© 2003 American Thoracic Society DOI: 10.1165/rcmb.2002-0156OC Reactive Oxygen Species and Extracellular Signal-Regulated Kinase 1/2 Mitogen-Activated Protein Kinase Mediate Hyperoxia-Induced Cell Death in Lung EpitheliumSection of Pulmonary and Critical Care Medicine, Yale University School of Medicine, New Haven, Connecticut; Eastern New Mexico Medical Center, Roswell, New Mexico; Department of Immunobiology, Yale University School of Medicine and Howard Hughes Medical Institute, New Haven, Connecticut; and Division of Pulmonary, Allergy, and Critical Care, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania Address correspondence to: Patty J. Lee, M.D., Section of Pulmonary and Critical Care Medicine, Yale University School of Medicine, 333 Cedar Street, LCI 105, New Haven, CT 06520. E-mail: patty.lee{at}yale.edu
Therapy with high oxygen concentrations (hyperoxia) is often necessary to treat patients with respiratory failure. However, hyperoxia may exacerbate the development of acute lung injury, perhaps by increasing lung epithelial cell death. Therefore, interrupting lung epithelial cell death is an important protective and therapeutic strategy. In the present study, hyperoxia (95% O2) results in murine lung epithelium cell death by DNA-laddering, terminal deoxynucleotidyltransferase dUTP nick end labeling, and Annexin V–fluorescein isothiocyanate flow cytometry assay. We show that hyperoxia increases superoxide production, as assessed by nicotinamide adenine dinucleotide phosphate reduced (NADPH) oxidase activity and flow cytometric assay, and increases phospho–extracellular signal-regulated kinase (ERK)1/2 by Western blot analysis. These processes are inhibited by a reactive oxygen species inhibitor, diphenylene iodonium (DPI), and by an inhibitor of the mitogen-activated protein (MAP) or ERK kinase (MEK)/ERK1/2 pathway, PD98059. ERK1/2 activation in hyperoxia is also inhibited by DPI. Hyperoxia-induced cell death is associated with cytochrome c release, subsequent caspase 9 and 3 activation, and poly (ADP-ribosyl) polymerase cleavage, which can all be suppressed by DPI and PD98059. However, the broad caspase inhibitor z-VAD-FMK protects cells from death without affecting superoxide generation and ERK1/2 activation. Taken together, our data suggest that hyperoxia, by virtue of activating NADPH oxidase, generates reactive oxygen species (ROS), which mediates cell death of lung epithelium via ERK1/2 MAPK activation, and functions upstream of caspase activation in lung epithelial cells.
Abbreviations: 2',7'-dichlorofluroescein, DCF • diphenylene iodonium, DPI • extracellular signal-regulated kinase, ERK • c-Jun NH2-terminal kinase, JNK1/2 • mitogen-activated protein kinase, MAPK • MAP or ERK kinase, MEK • murine lung epithelium, MLE12 • nicotinamide adenine dinucleotide phosphate reduced, NADPH • protein 38, p38 • poly (ADP-ribosyl) polymerase, PARP • phosphate-buffered saline, PBS • reactive oxygen species, ROS • terminal deoxynucleotidyltransferase dUTP nick end labeling, TUNEL
Reactive oxygen species (ROS), the collective term for superoxide anion (O2-.), hydrogen peroxide (H2O2), and hydroxyl radical (OH·), have been shown to induce various biological processes, including cell death (1, 2). Exposure to hyperoxia, or 95% O2, is associated with an accumulation of ROS (3). Therapy with hyperoxia is often necessary to treat newborns, older children, and adults with respiratory failure. However, supplemental O2 administered to patients with respiratory failure can in itself add to the oxidative burden already present in the inflamed lungs. Oxidant stress, such as hyperoxia, can lead to lung epithelial cell death, acute lung injury, and eventually, respiratory failure. Although the mechanisms by which hyperoxia mediates cell death are not well defined, the mitogen–activated protein kinase (MAPK) and caspase pathways have been implicated in cell death that is induced by a variety of oxidant stresses (4, 5). The MAPKs include extracellular signal-regulated kinase (ERK1/2), c-Jun N-terminal protein kinase (JNK1/2), and p38 kinase (6). Each MAPK is activated through dual phosphorylation via a specific phosphorylation cascade. ERK1/2 is generally considered to be a survival mediator involved in the protective action of growth factors against cell death, but it has also been reported that induction of cell death can be mediated via ERK1/2 (7, 8). The other MAPKs (JNK1/2, p38) are usually implicated in the induction of cell death and inflammation after exposure to