Platelet Protease Nexin-1, a Serpin That Strongly Influences Fibrinolysis and ThrombolysisClinical Perspective
Background—Protease nexin-1 (PN-1) is a serpin that inhibits plasminogen activators, plasmin, and thrombin. PN-1 is barely detectable in plasma, but we have shown recently that PN-1 is present within the α-granules of platelets.
Methods and Results—In this study, the role of platelet PN-1 in fibrinolysis was investigated with the use of human platelets incubated with a blocking antibody and platelets from PN-1–deficient mice. We showed by using fibrin-agar zymography and fibrin matrix that platelet PN-1 inhibited both the generation of plasmin by fibrin-bound tissue plasminogen activator and the activity of fibrin-bound plasmin itself. Rotational thromboelastometry and laser scanning confocal microscopy were used to demonstrate that PN-1 blockade or deficiency resulted in increased clot lysis and in an acceleration of the lysis front. Protease nexin-1 is thus a major determinant of the lysis resistance of platelet-rich clots. Moreover, in an original murine model in which thrombolysis induced by tissue plasminogen activator can be measured directly in situ, we observed that vascular recanalization was significantly increased in PN-1–deficient mice. Surprisingly, general physical health, after tissue plasminogen activator–induced thrombolysis, was much better in PN-1–deficient than in wild-type mice.
Conclusions—Our results reveal that platelet PN-1 can be considered as a new important regulator of thrombolysis in vivo. Inhibition of PN-1 is thus predicted to promote endogenous and exogenous tissue plasminogen activator–mediated fibrinolysis and may enhance the therapeutic efficacy of thrombolytic agents.
Vascular injury and subsequent thrombus formation are key events in the pathogenesis of atherothrombosis and venous thromboembolism. The serine proteases, urokinase plasminogen activator and tissue plasminogen activator (tPA), generate plasmin that drives fibrinolysis. The thrombolytic actions of these proteases are critical for clot dissolution. Their properties have numerous therapeutic applications, including fibrinolysis for ST-segment elevation myocardial infarction (STEMI). Direct recanalization of an occluded vessel by primary angioplasty became the preferred reperfusion strategy in STEMI patients. Thrombolysis remains, however, an option of reperfusion therapy in early STEMI presenters. Despite early administration of recombinant tPA (rtPA) in STEMI presenters, fibrinolysis fails to achieve myocardial reperfusion in 1 of 2 patients and is associated with poor clinical outcome.1 This phenomenon is of considerable clinical importance in the setting of acute myocardial infarction because early restoration of normal blood flow is strongly associated with improved survival. A few factors have been identified as being involved in this interindividual heterogeneity, such as age, delay between symptom onset and fibrinolytic therapy, smoking habit, infarct size, and site.2
Clinical Perspective on p 1334
Plasminogen activator inhibitor type 1 (PAI-1) is a serine protease inhibitor that is present in plasma and in platelet α-granules. An increased plasma concentration of PAI-1 has been associated with recurrent myocardial infarction.3,4 In humans, platelet PAI-1 is assumed to be a major contributor to the stabilization of the thrombus by inhibiting endogenous fibrinolysis.5,6 However, platelets have also been shown to inhibit fibrinolysis by PAI-1–independent mechanisms,7 and the individual role of other serpins in the thrombolytic process has not yet been defined. Protease nexin-1 (PN-1), also known as SERPINE2, deserves special attention because it has been shown in vitro to significantly inhibit urokinase plasminogen activator, tPA, and plasmin. PN-1 is barely detectable in plasma 8 but is produced by various cell types,9 and is stored in the α-granules of platelets.10 Because of its action on proteases of the plasminergic system, we hypothesized that platelet PN-1 may play a prominent role in the process of thrombolysis resistance.
