Increased circulating and tissue levels of plasminogen activator inhibitor 1 (PAI-1) are often present in severe inflammatory states associated with neutrophil activation and accumulation and correlate with poor clinical outcome from many of these conditions. The mechanisms by which PAI-1 contributes to inflammation have not been fully delineated. In the present experiments, we found that addition of PAI-1 to neutrophil cultures diminished the rate of spontaneous and TNF-related apoptosis-inducing ligand-induced apoptotic cell death. The effects of PAI-1 on cell viability were associated with activation of antiapoptotic signaling pathways, including upregulation of PKB/Akt, Mcl-1, and Bcl-xL. Although urokinase-plasminogen activator receptor, lipoprotein receptor-related protein, and vitronectin are primary ligands for PAI-1, these molecules were not involved in mediating its antiapoptotic properties. In contrast, blocking pertussis toxin-sensitive G protein-coupled receptors and selective inhibition of phosphatidylinositide 3-kinase reversed the ability of PAI-1 to extend neutrophil viability. The antiapoptotic effects of PAI-1 were also evident under in vivo conditions during LPS-induced acute lung injury, where enhanced apoptosis was present among neutrophils accumulating in the lungs of PAI-1−/− compared with PAI-1+/+ mice. These results demonstrate a novel antiapoptotic role for PAI-1 that may contribute to its participation in neutrophil-associated inflammatory responses.
- acute lung injury
although neutrophils play a beneficial role in the eradication of invading microorganisms, exaggerated proinflammatory activities of these cells can contribute to prolonged inflammation, tissue damage, and organ injury in inflammatory processes such as acute lung injury (ALI) or sepsis (1, 10). In such neutrophil-driven inflammatory conditions, apoptotic cell death and subsequent clearance to prevent the release of harmful proinflammatory mediators from dying cells are essential mechanisms for resolution of inflammation (19, 32). The prolonged presence of activated neutrophils in the lungs and other organs involved in inflammatory processes, including during ALI and myocardial ischemia, contributes to greater severity of organ injury in these settings (17, 20).
Apoptosis in neutrophils occurs through pathways similar to those observed in other cell populations and involves caspase activation, cell membrane blebbing, loss of mitochondrial membrane potential (Δψm), and externalization of phosphatidylserine (16). Previous studies have found that the lifespan of neutrophils is prolonged after exposure to LPS or proinflammatory cytokines, such as IL-1β, whereas cyclin-dependent kinase inhibitors or suppression of prosurvival signaling pathways, particularly those involving Akt or cAMP/PKA, increases neutrophil apoptosis (30, 42).
PAI-1 is a multifunctional protein that belongs to the SERPIN superfamily of serine protease inhibitors (25). Plasminogen activator inhibitor 1 (PAI-1) is an effective inhibitor of plasminogen activators, particularly urokinase-plasminogen activator (uPA) and tissue-type plasminogen activator, and binds to the LDL receptor (LRP) and vitronectin (25, 33). Although increased levels of PAI-1 and inhibition of plasmin generation are found in the lungs and circulation of critically ill patients with sepsis or ALI (28, 40), recent studies have also shown that PAI-1 can affect a broad range of cellular functions independent of its effects on coagulation, including cell adhesion, proliferation, migration, and viability (6, 21, 36). In particular, PAI-1 can facilitate neutrophil chemotaxis, enhance release of cytokines by Toll-like receptor 4-stimulated neutrophils, and diminish the uptake of apoptotic neutrophils by macrophages (23, 26).
Several studies have demonstrated that PAI-1 participates in regulating cell viability, with pro- and antiapoptotic effects being reported (6, 11). However, there is no information concerning a potential role for PAI-1 in modulating neutrophil viability and apoptosis. In the present experiments, we found a novel function for PAI-1 in diminishing apoptosis of neutrophils under in vitro and in vivo conditions.
MATERIALS AND METHODS
PAI-1 knockout mice backcrossed with C57BL/6 mice for at least nine generations and age- and sex-matched control C57BL/6 mice were obtained from Jackson Laboratory (Bar Harbor, ME). All experiments were conducted in accordance with protocols approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee.
Reagents and antibodies.