different agents; however, it also has been shown that p38 and JNK1/2 activation may protect against the induction of cell death (9, 10). Caspases are a family of specific cysteine proteases, and their activation is critical to the intracellular execution of programmed cell death (11). Although it is controversial whether hyperoxia induces a primarily apoptotic or nonapoptotic cell death (12), we chose to investigate the potential role of this important family of death proteases in hyperoxia-mediated death. Among the thirteen members in the caspase family that have been identified, caspase 3 is a major player in the effector phase of cell death induced by a variety of stimuli. It has been demonstrated that caspase 3 activation is regulated by at least two pathways (13). The first pathway involves the mitochondria ("mitochondrial pathway"). Cytotoxic agents cause the release of cytochrome c from the mitochondria into the cytosol. The release of cytochrome c activates caspase 9, and subsequently caspase 3 is activated via proteolytic processing. The second pathway is stimulated by cell surface death receptors such as tumor necrosis factor receptor 1 and Fas. Ligand-specific binding leads to caspase 8 activation, with subsequent activation of caspase 3 (death receptor pathway). Activated caspase 3 then cleaves caspase substrates, such as poly (ADP-ribosyl) polymerase (PARP), which leads to DNA fragmentation and cell death. The present study was conducted to determine the signal transduction pathway(s) involved in hyperoxia-induced lung epithelial cell death. We examined the production of ROS, activation of MAPK, mitochondrial cytochrome c release, caspase 9 and 3 activity, and cleavage of PARP. Our results indicate that the elevated level of ROS resulting from hyperoxia exposure may act as the initiator of death in lung epithelial cells. The increased level of ROS is involved in the activation of the ERK1/2 signaling pathway, which leads to the release of cytochrome c, activation of caspase 9 and 3, and cleavage of PARP to execute the cell death process.
Cell Culture and Hyperoxia Exposures Murine transformed lung epithelial cells (MLE12; American Type Culture Collection, Manassas, CA) were maintained in Dulbecco's Modified Eagle's Medium (DMEM; Life Technologies, Grand Island, NY) supplemented with 2% fetal bovine serum and 50 µg/ml gentamicin. Cells were cultured at 37°C in a humidified atmosphere containing 5% CO2. Hyperoxic conditions were achieved by placing confluent cells in 95% O2/5% CO2 at 37°C in a tightly sealed modular chamber (Billup-Rothberg, Del Mar, CA) for up to 72 h. All experiments were conducted in confluent, quiescent cells that form a monolayer to avoid cell density variability between control cells and those exposed to hyperoxia during the course of the experiment. Media was replaced daily during hyperoxia exposure.
Animals and Hyperoxia Exposures
Caspase Inhibitors
Cell Death Assays We also confirmed cell death with terminal deoxynucleotidyltransferase dUTP nick end-labeling (TUNEL) assay using the in situ cell death detection kit (Roche Molecular Biochemicals, Indianapolis, IN). Cultured MLE cells were split into a four-well chamber slide (Lab-Tec; Nalge Nunc International Corp, Naperville, IL). Following a 72-h hyperoxic exposure, cells were washed twice with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde for 60 min. Cells were then washed twice with PBS and permeabilized with 0.1% triton X-100 for 2 min. After two additional washes, the slides were incubated with TUNEL reaction mixture at 37°C for 1 h and then incubated with converter-AP at 37°C for 30 min. Cells were then washed and stained with NBT/BCIP substrate solution and counterstained with nuclear fast red. The dead and normal cells were observed under light microscope. Cells with purple nuclei were considered to be dead, cells with red nuclei were considered to be normal. For each sample 500 cells were counted and the number of dead cells was expressed as the percentage of the total cell population. For lung tissues, sections of formalin-fixed, paraffin-embedded lung tissue were deparaffinized and rehydrated, rinsed with PBS, and digested with proteinase K (Roche Molecular Biochemicals) at a concentration of 20 µg/ml for 20 min. After PBS washes, sections were incubated with TUNEL reaction mixture at 37°C for 1 h and then incubated with converter-AP at 37°C for 30 min. Sections were washed, stained with NBT/BCIP substrate solution, and counterstained with nuclear fast red.
