Plasminogen activator inhibitor-1 deficiency enhances flow-induced smooth muscle cell migration

Thrombosis Research (2004) 114, 57--65

intl.elsevierhealth.com/journals/thre

Regular Article

Plasminogen activator inhibitor-1 deficiency enhances flow-induced smooth muscle cell migration John P. Cullen a,1, Suzanne M. Nicholl a,1, Shariq Sayeed a, James V. Sitzmann a, S.Steve Okada a, Paul A. Cahill b, Eileen M. Redmond a,* a

Department of Surgery, University of Rochester Medical Center, Box SURG, 601 Elmwood Avenue, Rochester, NY 14642-8410, USA b Vascular Health Research Center, School of Biotechnology, Dublin City University, Dublin, Ireland Received 26 February 2004; received in revised form 22 April 2004; accepted 2 May 2004 Available online 4 June 2004

KEYWORDS Vascular smooth muscle; Pulse pressure; Migration; PAI-1; Matrix metalloproteinases; TIMP-2

ABSTRACT Introduction: We determined the role of smooth muscle cell (SMC)-derived plasminogen activator inhibitor-1 (PAI-1) in the flow-induced SMC migratory response. Materials and methods: Wild type (wt) or PAI-1 knockout SMC were cultured in the absence or presence of endothelial cells (EC) under static or pulsatile flow conditions in a perfused culture system. SMC migration was then assessed by Transwell assay. Results: Pulsatile flow significantly increased SMC PAI-1 mRNA and protein levels, f 4- and 3-fold respectively (n = 4, p < 0.05). In the absence, but not in the presence of EC, pulsatile flow significantly increased ( f 2.4-fold) the migration of wt SMC when compared to wt SMC cultured under static conditions. PAI1 / SMC migration was significantly increased under flow conditions as compared to wild-type controls (334 F 22% vs. 237 F 11%, n = 6, p < 0.05). This flow-induced migration was significantly attenuated, but not completely inhibited, when PAI-1 / SMC were cultured in the presence of EC (147 F 13%, n = 6, p < 0.05). The flowinduced PAI-1 / SMC migratory response was partially inhibited by an antiurokinase plasminogen activator (uPA) antibody (#1189), and completely inhibited by both 1189 and the matrix metalloproteinase (MMP) inhibitor BB3103. In parallel PAI1 / SMC cells, there was a greater flow-induced increase in proMMP-2 activity as

Abbreviations: EC, endothelial cells; ECM, extracellular matrix; HUSMC, human umbilical smooth muscle cells; MMP, matrix metalloproteinase; PA, plasminogen activator; PAI-1, plasminogen activator inhibitor-1; SMC, smooth muscle cell; TIMP, tissue inhibitors of matrix metalloproteinases; tPA, tissue plasminogen activator; uPA, urokinase plasminogen activator. * Corresponding author. Tel.: +1-585-275-2870; fax: +1-585-756-7819. E-mail address: [email protected] (E.M. Redmond). 1 Both authors contributed equally to this study.

0049-3848/$ - see front matter A 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2004.05.003

58

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65 compared to wild-type control cells. Moreover, under both static and flow conditions, tissue inhibitors of matrix metalloproteinases (TIMP)-2 activity was reduced in these PAI-1-deficient cells as compared to wild-type controls. Conclusions: These results suggest that SMC PAI-1 plays a role in limiting flowinduced SMC migration and thus may be an important mechanism for controlling the process of vascular remodelling. A 2004 Elsevier Ltd. All rights reserved.