The present article evaluates, by in vitro and ex vivo studies, the role of platelet PN-1 in platelet-rich clot (PRC) lysis. Moreover, we developed a murine model of thrombolysis and applied it to wild-type (WT) and PN-1–deficient mice to test the hypothesis that PN-1 inhibits thrombolysis initiated by recombinant tPA. Thus, PN-1 may be a potential target to improve the therapeutic applications of thrombolytic agents.
PN-1–deficient (PN-1−/−) mice come from the laboratory of D. Monard (FMI, Basel, Switzerland) and were back-crossed for 12 generations into the C57BL/6 line.11 Experimental animals were 8 to 16 weeks of age. Heterozygous mating generated PN-1−/− and WT mice. Mice were bred and maintained in our own laboratory (Paris, France). All animals were genotyped by polymerase chain reaction. All experiments were performed in accordance with European legislation on the protection of animals.
Preparation of Washed Platelets
Human blood from healthy adult volunteers was collected into 1/10 vol ACD-A (38 mmol/L citric acid, 60 mmol/L sodium citrate, 136 mmol/L glucose). Washed platelets were isolated as described previously.12
Blood was collected from anesthetized mice by cardiac puncture into syringes containing 1/10 vol ACD-C (130 mmol/L citric acid, 124 mmol/L sodium citrate, 110 mmol/L glucose). Washed platelets were isolated as described previously.10
Binding of tPA and Plasmin to Fibrin Matrices and Measurement of Plasmin Generation or Activity
Fibrin matrices in 96-well plates were prepared as described previously.13 The functionality of this fibrin surface was determined by measuring the activation of plasminogen by fibrin-bound tPA or the activity of fibrin-bound plasmin itself (see the online-only Data Supplement).
Sodium Dodecyl Sulfate–Polyacrylamide Gel Electrophoresis and Zymography
Platelets (5×108/mL in reaction buffer) were activated by PAR1-activating peptide (PAR1-AP; SFLLRN, NeoMPS) (50 μmol/L) for human platelets or by PAR4-activating peptide (PAR4-AP; AYPGKF, NeoMPS) (250 μmol/L) for mouse platelets for 30 minutes at 37°C. Control samples were obtained by incubating platelets for the same time with buffer. At the end of the incubation, samples were centrifuged, and the supernatants (secreted fraction) were removed for analysis. The secreted fractions were incubated with rtPA (10 IU/mL) or plasmin (0.25 μmol/L) for 30 minutes at 37°C in the presence or absence of the blocking anti-PN-1 (a generous gift from Dr D. Hantai, INSERM U582, Paris, France) or anti-PAI-1 immunoglobulin G (IgG) (MA-33B8–307, Molecular Innovations). Proteins were first separated on a 10% sodium dodecyl sulfate–polyacrylamide gel. After incubation with 2% Triton X-100, the gel was then overlaid on a fibrin-plasminogen (200 nmol/L)-agar gel for tPA activity measurement or on a fibrin-agar gel for plasmin activity, as described previously.14 Zymograms were allowed to develop at 37°C during 24 hours and photographed at regular intervals with dark-ground illumination. Zymograms were stained with Coomassie blue.15
Clot Formation and Fibrinolysis Ex Vivo
Human platelet-rich plasma (PRP) was obtained from citrated blood by centrifugation at 120 g for 15 minutes. PRP was adjusted at 108 platelets per milliliter in platelet-free plasma and supplemented with 75 μg/mL fluorescein isothiocyanate–fibrinogen. For mouse PRCs, citrated human platelet-free plasma was mixed with murine-washed platelets to a concentration of 8×108/mL. Samples were incubated with irrelevant IgG or the blocking anti-PN-1 IgG and/or anti-PAI-1 IgG, both at 100 μg/mL, and recalcified with 10 mmol/L CaCl2 in glass tubes. After retraction, clots were removed, blotted, and weighed. To assess fibrinolysis, clots were incubated in Hanks' buffer (Sigma) for 24 hours at 37°C. The supernatant was removed, and the fluorescence released from the clot was measured in a spectrofluorometer.16 The remaining clots were blotted and reweighed to calculate the loss of clot weight and then were totally dissolved to calculate the fluorescence remaining in the clot.