Annexin V–Fluorescein Isothiocyanate
Immunohistochemistry
Measurement of Caspase 3, 8, and 9 Activities
Western Blot Analysis Isolation of cytoplasmic proteins and release of cytochrome c. Cells were washed twice with PBS, the pellets collected by centrifugation at 600 x g for 10 min, resuspended in 500 µl buffer (20 mM HEPES, pH 7.5, 10 mM KCl, 1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.1 mM PMSF, and 5 µg/ml each aprotonin and leupeptin) containing 250 mM sucrose, and homogenized 15 s on ice. The homogenates were centrifuged at 800 x g for 10 min at 4°C and the supernatants centrifuged again at 10,000 x g for 20 min at 4°C. The resulting supernatants containing the free cytoplasmic proteins were used for detection of cytochrome c. Western blot for anti–cytochrome c (PharMingen, San Diego, CA) was then performed. Detection of ROS. The intracellular hydrogen peroxide (H2O2) levels were measured using 2',7'-dichlorofluroescein (DCF) as described previously (16). Briefly, after hyperoxic exposure, cells were loaded with 10 µM of 2',7'-dichlorodihydrofluorescein diacetate (H2DCF-DA; Molecular Probes, Eugene, OR) for 45 min at 37°C. In certain cases, cells were treated with 10 µM of NADH/nicotinamide adenine dinucleotide phosphate reduced (NADPH) oxidase inhibitor diphenyleneiodonium chloride (DPI; Calbiochem) for 1 h before hyperoxic exposure. H2DCF-DA is a nonfluorescent, cell membrane–permeable compound that is hydrolyzed to DCF and becomes fluorescent when it is oxidized by H2O2. Cells are then trypsinized, resuspended in PBS, and fluorescence measured at 520 nm following excitation with 488 nm light from an argon laser with a FACscan. Detection is based on mean fluorescence intensity of 10,000 cells. Superoxide anion (O2-.) production was also measured after DPI treatment and hyperoxic exposure by the cytochrome c reduction assay as described (17), with minor modifications. Briefly, cells were washed twice with cold PBS, cell pellets collected by centrifugation at 100 x g for 5 min, resuspended in 10 ml 340 mM sucrose, and homogenized 30 s on ice. The homogenates were mixed with 10 µl 500 mM EDTA, centrifuged at 13,000 x g for 30 min at 4°C, and the supernatants centrifuged again at 25,000 x g for 60 min at 4°C. The resulting pellet containing the membrane proteins was resuspended with 110 µl cold PBS, and protein concentration measured. Equal quantities of membrane protein were incubated with PBS, cytochrome c (12 mg/ml), and NADPH (1 mg/ml) in the presence or absence of SOD (10 mg/ml) for 30 min. Cytochrome c reduction was measured by reading absorbance at 550 nm on a spectrophotometer. The rate of superoxide anion generation was estimated by the difference in the rate of cytochrome c reduction in the presence and absence of SOD as previously described (17). The superoxide anion production was expressed in mM per mg protein.
Statistics
Hyperoxia Induces Cell Death in Lung Epithelial Cells To demonstrate that hyperoxia induces cell death, genomic DNA was isolated from MLE12 cells exposed to varying times in hyperoxia and analyzed for DNA laddering. Figure 1 shows marked DNA laddering at 72 h of hyperoxic exposure (lane 6) in contrast to cells exposed for 8 (lane 2), 16 (lane 3), 24 (lane 4), and 48 h (lane 5) to hyperoxia, and normoxia control (lane 1).