Migration of smooth muscle cells (SMC) from the media and accumulation in the intima of arteries is a key event in the development and progression of atherosclerosis and postangioplasty restenosis [1,2]. Cell migration and tissue remodeling require degradation of extracellular matrix (ECM). Indeed, several groups have reported the involvement of the fibrinolytic (plasminogen/plasmin) and matrix metalloproteinase (MMP) systems, which in concert can degrade most components of the ECM, in SMC migration [3--5]. The inactive proenzyme plasminogen is activated to the proteolytic enzyme plasmin by two plasminogen activators (PA), tissue plasminogen activator (tPA) and urokinase plasminogen activator (uPA) [6,7]. Plasminogen activator inhibitor-1 (PAI-1) is the primary inhibitor of uPA and tPA and a key regulator of the plasmin/plasminogen system. PAI-1 is present in plasma and platelets, and is synthesized by vascular endothelial and smooth muscle cells [6]. In addition to regulating fibrinolysis within the vessel wall, PAI-1 may also play an important role in controlling smooth muscle cell migration and vascular wall remodeling [8,9]. However, the biologically relevant source of PAI-1 for the pathogenesis of atherosclerosis is unclear. The MMPs constitute a tightly regulated family of zinc endopeptidases. The majority of MMPs are synthesized and secreted as inactive proenzymes [6,10]. The members of the MMP family are divided into five classes based on their structure and substrate specificities. Active MMPs are inhibited by tissue inhibitors (TIMPs). There are interactions between the fibrinolytic and MMP systems; for example, plasmin is a potent activator of most MMPs, promoting cleavage of the latent propeptides to the active molecule [6]. A potential role for increased proteolysis by plasmin or MMPs in atherosclerosis is suggested by the enhanced expression of tPA, uPA and several MMPs, including the gelatinases MMP-2 and MMP-9, in plaques [6,10]. Furthermore, mice deficient in uPA alone or both uPA and tPA show markedly reduced intimal thickening [11], whereas PAI-1 null mice demonstrated exaggerated vascular response to injury [8]. In addition, treatment with specific MMP inhibitors attenuates SMC migration in vivo [10,12].

Hemodynamic forces (pulse pressure, cyclic strain) play an important role in the control of blood vessel structure and remodeling by critically regulating SMC migration and proliferation [13--15]. However, the complex interactions between the mechanical deformation and the cellular and molecular events in the vessel wall are poorly understood. We have previously reported that pulse pressure due to pulsatile flow induces SMC migration via uPA- and MMP-dependent mechanisms [16], and that endothelial cells (EC), by a mechanism involving PAI-1, protect SMCs from the stimulatory effects of pulsatile flow on migration [17]. Here, using wild type and gene deficient SMC exposed to pulse pressure, in the absence or presence of endothelial cells, in an in vitro perfused cell culture system, we investigate the contribution of SMCderived PAI-1 in modulating the flow-induced migratory response and the effect of PAI-1 gene deletion on MMP and TIMP activity.

Materials and methods Animals and reagents Mice homozygous for a targeted mutation in the PAI-1 gene, (background strain C57BL/6J; strain name B6.129S2-Serpine1 tmlMlg ; stock number 002507), were purchased from The Jackson Laboratory (Bar Harbor, ME). These mice, originally generated and characterized by Carmeliet et al. [18], were backcrossed to C57BL/6J mice for > 8 generations. Age-matched mice on the same genetic background (C57BL/6J, stock number 000664), also obtained from The Jackson Laboratory, were used as wild-type controls. Mice were housed in specific pathogen-free rooms and fed a normal mouse laboratory diet. All procedures were approved by the Rochester University Animal Care and Use Committee. Monoclonal murine anti-uPA antibody (#1189) was obtained from American Diagnostica (Greenwich, CT) and rabbit antihuman PAI-1 antibody was purchased from Molecular Innovations (Royal Oak, MI). The matrix metalloprotei-

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65 nase inhibitor BB3103 was supplied by British Biotech Pharmaceuticals. Gelatin was purchased from Sigma (St. Louis, MO) and active MMP-2 enzyme was obtained from Oncogene Research Products (San Diego, CA). DMEM/Ham’s F-12 media, Medium 199, penicillin--streptomycin and Trizol were from GIBCO-BRL Life Technologies (Invitrogen, Carlsbad, CA). Fetal calf serum (FCS) was purchased from Gemini Bio-Products (Woodland, CA) and fibronectin and endothelial cell growth supplement were obtained from BD Biosciences (Palo Alto, CA).

Vascular smooth muscle cells (SMC) SMC cultures were prepared from (i) human umbilical veins and (ii) vena cavae of 10--12 wild type or PAI-1 / mice using the explant technique [16]. Briefly, the media of the vein was isolated surgically and the tissue minced into small pieces and pooled. The pieces were plated onto a fibronectin coated petri dish and cultured in DMEM/ Ham’s F-12 media with 10% heat-inactivated fetal calf serum (FCS), fungizone and 100 U/ml penicillin/100 Ag/ml streptomycin in a humidified atmosphere of 5% CO2, 95% air. Cells were subcultured using 0.125% trypsin--EDTA when there was adequate proliferation of cells beyond the explants. SMC, which represented >95% of the explant cells, displayed the typical spindle-shaped morphology and ‘hill and valley’ pattern of growth in culture and were further characterized by immunohistochemical staining for smooth muscle specific-actin. SMC, passages 2--4, from at least three independent isolates, were used in the perfused transcapillary cultures. The different isolates of smooth muscle cells behaved similarly in the assays.

