in the culture medium of cells incubated with increasing concentrations of plasminogen (Fig. 4C). The use of GM 6001, an MMP inhibitor did not prevent fibronectin degradation (Fig. 3B), cell retraction and detachment, nor did it modify plasmin generation in our cell culture system, at concentrations known to inhibit MMPs (26) (from 10 µM and up to 0.5 mM; data not shown). Inhibition of MMP-2 activity was verified by incubating gelatin zymography gels with 10 µM of GM 6001 (Fig. 3C, bottom).

VSMC-mediated plasminogen activation: a pericellular process

The mechanism of plasminogen activation by VSMC-bound t-PA requires binding of plasminogen to the cell surface as indicated by the presence of active cell-bound plasmin (Fig. 4A, !) and by the prevention of plasminogen activation by VSMCs in the presence of ε-amino-caproic-acid (EACA, a competitor for plasminogen binding to the cell surface via its lysine binding sites) (Fig. 5A). An IC50=0.9± 0.2 mM was calculated (concentration of EACA

necessary for 50% inhibition of plasmin generation). As shown in Fig. 5B, EACA inhibited the formation of plasmin at the cell surface, most probably by prevention of plasminogen binding, leading to plasminogen accumulation in the supernatant. EACA was also able to prevent cell detachment in a dose-dependent manner to reach complete protection for a dose of 0.5 mM at 0.5 µM plasminogen. At this concentration, EACA also inhibited all the morphologic features characterizing anoikis (data not shown). Plasmin and plasminogen were detected by Western blot as a double band corresponding to their two glycosylated forms (Fig. 4B, ←, ⇐) (27). Plasmin generated at the cell surface was also detected in the cell-conditioned medium (Fig. 4A, ", and 4B, ← top).

To further investigate the biological significance of plasmin generation by VSMCs, we tested the effects of α-2antiplasmin (α2AP), the main physiological inhibitor of plasmin. For this purpose, kinetics of plasmin generation from a saturating concentration of plasminogen (0.5 µM, Fig. 2A) on VSMCs were monitored for 16 h in presence of increasing concentrations of α2AP (Fig. 6A). Plasmin activity bound to the cell surface and released into the supernatant was also measured (Fig. 6A, inset). α2AP caused a concentration dependent decrease in plasmin activity with an IC50 = 64 nM. At a concentration of 31 nM, α2AP totally quenched the plasmin present in the supernatant (Fig. 6A, inset #,↓), but failed to prevent cell retraction and detachment (Fig. 6C, ↓). Incubation with α2AP led to the formation of inactive plasmin/α2AP complexes detectable as a 140 kDa by Western blot (Fig. 6B). No trace of such complexes could be detected at the cell surface (data not shown). These results indicate that plasmin released from the surface was immediately quenched by its specific inhibitor whereas cell-bound plasmin was still active (Fig. 6A, inset $). Anoikis of VSMCs could thus be attributed to pericellular proteolysis induced by plasmin generated at the surface of the cells by t-PA constitutively expressed by SMCs.

DISCUSSION

In addition to its fibrinolytic activity, plasmin is able to cleave other substrates such as latent MMPs (28, 29) or adhesive glycoproteins, such as fibronectin, vitronectin or laminin (30�32), that participate in cell anchorage and survival signaling via the integrin system. Using in vivo experimental models, it has been shown that plasmin generation is involved in pathological processes such as aneurysm formation and rupture (11, 33) and neuronal degeneration (34) which are, at least in part, characterized by a reduction in cellularity.

VSMCs produce their own ECM to which they adhere and are able to express t-PA and/or uPA that may generate pericellular plasmin (35). Our hypothesis is that under pathological conditions, plasmin could be generated in situ at the surface of the cells and be responsible for inducing anoikis by degrading pericellular adhesive glycoproteins. In the present study, we show that plasmin is indeed able to induce detachment of VSMCs both in vitro and ex vivo, leading to morphological and biochemical changes characteristic of anoikis. Incubation of VSMCs with various concentrations of plasminogen induced a dose-dependent generation of plasmin at the surface of the cells, leading to VSMC anoikis. No addition of exogenous plasminogen activators was required to obtain these effects. The conversion of plasminogen into plasmin can be mediated by t-PA and u-PA, and prevented by plasminogen activator inhibitor (PAI-1) largely secreted by VSMCs in culture (36, 37). Expression of t-PA and u-PA by VSMCs depends upon various factors such as their phenotype, potential extracellular stimulation, or a pathological

context such as atherosclerosis (35, 38�40). Exogenous stimuli affect differently the expression of u-PA and t-PA (41). The cell fraction of VSMCs that we used for our study exhibited tPA activity. The presence of t-PA/PAI-1 complexes and the absence of t-PA activity in the supernatant suggests that PAI-1 is secreted in excess as compared with t-PA. The excess of PAI-1 in the supernatant did not affect the activity of membrane-bound t-PA, confirming previously reported binding experiments (12).