Fibrinolysis Experiments: Microscopic Lysis Velocity by Laser Scanning Confocal Microscopy
Citrated human or mice PRP was adjusted at 108 platelets per milliliter and supplemented with Alexa 488-fibrinogen (Invitrogen). Human PRP was incubated with control IgG (Jackson ImmunoResearch) or blocking anti-PN-1 IgG (100 μg/mL), and PRCs were obtained by adding tissue factor (Innovin 1/5 vol/vol, Diagnostica Stago) and 10 mmol/L CaCl2 in microchambers as described previously.17 PRP from WT or PN-1−/− mice was clotted in the same conditions. Clots were scanned with a Leica confocal laser scanning microscope linked to a Leica inverted microscope equipped with a ×63 water immersion objective. Scans were collected in a format of 512×512 pixels with 1024 gradations of intensity. rtPA (26 nmol/L) (Alteplase, Boerhinger) was loaded at the edge of the labeled PRC. The edge of the clot was visualized with the confocal microscope set up in the reflection mode. Scanning was performed at a magnification of 125×125 μm every 15 seconds for 30 minutes. The velocity of the lysis front was determined from confocal microscope images and analyzed with image J software.
ROTEM Modified Rotational Thromboelastography Analyzer
Citrated PRP was obtained and adjusted at 108 platelets per milliliter as described above. ROTEM analysis was performed in a prewarmed ROTEM cup containing 300 μL of PRP in the presence of control IgG or the blocking anti-PN-1 and/or anti-PAI-1 IgG, both at 100 μg/mL. Clotting was initiated by the addition of tissue factor (Innovin 1/5 vol/vol) and CaCl2 (10 mmol/L). Fibrinolysis was initiated by the addition of human rtPA (0.5 nmol/L) (Alteplase, Boerhinger) or mouse rtPA (30 nmol/L) (Molecular Innovation). The fibrinolytic response by rtPA was assessed with the use of ROTEM software, thereby providing the lysis rate at 60 minutes in each condition.
Dorsal Skinfold Chamber
Dorsal skinfold chambers were implanted in 10- to 12-week-old mice (25 to 30 g body weight) anesthetized by intraperitoneal injection of 100 mg/kg ketamine and 10 mg/kg xylazine in saline solution as described previously.18 Briefly, a patch of dorsal hair was removed, and 2 titanium frames were positioned to sandwich the extended double layer of skin. One layer of Betadine-cleaned skin was completely removed in a circular area of 13 mm in diameter, and the remaining layer, consisting of epidermis, subcutaneous tissue, and striated skin muscle, was covered with a 12-mm glass coverslip incorporated in the frame. Mice were injected subcutaneously with buprenorphine (0.05 mg/kg) after surgery and then again 8 to 12 hours later. The animals tolerated the chambers well and showed no sign of discomfort. After a 48-hour period of recovery from surgery, preparations fulfilling the criteria of intact microcirculation and showing no signs of inflammation were utilized for thrombosis and thrombolysis experiments.
Real-Time Intravital Imaging of Thrombus Formation and Thrombolysis
Mice were anesthetized with 100 mg/kg ketamine and 10 mg/kg xylazine, and vascular injury was induced by placing a Whatman filter paper strip (1×0.5 mm) saturated with 15% FeCl3 (Sigma) over venules (ranging from 130 to 160 μm in diameter) in dorsal skinfold chambers for 3 minutes. Thrombus formation after vessel injury was examined in real time by monitoring the accumulation of rhodamine 6G (Sigma) (3 mg/kg mouse)–labeled platelets with the use of an inverted fluorescence microscope (Axio Observer, Carl Zeiss MicroImaging GmbH, Germany) with a ×5 objective connected to a Hamamatsu Orca-R2 charge-coupled device video camera. Platelet deposition and thrombus growth in injured venules were monitored until vessel occlusion, defined as a complete arrest of blood flow for at least 5 minutes. Immediately after vessel occlusion, 20 μL of saline containing rtPA (80 μmol/L) and hirudin (10 μmol/L) (Serbio) was applied topically in the chamber to enhance thrombolysis and prevent rethrombosis. Thrombolysis was analyzed by measuring the occurrence of recanalization of occluded venules, the time to recanalization, and the decrease in thrombus area at 30 minutes and 1 hour after rtPA treatment. A total of 13 venules in 7 PN-1−/− mice and 13 venules in 7 WT mice were studied. Data acquisition and analysis were performed with the use of Axiovision software (Carl Zeiss MicroImaging GmbH, Germany).