59

Perfused transcapillary cultures Perfused cocultures of SMC and EC were established as described in detail previously [16,21,22]. The CELLMAXk QUAD Artificial Capillary Cell Culture System (Spectrum Laboratories, Rancho Dominquez, CA) was utilized. This apparatus consisted of an enclosed bundle of 50 permeable, Pronectin-F TM coated polypropylene capillaries (capillary length 13 cm; internal diameter 330 AM; wall thickness 150 AM; pore size 0.3 AM; extracapillary surface area 100 cm2; luminal surface area 70 cm2) through which media from a reservoir is pumped in a pulsatile fashion, at a chosen flow rate, via silicone rubber tubing. Pronectin-F is a synthetic protein polymer that incorporates multiple copies of the RGD cell attachment ligand of human fibronectin. In the current study, unless otherwise specified, the intraluminal flow rate used was 25 ml/min corresponding to a shear stress of 23 dyne/cm2, an intraluminal pulse pressure of 149/48 mm Hg experienced by EC grown in the intraluminal space and an extraluminal pulse pressure of 106/50 mm Hg with a frequency of 110 cycles/min experienced by SMC grown in the extracapillary (ECS) space.

Seeding of SMC and EC experimental protocol SMC ( f 5  106 cells) in DMEM/Ham’s F-12 supplemented with 10% FBS and antibiotics (SMC media), were seeded into the extracapillary space via the side ports. The SMC were allowed to adhere and establish themselves on the outer surface of the capillaries which were perfused at the lowest flow rate, 0.3 ml/min. For cocultures, EC ( f 2  106 cells) were seeded intraluminally via the endports essentially as described previously [21,22].

Endothelial cells (EC) Experimental protocol EC cultures were prepared from murine vena cava of wild-type animals, by established methods as previously described [19]. EC were isolated from the inferior vena cava of mice (20 mice were sacrificed) using a multistep isolation scheme previously described by Ojeifo et al. [20]. The cells were grown to confluence in Medium 199 supplemented with 10% heat inactivated FCS, fungizone, 100 U/ml penicillin/100 Ag/ml streptomycin and endothelial cell growth supplement. Cells were assessed for endothelial cell phenotype by morphology, expression of von Willebrand Factor antigen and PECAM. EC between passages 2--5 were used in the coculture system as described below.

A series of perfused transcapillary cultures was examined in parallel. The cultures were designated as either ‘static’ or ‘flow’. Following the 48-h stabilization period, the static group were disconnected from the pulsatile pump and experienced static conditions (0 ml/min), whereas the ‘flow’ group were exposed to a single step increase in flow up to 25 ml/min and maintained for 24 h (Fig. 1). At the end of the experimental period either the SMC were harvested using trypsin and used in migration and zymography assays or total SMC RNA and protein were extracted after treatment of cultures with Trizol for Northern and Western blot analyses.

60

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65

Fig. 1 Experimental protocol. SMC and EC were isolated from human umbilical veins or vena cavae of wild type or PAI-1 gene deficient mice and grown in petri dishes under standard culture conditions. At passage 2--5, wild type or PAI-1 / SMC were cocultured in the absence or presence of wild-type EC in the perfused transcapillary culture system and divided into two groups; ‘static’ (a) and ‘flow’ (b). Following a 48-h stabilization period at a flow rate of 0.3 ml/min, the static group were disconnected from the pulsatile pump and experienced no-flow ‘static’ conditions (0 ml/min), whereas the ‘flow’ group were exposed to a single step increase in flow up 25 ml/min, maintained for 24 h. At the end of the experimental period, SMC were harvested and their migration assessed by filter migration assay. In some cases, total SMC RNA, protein and conditioned media were harvested and used for Northern blot analysis, Western blot analysis and zymography, respectively.

Northern blot analysis Total RNA was isolated from ‘static’ and ‘flow’ group SMC using TRIzol Reagent (Gibco BRL) as described previously [16]. Aliquots (10 Ag) of the total RNA samples were separated on formaldehyde-agarose gels. The RNAs were transferred and UV cross-linked to nylon membranes and hybridized with 32P-labeled cDNA probes for human PAI-1 and GAPD (ATCC) which were prepared by random priming. Transcripts were quantitated by Kodak 1D Image Analysis software (Kodak, Rochester, NY) and normalized using GAPD levels for equal loading.

Immunoblots SMC lysates were prepared and analyzed for PAI-1 expression by Western Blot analysis essentially as described previously [16]. Rabbit antihuman recombinant PAI-1 was used at 1--5000 dilution.