Activation of plasminogen at the cell surface and anoikis were also both inhibited by EACA, a competitor of plasminogen binding, suggesting that specific binding to cellular acceptors via carboxy-terminal lysine residues is necessary to convert plasminogen into plasmin. In our cell system, assembly of plasminogen and t-PA on the cell membrane is therefore required for the generation of plasmin in situ. Plasmin released in the conditioned medium could be trapped by addition of α2AP, the physiological circulating inhibitor of plasmin. At a concentration of 0.5 µM plasminogen, 31 nM α2AP completely inhibited the plasmin in the supernatant but failed to block cell-bound plasmin activity and cell anoikis, arguing for a cellular process. At higher concentrations of α2AP (>125 nM) a trend to a decrease in cell-bound plasmin activity and protection from anoikis were however observed. Since cell-bound plasmin is protected from inhibition by α2AP (42), these effects are most probably related to a reduced positive feedback for plasminogen binding, as has been described in a fibrin surface system (43, 44).

The observation of anoikis in the absence of active plasmin in the supernatant adds further support to the concept of pericellular proteolysis as a surface-controlled process. Smooth muscle cells are dependent on survival signals coming from the ECM. By degrading ECM components, plasmin could lead to the disruption of survival signals and eventually trigger programmed cell death (45). Fibronectin is implicated in cell survival through activation of the focal adhesion kinase signaling pathway, and could regulate expression of anti-apoptotic genes such as bcl-2 (46). Disruption of fibronectin/integrin signaling has been reported to be the cause of serum-deprived thyroid cell anoikis (47), and serum starvation-induced apoptosis of SMCs was associated with cellular fibronectin degradation (48). Soluble fibronectin peptides induce fibroblast apoptosis (49) and fibroblast survival requires interaction of integrins with the cell- (via RGD) and the heparin binding domain of fibronectin (50). Intracellular signaling linking fibronectin degradation to anoikis is not the focus of the present work, and remains to be investigated in comparison with the well-explored intracellular pathways leading to spontaneous physiological anoikis (2). We cannot exclude that plasmin could cleave cell-adhesion molecules as shown for E-Cadherin (51) or neural cell adhesion molecule L1, which undergoes cleavage in the middle of fibronectin-like domain after addition of plasminogen to neurons and melanoma cells (52). Plasmin is a wide-spectrum protease also able to cleave and activate directly or indirectly the matrix metalloproteinases (11, 28). Therefore, active MMPs could be the molecular effectors of plasmin-induced anoikis. In our model, we demonstrated that plasmin generated at the surface of the cell could in turn generate an active truncated form of MMP2, as detected by gelatin zymography. However, GM6001, a broad range inhibitor of MMPs, did not prevent plasmin-induced fibronectin fragmentation, cell detachment and anoikis of VSMCs, in agreement with a recent report (53). This result suggests that plasmin might act as a pericellular proteolytic factor independently of the MMP activation system.

In conclusion, we provide evidence that constitutively expressed t-PA can generate plasmin at the surface of VSMCs in primary culture, at concentrations sufficient to induce pericellular

proteolysis and subsequent anoikis. Pathological factors influencing expression of plasminogen activators by VSMCs, and conditions favoring the bio-availability of plasminogen as a pericellular substrate, remain to be further investigated. Such a cell-dependent phenomenon could be involved as an extracellular trigger capable of inducing cell disappearance in different vascular pathologies.

ACKNOWLEDGMENTS

Authors would like to thank Liliane Louedec, Stéphane Loyau, and Adrien Torelli for their excellent technical assistance and Mary Osborne-Pellegrin for editing this paper. This work was supported by INSERM and grants from �Pfizer-Société Française d�Athérosclérose,� ARCOL, and the Leducq Foundation.