Results are shown as mean±SEM. Student t test was used for in vitro experiments with recombinant PN-1, for in vitro experiments of WT and PN-1–deficient mice, and for lysis front velocity experiments. One-way ANOVA followed by the Dunnett test was used when comparisons of anti-PN-1 IgG or anti-PAI-1 IgG groups versus control IgG were performed. A linear mixed-effects model was used for the analysis of in vivo thrombolysis. A P value ≤0.05 was considered significant.
PN-1 Inhibits Plasminogen Activation by Fibrin-Bound tPA
Plasminogen activation by tPA was measured on a fibrin surface in the presence or absence of recombinant PN-1. First, tPA was incubated for 1 hour on fibrin-coated plates, and the excess of unbound tPA was eliminated. PN-1 was subsequently added to the fibrin-coated plates, and the excess was discarded. Plasmin generation induced by the residual fibrin-bound tPA was then determined after addition of plasminogen with the chromogenic substrate CBS0065. The initial rate of plasmin generation by tPA decreased by ≈2-fold in the presence of PN-1 (Figure 1A): 2.7±0.3 nmol/L and 1.3±0.1 nmol/L plasmin were generated in the absence and presence of PN-1, respectively (Figure 1B).
tPA-induced fibrin degradation was measured by fibrin-plasminogen-agar zymography with platelet releasates. Recombinant tPA induces a lysis area reflecting fibrinolytic activity relative to the amount of plasmin converted from plasminogen by tPA. As expected, the fibrin zymography lysis band corresponding to tPA was reduced by recombinant PN-1 (Figure 1C). No reduction in tPA-induced lysis area was observed after tPA incubation with the supernatant of resting human platelets. In contrast, when tPA was incubated with the secretion products of activated human platelets, the fibrin zymography lysis tPA band was barely detected, indicating the secretion of tPA inhibitor(s) by activated platelets (Figure 1C). To determine whether PN-1 contributed to fibrinolysis inhibition, zymography experiments were performed in the presence of a PN-1–blocking antibody. Tissue plasminogen activator activity was restored in the presence of the anti-PN-1 IgG (Figure 1C) but not in the presence of an irrelevant IgG (not shown). To confirm these findings, the same experiments were performed with platelets from PN-1–deficient mice and their littermate controls. Incubation of tPA with the secretion products of activated platelets from WT mice resulted in an almost complete inhibition of lysis (Figure 1D). On the contrary, the products secreted by platelets from PN-1−/− mice did not decrease the tPA-induced lysis area (Figure 1D). Together, these data demonstrate that PN-1 has the remarkable capacity to inhibit the generation of plasmin induced by tPA bound to fibrin.
PN-1 Inhibits Fibrin-Bound Plasmin
Degradation of fibrin by the serine protease plasmin is a step in the fibrinolysis process in which PN-1 can also play an important role. To test this hypothesis, plasmin activity was measured on a fibrin surface, in the presence or absence of recombinant PN-1. The initial rate of substrate hydrolysis induced by fibrin-bound plasmin decreased by ≈10-fold in the presence of PN-1 (Figure 2A and Figure 2B). Fibrin-bound plasmin activity was thus drastically inhibited by PN-1.
