Urokinase but Not Tissue Plasminogen Activator Mediates Arterial Neointima Formation in Mice
Abstract To define the role of the plasminogen activators (PAs) tissue PA (t-PA) and urokinase PA (u-PA) in vascular wound healing, neointima formation and reendothelialization were evaluated after electric or mechanical arterial injury in mice with a single or combined deficiency of t-PA (t-PA−/−) and/or u-PA (u-PA−/−). In both models, neointima formation and neointimal cell accumulation were reduced in u-PA−/− and in t-PA−/−/u-PA−/− arteries but not in t-PA−/− arteries. The electric injury model was used to characterize the underlying cellular mechanisms. Topographic analysis of vascular wound healing in electrically injured wild-type and t-PA−/− arteries revealed a similar degree of migration of smooth muscle cells from the noninjured borders into the necrotic center. In contrast, in u-PA−/− and t-PA−/−/u-PA−/− arteries, smooth muscle cells accumulated at the uninjured borders but failed to migrate into the necrotic center. Cultured u-PA−/− but not t-PA−/− smooth muscle cells also failed to migrate in vitro after scrape wounding. Proliferation of smooth muscle cells was not affected by PA deficiency. Reendothelialization after electric injury was similar in all genotypes. In situ analysis revealed markedly elevated u-PA zymographic activity, mRNA, and immunoreactivity in smooth muscle cells, endothelial cells, and leukocytes within 1 week after injury, eg, when cells migrated into the wound. Thus, u-PA plays a significant role in vascular wound healing and arterial neointima formation after injury, most likely by affecting cellular migration.
Proteinases have been implicated in luminal stenosis, complicating vascular reconstructions for the treatment of atherothrombosis.1 2 3 4 5 Although vascular remodeling appears to be a major determinant of luminal stenosis after balloon angioplasty, intimal thickening can also contribute to the luminal narrowing, especially when vascular remodeling is prevented by placement of intraluminal stents. The latter results from accumulation of smooth muscle cells that proliferate in the media, migrate across the elastic membrane, and contribute to matrix deposition.3 4 5 Recent indirect evidence suggests an involvement of two proteinase systems, the plasminogen and metalloproteinase systems, in the process of smooth muscle cell migration.1 2 3 4 5 6 7 8 9 10 11
The plasminogen system is composed of an inactive proenzyme, plasminogen, that is converted to active plasmin by two physiological PAs, t-PA and u-PA.12 Their action is controlled by PAIs, of which PAI-1 appears to be the predominant physiological inhibitor. Whereas t-PA is primarily involved in clot dissolution, u-PA, which binds to a membrane-anchored receptor (u-PAR), has been implicated in pericellular proteolysis during cell migration or tissue remodeling.13 Plasmin has been presumed to play a role in tissue remodeling during wound healing and inflammation in a variety of processes, including glomerulonephritis, skin healing, atherosclerosis, and arterial neointima formation, via proteolysis of extracellular matrix components or activation of growth factors.14 15 16 Recent observations further suggest that the role of plasmin may largely depend on the breakdown of fibrin.17
In an uninjured artery, t-PA production by quiescent endothelial cells may promote vascular patency, whereas PAI-1 synthesis by medial smooth muscle cells has been proposed to provide a hemostatic barrier.1 10 After injury, expression of t-PA, u-PA, and u-PAR by smooth muscle, endothelial, and inflammatory cells is significantly induced, suggesting that a hyperfibrinolytic response may participate in the migration and/or proliferation of these cells.7 8 9 11 13 18 19 Indeed, smooth muscle cells use proteinases to degrade the extracellular matrix that encages them, preventing them from migration into the wound.1 3 Plasmin may trigger this process, since it can directly degrade fibrin and matrix and also activate other matrix-degrading proteinases, including the metalloproteinases.1 14 Indirect evidence has been provided that the plasminogen system participates in vascular wound healing in several species, including humans.20 21 22 In addition, when a perivascular electric injury model was used in plasminogen-deficient mice, it was demonstrated that plasmin proteolysis was essential for normal arterial wound healing and contributed significantly to the formation of a neointima.23 Studies in the rat with the plasmin inhibitor tranexamic acid have also suggested a role for plasmin in cellular migration following balloon injury of the carotid artery.8 Although increased expression of both plasminogen activators suggested their involvement in this process,1 7 8 9 direct functional proof for a causal and possible differential role in neointima formation remained to be determined. In addition, it was unknown whether t-PA and/or u-PA are essential for reendothelialization.
Using two models of arterial injury, the present study involving mice with inactivation of the genes encoding t-PA (t-PA−/−) and/or u-PA (u-PA−/−)24 provides direct genetic proof, of a significant role of u-PA–mediated but not t-PA–mediated plasmin proteolysis in arterial neointima formation, most likely via regulation of smooth muscle cell migration.