Antibodies to phosphorylated and total Akt, Erk1/2, cAMP response element-binding protein (CREB), Bcl-xL, and Mcl-1 were purchased from Cell Signaling (Beverly, MA); antibody to actin and polymyxin B from Sigma Chemical (St. Louis, MO); and propidium iodide and antibodies to annexin V from EMD Chemicals (Gibbstown, NJ). Recombinant mouse PAI-1 and mutant PAI-1 (R76E) or PAI-1 (R101A/Q123K) proteins were gifts from Dr. Victoria Ploplis (Notre Dame, IN). Mouse vitronectin was purchased from Abcam (Cambridge, MA); recombinant TRAIL/TNFSF10 from R & D Systems (Minneapolis, MN); Hoechst 33342 from Invitrogen (Carlsbad, CA); and CD11b-phycoerythrin (PE) antibody from BD Pharmingen (San Diego, CA).
Neutrophil isolation and culture.
Bone marrow neutrophils were isolated as previously described (50). Neutrophil purity was consistently >97%, as determined by Wright-Giemsa-stained cytospin preparations. Neutrophils were cultured in RPMI 1640 medium containing 5% FBS and treated as indicated.
Measurement of neutrophil apoptosis.
The percentage of early and late apoptotic cells was determined by staining the cells with annexin V-FITC and propidium iodide and studying the cells by flow cytometry (13). Caspase-3 and -9 activities were measured using standard fluorogenic substrates (Calbiochem, La Jolla, CA), in conjunction with epifluorescent microscopy. Briefly, neutrophils (1.5 × 106/well) were treated as indicated, transferred to 1.5-ml centrifuge tubes, and washed twice with serum-free RPMI 1640 medium. The neutrophils were incubated with cell-permeable caspase-3 or -9 fluorogenic substrate for 60 min at 37°C, and Hoechst 33342 (1 μg/ml) was added for an additional 15 min to stain the nuclei. The cells were washed with RPMI 1640 medium and transferred to a four-well chambered coverslip (Nalge, Naperville, IL), and images were acquired using a Leica DM IRBE inverted epifluorescence/Nomarski microscope outfitted with laser confocal optics (TCS NT, Leica, Exton, PA). The level of the fluorescence product (for caspase-3 or -9) and number of nuclei (2–3 × 105 cells/field) were measured in randomly acquired fields using SimplePCI software (Compix, Cranberry Township, PA). Data were calculated as the ratio of green fluorescence to number of nuclei. Final images were processed and stored using IPLab Spectrum and Adobe Photoshop software (Adobe Systems, San Jose, CA). Δψm was determined using the cell-permeable probe JC-1 and confocal microscopy, as previously described (34, 49).
Western blot analysis.
Cells were prepared in lysis buffer [50 mM Tris, pH 7.4, 150 mM NaCl, 0.5% (vol/vol) NP-40, 1 mM EDTA, 1 mM EGTA, 1 mM Na3VO4, 50 mM NaF, and protease inhibitors] and then sonicated and centrifuged at 10,000 g for 15 min at 4°C. The protein concentration in supernatants was determined using Bradford reagent (Bio-Rad, Hercules, CA) with BSA as a standard (46, 48). Samples were mixed with Laemmli sample buffer and boiled for 5 min. Equal amounts of protein were resolved by 8% SDS-PAGE and transferred onto polyvinylidene difluoride membranes (Immobilon P, Millipore, Billerica, MA). The membranes were probed with specific antibodies to detect total and phosphorylated Akt, Erk1/2, and CREB, as well as Mcl-1 or Bcl-xL. Bands were visualized by enhanced chemiluminescence (SuperSignal, Pierce Biotechnology, Rockford, IL). Each experiment was carried out two or more times using neutrophils obtained from separate groups of mice.
ALI was induced by intratracheal administration of 1 mg/kg LPS in 50 μl of PBS, as previously described (15, 44, 47, 48). Briefly, mice were anesthetized with isoflurane and then suspended by their upper incisors on a 60° incline board. The tongue was gently extended, and LPS solution was deposited into the pharynx (9, 48). Lungs were harvested 24 h after LPS administration.
Measurement of neutrophil apoptosis in vivo.
Infiltrating leukocytes were isolated from minced lung tissue by treatment with collagenase B (2 mg/ml; Roche) and DNase I (0.02 mg/ml; Sigma Chemical) in Iscove's modified Dulbecco's medium supplemented with 1 mM sodium pyruvate, 2 mM l-glutamine, 10 μg/ml penicillin-streptomycin, 25 μM 2-mercaptoethanol, and 0.1 mM nonessential amino acids (Life Technologies) at 37°C for 30 min. Then an equal volume of Iscove's modified Dulbecco's medium containing 20% FBS was added. Cell suspensions were passed through a 40-μm cell strainer and washed with PBS. To determine the percentage of apoptotic neutrophils, the cells were incubated with annexin V-FITC and CD11b-PE antibodies and then analyzed by flow cytometry. In selected experiments, cells isolated from lung digests were incubated with caspase-9 fluorogenic substrate for 60 min and with CD11b-PE antibody for an additional 30 min at 37°C and then subjected to flow cytometry.