Plasmin-induced fibrin degradation was measured with the use of fibrin-agar zymography. Similarly to the results obtained with tPA, we observed that the secretion products of activated platelets inhibited plasmin-induced lysis. This inhibition was completely prevented by the blocking anti-PN-1 antibody (Figure 2C). Fibrin-agar zymography was also performed with platelets from PN-1–deficient mice and their littermate controls. Incubation of plasmin with the secretion products of activated platelets from WT mice resulted in an almost complete inhibition of lysis (Figure 2D). On the contrary, the products secreted by platelets from PN-1−/− mice did not reduce plasmin-induced lysis area (Figure 2D). Our results thus demonstrate that PN-1 secreted by activated platelets is able to inhibit the fibrinolysis induced by fibrin-bound plasmin.
Platelet PN-1 Limits PRC Lysis
To test the functional effect of PN-1 on endogenous clot lysis, human PRP containing fluorescein isothiocyanate–fibrinogen was preincubated with a control IgG or the blocking anti-PN-1 IgG before clotting. Fibrinolysis was then assessed by clot weight loss and fluorescence release from the clot after 24 hours at 37°C. In the presence of a control IgG, clot weight loss was 7±1 mg, whereas preincubation with the anti-PN-1 IgG resulted in a large increase in clot weight loss, reaching 27±2 mg (Figure 3A). A blocking anti-PAI-1 IgG also enhanced clot weight loss by 17±5 mg, although this increase was not statistically significant. The combination of both blocking antibodies resulted in a large increase in clot weight loss, reaching 46±10 mg (Figure 3A). The percentage of fluorescein isothiocyanate released from the clots was also significantly higher in the presence of the anti-PN-1 IgG (37±2%) than in the presence of an irrelevant IgG (26±1%) (Figure 3B). The same experiments were performed with PRP from WT and PN-1–deficient mice. Clot weight loss was greater for fibrinolysis with PN-1−/− clots (55±6 mg) than with WT clots (31±2 mg) (Figure 3C), and the percentage of released fluorescence was higher for PN-1−/− (89±3%) than for WT clots (64±3%) (Figure 3D). Together, these results show that, in the absence of PN-1, endogenous tPA-induced clot lysis is enhanced within 24 hours, indicating that platelet PN-1 is a regulator of endogenous clot lysis.
The effect of PN-1 inhibition or PN-1 deficiency on clot lysis was further investigated with the use of a ROTEM analyzer. An exogenous supplement of a subthreshold lytic concentration of tPA (0.5 nmol/L) was used to induce clot lysis. As shown in Figure 4A and 4B, the percentage of tPA-induced clot lysis was minimal in the presence of a control IgG, reaching 16±2%, whereas it was greatly increased in the presence of the anti-PN-1 IgG, reaching 42±5%. A blocking anti-PAI-1 IgG also has an increased tendency for clot lysis by 28±5%, although it was statistically insignificant. The combination of both blocking antibodies resulted in an almost complete clot lysis at 60 minutes (91±1% of lysis). To substantiate these results, experiments were also performed with mouse platelets (Figure 4C). The ROTEM tracing showed that a subthreshold concentration of tPA induced lysis of WT clots by 56±7%, whereas lysis of PN-1–deficient clots was almost complete (84±8%) under our experimental conditions (Figure 4D). These results indicate that both PN-1 and PAI-1 released by activated platelets contribute to inhibit tPA-induced clot lysis.