Hyperoxia Induces Production of ROS in Lung Epithelial Cells Given published evidence of ROS mediating cell death (18), we confirmed ROS production in MLE12 cells after hyperoxia using assays for O2-. and H2O2. In Figure 2A, a NADPH oxidase activity assay was used to detect O2-. production. O2-. production was markedly increased within 5 min (lane 2) of hyperoxia and lasted to 1 h (lane 6). In addition, H2O2 production, assessed by a DCF fluorescence assay, was also dramatically increased at 15 min of hyperoxia (Figure 2B). Figure 2C shows the quantitation of fluorescence generated by H2O2 after hyperoxia. To attempt to localize the source of the ROS, we used DPI, an inhibitor of NADH/NADPH oxidase, which catalyzes O2-. generation. MLE12 cells were pretreated with 10 µM DPI for 1 h before exposure to hyperoxia. As shown in Figures 2A–2C, DPI inhibited hyperoxia-induced generation of O2-. and H2O2. This indicated that NADH/NADPH oxidase might be a potential source of hyperoxia-induced generation of ROS.
MEK/ERK1/2 Pathway Is Activated by Hyperoxia in Lung Epithelial Cells The MAPKs are activated by ROS (19, 20) and have been implicated in a variety of cellular functions, including cell death. Therefore, we examined the effect of hyperoxia on MAPK (MEK, ERK1/2, JNK1/2, and p38) activation using Western blot analyses. Phospho-antibodies to MEK, ERK1/2, JNK1/2, and p38 MAPK were used to detect activated MAPKs. As shown in Figure 3, hyperoxia leads to strong activation of ERK1/2. Activation becomes apparent from 15 min of hyperoxia and lasts to 1 h. MEK, which is upstream of ERK1/2 and reported to be responsible for ERK1/2 activation, is also phosphorylated by hyperoxia. In contrast, hyperoxia does not activate the JNK1/2 or p38 MAPK pathway in MLE12 cells (data not shown).
ROS, MEK/ERK1/2, Or Caspase Inhibitors Block Hyperoxia-Induced Cell Death in Lung Epithelial Cells Based on our observations that hyperoxia induces cell death, generates ROS, and activates MEK/ERK1/2 MAPKs in MLE12 cells, we hypothesized that both ROS and MEK/ERK1/2 may mediate hyperoxia-induced cell death. To test this hypothesis, PD98059, a specific inhibitor of MEK/ERK1/2, was used to evaluate whether ERK1/2 activation is required for hyperoxia-induced cell death. DPI, a ROS generation inhibitor, was used to test whether ROS were required for hyperoxia-induced cell death. MLE12 cells were pretreated with PD98059 or DPI for 1 h before hyperoxic exposure. As seen in Figure 4, MLE12 cells exposed to 72 h of hyperoxia alone exhibit marked DNA laddering, whereas cells pretreated with PD98059 or DPI exhibit attenuation of DNA laddering. Given that a functional link between MAPK and caspases in modulating cell death has been observed by other investigators (21, 22), we chose to investigate the role of caspases in our model. The broad caspase inhibitor, Z-VAD-FMK, blocked hyperoxia-induced DNA laddering, which implicates caspase-mediated cell death in this process. TUNEL assay was also used to confirm the protective effects of PD98059 and DPI (Figure 5A). Figure 5A, panel B shows increased TUNEL staining after 72 h of hyperoxia compared with DPI-treated (panel C) or PD98059-treated (panel D) cells. Quantitation of TUNEL staining (Figure 5B) shows a significant increase in the number of dead cells after 72 h of hyperoxia with virtually complete attenuation of cell death when ROS is inhibited using DPI (Bar 3) or MEK/ERK1/2 is inhibited using PD98059 (Bar 4). We confirmed the attenuation of hyperoxia-induced cell death with ROS, MEK/ERK1/2, or caspase inhibitors using FACs analysis as shown in Figure 6. There is a significant increase in signs of both early apoptosis, as designated by Quadrant II, and late apoptosis/necrosis (Quadrant III) after 72 h of hyperoxia (Figure 6A). The percentage of total cell death (Quadrants II + III) decreased from a mean of 22% after hyperoxia to  5% with the inhibition of ROS with DPI or with the inhibition of MEK/ERK1/2 with PD98059 (Figure 6B). Given the previous observation that the inhibition of caspases can also lead to necrosis (23), we assessed the degree of apoptosis and necrosis in the presence of the broad caspase inhibitor, Z-VAD-FMK, and the caspase 3–specific inhibitor, Z-DQMD-FMK. Figure 6B shows that caspase inhibition attenuates and apoptotic and necrotic cell death to < 4%.