Materials and Methods
Arterial Injury Model
Six- to 8-week-old WT, t-PA−/−, or u-PA−/− mice24 of either sex with a 75% C57Bl6 and 25% 129 genetic background were used. Groups of 8 to 14 animals were studied. Perivascular electric injury of left femoral arteries and analysis of the neointima was performed as described elsewhere.25 Although neointima formation occurs in both femoral and carotid arteries, the response in the femoral arteries was analyzed, since it occurs somewhat more rapidly and extensively (authors’ unpublished data, 1996). Intraluminal mechanical injury of the carotid arteries was performed according to a recently described method by Lindner et al,26 with minor modifications. Briefly, the left internal carotid artery was exposed by blunt-end dissection, tied off distally, and looped proximally with 7.0 Mersilene nylon suture for vascular control during the procedure. A transverse arteriotomy was made in the proximal portion of the internal carotid artery, and a 380-μm flexible guide wire (C-SF 15-15, Cook Belgium NV) was passed toward the aortic arch over ≈15 mm and withdrawn three times with a rotating motion. After removal of the guide wire, the proximal portion of the internal carotid artery was tied off, and the skin incision was closed. This protocol allowed reproducible injury, resulting in rupture of the internal elastic lamina and neointimal accumulation of α-smooth muscle actin–positive cells within 3 weeks after injury. The carotid artery was used because of technical limitations to insert the guide wire in the smaller femoral artery.
Tissue Harvest, Histology, and Immunocytochemistry
Tissue harvest, fixation, embedding, and sectioning were performed as described.25 For t-PA or u-PA immunostaining, the injured vessels were perfusion-fixed with 1% phosphate-buffered paraformaldehyde (pH 7.0) for 10 minutes without postfixation and cryoembedded. Smooth muscle and inflammatory cells were immunostained for smooth muscle α-actin or CD45 as described.25 Macrophages were visualized using a rat monoclonal antibody against Mac-3 (Pharmingen). Replication was determined by labeling cells with BrdU and immunostaining as described.25 For t-PA staining, a polyclonal rabbit anti-murine t-PA24 or goat anti-human t-PA (10 μg/mL, American Diagnostica No. 387/1) antiserum was used with similar results. For u-PA staining, a biotinylated mouse anti-murine u-PA monoclonal antibody (clone No. H77A10, 30 μg/mL),27 a rabbit anti-murine u-PA,24 a goat anti-human u-PA (100 μg/mL, American Diagnostica No. 398), or rabbit anti-human u-PA (100 μg/mL, American Diagnostica No. 389) antiserum was used with similar results, although background staining was somewhat different. Colocalization of PAs with smooth muscle cells or macrophages was established by use of a double–immunofluorescence labeling approach as previously described.28 With this procedure, PA-positive cells appear green, α-actin– or Mac-3–positive cells appear red, and cells containing PA and α-actin or PA and Mac-3 appear yellow.
Morphometric Analysis and Reendothelialization
Morphometric measurements of cross-sectional areas, cell counts, and proliferation rates and topographic analysis of neointima formation in electrically injured arteries were performed as described.25 Morphometric analysis of mechanically injured arteries was performed on five sections equally spaced across the injured segment. Reendothelialization was evaluated after staining the denuded blood vessels with Evans blue as previously described.25
Zymographic analysis of PA activities in arterial extracts was performed as described.23 In situ zymography on 7-μm arterial cryosections was performed by fibrin overlay, using a gel of fixed thickness prepared by clotting a mixture of human fibrinogen (final concentration, 4 mg/mL), plasminogen (final concentration, 10 μg/mL), and agarose (final concentration, 0.5%) with thrombin (final concentration, 0.3 NIH units/mL). The fraction of the lysis due to t-PA or u-PA activity was determined by including in the fibrin gel 50 μg/mL polyclonal neutralizing rabbit anti-murine t-PA– and/or u-PA–specific IgGs, respectively.24 A standard curve for t-PA and u-PA activity was obtained by spotting different amounts of purified murine t-PA and u-PA24 on the fibrin gels and quantification of the lysis at different time intervals. The amount of lysis, representing the lysis area multiplied by the intensity of lysis and thus the total fibrinolytic activity per section, was quantitatively analyzed using the Quantimed 600 image analysis software and expressed in arbitrary units.
In Situ Hybridization
In situ hybridization was performed as previously described.28 The following cRNA probes were used: a 600-bp fragment of the human u-PA cDNA or a 159-bp fragment of the murine u-PA cDNA, subcloned in the Bluescript M13+ vector (Stratagene Inc) or in the pGEM-7 vector (Promega), respectively, and labeled by run-on transcription using 35S-labeled UTP (specific activity, 1300 Ci/mmol, New England Nuclear). Sense and antisense probes were prepared by linearizing the constructs with the appropriate restriction endonucleases and by the use of either T7/T3 RNA or T7/SP6 RNA polymerases (Boehringer-Mannheim). RNase digestion was carried out after probe hybridization to reduce the background. Both human and murine u-PA probes yielded similar results.