Platelet PN-1 Reduces the Velocity of Clot Lysis
We visualized the lysis front of PRC by using laser scanning confocal microscopy (Figure 5). Addition of tPA at the edge of the microchambers of PRC initiated lysis with a straight and sharp front moving across the entire fibrin surface. A significant increase in the lysis front velocity was observed in the presence of the blocking anti-PN-1 IgG with an average rate of 22.5±2.8 μm/min compared with the control IgG rate of 11.8±1.6 μm/min (P<0.01; n=5) (Figure 5A). To confirm these findings, the same experiments were performed with clots from PN-1–deficient mice and their littermate controls. As observed with human clots, addition of tPA resulted in an acceleration of the lysis front in PN-1–deficient clots, with a rate of 16.0±1.5 μm/min versus 10.3±0.9 μm/min with the WT clots (P<0.05; n=5) (Figure 5B).
tPA-Induced Thrombolysis Is Enhanced in PN-1−/− Mice
To determine whether the antifibrinolytic effect of PN-1 is of in vivo relevance, we have developed a method by which thrombolysis can be measured by intravital microscopy using the dorsal skinfold chamber model in mice. We compared the efficiency of tPA-induced thrombolysis in WT and PN-1−/− mice (Figure 6A). Topical application of FeCl3 over venules ranging from 130 to 160 μm in diameter was used to induce vascular injury leading to occlusive thrombosis. Although there was no significant difference in the occlusive thrombus size/area between WT and PN-1−/− mice (34 163±5459 μm2 versus 31 656±4709 μm2; n=13 vessels from 7 mice per group), the time to reach complete occlusion was significantly reduced in PN-1−/− mice compared with WT mice, in agreement with data obtained in the mesenteric vessel thrombosis model.10 During the 24 hours after complete arrest of blood flow, spontaneous recanalization was observed in only 2 of 13 vessels of 7 PN-1−/− mice and in none of the 13 occluded vessels from WT mice. This indicates that spontaneous thrombolysis after FeCl3 injury is a slow process in both WT and PN-1−/− mice. To accelerate thrombolysis, tPA was directly added to the chamber 5 minutes after complete vessel occlusion. Hirudin was simultaneously added to prevent rethrombosis. In WT mice, the mean time to recanalization after tPA treatment was >1 hour, whereas it was 13±2 minutes in PN-1−/− mice (n=7 mice) (Figure 6A). Furthermore, 1 hour after tPA treatment, the incidence of recanalization was 15% (2 of 13 vessels) in WT mice and reached 92% (12 of 13 vessels) in PN-1−/− mice (Figure 6B). Thirty minutes after tPA treatment, thrombus size remained unchanged in WT mice (101.6±7.2% of initial size), whereas it was significantly reduced in PN-1−/− mice (56.1±8.5% of initial size). At 1 hour after tPA treatment, the thrombus size was reduced in WT mice, but this reduction was less important than in PN-1−/− mice (76.7±6.3% versus 42.8±9.5% of initial size) (Figure 6C). Considered together, these results confirm that PN-1 is a potent inhibitor of tPA-induced thrombolysis in vivo.
After the thrombolysis experiments, mice were kept under observation for 24 hours and euthanized. Four hours after tPA treatment, all vessels occluded by FeCl3 injury were recanalized in both WT and PN-1−/− mice. Interestingly, all PN-1–deficient mice (7 of 7) remained healthy the day after thrombolytic treatment, whereas 71% (5 of 7) of WT mice were apathetic and showed signs of respiratory distress.