Student's t-test was used for comparisons between two groups. Tukey's test was used for comparisons between more than two groups. P < 0.05 was considered to be statistically significant.
Exposure to PAI-1 diminishes spontaneous apoptosis in neutrophils.
In initial experiments, the percentage of viable and apoptotic cells was determined among neutrophils cultured with or without recombinant PAI-1 for 24, 48, or 72 h. Although neutrophil apoptosis increased upon culture of cells for 24 h, the difference between control and PAI-1-treated cells was not significant. However, as shown in Fig. 1, neutrophils cocultured with PAI-1 for 48 or 72 h demonstrated significantly decreased apoptosis compared with cells cultured in the absence of PAI-1. One application of PAI-1 was sufficient to increase neutrophil viability (Fig. 2B). The activity of caspase-3 and -9 was decreased in neutrophils exposed to PAI-1 (Fig. 3, A and C). Exposure to PAI-1 preserved Δψm, which normally declines in untreated cells during apoptosis (Fig. 3, B and C). To confirm that the antiapoptotic effects of PAI-1 were not due to contamination with LPS, we added polymyxin B to the cultures and found that whereas polymyxin B prevented LPS-mediated increases in neutrophil viability, addition of polymyxin B to cultures containing PAI-1 did not affect the antiapoptotic actions of PAI-1 (Fig. 2A). The concentrations of PAI-1, i.e., 100–1,000 ng/ml, that resulted in diminished apoptosis of neutrophils are physiologically relevant, with similar PAI-1 levels having been reported in the lungs and in the circulation of critically ill patients with sepsis or ALI (28, 40).
To determine the temporal relationship between exposure to PAI-1 and inhibition of apoptosis, neutrophils were cultured with PAI-1 for 0, 4, 8, or 24 h, washed, and cultured without PAI-1 for a total of 48 h. In parallel experiments, neutrophils were cultured without PAI-1 for 4, 8 or 24 h, and PAI-1 was added to the medium before the cells were collected after a total culture period of 48 h. As shown in Fig. 4, A and B, addition of PAI-1 to the neutrophil cultures after 8 h or withdrawal of PAI-1 after 4 h of culture resulted in lack of any effect on neutrophil apoptosis. These results indicate that exposure of neutrophils to PAI-1 4–8 h after initiation of culture was essential for its antiapoptotic effects.
Exposure of neutrophils to PAI-1 induced activation of antiapoptotic signaling pathways. In particular, as shown in Fig. 4C, culture of neutrophils with PAI-1 for 4, 8, or 16 h resulted in enhanced activation of Akt, phosphorylation of CREB, and increased intracellular levels of the antiapoptotic proteins Mcl-1 and Bcl-xL. These results are consistent with previous studies in which Mcl-1 was shown to be a major antiapoptotic factor in neutrophils (2, 39).
PAI-1-dependent activation of phosphatidylinositol 3-kinase is essential for diminished apoptosis in neutrophils.
Although enhanced phosphorylation of Akt and CREB and increased levels of Mcl-1 were found in PAI-1-treated neutrophils, it is not clear whether activation of phosphatidylinositol 3-kinase (PI3-K) or PKA signaling pathways was a predominant mechanism involved in the antiapoptotic effects of PAI-1. To examine this issue, neutrophils were cultured with or without the PKA inhibitor H-89 or the PI3-K inhibitor wortmannin. As shown in Fig. 5A, whereas inhibition of PKA or PI3-K resulted in enhanced neutrophil apoptosis, coculture with PAI-1 increased the viability of neutrophils treated with H-89, but not with wortmannin. These results suggest that activation of PI3-K, but not PKA, is required for the inhibition of neutrophil apoptosis by PAI-1. In addition, treatment of neutrophils with PAI-1 and H-89 diminished phosphorylation of CREB and the levels of Mcl-1 but had no effect on activation of Akt (Fig. 5B). In contrast, exposure of neutrophils to wortmannin and PAI-1 blocked phosphorylation of Akt but did not completely prevent the increases in Mcl-1 or phosphorylated CREB in PAI-1-treated neutrophils. These experiments demonstrate that activation of PI3-K is a major contributor to the antiapoptotic effects of PAI-1.
Antiapoptotic effects of PAI-1 are dependent on G protein-coupled receptor but are independent of uPAR, LRP, and vitronectin.