Smooth Muscle Cell Cultures and In Vitro Migration Assay
After saline perfusion and removal of the adventitia, the thoracic and upper part of the abdominal aorta were rinsed in saline and cut into small pieces (1 mm2), and the fragments were enzymatically dispersed by incubation for 16 hours at 37°C in DMEM containing 0.15% collagenase (type II, Sigma) and 5% fetal calf serum. Smooth muscle cells were cultured in DMEM containing 10% fetal calf serum and passaged one to three times before analysis. Wound assays were performed in vitro as described.29 Briefly, confluent monolayers of smooth muscle cells, seeded 7 days earlier in DMEM containing 10% fetal calf serum, in 35-mm dishes were wounded with a razor blade. After wounding, the cells were washed with PBS and further incubated for 72 hours at 37°C in DMEM containing 0.1% gelatin and 10% fetal calf serum. After fixation with absolute methanol, the cells were stained with Giemsa. Cells that had migrated from the edge of the wound into the denuded area were counted in seven successive 125-μm increments at ×100 magnification.
Experimental values were expressed as mean±SEM. Statistically significant differences between groups were calculated by ANOVA followed by Bonferroni correction or by χ2 analysis, as indicated in the text.
Vascular Wound Healing in WT Mice
Electric Injury Model
Since the electric injury model differs from the mechanical injury model,25 its characteristics are briefly discussed (Fig 1⇓). Electric injury denudes the endothelium and destroys all cells in the media (<1% surviving medial cells), resulting in a wound-healing response characterized by transient thrombosis, removal of necrotic debris, reendothelialization, repopulation of the media by smooth muscle cells, and the formation of a neointima with multiple layers of smooth muscle cells within 3 weeks after injury (Fig 2a⇓ to 2d).25 A neoadventitia that was rich in leukocytes and fibroblast-like cells developed.
Mechanical Injury Model
Mechanical injury resulted in less severe medial cell necrosis (52±8% surviving medial cells, P<.05 versus uninjured arteries) than did electric injury. Since residual viable smooth muscle cells persisted across the entire injured segment, neointima formation occurred more uniformly across the injured segment, as described previously for similar mechanical injury models.30 In Fig 2⇑, panel g displays a representative neointima from a WT carotid artery, containing smooth muscle cells within 3 weeks after mechanical injury. Since the electric injury model permits the differentiation between proliferation and migration of smooth muscle cells and is technically easier than the mechanical injury model, it was used more extensively throughout the present study.
Vascular Wound Healing in Gene-Inactivated Mice
Electric Injury Model
After electric injury, vascular wound healing was similar in t-PA−/− and WT arteries. In contrast, healing was significantly impaired in u-PA−/− and t-PA−/−/ u-PA−/− arteries, as evidenced by the reduced accumulation of cells within the media and neointima (Fig 2e⇑ and 2f⇑; see below for quantification) and by the impaired removal of the necrotic debris in the media (as judged by microscopic analysis, Fig 2e⇑ and 2f⇑). Compared with WT and t-PA−/− arteries, which revealed a neointima across the entire injured segment (from location 1 to 10 in schematically represented arteries in Fig 1⇑) within 3 weeks after injury, a neointima was present only in u-PA−/− and t-PA−/−/u-PA−/− arteries at the uninjured borders (location 1 or 10) (Fig 2e⇑) and failed to progress into the necrotic center (location 5) (Fig 2f⇑) (see also below). Although adventitial inflammation was observed in all genotypes, it was less extensive in u-PA−/− and t-PA−/−/u-PA−/− arteries than in WT or t-PA−/− arteries (compare Fig 2c⇑ with 2e).
Quantitative morphometry revealed that the neointima, as deduced from measuring the cross-sectional area between the internal elastic lamina and the lumen (Fig 3⇓), and the number of neointimal cells (Table 1⇓) were significantly reduced in u-PA−/− and t-PA−/−/u-PA−/− arteries but not in t-PA−/− arteries. Within 4 weeks after injury, the intima-to-media ratio was 1.60±0.25 in WT mice, 1.48±0.33 in t-PA−/− mice (P=NS), 0.57±0.15 in u-PA−/− mice (P=.003 versus WT mice), and 0.29±0.17 in t-PA−/−/u-PA−/− mice (P=.011 versus WT mice). Repopulation of the media was also impaired in u-PA−/− and in t-PA−/−/u-PA−/− mice compared with WT and t-PA−/− mice (Table 1⇓).