Transwell filter migration assay Fibronectin-coated Transwell filters, 12 AM pore size (Costar) were used for migration assays as

described previously [16,17]. SMC harvested from the transcapillary cultures were seeded at a density of 1.0  104 cells/filter. Cells were allowed to migrate for 10 h with conditioned media from the respective transcapillary cultures in both upper and lower chambers. In this way, random migration or chemokinesis was being measured as there was no concentration gradient between the upper and lower chambers. The number of cells that had migrated through the filter pores was manually counted per high power field (hpf) (20  magnification) using a microscope (Nikon Diaphot). Data are reported as the number of SMC counted per 12 hpf, expressed as percentage of control, where control is SMC monocultures exposed to static conditions unless otherwise stated. Zymography: The secretion and activity of MMP2/MMP-9 in cell lysates and TIMP-2 in conditioned media was assayed by gelatin zymography and reverse gelatin zymography, respectively. Briefly, equal amounts of protein were mixed with nonreducing SDS sample buffer and subjected to gel electrophoresis with 10% running gels containing 0.12 mg/ml gelatin, and, active MMP-2 (1.2 Ag/ ml) in the case of reverse zymography. The gel was washed twice in 2.5% Triton X100 for 30 min fol-

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65

61

lowed by incubation for 18 h at 37 jC in incubation buffer (50 mM Tris--HCl, pH 7.4, 10 mM CaCl2). A staining solution was prepared by mixing 10 ml stain stock solution (0.2% Coomassie brilliant blue) with 200 ml destain (1:3:6 glacial acetic acid:methanol:distilled water) and gels were stained for 2 h on a shaker at room temperature. After completion gels were briefly rinsed in destain solution, photographed and densitometric analysis performed using Kodak 1D Image Analysis software.

Statistical analysis Results are expressed as mean F S.E.M. n = number of individual perfused transcapillary experiments from which cells were harvested. Experimental points were performed in triplicate with a minimum of three independent experiments. Comparisons between experimental and control cells in migration assays were done using unpaired two tailed Student’s t-tests. When more than two groups were present, an ANOVA (factorial design) was used (Statview). A probability value of P < 0.05 was considered significant. Fig. 2 SMC PAI-1 under static and pulsatile flow conditions. HUSMC monocultures were exposed to pulsatile flow (25 ml/min) or static (no flow) conditions for 24 h whereupon total SMC RNA and protein were isolated and used for (a) Northern blot analysis and (b) Western blot analysis, respectively. PAI-1 mRNA and protein levels were quantified with optical densitometry and image analysis software. Arbitrary values are expressed relative to static SMC. Representative blots are shown, together with the cumulative densitometric data, n = 4. *p < 0.05.

Results Effect of pulsatile flow on SMC PAI-1 mRNA and protein levels Compared with the no-flow static group, monocultured human umbilical smooth muscle cells

Fig. 3 Effect of pulsatile flow on migration of monocultured or cocultured wild type and PAI-1 / SMC. Wild type (wt) or PAI-1 / SMC, cultured in the absence or presence of wild-type mouse EC, were exposed to either static or pulsatile flow conditions (25 ml/min, 24 h) before random migration or chemokinesis was assessed by Transwell assay. The number of SMC migrating through the filter was expressed as a percentage of control, static wt SMC monocultures. Data are mean F S.E.M., n = 6. *p < 0.05 vs. static SMC, + p < 0.05 vs. wt SMC flow.

62

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65

Fig. 4 Effect of uPA and MMP inhibition on flow-induced SMC migration. PAI-1 / SMC were exposed to either static or pulsatile flow conditions (25 ml/min, 24 h) in the presence or absence of anti-uPA antibody #1189 (25 Ag/ ml) F the MMP inhibitor, BB3103 (4 AM). Data are expressed as mean F S.E.M., n = 3. *p < 0.05 vs. respective static SMC, + p < 0.05 vs. PAI-1 / SMC flow.

(HUSMC) exposed to pulsatile flow (25 ml/min, 24 h, corresponding to a pulse pressure of 149/48 mm Hg intraluminally and 106/50 extraluminally) demonstrated a significant increase in PAI-1 mRNA and protein levels; f 4- and 3-fold increases, respectively (Fig. 2).