ROS Mediate the Activation of MEK/ERK1/2 Pathway in Lung Epithelial Cells after Hyperoxia Thus far, we have shown that ROS generation and the activation of MEK/ERK1/2 MAPKs occur after hyperoxia and inhibition of either ROS or MEK/ERK1/2 attenuates hyperoxia-induced cell death in MLE12 cells. To establish a link between ROS production and MEK/ERK1/2 activation, we determined whether DPI, an ROS inhibitor, modulates MEK/ERK1/2 activation. As shown in Figure 7, pretreatment of MLE12 cells with DPI for 1 h before hyperoxia dramatically inhibited phospho-MEK and phospho-ERK1/2 levels. Taken together, these results demonstrate that ROS may be upstream of MEK/ERK1/2, which, in turn, modulates cell death.
ROS and MEK/ERK1/2 Inhibitors Block Downstream Cytochrome c Release and Cleavage of PARP in Lung Epithelial Cells after Hyperoxia To delineate the events downstream of ROS and MEK/ERK1/2 activation, we next studied the release of cytochrome c and cleavage of PARP. Recent studies have indicated that cytochrome c participates in activating cell death (24). Cytochrome c normally resides in mitochondria, but is released into the cytoplasm following certain cellular stresses. In the cytoplasm, cytochrome c binds to Apaf-1, resulting in the activation of caspase 9 and other downstream caspases such as caspase 3. Subsequently, caspase 3 cleaves PARP, which leads to DNA fragmentation and cell death. MLE12 cells were exposed to hyperoxia in the presence and absence of DPI or PD98059, and cytochrome c levels in cytoplasm were detected with immunoblots. We first confirmed the presence of hyperoxia-induced cytochrome c release and PARP cleavage. As seen in Figure 8A, cytoplasmic cytochrome c levels were elevated by 30 min of hyperoxia, which correlated with a subsequent increase in cleaved PARP levels after 1 h of hyperoxia. However, both cytochrome c release and cleaved PARP levels were all markedly diminished in the presence of DPI or PD98059 (Figure 8B). These findings imply that ROS and MEK/ERK1/2 are upstream of cytochrome c release as well as upstream of PARP cleavage, and that the anti–cell death effects of DPI and PD98059 may involve the inhibition of cytochrome c release and PARP cleavage.
Hyperoxia Activates Caspases 9 and 3 in Lung Epithelial Cells and ROS and MEK/ERK1/2 Are Upstream of Caspase Activation Thus far, we have shown evidence for caspase involvement by using the broad caspase inhibitor, Z-VAD-FMK, to attenuate hyperoxia-induced DNA laddering (Figure 4). We next focused upon caspase 3 and its upstream protease, caspase 9, because of the central role caspase 3 has been shown to have in various cell death models. Figure 9 shows that caspase 9 activity was increased after 30 min of hyperoxia. Pretreatment of MLE12 cells with DPI or PD98059 dramatically blocked caspase 9 activity after hyperoxia. Furthermore, hyperoxia increased caspase 3 activity that was also effectively blocked by DPI, PD98059, or Ac-LEHD-CMK, a caspase 9–specific inhibitor (Figure 10). We also tested for caspase 8 activity, which was not increased with hyperoxia (Figure 11). These results suggest that ROS production and MEK/ERK1/2 activation precede caspase 9 activation, which in turn precedes caspase 3 activation in MLE12 cells after hyperoxia. The specificity of the respective caspase inhibitors has been previously validated and published (25, 26).