REFERENCES

1. Frisch, S. M., and Francis, H. (1994) Disruption of epithelial cell-matrix interactions induces apoptosis. J. Cell Biol. 124, 619�626

2. Frisch, S. M., and Screaton, R. A. (2001) Anoikis mechanisms. Curr. Opin. Cell Biol. 13, 555�562

3. Walsh, K., Smith, R. C., and Kim, H. S. (2000) Vascular cell apoptosis in remodeling, restenosis, and plaque rupture. Circ. Res. 87, 184�188

4. Kockx, M. M., and Knaapen, M. W. (2000) The role of apoptosis in vascular disease. J. Pathol. 190, 267�280

5. Lopez-Candales, A., Holmes, D. R., Liao, S., Scott, M. J., Wickline, S. A., and Thompson, R. W. (1997) Decreased vascular smooth muscle cell density in medial degeneration of human abdominal aortic aneurysms. Am. J. Pathol. 150, 993�1007

6. Holderbaum, D., and Ehrhart, L. A. (1986) Substratum influence on collagen and fibronectin biosynthesis by arterial smooth muscle cells in vitro. J. Cell. Physiol. 126, 216�224

7. Gonzalez, W., Fontaine, V., Pueyo, M. E., Laquay, N., Messika-Zeitoun, D., Philippe, M., Arnal, J. F., Jacob, M. P., and Michel, J. B. (2000) Molecular plasticity of vascular wall during N(G)-nitro-L-arginine methyl ester-induced hypertension: modulation of proinflammatory signals. Hypertension 36, 103�109

8. Badier-Commander, C., Couvelard, A., Henin, D., Verbeuren, T., Michel, J. B., and Jacob, M. P. (2001) Smooth muscle cell modulation and cytokine overproduction in varicose veins. An in situ study. J. Pathol. 193, 398�407

9. Luttun, A., Lupu, F., Storkebaum, E., Hoylaerts, M. F., Moons, L., Crawley, J., Bono, F., Poole, A. R., Tipping, P., Herbert, J. M., et al. (2002) Lack of plasminogen activator inhibitor-1 promotes growth and abnormal matrix remodeling of advanced atherosclerotic plaques in apolipoprotein E-deficient mice. Arterioscler. Thromb. Vasc. Biol. 22, 499�505

10. Michel, J. B. (2001) Contrasting outcomes of atheroma evolution: intimal accumulation versus medial destruction. Arterioscler. Thromb. Vasc. Biol. 21, 1389�1392

11. Carmeliet, P., Moons, L., Lijnen, R., Baes, M., Lemaitre, V., Tipping, P., Drew, A., Eeckhout, Y., Shapiro, S., Lupu, F., et al. (1997) Urokinase-generated plasmin activates matrix metalloproteinases during aneurysm formation. Nat. Genet. 17, 439�444

12. Ellis, V., and Whawell, S. A. (1997) Vascular smooth muscle cells potentiate plasmin generation by both urokinase and tissue plasminogen activator-dependent mechanisms: evidence for a specific tissue-type plasminogen activator receptor on these cells. Blood 90, 2312�2322

13. Herron, G. S., Unemori, E., Wong, M., Rapp, J. H., Hibbs, M. H., and Stoney, R. J. (1991) Connective tissue proteinases and inhibitors in abdominal aortic aneurysms. Involvement of the vasa vasorum in the pathogenesis of aortic aneurysms. Arterioscler. Thromb. 11, 1667�1677

14. Fleury, V., and Angles-Cano, E. (1991) Characterization of the binding of plasminogen to fibrin surfaces: the role of carboxy-terminal lysines. Biochemistry 30, 7630�7638

15. Wiman, B., and Wallen, P. (1973) Activation of human plasminogen by an insoluble derivative of urokinase. Structural changes of plasminogen in the course of activation to plasmin and demonstration of a possible intermediate compound. Eur. J. Biochem. 36, 25�31

16. Battle, T., Arnal, J. F., Challah, M., and Michel, J. B. (1994) Selective isolation of rat aortic wall layers and their cell types in culture�application to converting enzyme activity measurement. Tissue Cell 26, 943�955

17. Denizot, F., and Lang, R. (1986) Rapid colorimetric assay for cell growth and survival. Modifications to the tetrazolium dye procedure giving improved sensitivity and reliability. J. Immunol. Methods 89, 271�277