Mechanical Injury Model
Deficiency of u-PA also impaired neointima formation after mechanical injury, as revealed by the histological analysis in Fig 2h⇑ and 2i⇑. Within 3 weeks after injury, the neointimal cross-sectional area, the luminal stenosis, the intima-to-media ratio, and the number of neointimal cells were lower in u-PA−/− than in WT or t-PA−/− mice (Table 2⇑). Thus, deficiency of u-PA impaired vascular wound healing and neointima formation in both injury models. The mechanisms underlying the improved neointima formation in u-PA−/− mice were further investigated using the electric injury model.
Proliferation of Smooth Muscle Cells
To evaluate whether deficiency of t-PA or u-PA affects cellular proliferation, incorporation of BrdU into replicating cells was determined (Table 3⇑). Within 1 week after electric injury, the total BrdU-labeling index represents proliferation of leukocytes and, to a lesser extent, smooth muscle cells, whereas within 2 weeks after injury, the total BrdU-labeling index reflects proliferation of smooth muscle cells, since they constitute >95% of the intimal and medial cell population.25 As shown in Table 3⇑, medial and neointimal cell proliferation were not significantly different between the different genotypes, except for a somewhat lower proliferation rate of medial smooth muscle cells in u-PA−/− arteries within 1 week after injury.
Migration of Smooth Muscle Cells In Vivo
Smooth muscle cells after electric injury migrated within the media and alongside the lumen from the uninjured borders into the necrotic center.25 This process was quantified by measuring the luminal narrowing (percent stenosis) and cell accumulation at equally spaced positions across the injured segment. Within 3 weeks after injury, a significant neointima was present throughout the entire injured segment in WT and t-PA−/− arteries (Fig 4⇓). In contrast, a neointima was initiated at the borders of the injury in u-PA−/− and t-PA−/−/u-PA−/− arteries but failed to progress into the necrotic center (Fig 4⇓).
These findings were extended by counting the medial and neointimal cell nuclei at three equally spaced positions across the injured segment (locations 1, 5, and 10 in the arteries of Fig 1⇑) within the first 2 weeks after injury, when cells start to migrate and accumulate in the intima. As shown in Table 4⇑, cells accumulated at the borders and progressed subsequently into the center of the injured segment in WT and t-PA−/− arteries. In contrast, cells accumulated at the borders but failed to migrate into the center of the injured segment of u-PA−/− arteries (Table 4⇑). Thus, these data indicate that smooth muscle cells failed to migrate as far in u-PA−/− arteries as they did in t-PA−/− or WT arteries. This process is schematically represented in Fig 1⇑.
Migration of Smooth Muscle Cells In Vitro
To substantiate the role of u-PA in smooth muscle cell migration, cultured smooth muscle cells from WT, t-PA−/−, and u-PA−/− mice were scrape-wounded, and their accumulation into the denuded area was quantified. WT and t-PA−/− but not u-PA−/− cells accumulated into the denuded area within 72 hours after wounding (Fig 5⇓). In addition, accumulation of WT cells was reduced in the presence of neutralizing u-PA antibodies (50 μg/mL of the polyclonal rabbit anti-murine u-PA) but not of t-PA antibodies (50 μg/mL of the polyclonal rabbit anti-murine t-PA). Within 6 hours after wounding, 25±4 WT and 19±1 t-PA−/− but only 3±0.5 u-PA−/− cells (P<.05 versus WT cells) accumulated in the denuded area. Since this occurs before the cells have a chance to divide (estimated to be 8 to 12 hours in these cultured murine smooth muscle cells), the reduced accumulation of cells is most likely due to impaired cellular migration.
Electric injury completely denuded the injured segment of intact endothelium, as revealed by Evans blue staining immediately after injury.25 Reendothelialization was initiated from the uninjured borders into the necrotic center and was almost complete within 1 week after injury in WT mice. Endothelial regrowth within 7 days after injury was not affected, either in its extent or in its rate, by deficiency of either PA (P=NS for t-PA−/− and u-PA−/− compared with WT mice) (Fig 6⇓).
Electric injury of WT arteries resulted in transient mural thrombosis during the first week after injury, ie, before the appearance of the first smooth muscle cells in the neointima.25 Since loss of PA function increases the susceptibility to (venous) thrombosis and results in a decreased ability to spontaneously lyse 125I-fibrin–labeled pulmonary plasma clots,24 the incidence and extent of arterial thrombosis in PA−/− arteries was semiquantitatively evaluated. Within 1 week after injury, approximately one quarter of the arteries contained a mural thrombus, occluding the lumen between 5% and 25%. Within 2 weeks after injury, most WT arteries were completely devoid of thrombus, whereas mural thrombosis persisted somewhat longer in t-PA−/− arteries but persisted the longest in u-PA−/− arteries (Table 5⇑).