Caspase 9 and 3 Inhibitors Attenuate Hyperoxia-Induced Parp Cleavage in Lung Epithelial Cells We have shown that caspase 9 and 3, but not caspase 8, activities were increased and that PARP was cleaved in MLE12 cells after hyperoxia. To clarify the relationship between caspase activation and PARP cleavage, we pretreated cells for 1 h with either 50 µM Ac-IETD-CHO (a caspase 8–specific inhibitor), Ac-LEHD-CMK (a caspase 9–specific inhibitor [25]), or 50 µM Z-DQMD-FMK (a caspase 3–specific inhibitor [27]), then exposed cells to hyperoxia, and assessed for PARP cleavage. Figure 12 shows that Ac-LEHD-CMK or Z-DQMD-FMK completely blocked the cleavage of PARP, whereas Ac-IETD-CHO did not, which implied that caspase 9 and 3 activation, but not caspase 8 activation, precede PARP cleavage after hyperoxia.
Hyperoxia-Induced Cell Death in Mouse Lungs Involves MEK/ERK1/2 and Caspase 3 Pathways We have shown that caspase 3 is activated after hyperoxia, and that the inhibition of caspases attenuates hyperoxia-induced cell death in MLE12 cells (Figures 10 and 4, respectively). To confirm our in vitro findings in vivo, we exposed wild-type littermates and caspase 3-/- mice to hyperoxia. In Figure 13B, there is significantly increased TUNEL staining in the lung epithelial and bronchial epithelial cells of wild-type mice exposed to 72 h of hyperoxia. The TUNEL staining is attenuated with MEK/ERK1/2 inhibition (Figure 13C) or in caspase 3-/- mice exposed to 72 h hyperoxia (Figure 13D). Figure 14 shows that hyperoxia induces caspase 3 activation in wild-type mice after 72 h (panel B). We were able to effectively block caspase 3 staining by MEK/ERK1/2 inhibition with PD98059 (panel C) to the minimal levels observed in caspase 3-/- mice (panel D). Therefore, similar to our in vitro results, we show that hyperoxia leads to caspase 3 activation and cell death in mouse lungs. Furthermore, hyperoxia-induced cell death is attenuated by blocking MEK/ERK1/2 or in the genetic absence of caspase 3 in mouse lungs.
The pathologic changes in pulmonary cells during exposure to hyperoxia have been extensively studied in both animal models and in cell culture systems. In many animal models, prolonged exposure to hyperoxia results in cell injury and death in the lung (28–30). The issue of whether hyperoxia induces an apoptotic or nonapoptotic death in vivo and in vitro is complex and unclear, as recently reviewed by Albertine and Plopper (31). There are likely multiple death pathways occurring in response to oxygen toxicity with features of both necrosis and apoptosis and the various assays used, including DNA laddering, TUNEL, and caspase activation, are not specific enough to reliably distinguish between the two modes of death. Several investigators have demonstrated that interruption of pulmonary cell death may lead to an improved outcome from hyperoxic lung injury (29, 32). However, the precise mechanisms of hyperoxia-induced cell death remain an area of active investigation. The goal of our current study was to delineate potential signaling mechanisms in hyperoxia-induced lung epithelial cell death. Our data demonstrate the coordinated roles of ROS, MAPK, cytochrome c, and caspases in hyperoxia-induced lung injury in mouse lung epithelial cells. Hyperoxia generates ROS that can directly injure pulmonary cells via lipid, protein, and nucleic acid oxidation. Our results indicate that exposure of MLE12 cells to hyperoxia generates ROS, and that DPI, an inhibitor of NADH/NADPH oxidase, inhibits hyperoxia-induced generation of ROS. This implicates the NADH/NADPH oxidase system as a major source of ROS generation in MLE12 cells. NADPH oxidase is a multi-component enzyme complex present in the membranes of various cells. Functional assembly of the oxidase catalyzes the transfer of one electron from cytosolic NADPH to molecular oxygen, generating superoxide anion (33). Although NADPH oxidase is known to play a critical role in host protection against infection, recent works show that virtually all cell types have a similar oxidase system. Furthermore, in a variety of nonphagocytic cells, NADPH oxidase-dependent O2-/H2O2 generation is observed in response to divergent extracellular stimuli (34). These observations strongly suggest that, similar to the NADPH oxidase system in phagocytic cells, a