18. Montes, R., Paramo, J. A., Angles-Cano, E., and Rocha, E. (1996) Development and clinical application of a new ELISA assay to determine plasmin-alpha2-antiplasmin complexes in plasma. Br. J. Haematol. 92, 979�985

19. Badier-Commander, C., Verbeuren, T., Lebard, C., Michel, J. B., and Jacob, M. P. (2000) Increased TIMP/MMP ratio in varicose veins: a possible explanation for extracellular matrix accumulation. J. Pathol. 192, 105�112

20. Gaussem, P., Grailhe, P., and Angles-Cano, E. (1993) Sodium dodecyl sulfate-induced dissociation of complexes between human tissue plasminogen activator and its specific inhibitor. J. Biol. Chem. 268, 12150�12155

21. Angles-Cano, E., Boutiere, B., Arnoux, D., Masson, C., Contant, G., Benchimol, P., and Sampol, J. (1987) Release pattern of the vascular plasminogen activator and its inhibitor in human postvenous occlusion plasma as assessed by a spectrophotometric solid-phase fibrin-tPA activity assay. Thromb. Haemost. 58, 843�849

22. Fleury, V., Lijnen, H. R., and Angles-Cano, E. (1993) Mechanism of the enhanced intrinsic activity of single-chain urokinase- type plasminogen activator during ongoing fibrinolysis. J. Biol. Chem. 268, 18554�18559

23. Frisch, S. M., and Ruoslahti, E. (1997) Integrins and anoikis. Curr. Opin. Cell Biol. 9, 701�706

24. Weiss, R. E., and Reddi, A. H. (1981) Isolation and characterization of rat plasma fibronectin. Biochem. J. 197, 529�534

25. Sinniger, V., Merton, R. E., Fabregas, P., Felez, J., and Longstaff, C. (1999) Regulation of tissue plasminogen activator activity by cells. Domains responsible for binding and mechanism of stimulation. J. Biol. Chem. 274, 12414�12422

26. Jones, P. L., Crack, J., and Rabinovitch, M. (1997) Regulation of tenascin-C, a vascular smooth muscle cell survival factor that interacts with the alpha v beta 3 integrin to promote epidermal growth factor receptor phosphorylation and growth. J. Cell Biol. 139, 279�293

27. Angles-Cano, E. (1994) Overview on fibrinolysis: plasminogen activation pathways on fibrin and cell surfaces. Chem. Phys. Lipids 67-68, 353�362

28. Baricos, W. H., Cortez, S. L., el-Dahr, S. S., and Schnaper, H. W. (1995) ECM degradation by cultured human mesangial cells is mediated by a PA/plasmin/MMP-2 cascade. Kidney Int. 47, 1039�1047

29. Davis, G. E., Pintar Allen, K. A., Salazar, R., and Maxwell, S. A. (2001) Matrix metalloproteinase-1 and -9 activation by plasmin regulates a novel endothelial cell-mediated mechanism of collagen gel contraction and capillary tube regression in three-dimensional collagen matrices. J. Cell Sci. 114, 917�930

30. Ehrlich, M. I., Krushell, J. S., Blumenstock, F. A., and Kaplan, J. E. (1981) Depression of phagocytosis by plasmin degradation products of plasma fibronectin. J. Lab. Clin. Med. 98, 263�271

31. Chain, D., Kreizman, T., Shapira, H., and Shaltiel, S. (1991) Plasmin cleavage of vitronectin. Identification of the site and consequent attenuation in binding plasminogen activator inhibitor-1. FEBS Lett. 285, 251�256

32. Liotta, L. A., Goldfarb, R. H., and Terranova, V. P. (1981) Cleavage of laminin by thrombin and plasmin: alpha thrombin selectively cleaves the beta chain of laminin. Thromb. Res. 21, 663�673

33. Anidjar, S., Salzmann, J. L., Gentric, D., Lagneau, P., Camilleri, J. P., and Michel, J. B. (1990) Elastase-induced experimental aneurysms in rats. Circulation 82, 973�981

34. Chen, Z. L., and Strickland, S. (1997) Neuronal death in the hippocampus is promoted by plasmin-catalyzed degradation of laminin. Cell 91, 917�925