Effect of PAI-1 gene deletion on flow-induced SMC migration We compared the effect of pulse pressure due to pulsatile flow on random migration (i.e., chemo-

kinesis) of murine wild type and PAI-1 / SMC. Confirming our previously reported finding [16,17], wild-type SMC increased ( f 2.4-fold) their migration after exposure to pulsatile flow compared with wt SMC under static (0 ml/min) conditions (Fig. 3). The flow-induced migration response of these wildtype cells was completely inhibited when they were cocultured in the presence of EC (Fig. 3). While there was no difference in the migration of PAI-1 / SMC compared to wt SMC under static conditions, PAI-1 / SMC migration increased to a greater extent under pulsatile flow conditions as compared to wild-type controls (334 F 22% vs. 237 F 11%, n = 6, p < 0.05) (Fig. 3). This flow-induced migration was significantly attenuated, but not completely inhibited, when PAI-1 / SMC were cultured in the presence of EC (147 F 13%, n = 6, p < 0.05) (Fig. 3).

Effect of uPA and matrix metalloproteinase inhibition on flow-induced PAI-1 / SMC migration We investigated the role of uPA and MMP in mediating the flow-induced migration response in PAI-1 gene deficient SMC. PAI-1 / SMC were exposed to static or pulsatile flow conditions in the absence or presence of an anti-uPA antibody (#1189; 25 Ag/ ml) alone, or in combination with the MMP inhibitor BB3103 (4 AM). The PAI-1 / SMC migratory response was significantly attenuated by 1189,

Fig. 5 Effect of pulsatile flow on MMP-2 and MMP-9 activity in wild type and PAI-1 / SMC. Wild type or PAI-1 / SMC were exposed to either static or flow conditions (25 ml/min, 24 h) and MMP-2 and MMP-9 activity in cellular protein was analyzed by zymography as described in Materials and methods. Representative zymograph (top) with cumulative densitometric data of five separate experiments. Data are expressed as mean F S.E.M. *p < 0.05 vs. respective wild type, #p < 0.05 vs. respective static.

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65 and completely inhibited in the presence of both 1189 and BB3103 (Fig. 4).

63

there was no effect of flow on TIMP-2 activity in PAI-1 / SMC (Fig. 6).

Effect of PAI-1 gene deletion on MMP activity Under static conditions, MMP-9 activity was significantly greater (173 F 40%, n = 5) in monocultured PAI-1 / SMC, when compared to wt controls (Fig. 5). However, there was no flow-induced change in MMP-9 activity in either wt or PAI-1 / SMC (Fig. 5). In contrast, while there was no difference in MMP-2 activity between wt and PAI-1 / SMC under static conditions, there was a greater flow-induced increase in MMP-2 activity in PAI-1 / SMC, as compared to wt SMC (236 F 74% vs. 125 F 9%, n = 5) (Fig. 5). EC-coculture had no effect on the flow-induced increase in MMP-2 activity in wild type or PAI-1 / SMC (data not shown).

Effect of PAI-1 gene deletion on TIMP-2 activity under static and flow conditions TIMP-2 activity was measured in conditioned media by reverse gelatin zymography. TIMP-2 activity was greater in wild type vs. PAI-1 / SMC under static conditions (Fig. 6). While there was a flowdependent increase in TIMP-2 activity in wt SMC,

Fig. 6 Effect of pulsatile flow on TIMP activity in wild type and PAI-1 / SMC. Wild type or PAI-1 / SMC were exposed to either static or flow conditions and TIMP-2 activity in conditioned media was analyzed by reverse zymography as described in Materials and methods. Representative reverse zymograph (top) with cumulative densitometric data of four separate experiments. Data are expressed as mean F S.E.M. *p < 0.05 vs. respective static SMC, #p < 0.05 vs. wt SMC flow.