Previous studies showed that PAI-1 can bind to the uPA (uPAR) and LDL (LRP) receptors and also forms complexes with vitronectin, resulting in modulation of signaling pathways involving αVβ3- and other integrins (22, 35). Given the ability of uPAR, LRP, and vitronectin to influence cellular viability in many cell populations (4, 8), we examined their role in mediating the delay in neutrophil apoptosis induced by PAI-1.
In initial experiments, recombinant [wild-type (WT)] PAI-1 or the PAI-1 mutants R76E and PAI-R101A/Q123K, which do not bind to LRP or vitronectin, respectively, were used (37, 43). Apoptosis was inhibited to a similar degree by culture of neutrophils with R76E or R101A/Q123K and with WT PAI-1 (Fig. 6A). Exposure to PAI-1 decreased apoptosis to approximately the same extent in neutrophils from vitronectin-deficient (vtn−/−) mice and neutrophils from control vtn+/+ mice (Fig. 6B). Preincubation of PAI-1 with vitronectin resulted in loss of the ability of PAI-1 to inhibit neutrophil apoptosis (Fig. 6B). These results suggest that whereas preexposure to vitronectin can sequester PAI-1 and, therefore, prevent the antiapoptotic effects of PAI-1, binding to vitronectin on the cell surface is not required for the antiapoptotic effects of PAI-1.
We next determined if the antiapoptotic effects of PAI-1 were mediated by interaction with uPAR. As shown in Fig. 6C, PAI-1 diminished spontaneous apoptosis in uPAR-deficient (upar−/−) neutrophils, indicating that uPAR was not required for the inhibitory effects of PAI-1 on apoptosis.
Consistent with previous studies (12, 38), addition of anti-CD11b antibody to neutrophils resulted in enhanced apoptosis (Fig. 6D). However, PAI-1 was still able to inhibit neutrophil apoptosis, even after incubation of the cells with antibodies to CD11b (Fig. 6D). These results indicate that the antiapoptotic effects of PAI-1 are independent of the Mac-1 signaling pathway.
While the above-described experiments showed that PAI-1-mediated inhibition of neutrophil apoptosis is independent of interactions with uPAR, LRP, Mac-1, and vitronectin, inhibition of G protein-coupled receptors (GPCRs) with pertussis toxin diminished the inhibitory effect of PAI-1 on apoptosis (Fig. 6E). The concentrations of pertussis toxin used in these experiments did not adversely affect the viability of neutrophils in the absence of PAI-1. These results demonstrate that specific signaling pathways involving the GPCR are required for the antiapoptotic effects of PAI-1.
Exposure of neutrophils to TNF-related apoptosis-inducing ligand (TRAIL) resulted in enhanced apoptosis (Fig. 6F), consistent with previous studies showing the importance of extrinsic induced apoptotic pathways in neutrophil death (27, 31, 41). Inclusion of PAI-1 in the neutrophil cultures diminished the proapoptotic effects of TRAIL (Fig. 6F). These results demonstrate that PAI-1 diminishes spontaneous and extrinsically mediated neutrophil apoptosis.
PAI-1 deficiency is associated with increased neutrophil apoptosis under in vivo conditions during acute inflammatory lung injury.
Although our results with cultured neutrophils demonstrate that PAI-1 has potent antiapoptotic properties, the in vivo relevance of this finding was unclear. To examine this question, LPS-induced ALI was generated in control WT (PAI-1+/+) and PAI-1 deficient (PAI-1−/−) mice. As shown in Fig. 7, A and B, apoptosis was significantly increased (i.e., as defined by the cells being annexin V-positive) in neutrophils isolated from the lungs of LPS-exposed PAI-1−/− compared with PAI-1+/+ mice. Consistent with these results, caspase-9 activation was significantly increased among neutrophils isolated from the lungs of PAI-1−/− compared with PAI-1+/+ mice (Fig. 7C). The number of neutrophils in bronchoalveolar lavage fluid from PAI-1−/− mice was also decreased compared with bronchoalveolar lavage fluid from PAI-1+/+ mice (Fig. 7D), consistent with enhanced apoptotic death and clearance of pulmonary neutrophils during LPS-induced ALI in PAI-1−/− mice.