In humans, platelet PAI-1 released locally after platelet activation is assumed to be a major contributor to the stabilization of the thrombus by inhibiting endogenous fibrinolysis.5,6 However, PAI-1–independent mechanisms have also been proposed to contribute to platelet-dependent inhibition of fibrinolysis.7 The existence of other non–PAI-1 proteinase inhibitors able to reduce plasminogen activation and/or plasmin activity has been suggested previously.19 Our study also suggests a less important role for PAI-1 and reveals that an additional serpin plays an important role in inhibiting plasminogen activators and plasmin. Indeed, we show here for the first time that PN-1, which can accumulate at the site of vascular injury because of its presence in platelets,10 is an important player in the control of fibrinolysis. The fact that PN-1 can downregulate both plasmin generation and plasmin activity on the fibrin matrix highlights the potential influence of PN-1 on fibrinolysis. Indeed, the fibrin matrix is largely recognized as an essential actor in the fibrinolysis process. It is well known that tPA-mediated plasminogen activation is dependent on fibrin, which restricts fibrinolysis to the site of thrombus.20 Importantly, when bound to fibrin, tPA is protected from inhibition by PAI-1.21,22 The inhibition of tPA by PAI-1 is decreased by 80% to 90% in the presence of fibrin because PAI-1 has no access to the catalytic domain of fibrin-bound tPA.23 Moreover, the rate of inactivation of plasmin by α2-antiplasmin slows down very significantly when plasmin is bound to fibrin.24 Thus, whereas serine proteases of the fibrin-bound plasminergic system are protected from their principal inhibitors, platelet PN-1 appears to be an inhibitor capable of blocking them in situ. The blocking PAI-1 antibody alone led to a nonsignificant increase in clot lysis, in agreement with previous data demonstrating that PAI-1 deficiency induced only mild hyperfibrinolysis.19 This suggests that PAI-1 alone is not sufficient in regulating the lysis of PRCs. The higher fibrinolytic capacity observed in the presence of both PN-1 and PAI-1 blocking antibodies supports a synergic involvement of both proteins in the regulation of clot lysis. Moreover, platelet PN-1 can influence the lysis of fibrin clots generated spontaneously from PRP, without any exogenous tPA, but also after the addition of rtPA, indicating that PN-1 is inhibitory not only on endogenous but also on exogenous tPA-mediated lysis. These points are of clinical relevance (1) because endogenous fibrinolysis is known to play a pivotal role in the evolution of thrombotic cardiovascular diseases and (2) because this may relate to the failure of optimal reperfusion in approximately one half of STEMI patients who are treated with fibrinolytic agents. A polymorphism in PN-1 could possibly explain the heterogeneity in the therapeutic efficacy of thrombolytic agents. Moreover, the fact that the lysis front moves faster when the PRC is devoid of PN-1 may imply that PRCs are refractory to tPA-induced lysis in a PN-1–dependent manner and that platelet PN-1 may have a critical impact at the level of fibers in the fibrin clot. Further experiments are needed to clarify this potential implication of platelet PN-1 on clot structure.
PN-1 appears to be a particularly important actor both in the development and in the dissolution of a thrombus. Indeed, PN-1 is involved in thrombus generation and extension by its capacity to inhibit thrombin-mediated fibrin formation and platelet activation,10 and we demonstrate here that PN-1 is also involved in thrombolysis by its capacity to inhibit the local generation and activity of plasmin. Because of these opposing effects, it was of great interest to analyze the effect of PN-1 deficiency in the process of thrombus dissolution in vivo. For this purpose, we have developed an original murine model of in vivo thrombolysis associating ferric chloride injury and the dorsal skinfold chamber model. This approach is a reproducible method to quantify thrombus formation and lysis induced by a topical application of tPA. This device has the great advantage of allowing direct visualization, via intravital videomicroscopy, of thrombus formation and also of thrombus lysis in living animals, which demonstrates the originality of our model. We observed that tPA-triggered PRCs are more readily lysed in PN-1–deficient mice than in WT mice, with both the rate and the extent of recanalization being increased in PN-1−/− mice. Our data thus demonstrate the important role of PN-1 in mediating the resistance of PRC to lysis. We also observed that WT mice poorly survived thrombolysis and exhibited a global organ failure syndrome, in contrast to PN-1–deficient mice, which resisted well after the procedure without exhibiting any clinical manifestations.
The fact that platelet PN-1 is so important to protect the developing thrombus from premature lysis may explain the reason why the role of PAI-1 in thrombolysis resistance is a subject of controversy. Indeed, none of the previous investigations studying the role of PAI-1 in thrombolysis failure took into account the contribution of PN-1. We suggest here that endogenous PN-1 can play an important role in the failure of thrombolytic therapy to restore arterial blood flow. Clearly, our findings should be considered in the design of new therapeutic strategies, which should include the inhibition of PN-1 by antibodies or synthetic compounds to improve the therapeutic efficacy of thrombolytic agents.