35. Salame, M. Y., Samani, N. J., Masood, I., and deBono, D. P. (2000) Expression of the plasminogen activator system in the human vascular wall. Atherosclerosis 152, 19�28

36. van Leeuwen, R. T., Kol, A., Andreotti, F., Kluft, C., Maseri, A., and Sperti, G. (1994) Angiotensin II increases plasminogen activator inhibitor type 1 and tissue-type plasminogen activator messenger RNA in cultured rat aortic smooth muscle cells. Circulation 90, 362�368

37. Takeda, K., Ichiki, T., Tokunou, T., Iino, N., Fujii, S., Kitabatake, A., Shimokawa, H., and Takeshita, A. (2001) Critical role of Rho-kinase and MEK/ERK pathways for angiotensin II- induced plasminogen activator inhibitor type-1 gene expression. Arterioscler. Thromb. Vasc. Biol. 21, 868�873

38. Lau, H. K. (1999) Regulation of proteolytic enzymes and inhibitors in two smooth muscle cell phenotypes. Cardiovasc. Res. 43, 1049�1059

39. Clowes, A. W., Clowes, M. M., Au, Y. P., Reidy, M. A., and Belin, D. (1990) Smooth muscle cells express urokinase during mitogenesis and tissue- type plasminogen activator during migration in injured rat carotid artery. Circ. Res. 67, 61�67

40. Lupu, F., Heim, D. A., Bachmann, F., Hurni, M., Kakkar, V. V., and Kruithof, E. K. (1995) Plasminogen activator expression in human atherosclerotic lesions. Arterioscler. Thromb. Vasc. Biol. 15, 1444�1455

41. Shireman, P. K., McCarthy, W. J., Pearce, W. H., Shively, V. P., Cipollone, M., Kwaan, H. C., and Yao, J. S. (1996) Plasminogen activator levels are influenced by location and varicosity in greater saphenous vein. J. Vasc. Surg. 24, 719�724

42. Lijnen, H. R. (2001) Gene targeting in hemostasis. Alpha2-antiplasmin. Front. Biosci. 6, D239�D247

43. Suenson, E., Lutzen, O., and Thorsen, S. (1984) Initial plasmin-degradation of fibrin as the basis of a positive feed- back mechanism in fibrinolysis. Eur. J. Biochem. 140, 513�522

44. Angles-Cano, E., Rouy, D., and Lijnen, H. R. (1992) Plasminogen binding by alpha 2-antiplasmin and histidine-rich glycoprotein does not inhibit plasminogen activation at the surface of fibrin. Biochim. Biophys. Acta 1156, 34�42

45. Isik, F. F., Gibran, N. S., Jang, Y. C., Sandell, L., and Schwartz, S. M. (1998) Vitronectin decreases microvascular endothelial cell apoptosis. J. Cell. Physiol. 175, 149�155

46. Matter, M. L., and Ruoslahti, E. (2001) A signaling pathway from the alpha5beta1 and alpha(v)beta3 integrins that elevates bcl-2 transcription. J. Biol. Chem. 276, 27757�27763

47. Di Matola, T., Mueller, F., Fenzi, G., Rossi, G., Bifulco, M., Marzano, L. A., and Vitale, M. (2000) Serum withdrawal-induced apoptosis in thyroid cells is caused by loss of fibronectin-integrin interaction. J. Clin. Endocrinol. Metab. 85, 1188�1193

48. Ikari, Y., Mulvihill, E., and Schwartz, S. M. (2001) alpha 1-Proteinase inhibitor, alpha 1-antichymotrypsin, and alpha 2- macroglobulin are the antiapoptotic factors of vascular smooth muscle cells. J. Biol. Chem. 276, 11798�11803

49. Hadden, H. L., and Henke, C. A. (2000) Induction of lung fibroblast apoptosis by soluble fibronectin peptides. Am. J. Respir. Crit. Care Med. 162, 1553�1560

50. Kapila, Y. L., Wang, S., and Johnson, P. W. (1999) Mutations in the heparin binding domain of fibronectin in cooperation with the V region induce decreases in pp125(FAK) levels plus proteoglycan-mediated apoptosis via caspases. J. Biol. Chem. 274, 30906�30913