Discussion This study demonstrates that smooth muscle cellderived PAI-1 plays a role in limiting flow-induced SMC migration and may thus be an important mechanism for controlling the process of vascular remodeling. Seemingly, paradoxical effects of PAI-1 on cellular migration and plaque development have been reported. Increased levels of PAI-1 mRNA within the thickened intima of human atherosclerotic arteries have been observed [23], together with an exaggerated neointimal response to arterial injury in PAI-1 deficient mice and inhibition of vascular wound healing and neointima formation following adenoviral-delivered PAI-1 [8]. Similarly, implantation of SMC genetically engineered to overexpress PAI-1 transiently limited neointimal growth in rats reinforcing the importance of PAI-1 in the process of neointimal growth [9]. Indeed, active PAI-1 has been shown to impair SMC adhesion and migration by limiting the binding of vitronectin to the integrin receptor v3 [24]. On the other hand, PAI-1 can enhance SMC migration under selective circumstances [25], while Sjoland et al. [26] reported that genetic modification of PAI-1 had no effect on the progression of atherosclerosis in LDL receptor-deficient and apolipoprotein E-deficient mice models of atherosclerosis. Furthermore, in apparent contrast to the study by Carmeliet et al. [8], Zhu et al. [27] demonstrated that PAI-1 deficiency was protective against neointima formation after ferric chloride-induced vascular injury in atherosclerosis-prone mice. Interestingly, Eitzman et al. [28] reported that PAI-1 deficiency protected against atherosclerosis preferentially at sites of turbulent flow, e.g., at the carotid bifurcation, suggesting a hemodynamic component. Moreover, a potentially key role for PAI-1 in relation to its effects on SMC migration in dictating plaque stability has emerged [29]. This may be of considerable clinical importance as the most relevant manifestation of atherosclerosis is plaque rupture and formation of an occlusive thrombus. Taken together, these studies implicate PAI-1 in the setting of restenosis and atherosclerosis. In healthy human arteries PAI-1 is found in both EC and SMC of the arterial media [30,31]. PAI-1 is associated with several cellular components of the developing atherosclerotic plaque, including EC, SMC and macrophages, in addition to being associ-

64

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65

ated with the extracellular matrix. In situ hybridization and immunohistochemistry studies in both human and mouse indicate that expression of PAI-1 by activated endothelium is prevalent in the early stages of atherosclerosis, whereas in advanced lesions SMC and macrophage PAI-1 expression is predominant [30,32]. Platelets do not appear to be a major source of circulating PAI-1 during atherogenesis [32]. Thus, SMC are an important source of PAI-1 in normal and atheromatous vascular tissue. Given the importance of hemodynamic forces in atherosclerotic plaque development, our in vitro studies, using cells in culture exposed to physiologically relevant mechanical forces, are pertinent to the potential role and cellular source of PAI-1 in atherosclerosis development. We have previously demonstrated that pulse pressure increases SMC migration by a mechanism mediated in part by uPA [16]. Furthermore, endothelial cell-derived PAI-1, or exogenously added PAI-1, inhibited the flow-induced migration response [17]. Our current study suggests that SMC-derived PAI-1, in part, by inhibiting MMP and stimulating TIMP activity, also plays an important role in limiting hemodynamic force-induced SMC migration as PAI-1 deficient SMC had a greater MMP-2 activity and reduced TIMP-2 activity concomitant with a greater flow-induced migratory response, compared to wild-type cells. A likely explanation for our results supporting an inhibitory effect of SMC PAI-1 on MMP activity and flow-induced SMC migration is that PAI-1 inhibits uPA-and tPA-mediated activation of plasmin, which in turn limits MMP activity and ECM degradation. These results are in agreement with Proia et al. [33] who showed that endothelial cell PAI-1 overexpression decreased SMC MMP-2 and MMP-9 activity and SMC migration. However, a direct inhibitory effect of PAI-1 on SMC migration also remains a possibility [25,34]. Further work is required to delineate the precise mechanism(s) for the inhibitory effect of SMC PAI-1 on flow-induced SMC migration. The activities of MMPs are controlled at three levels; gene expression, activation of the latent proenzyme forms of the MMPs and inhibition by complexing with their specific TIMPs [10]. TIMP-2 activity was decreased in PAI-1 / SMC, especially under flow conditions. A PAI-1-mediated increase in TIMP2 activity could also result in MMP inhibition. While there are few previous reports to our knowledge of interaction between the plasminogen/plasmin and MMP systems at the level of their respective inhibitors (i.e., PAI-1 and TIMP-2), Hasenstab et al. [35] found that first PAI-1 activity, then TIMP-2 activity, increased following arterial injury in rats, supporting potential crosstalk be-

tween the two inhibitors. Interestingly, while most of the cell biological effects of TIMPs are believed to be mediated indirectly by inhibition of MMPs, direct effects of TIMPs on SMC migration, proliferation and apoptosis have been described [36]. Following endothelial dysfunction or denudation, dramatic changes in SMC migration and proliferation can occur [1,2]. Indeed, when wild-type SMC were cocultured with EC, there was no effect of pulsatile flow on SMC migration, underscoring the ‘protective’ capacity of the endothelium. However, the presence of endothelial cells, while significantly diminishing, did not completely inhibit the migratory response to flow of PAI-1 / SMC, suggesting that SMC-derived PAI-1 is also necessary to fully control the flow-induced SMC migratory response, at least in vitro. In conclusion, the current study demonstrates for the first time that smooth muscle cells deficient in the PAI-1 gene have an exacerbated migratory response to pulsatile flow as compared with wild-type cells. Taken together with our previous reports [16,17], this data suggests that in addition to endothelial PAI1, smooth muscle cell-derived PAI-1 may play an important role in hemodynamic force-induced remodeling in vivo.