Sources of Funding
This work was supported by INSERM, Université Paris 7, and Fondation de France (grant No. 2009002497). Dr Boulaftali was supported by the Fondation pour la Recherche Médicale.
We thank Liliane Louedec for technical assistance, Julien Labreuche for the statistical analysis, and Dr Mary Osborne-Pellegrin for editing this article.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.110.000885/DC1.
- Received October 8, 2010.
- Accepted January 21, 2011.
- © 2011 American Heart Association, Inc.
- Armstrong PW,
- Collen D
- Smith A,
- Patterson C,
- Yarnell J,
- Rumley A,
- Ben-Shlomo Y,
- Lowe G
- Levi M,
- Biemond BJ,
- van Zonneveld AJ,
- ten Cate JW,
- Pannekoek H
- Fay WP,
- Eitzman DT,
- Shapiro AD,
- Madison EL,
- Ginsburg D
- Mansuy IM,
- van der Putten H,
- Schmid P,
- Meins M,
- Botteri FM,
- Monard D
- Boulaftali Y,
- Adam F,
- Venisse L,
- Ollivier V,
- Richard B,
- Taieb S,
- Monard D,
- Favier R,
- Alessi MC,
- Bryckaert M,
- Arocas V,
- Jandrot-Perrus M,
- Bouton MC
- Luthi A,
- Van der Putten H,
- Botteri FM,
- Mansuy IM,
- Meins M,
- Frey U,
- Sansig G,
- Portet C,
- Schmutz M,
- Schroder M,
- Nitsch C,
- Laurent JP,
- Monard D
- Jandrot-Perrus M,
- Lagrue AH,
- Okuma M,
- Bon C
- Kufrin D,
- Eslin DE,
- Bdeir K,
- Murciano JC,
- Kuo A,
- Kowalska MA,
- Degen JL,
- Sachais BS,
- Cines DB,
- Poncz M
- Collet JP,
- Montalescot G,
- Lesty C,
- Weisel JW
- Hoylaerts M,
- Rijken DC,
- Lijnen HR,
- Collen D
- Chmielewska J,
- Ranby M,
- Wiman B
- Keijer J,
- Linders M,
- van Zonneveld AJ,
- Ehrlich HJ,
- de Boer JP,
- Pannekoek H
- Kolev K,
- Lerant I,
- Tenekejiev K,
- Machovich R
Mechanical desobstruction of the occluded coronary artery by primary angioplasty has become the gold standard method for myocardial reperfusion in patients with ST-segment elevation myocardial infarction. Indeed, early coronary reocclusion or failure to restore coronary patency after thrombolysis is a major limitation leading to frequent percutaneous coronary intervention. As a consequence, thrombolysis has become the first-line reperfusion therapy in early presenters with ST-segment elevation myocardial infarction (<3 hours) who cannot be transferred quickly to the catheterization laboratory (<2 hours). Plasminogen activator inhibitor type 1 is assumed to play a role in thrombolysis failure. The data reported in this article demonstrate for the first time that another protein present in platelets, protease nexin-1 (PN-1), which is known to significantly inhibit urokinase plasminogen activator, tissue plasminogen activator, and plasmin, possesses efficient antifibrinolytic properties. In the present article, platelet PN-1 is shown to inhibit both the generation of plasmin by fibrin-bound tissue plasminogen activator and the activity of fibrin-bound plasmin itself. The remarkable protective effects of PN-1 toward premature lysis of the developing thrombus may thus represent an unknown and underestimated mechanism of thrombolysis regulation in vivo. PN-1 provides a further dimension in our understanding of the regulation of the plasminergic system and opens an interesting perspective for future clinical investigations examining pharmacological thrombolysis. Protease nexin-1 may be a potential target to improve the therapeutic applications of thrombolytic agents.