Electrically injured arteries become transiently infiltrated by leukocytes during the first week after injury.25 Since the plasminogen system significantly affects the migration of leukocytes,31 the number of leukocytes in the media and neointima were counted after staining for the pan-leukocyte marker CD45. Fewer leukocytes infiltrated into u-PA−/− than t-PA−/− or WT arteries. In the neointima, 15±3 leukocytes (16±3% of the total number of neointimal cells) accumulated in WT arteries, 12±4 leukocytes (23±10%) accumulated in t-PA−/− arteries (P=NS), but only 1.3±0.3 leukocytes (4±2%) accumulated in u-PA−/− arteries (P<.05 versus WT arteries). Although there were too few cells present in the media for appropriate statistical analysis, there was a similar trend of reduced leukocyte accumulation in the media of u-PA−/− arteries: 2±1 leukocytes (5±1% of total number of medial cells) accumulated in WT arteries, 1±1 leukocytes (13±7%) accumulated in t-PA−/− arteries, and only 0.2±0.2 leukocytes (1±1%) accumulated in u-PA−/− arteries.
Zymographic Analysis of PA Expression
PA activities of arteries were evaluated by in situ zymography on sections using fibrin overlay, since this technique allowed precise discrimination between uninjured and electrically injured regions as deduced from histological analysis on adjacent sections. Lysis is expressed in arbitrary units (see “Materials and Methods”).
In uninjured arteries from WT mice, lysis of the fibrin gel appeared within 2 hours (0.26±0.029, mean±SEM, n=4) and was inhibited for ≥95% (0.01±0.004) by inclusion of neutralizing t-PA antibodies but not u-PA antibodies (0.26±0.055), indicating that lysis was due to t-PA activity. As deduced from the immunocytochemical analysis (see below), only endothelial cells expressed t-PA in quiescent arteries.
In injured arteries, PA activities were expressed as a ratio of the lysis present in injured versus noninjured segments to normalize for variations in the absolute levels of lysis between experiments. In addition, since the specific activities of t-PA and u-PA in this assay were significantly different (because of their different affinity for and activation by fibrin), expressing the PA activities as ratios allowed us to compare their individual levels of induction after injury. In WT arteries, the ratio of lysis in injured versus noninjured sections was 0.28±0.06 within 2 days after injury (n=20) and restored to 1.0±0.09 (n=17) within 1 week after injury coincident with regeneration of the endothelium over the denuded surface. In u-PA−/− arterial sections, the ratio was 2.3±0.15 (n=26, P<.05 versus WT sections) within 1 week of healing. Since lysis by WT or u-PA−/− arteries at 2 and 7 days after injury was inhibited by >95% by neutralizing t-PA antibodies, lysis was essentially due to t-PA activity. There was no lysis within 2 to 5 hours over t-PA−/− sections (not shown).
In order to evaluate u-PA–mediated lysis, t-PA antibodies were incorporated in the fibrin overlay to neutralize t-PA activity. Lysis over sections from uninjured WT arteries in the presence of t-PA antibodies appeared only after prolonged incubation (24 hours at 37°C) but was ≈20-fold higher than lysis in uninjured t-PA−/− arterial sections analyzed side by side (not shown). Thus, it is likely that the t-PA antibodies incompletely blocked the residual t-PA in these sections upon prolonged overlay, thereby precluding accurate quantitative analysis of u-PA activity. Therefore, u-PA–mediated lysis of the fibrin overlay was evaluated in t-PA−/− arteries.
In uninjured t-PA−/− arteries, u-PA–mediated lysis appeared only after prolonged incubation (24 hours at 37°C) and was minimal (0.002±0.0006, n=8), consistent with the minimal expression of u-PA mRNA or immunoreactivity (see below). In sharp contrast, within 1 week after injury, fibrin overlay on arterial sections of t-PA−/− mice indicated that u-PA–mediated lysis was dramatically enhanced in injured segments compared with the noninjured border segments. Indeed, the ratio of lysis observed in injured versus noninjured sections was 470±120 (n=19). Approximately 50% of this lysis was mediated by u-PA, as evidenced by inclusion of neutralizing u-PA antibodies in the fibrin overlay, suggesting that other proteinases, possibly metalloproteinases, are responsible for the residual lysis. Lysis in t-PA−/− arteries was present in microdissected arterial fragments containing either the media and neointima (primarily smooth muscle cells and some leukocytes) (Fig 7a⇓) or only the adventitia (leukocytes and fibroblasts) (Fig 7b⇓), consistent with the in situ expression of u-PA mRNA and immunoreactive protein in these cell types (see below). Taken together, these data demonstrate that u-PA expression was minimal in quiescent arteries but that it markedly increased (eg, much more than t-PA) after arterial injury.
In Situ Hybridization and Immunostaining
These results were extended by in situ identification of the PA-producing cell types in WT arteries (n=3) within 1 week after injury (eg, when cells actively migrate).