Acknowledgements This work was supported in part by grants from the National Institutes of Health (AA-12610 and HL59696 to EMR, DK47067 to JVS, HL-64971 to SSO), and by grants from the Wellcome Trust and the Health Research Board of Ireland (PAC). JPC and SMN are both recipients of Postdoctoral Fellowship Awards from the American Heart Association, New York State Affiliate. We thank Dr. Nicholas Theodorakis for helpful discussion of the manuscript.

References [1] Schwartz SM. Smooth muscle migration in atherosclerosis and restenosis. J Clin Invest 1997;100(Suppl. 11): 87--9. [2] Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990’s. Nature 1993;362:801--9. [3] Kenagy RD, Vergel S, Mattsson E, Bendeck M, Reidy MA, Clowes AW. The role of plasminogen, plasminogen activators and matrix metalloproteinases in primate arterial smooth muscle migration. Arterioscler Thromb Vasc Biol 1996;16:1373--82. [4] Clowes AW, Clowes MM, Au YP, Reidy MA, Belin D. Smooth muscle cells express urokinase during mitogenesis and tissue-type plasminogen activator during migration in injured rat carotid artery. Circ Res 1990;67:61--7. [5] Cho A, Reidy MA. MMP-9 is necessary for the regulation of

J.P. Cullen et al. / Thrombosis Research 114 (2004) 57--65

[6] [7]

[8]

[9]

[10] [11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

smooth muscle cell replication and migration after arterial injury. Circ Res 2002;91:845--51. Lijnen HR. Plasmin and matrix metalloproteinases in vascular remodeling. Thromb Haemost 2001;86:324--33. Carmeliet P, Bouche A, de Clercq C, Janssen S, Pollefeyt S, Wyns S, et al. Biological effects of disruption of the tissuetype plasminogen activator, urokinase-type plasminogen activator and plasminogen activator inhibitor-1 genes in mice. Ann N Y Acad Sci 1995;748:367--81. Carmeliet P, Moons L, Lijen R, Janssens S, Lupu F, Collen D, et al. Inhibitory role of plasminogen activator inhibitor-1 in arterial wound healing and neointima formation: a gene targeting and gene transfer study in mice. Circulation 1997;96:3180--91. Hasenstab D, Lea H, Clowes AW. Local plasminogen activator inhibitor type 1 overexpression in rat carotid artery enhances thrombosis and endothelial regeneration while inhibiting intimal thickening. Arterioscler Thromb Vasc Biol 2000;20:846--52. Dollery CM, McEwan JR, Henney AM. Matrix metalloproteinases and cardiovascular disease. Circ Res 1995;77:863--8. Carmeliet P, Moons L, Herbert J, Crawley J, Lupu F, lijnen R, et al. Urokinase but not tissue plasminogen activator mediates arterial neointima formation in mice. Circ Res 1997;81:829--39. Bendeck MP, Irvin C, Reidy MA. Inhibition of matrix metalloproteinase activity inhibits smooth muscle cell migration but not neointimal thickening after arterial injury. Circ Res 1996;78:38--43. Wilson E, Sudhir K, Ives HE. Mechanical strain induces growth of vascular smooth muscle cells via autocrine production of PDGF. J Cell Biol 1993;123:741--7. Cappadona C, Redmond EM, Theodorakis NG, McKillop IH, Hendrickson R, Chhabra A, et al. Phenotype dictates the growth response of vascular smooth muscle cells to pulse pressure in vitro. Exp Cell Res 1999;250:174--86. Heistad DD, Lopez JA, Baumbach GL. Hemodynamic determinants of vascular changes in hypertension and atherosclerosis. Hypertension 1991;17:7--11. Redmond EM, Cahill PA, Hirsch M, Wang Y.-N., Sitzmann JV, Okada SS. Effect of pulse pressure on vascular smooth muscle cell migration: the role of urokinase and matrix metalloproteinase. Thromb Haemost 1999;81:293--300. Redmond EM, Cullen JP, Cahill PA, Sitzmann JV, Stefansson S, Lawrence DA, et al. Endothelial cells inhibit flow-induced smooth muscle cell migration: role of plasminogen activator inhibitor-1. Circulation 2001;103:597--603. Carmeliet P, Kieckens L, Schoojans L, Ream B, van Nuffelen A, Prendergast G, et al. Plasminogen activator inhibitor-1 gene-deficient mice. Generation by homologous recombination and characterization. J Clin Invest 1993;92:2746--55. Grobmyer SR, Kuo A, Orishimo M, Okada S, Cines DB, Barnathan ES. Determinants of binding and internalization of tissue-type plasminogen activator by human vascular smooth muscle and endothelial cells. J Biol Chem 1993; 268:13291--300. Ojeifo JO, Su N, Ryan US, Verma UN, Mazumder A, Zwiebel JA. Towards endothelial-cell-directed cancer immunotherapy: in vitro expression of human recombinant cytokine genes by human and mouse primary endothelial cells. Cytokines Mol Ther 1996;2(2):89--101. Redmond EM, Cahill PA, Sitzmann JV. Perfused transcapil-