NADPH oxidase-like system may function as a ROS-generating system in nonphagocytic cells. Our results suggest that hyperoxia activates such a NADPH oxidase-like system in MLE12 cells. Evidence suggests that ROS contribute to the activation of the ERK1/2 pathway (19). Although it is generally thought that activation of ERK1/2 can confer a survival advantage to cells, there is growing evidence suggesting that the activation of ERK1/2 also contributes to cell death (7). In our present study, ERK1/2 is activated after hyperoxic exposure. Inhibiting ERK1/2 activation with PD98509 significantly attenuates hyperoxia-induced cell death in vitro and in vivo. Our data show that ERK1/2 activation plays an important role in the genesis of hyperoxia-induced cell death. In addition, ROS appeared to be a key trigger for ERK1/2 activation in hyperoxia-induced cell death because the pretreatment with DPI, an ROS inhibitor, virtually completely inhibited both ERK1/2 activation and subsequent MLE cell death. At present the downstream effectors of cell death initiated by ROS and MAPK activation remain to be elucidated. There has been evidence of crosstalk between MAPKs and caspase activation (21). Our results show that caspase 3-/- is activated in response to hyperoxia in MLE cells. Activation of caspase 3 was also confirmed by the cleavage of PARP, a major substrate of active caspase 3. Furthermore, MLE12 cells did not exhibit DNA laddering when treated with Z-VAD-FMK, a broad caspase inhibitor. Caspase 3-/- mice also do not exhibit TUNEL staining after hyperoxia. Thus, we provide strong evidence that caspase 3 activation may be one of the critical steps in hyperoxia-induced cell death. Of note, Barrazone and coworkers observed that at 24, 48, and 90 h, hyperoxia-induced caspase activities in whole lung lysates were not increased (35). Interestingly at 72 h hyperoxia, which is the time point we use, they show a 2-fold increase in caspase 3 activity (compared with control) that may correlate to our finding that caspase 3 staining in lung is increased after 72 h hyperoxia (Figure 13B). The mechanisms by which hyperoxia activates caspase 3 remain unknown, but we have evidence that cytochrome c is a potential candidate. Cytochrome c released from mitochondria forms an apoptosome which is composed of Apaf-1 and caspase 9, resulting in the activation of caspase 9. Caspase 9 activates the effector procaspases, including caspase 3, to execute the process of cell death (36). We observed increased levels of cytochrome c in the cytoplasm of hyperoxia-exposed MLE12 cells relative to untreated cells, suggesting that release of cytochrome c may play a role in mediating hyperoxia-induced cell death. The ability of DPI and PD98059 to diminish cytochrome c release suggests that ROS and the ERK1/2 signaling pathway, respectively, function upstream of cytochrome c in the induction of cell death. ROS generation is likely the more upstream signal given the ability of DPI, a ROS generation inhibitor, to ablate ERK1/2 activation. We have preliminary evidence that hyperoxia does not modulate Bcl-2 family members, Bad, or bfl1 at the level of mRNA in MLE12 cells. However, we recognize that post-transcriptional modifications may be important in hyperoxia signaling, and this will be a focus for future studies. Collectively, our results suggest that the elevated level of ROS generated from hyperoxia exposure may act as the initiator of cell death in MLE12 cells. The ROS then activate ERK1/2, which leads to the release of cytochrome c, activation of caspases 9 and 3, and subsequent cleavage of PARP to execute the cell death process. Knowledge of such a signaling pathway will not only help to obtain a better understanding of the underlying mechanism in the development of hyperoxia-induced cell death, but may also help develop strategies to interrupt the cell death cascade, and thereby abrogate potentially lethal organ injury.
P.J.L. was supported by NIH KO8 HL04034 and ALA of Connecticut. A.M.K.C. was supported by NIH HL-55330, AHA EIA, HIN AI42365, and NIH HL-60234. R.A.F. is an investigator of the Howard Hughes Medical Institute. G.L.C. was supported by NIH HL04007. Received in original form August 15, 2002 Received in final form September 16, 2002
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