In uninjured arteries, a low level of u-PA expression was detectable by in situ hybridization but not by immunocytochemistry (presumably reflecting differences in sensitivity of the assays) (Figs 8a⇓ and 9b⇓). In contrast, within 1 week after injury, there was a marked upregulation of u-PA mRNA and antigen expression in all layers of the injured vessel. The uninjured border from which smooth muscle cells start to migrate into the wound (location I in the schematically represented artery in Fig 9a⇓) displayed increased u-PA staining in endothelial cells in the intima, in smooth muscle cells in the media, and, at a lower level, in fibroblast-like cells and leukocytes in the enlarged adventitia (Fig 9c⇓). Significantly induced u-PA antigen and mRNA expression were detected in endothelial cells in the intima (Fig 8b⇓, 8c⇓, and 9d⇓) and in leukocytes (Fig 8b⇓, 8c⇓, and 9e⇓) and fibroblast-like cells (Fig 8b⇓, 8c⇓, and 9f⇓) in the adventitia at the site of maximal neointima formation (location II) and at the leading front of cellular migration (location III). In addition, double immunostaining revealed that smooth muscle cells in the intima and media (Fig 9h⇓) and Mac-3–positive macrophages in the adventitia (Fig 9i⇓) expressed increased amounts of u-PA. u-PA immunoreactivity was absent in u-PA−/− arteries (Fig 9g⇓) and in WT arteries after omission of the primary antisera (not shown). Only background hybridization was present using a sense u-PA probe (Fig 8d⇓). Thus, u-PA is expressed by endothelial cells, smooth muscle cells, leukocytes, and myofibroblast-like cells when they repopulate the wound.
In uninjured femoral arteries, only weak t-PA immunoreactivity was observed in intimal endothelial cells (Fig 10b⇓) but at a much lower level than in endothelial cells from the aorta (insert, Fig 10b⇓), suggesting topographic differences in t-PA expression across the vasculature. Immediately after injury, when the endothelium was denuded, t-PA staining was lost in the necrotic region (not shown), consistent with the drop in t-PA activity as analyzed by in situ zymography. Beyond 1 week after injury, t-PA staining was present in intimal endothelial cells, in smooth muscle cells in the media and intima, and in adventitial myofibroblast-like cells (Fig 10c⇓ to 10e). However, compared with the u-PA staining, the intensity of the t-PA staining was significantly lower, and there were fewer immunoreactive cells. In addition, topographic analysis revealed that t-PA was present in a majority of medial smooth muscle cells in the border region, adjacent to the injury (location I) (Fig 10c⇓), but that the fraction of t-PA–positive cells progressively diminished from the uninjured border to the cellular migration front; eg, the presence of t-PA immunoreactive cells diminished from ±50% of cells at location II (Fig 10d⇓) to only a minor fraction (<10%) at the leading edge of the migration front (location III) (Fig 10e⇓). Double immunostaining revealed that t-PA colocalized within α-actin smooth muscle cells (Fig 10f⇓) but not in Mac-3–positive macrophages (not shown). The weak α-actin immunostaining within 1 week after injury is due to its reduced expression, frequently observed in proliferating and migrating smooth muscle cells.25
Role of PAs in Cellular Migration and Tissue Remodeling
Deficiency of u-PA, but not t-PA, greatly reduces neointima formation, suggesting that u-PA mediates vascular wound healing after injury. In addition, combined deficiency of t-PA and u-PA resulted in a similar impairment of vascular wound healing. These data, together with our recent observations that deficiency of plasminogen also impaired arterial neointima formation,23 indicate that u-PA acts by conversion of plasminogen to plasmin.
The data further indicate that smooth muscle and possibly also inflammatory cells require u-PA–mediated plasmin proteolysis to migrate into the wound. Indeed, the observations, in u-PA−/− mice, that neointima formation was reduced, that cells accumulated at the borders but failed to progress into the necrotic center in vivo, and that u-PA−/− cultured smooth muscle cells failed to migrate into the denuded area after scrape wounding in vitro indicate that u-PA deficiency impaired smooth muscle cell and possibly also inflammatory cell migration (also see Fig 1⇑). Proliferation of smooth muscle cells was not affected by deficiency of u-PA. The plasmin inhibitor tranexamic acid and metalloproteinase inhibitors also reduce smooth muscle cell migration in the rat carotid artery without affecting smooth muscle cell proliferation.6 8 Thus, u-PA–mediated plasmin proteolysis appears to play a central role in proteolytic degradation of the extracellular matrix, allowing the cells to migrate to distant sites. Whether u-PA affects neointima formation/smooth muscle cell migration by modulating cell adhesion via interaction with the αvβ3-integrin receptor32 remains to be defined.
Besides its role in cellular migration, u-PA may also be involved in tissue remodeling during wound healing. Indeed, electron microscopy of WT arteries has revealed that the media after electric injury consists of necrotic smooth muscle cell remnants embedded in a fibrin-rich extracellular matrix.25 Thus, healing of the necrotic media must involve proteolytic degradation and removal of this debris. Impaired migration of u-PA−/− leukocytes and fibroblasts may also have contributed to the reduced formation of a neoadventitia in u-PA−/− or combined t-PA−/−/u-PA−/− mice.