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

65

lary smooth muscle cell and endothelial cell co-culture: a novel in vitro model. In Vitro Cell Dev Biol 1995;31:601--9. Redmond EM, Cahill PA, Sitzmann JV. Flow-mediated regulation of endothelin receptors in co-cultured vascular smooth muscle cells: an endothelium-dependent effect. J Vasc Res 1997;34:425--35. Schneiderman J, Sawdey MS, Keeton MR, Bordin GM, Bernstein EF, Dilley RB, et al. Increased type 1 plasminogen activator inhibitor gene expression in atherosclerotic human arteries. PNAS 1992;89:6998--7002. Stefansson S, Lawrence DA. The serpin PAI-1 inhibits cell migration by blocking integrin v3 binding to vitronectin. Nature 1996;383:391--441. Tanaka S, Koyama H, Ichii T, Shioi A, Hosoi M, Raines EW, et al. Fibrillar collagen regulation of plasminogen activator inhibitor-1 is involved in altered smooth muscle cell migration. Arterioscler Thromb Vasc Biol 2002;22:1573--8. Sjoland H, Eitzman DT, Gordon D, Westrick R, Nabel EG, Ginsburg D. Atherosclerosis progression in LDL receptor-deficient and apolipoprotein E deficient mice is independent of genetic alterations in plasminogen activator inhibitor-1. Arterioscler Thromb Vasc Biol 2000;20:846--52. Zhu Y, Farrehi PM, Fay WP. Plasminogen activator inhibitor type 1 enhances neointima formation after oxidative vascular injury in atherosclerosis-prone mice. Circulation 2001;103:3105--10. Eitzman DT, Westrick RJ, Xu Z, Tyson J, Ginsburg D. Plasminogen activator inhibitor-1 deficiency protects against atherosclerosis progression in the mouse carotid artery. Blood 2000;96:4212--5. Sobel BE, Taatjes DJ, Schneider DJ. Intramural plasminogen activator inhibitor type-1 and coronary atherosclerosis. Arterioscler Thromb Vasc Biol 2003;23:1979--89. Lupu F, Bergonzelli GE, Heim DA, Cousin E, Genton CY, Bachmann F, et al. Localization and production of plasminogen activator inhibitor-1 in human healthy and atherosclerotic arteries. Arterioscler Thromb 1993;13:1090--100. Salame MY, Samani NJ, Masood I, deBono DP. Expression of the plasminogen activator system in the human vascular wall. Atherosclerosis 2000;152:19--28. Schafer K, Muller K, Hecke A, Mounier E, Goebel J, Loskutoff DJ, et al. Enhanced thrombosis in atherosclerosis-prone mice is associated with increased arterial expression of plasminogen activator inhibitor-1. Arterioscler Thromb Vasc Biol 2003;23:2097--103. Proia RR, Nelson PR, Mulligan-Kehoe MJ, Wagner RJ, Kehas AJ, Powell RJ. The effect of endothelial cell overexpression of plasminogen activator inhibitor-1 on smooth muscle cell migration. J Vasc Surg 2002;36:164--71. Kjoller L, Kanse SM, Kirkegaard T, Rodenburg KW, Ronne E, Goodman SL, et al. PAI-1 represses integrin- and vitronectin-mediated cell migration independently of its function as an inhibitor of plasminogen activation. Exp Cell Res 1997;232:420--9. Hasenstab D, Forough R, Clowes AW. Plasminogen activator inhibitor type 1 and tissue inhibitor of metalloproteinases2 increase after arterial injury in rats. Circ Res 1997;80: 490--6. Baker AH, Zaltsman AB, George SJ, Newby AC. Divergent effects of TIMP-1, -2 or -3 overexpression on rat vascular smooth muscle cell invasion, proliferation and death in vitro. J Clin Invest 1998;101:1478--87.