51. Ryniers, F., Stove, C., Goethals, M., Brackenier, L., Noe, V., Bracke, M., Vandekerckhove, J., Mareel, M., and Bruyneel, E. (2002) Plasmin produces an E-cadherin fragment that stimulates cancer cell invasion. Biol. Chem. 383, 159�165

52. Nayeem, N., Silletti, S., Yang, X., Lemmon, V. P., Reisfeld, R. A., Stallcup, W. B., and Montgomery, A. M. (1999) A potential role for the plasmin(ogen) system in the posttranslational cleavage of the neural cell adhesion molecule L1. J. Cell Sci. 112, 4739�4749

53. Ugwu, F., Lemmens, G., Collen, D., and Lijnen, H. R. (2001) Matrix metalloproteinase deficiencies do not impair cell-associated fibrinolytic activity. Thromb. Res. 102, 61�69

Received October 21, 2002; accepted March 10, 2003.

Fig. 1

Figure 1. Plasmin induces anoikis of smooth muscle cells. A) Confluent serum-deprived primary cultures of rat VSMCs were incubated (2–4) or not (1) with 0.25 µM of plasmin for 16 h. Preparations were either submitted to TUNEL reaction for detection of DNA fragmentation (1–3) or stained by H/E (4). Plasmin-treated floating cells were cytospun (3, 4) before treatment, whereas remaining adherent cells were treated directly on the slide. (4: black arrows: nuclear condensation, white arrow: nuclear fragmentation, circle: membrane blebbing). B) Visualization of internucleosomal DNA fragmentation after electrophoresis of genomic DNA extracted from plasmin-treated cells (floating and adherent cells were pooled) at indicated concentrations for 16 h (M: φ174 DNA marker). Results depicted in (A) and (B) are representative of one out of at least three separate experiments. C) Viable adherent VSMCs were quantified by using the MTT test after treatment by increasing concentrations of plasmin (0 to 1 µM). Results are expressed as percentage of untreated control cells. Each column represents the mean ± SD of three separate experiments performed in triplicate (*P<0.05). D) Adventitia-free rat abdominal aorta segments were incubated (3, 4) or not (1, 2) with 0.25 µM of plasmin for 16 h. TUNEL reaction was performed on paraffin sections (2, 4) and counterstained with DAPI (1, 3). Photomicrographs are representative fields out of three sections. This experiment has been performed twice with similar results. E) Corresponding conditioned media were analyzed for fibronectin fragmentation by Western blot (0: control untreated aorta, Pn 0.25: aorta treated with 0.25 µM plasmin; arrows: Fn: 0.5 ng of purified cellular fibronectin, Fn+Pn: 1 ng of cellular fibronectin incubated with 0.25 µM plasmin for 1 h at 37°C, Pn: 0.25 µM plasmin).

Fig. 2

Figure 2. Plasminogen activation by VSMCs. A) Plasminogen dose-response and kinetics of plasmin generation: 0 to 1 µM of plasminogen was added to VSMCs in the presence of 0.75 mM CBS0065, a chromogenic substrate for plasmin. Absorbance was monitored at 405 nm during 16 h. Plasmin generated as a function of time for different concentrations of plasminogen is indicated in the inset (�: 0 µM, �: 0.12 µM, �: 0.25 µM, �: 0.5 µM). The main curve is fitted to the standard Michaelis-Menten equation by non-linear regression analysis, and represents the rate of plasmin formed as a function of plasminogen concentration ± SD (�). The results depicted are representative of three separate experiments performed in triplicate. B) Fibrin-agar zymography: VSMCs extract or supernatant (SN) from serum-deprived cells were separated by SDS-PAGE and then overlaid onto a fibrin agarose gel containing (bottom) or not (top) purified anti-t-PA IgG (100 µg/mL). Purified t-PA and u-PA were used as positive controls. This experiment has been performed three times with similar results. C) Inhibition of plasmin generation by dGGACK or amiloride. Cells were incubated with 0.5 µM of plasminogen for 16 h and supplemented with dGGACK (0.6 to 40 µM) or amiloride (3.9 to 125 µM). Results are expressed as a percentage of the amount of plasmin converted from plasminogen without any inhibitor (considered as 100%). Column represents mean ± SD of three separate experiments performed in triplicate (*P<0.01).