Expression of PAs During Vascular Wound Healing
t-PA was expressed by quiescent endothelial cells but was lost immediately after denudation, similar to expression in the balloon-injured rat carotid artery.7 Within 1 week after injury, t-PA was expressed in regenerating endothelial cells and in some smooth muscle cells, restoring t-PA activity in injured regions to preinjury levels. The most significant difference in expression was, however, observed for u-PA, which was minimal in quiescent arteries but markedly induced during wound healing (eg, much more than t-PA). Increased expression of u-PA during vascular wound healing has also been observed in several other species, including humans.7 19 20 21 Notably, net u-PA–mediated fibrinolytic activity in injured arterial segments was >100-fold induced. Since u-PA mRNA and immunoreactivity were maximal at the leading edge of the migration front, a role for u-PA in cell migration is suggested. Since smooth muscle cells produce u-PA (the present study and References 1, 7, 9, 11, 191 7 9 11 19 -22, 33, and 34), they may control their own migration in an autocrine way. Alternatively, inflammatory cells in the wound (the present study and Reference 55 ) may assist in migration of smooth muscle cells by providing u-PA (similar to the paracrine regulation of tumor cell migration by stromal cells35 ) or by “clearing a path” for them. In this respect, it is noteworthy that infiltration of the vascular wound by leukocytes was reduced in u-PA−/− mice.
Role of PAs in Thrombosis
Loss of PA gene function increases the susceptibility of mice to injury- or inflammation-associated thrombosis, but almost exclusively in the venous system.24 Notably, u-PA−/− mice were more susceptible to fibrin deposition than were t-PA−/− mice.24 The present findings of prolonged arterial thrombosis, most significant in u-PA−/− mice, further underline the role of u-PA in fibrin surveillance. Endothelial and inflammatory cells may participate in this process, since they both can produce u-PA in vivo. Thrombosis may promote neointima formation by providing a provisional matrix for smooth muscle cell migration and supplying a variety of growth factors able to modulate proliferation and migration of smooth muscle cells.3 4 5 Interestingly, neointima formation was smaller but thrombosis was more persistent in u-PA−/− arteries, suggesting that the presence of a mural thrombus was not sufficient and even may have impeded neointima formation in u-PA−/− arteries.
Role of PAs in Endothelial Cell Regeneration
Endothelial cell regrowth was not impaired by t-PA or u-PA deficiency. This was not anticipated in view of the induced expression of PAs in regenerating endothelium in vivo and in vitro after injury (the present study and References 9 and 139 13 ). However, to date, no abnormalities in vascular development have been described in mice with inactivated genes of plasminogen system components during embryogenesis,15 suggesting an accessory role for the plasminogen system and/or compensation by other proteinase systems in endothelial cell migration. Endothelial cell migration alongside a denuded surface may, however, require proteolytic mechanisms other than invasion through anatomic barriers.
Role of t-PA in Vascular Wound Healing
Neointima formation in t-PA−/− mice was normal and not more impaired in combined t-PA−/−/u-PA−/− than in u-PA−/− mice, except for a transient reduction of neointimal cell accumulation within 1 week after injury in t-PA−/− mice. Interestingly, t-PA expression in the medial smooth muscle cells at the uninjured borders was increased shortly after injury, possibly indicating that t-PA participates in the initiation of smooth muscle cell migration. Such a role for t-PA was suggested previously in the balloon-injured rat carotid artery.7 8 9 However, t-PA appears to be inefficient in mediating sustained cellular migration in contrast to u-PA, presumably related to its lower level of expression. Unfortunately, we could not assess whether u-PA expression was compensatorily increased in t-PA−/− mice.
Electric Versus Mechanical Injury Model
Deficiency of u-PA impaired neointima formation, whereas deficiency of t-PA did not affect this process in either the electric or the mechanical injury model, indicating a general role of u-PA in the response to arterial injury. Whereas the mechanical injury model reflects more closely the injury inflicted in patients undergoing balloon angioplasty,5 the electric injury model is technically easier and has the advantage of providing a means to quantify the migration of smooth muscle cells from the uninjured borders into the necrotic center. Both models differ from each other in that the electric injury induces transient thrombosis and inflammation of the injured vessel wall. Since thrombosis and leukocytes accompany and modulate the wound-healing response in patients undergoing vascular reconstructions,5 the present study on wound-healing response after electric injury may bear some clinical relevance in our understanding of how PAs function during this process.
In conclusion, the present study provides direct evidence that u-PA–mediated plasmin proteolysis promotes vascular wound healing and associated neointima formation in mice, most likely by promoting migration of smooth muscle cells into the wound.
Selected Abbreviations and Acronyms
|t-PA−/−/u-PA−/−||=||t-PA and u-PA deficient|
|t-PA, u-PA||=||tissue and urokinase PA|
- Received March 24, 1997.