Recent studies have reported that PAI-1 can induce pro- and antiapoptotic responses in nonmyeloid cell populations, including fibroblasts and endothelial cells (3, 5, 29). We have extended those observations and shown that exposure of neutrophils to PAI-1 diminished their progression to apoptotic cell death. Incubation of neutrophils with PAI-1 was accompanied by activation of pathways involving antiapoptotic mediators, including Akt, Mcl-1 and Bcl-xL, inhibition of caspases-3 and -9, and preservation of Δψm. Although previous studies have reported that PAI-1 can bind to and inhibit caspase-3 (11), this does not appear to be the mechanism for the antiapoptotic effects of PAI-1, since caspase-9, which is upstream to caspase-3, was also inhibited in PAI-1-treated neutrophils. Whereas signaling events that involve activation of PKA and CREB, as well as CREB-dependent enhanced expression of Mcl-1, can increase neutrophil viability (24), we did not find a contributory role for these events in the PAI-1-induced reduction of neutrophil apoptosis. Rather, the present experiments identified PI3-K activation as a central event in PAI-1-associated preservation of neutrophil viability. These results are consistent with recent findings of decreased viability of neutrophils isolated from PI3-Kγ-deficient mice and increased viability of neutrophils from phosphatase and tensin homolog (PTEN)-deficient mice (45).
Although interactions between PAI-1 and uPAR/uPA, LRP, and vitronectin are well described, none of these appear to be responsible for the antiapoptotic effects of PAI-1 among neutrophils. In particular, PAI-1 maintained its antiapoptotic effects and significantly increased the viability of neutrophils isolated from uPAR- or vitronectin-deficient mice. PAI-1 mutants unable to interact with vitronectin or LRP still were able to decrease neutrophil apoptosis. In contrast, treatment of neutrophils with pertussis toxin, which inhibits GPCR signaling, eliminated the antiapoptotic effects of PAI-1. The involvement of interactions between PAI-1-, PI3-K-, and GPCR-related signaling in modulating neutrophil viability is consistent with a recent study (42) showing that GPCR-mediated regulation of Akt activation plays an important role in regulating spontaneous death in neutrophils.
Clearance of activated neutrophils through activation of cellular programs involving constitutive cell death plays a central role in the resolution of inflammation. Our results, showing that PAI-1 has potent antiapoptotic effects in neutrophils, demonstrate a novel proinflammatory property for PAI-1. This antiapoptotic effect of PAI-1 is likely to be particularly important in clinical settings, such as ALI and sepsis, in which PAI-1 levels are increased and neutrophil accumulation in the lungs and other anatomic sites contributes to organ dysfunction and death. In particular, by prolonging the viability of neutrophils activated to produce proinflammatory mediators, PAI-1 may contribute to exacerbation of inflammatory processes in which neutrophils play a role.
The enhanced production of PAI-1 during infectious and noninfectious inflammatory conditions can have beneficial and deleterious consequences for the host. In particular, upregulation and secretion of PAI-1 in endothelial and epithelial cells, by enhancing neutrophil chemotaxis (23), can facilitate trafficking of neutrophils into infectious foci and subsequent clearance of bacteria. In contrast, extended exposure of neutrophils to PAI-1 may inhibit resolution of inflammation through prolongation of the presence in the lungs and other organs of neutrophils activated to produce tissue-damaging proinflammatory mediators such as reactive oxygen species and cytokines. Additionally, PAI-1 also functions as a potent “don't eat me” signal, reducing uptake of neutrophils by macrophages. The inhibitory effect of PAI-1 on efferocytosis, coupled with the ability of PAI-1 to diminish constitutive cell death in neutrophils, is likely to potentiate the magnitude and duration of neutrophil-driven inflammatory responses. Indeed, previous studies have shown more severe pulmonary injury after LPS exposure or hyperoxia in PAI-1+/+ than PAI-1−/− mice (7), a finding consistent with our results demonstrating increased numbers of apoptotic neutrophils in the lungs of PAI-1-deficient mice after pulmonary exposure to LPS. Previous studies have shown increased severity of bleomycin-induced lung injury and fibrosis in transgenic mice overexpressing PAI-1 (14). As the early stages of lung injury due to bleomycin are characterized by neutrophil infiltration (18), it is possible that reduced neutrophil apoptosis and clearance contribute to the enhanced susceptibility of PAI-1-overexpressing mice to bleomycin-associated lung injury. Thus pharmacological inhibition of PAI-1 may be beneficial in accelerating the resolution of inflammatory responses characterized by a prominent neutrophilic component.
This work was supported in part by National Institutes of Health Grants HL-76206 and GM-87748 to E. Abraham. The Flow Cytometry Core, Arthritis and Musculoskeletal Center at the University of Alabama at Birmingham is supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant P30 AR-48311.
No conflicts of interest, financial or otherwise, are declared by the authors.
We thank Celeste P. Bell and Enid Keyser for technical support and the Flow Cytometry Core, Arthritis and Musculoskeletal Center at the University of Alabama at Birmingham.
- Copyright © 2011 the American Physiological Society