Fig. 3

Figure 3. Pericellular effects of VSMC-generated plasmin. A) VSMCs were incubated with increasing concentrations of plasminogen for 16 h. After washing with PBS, viability was assessed by the MTT test, which reveals the remaining adherent cells. Untreated control cells represent 100% of viable adherent cells and the results are expressed as a percentage of control (mean ± SD of three separate experiments performed in triplicate, *P<0.05) . Inset shows H/E staining on adherent untreated cells (0) and cytospun floating cells after treatment by 0.5 µM plasminogen (0.5). B) Detection of fibronectin fragments in VSMC-conditioned medium by Western blot. VSMCs were treated or not with 0.25 or 0.5 µM plasminogen (Pg) with or without 10 µM of GM6001 (GM). Purified plasminogen (Pg 0.5 µM, arrow) is used in order to challenge the specificity of the antibody used. C) Zymogram showing gelatinolytic activity of VSMC-conditioned medium (19). Pro-forms of metalloproteinase 2 (pro-MMP2) are constitutively expressed, whereas the MMP2 appears gradually with increasing concentrations of plasminogen (top zymogram). The presence of higher molecular weight active bands when VSMCs were incubated with high concentrations of plasminogen (0.25 to 1 µM) could be attributed to active fragments of plasmin. Incubation of a zymography gel with 10 µM of GM6001 inhibited totally MMP-2 activity (bottom zymogram, left) as compared with the corresponding zymogram without MMP inhibitor (bottom zymogram, right). Remaining bands in left panel correspond to plasmin activity. (0: VSMC-conditioned medium, Pn: 0.12 µM plasmin-treated VSMC conditioned medium). Experiments were performed at least three times with reproducible results.

Fig. 4

Figure 4. Localization of plasmin generated by VSMCs. Plasminogen activation was performed as indicated in Fig. 2. After 16 h of incubation with 0 to 2 µM plasminogen, VSMCs were centrifuged and the supernatant was assayed for plasmin activity. Cells were washed briefly with PBS before addition of CBS0065 for determination of membrane-associated plasmin activity. A) Plasmin accumulated in the supernatant (!) and bound to the cell fraction ("). Plasmin concentration was determined as described in Materials and Methods. B) Western blot was performed on the cell extract (bottom) and cell-conditioned medium (top), using CPL15 monoclonal antibody (directed against kringle 1 of plasmin(ogen)). Bands corresponding to plasmin (←) and plasminogen (⇐) are indicated. Purified plasmin and plasminogen were used as positive controls (0.5 µM). All experiments have been performed thrice with similar results.

Fig. 5

Figure 5. Effect of EACA on plasminogen activation by VSMC-bound t-PA. A) The activation of plasminogen (0.5 µM) was evaluated as indicated in Fig. 2, in the presence of increasing concentrations of EACA (0 to 10 mM) used as a competitor of C-terminal lysine residues on the cell surface. The IC50 was calculated by fitting data to the hyperbolic decay equation. B) Western blot analysis of cell extracts (right) or supernatants (SN, left) of VSMCs treated with 0.5 µM plasminogen supplemented or not with 25 mM EACA. CPL15 monoclonal antibody was used to detect plasmin (←) and plasminogen (⇐). Plasminogen and plasmin incubated in the same conditions without cells are used as positive controls (Pg and Pn). Electrophoresis was performed under reducing conditions. The band of plasminogen present in the cell extract in the presence of EACA correspond to Lys-independent binding of plasminogen.

Fig. 6

Figure 6. Plasmin-induced anoikis: a pericellular process. A) Curve of inhibition of plasmin generation by increasing concentrations of α2-antiplasmin (α2AP) for a constant concentration of plasminogen (0.5 µM). Plasmin formation on cells was monitored during 16 h by addition of CBS0065 (�). Cells were washed, and membrane-bound plasmin was quantified (inset, �). In parallel, plasmin accumulated in the supernatant was assessed (inset, �). B) Western blot analysis was performed under non-reducing conditions for detection of α2AP in its free form (←) or as a complex with plasmin (⇐). VSMC-conditioned medium alone or supplemented with 0.5 µM plasminogen and indicated concentrations of α2AP were analyzed. Complexes of α2AP:plasmin (125nM:250nM) were prepared and used as positive control. C) Viable adherent cells were quantified by using the MTT test. The vertical arrow in (A, C) indicates the concentrations of α2AP at which the accumulated amount of cell-bound plasmin is still able to produce cell detachment in the absence of free plasmin.