- Accepted August 7, 1997.
- © 1997 American Heart Association, Inc.
van Leeuwen RTJ. Extracellular proteolysis and the migrating vascular smooth muscle cell. Fibrinolysis. 1996;10:59-74.
Dollery CM, McEwan JR, Henney AM. Matrix metalloproteinases and cardiovascular disease. Circ Res. 1995;77:863-868.
Reidy MA, Jackson D, Lindner V. Neointimal proliferation: control of vascular smooth muscle cell growth. Vasc Med Rev. 1992;3:156-167.
Libby P, Schwartz D, Brogi E, Tanaka H, Clinton SK. A cascade model for restenosis: a special case of atherosclerotic progression. Circulation. 1992;86(suppl III):III-47-III-52.
Bendeck MP, Zempo N, Clowes AW, Galardy RE, Reidy MA. Smooth muscle cell migration and matrix metalloproteinase expression after arterial injury in the rat. Circ Res. 1994;75:539-545.
Clowes AW, Clowes MM, Au YPT, 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;67:61-67.
Reidy MA, Irvin C, Lindner V. Migration of arterial wall cells: expression of plasminogen activators and inhibitors in injured rat arteries. Circ Res. 1996;78:405-414.
Fearns C, Samad F, Loskutoff DJ, van Hinsbergh VWM, ed. Vascular control of hemostasis. In: Vadas MA, Harlan J, series eds. Advances in Vascular Biology. London, UK: Harwood Academic Publishing; 1996:207-227.
Collen D, Lijnen HR. Basic and clinical aspects of fibrinolysis and thrombolysis. Blood. 1991;78:3114-3124.
Vassalli JD. The urokinase receptor. Fibrinolysis. 1994;8(suppl 1):172-181.
Estreicher A, Müllhauser J, Carpentier GL, Orci L, Vassalli JD. The receptor for urokinase type plasminogen activator polarizes expression of the protease to the leading edge of migrating monocytes and promotes degradation of enzyme inhibitor complexes. J Cell Biol. 1990;111:783-792.
Reuning U, Bang N. Regulation of the urokinase-type plasminogen activator receptor on vascular smooth muscle cells is under the control of thrombin and other mitogens. Arterioscler Thromb. 1992;12:1161-1170.
Salame M, Martin W, More RS, Mahadeva H, deBono DP. Time sequence of changes in vessel wall tissue-type plasminogen activator, urokinase-type plasminogen activator and its receptor following experimental angioplasty. Fibrinolysis. 1996;10(suppl 3):249a. Abstract.
Kanthou C, Das S, Kakkar VV, Lupu F. Expression of some components of plasminogen activation (PA) pathway by arterial segments in organ culture after in vitro injury. Fibrinolysis. 1996;10(suppl 3):11a. Abstract.
Villaverde CA, Miralles M, Salvat N, Juan O, Tibau N, Pena E, Vigo P. Regulation of plasmin-generating activity in human arteries. Fibrinolysis. 1996;10(suppl 3):148a. Abstract.
Lindner V, Fingerle J, Reidy MA. Mouse model of arterial injury. Circ Res. 1993;73:792-796.
Declerck PJ, Carmeliet P, Verstreken M, De Cock F, Collen D. Generation of monoclonal antibodies against autologous proteins in gene-inactivated mice. J Biol Chem. 1995;270:8397-8400.
Lupu F, Heim DA, Bachmann F, Hurni M, Kakkar VV, Kruithof EKO. Plasminogen activator expression in human atherosclerotic lesions. Arterioscler Thromb Vasc Biol. 1995;15:1444-1455.
Sato, Rifkin D. Autocrine activities of basic fibroblast growth factor: regulation of endothelial cell movement, plasminogen activator synthesis, and DNA synthesis. J Cell Biol. 1988;107:1199-1205.
Ploplis V, French E, Carmeliet P, Collen D, Plow E. The plasminogen system and cell migration during an inflammatory response. Fibrinolysis. 1996;10(suppl 3):55. Abstract.
Dechend R, Mo S, Tamaschcke C, Schulz W, Gross M, Dietz R, Gulba D. Thrombin receptor and urokinase-type plasminogen activator are colocalized in vascular smooth muscle cells derived from human carotid atherosclerotic plaques. Fibrinolysis. 1996;10(suppl 3):254a. Abstract.
Padro T, Steins M, Li CX, Kienast J. Urokinase plasminogen activator (u-PA) in human atherosclerotic coronary arteries. Fibrinolysis. 1996;10(suppl 3):10a. Abstract.
Dano K, Behrendt N, Brünner N, Ellis V, Ploug M, Pyke C. The urokinase receptor: protein structure and role in plasminogen activation and cancer invasion. Fibrinolysis. 1994;8(suppl 1):189-203.