Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The repair mechanisms that follow lung injury involve the process of wound healing by mesenchymal cells (1). In this regard, the biologic basis of pulmonary fibrosis is akin to the process of normal wound healing, in which injury to normal tissue is followed by inflammation and then repair by the mechanism of scar formation (1). Wound healing is a complex process requiring different cellular behaviors, such as migration, proliferation, and probably apoptosis (2, 3). Fibroblasts in tissue surrounding the wound are thought to play a central role in the processes of wound repopulation and subsequent scar formation. After wound injury, fibroblasts at the wound margin begin to proliferate and migrate into the denuded area, where they lay down their own collagen-rich matrix and form granulation tissue. During granulation-tissue formation, many fibroblasts transform into myofibroblasts that generate strong contractile forces to draw the wound margins toward one another (4). This process ends with the development of a permanent scar; as the wound closes and the scar forms, fibroblasts, particularly the myofibroblasts, disappear (5). This disappearance of fibroblasts is achieved to a great extent through apoptosis (6).

Apoptosis provides a vital mechanism for eliminating unneeded cells (7). Conversely, excessive apoptosis may cause extensive cell loss. In normal wound healing, fibroblasts appear to undergo apoptosis after they have finished repopulating the wound. However, if fibroblasts undergo apoptosis while repopulating the wound, the wound may not heal. Thus, regulation of apoptosis of wound fibroblasts could be of importance for normal wound healing. The question remains: what is the signal that regulates apoptosis of wound fibroblasts?

One candidate is intracellular oxidative stress. It is well established that intracellular reactive oxygen species (ROS) generated by environmental stresses can trigger apoptotic pathways in many cell types, including fibroblasts, although they are not obligatory for all apoptosis induction protocols (8, 9). Intracellular ROS are thus tightly controlled by antioxidant defense mechanisms including thiol compounds. The most abundant thiol in cells is glutathione (GSH) that acts as a cosubstrate in the GSH peroxidase- catalyzed reduction of hydrogen peroxide (H₂O₂) or lipid peroxides (10). Reduction of cellular thiol levels, which allow ROS accumulation, has been shown to cause apoptosis in Jurkat T cells (11), neutrophils (12), and neural cells (13). These data suggest that cellular thiols as well as ROS regulate apoptosis, although the molecular mechanism is largely unknown.

A possible mechanism for thiol regulation of apoptosis is modulation of protein kinase activities. Cellular thiols have been shown to modulate tyrosine phosphorylation of many signaling molecules (14). Recent studies have identified the two different mitogen-activated protein kinase (MAPK) isoforms p38-MAPK and c-Jun N-terminal kinase (c-JNK), which are activated by dual phosphorylation on both a tyrosine and a threonine and become involved in apoptotic induction by environmental stresses such as radiation, DNA-damaging agents, heat shock, and inflammatory cytokines (17). However, it is not known whether these MAPKs are involved in thiol regulation of apoptosis. Another possible mechanism could be arachidonic acid metabolism, which has been shown to be involved in some apoptosis induction protocols (18, 19).

Recent studies have demonstrated the role of thiol antioxidants in wound healing (20). Wounds, particularly inflamed or ischemic wounds, are frequently exposed to oxidative stress and thiol antioxidant depletion, both of which are believed to impair wound healing (20, 24). To protect themselves from oxidative damage at the wound site, cells have been shown to enhance antioxidant defense mechanisms such as GSH-recycling systems (23). Thus, thiol antioxidants within fibroblasts may play a role in wound healing after injury to tissues, including the lung.

In the present study we examined whether reduction of cellular thiols induces apoptosis of lung fibroblasts. We report here that reduced cellular thiol levels induce fibroblast apoptosis by activating an ordered cell-death pathway composed of ROS accumulation, leukotriene (LT) production, and p38-MAPK phosphorylation. Using an in vitro scratch wound model, we show that fibroblasts that are stimulated to repopulate the wound, but not quiescent fibroblasts at the intact site, selectively undergo thiol depletion-induced apoptosis that is completely protected by application of inhibitors of the 5-lipoxygenase (5-LO) and p38-MAPK pathways.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

All reagents for cell culture were obtained from GIBCO Life Technologies, Inc. (Gaithersburg, MD). Diethyl maleate, chlorodinitrobenzene (CDNB), aminotriazole, GSH, N-acetylcysteine, catalase, ascorbic acid (AA), deferoxamine mesylate, N-nitro-L-arginine methyl ester (L-NAME), maleic acid diethyl ester, indomethacin, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT), bisBenzimide (Hoechst33342), propidium iodide, RNAse A, leupeptin, phenylmethylsulfonyl fluoride (PMSF), aprotinin, and sodium vanadate were purchased from Sigma Chemical Co. (St. Louis, MO). The ApopTag Plus Peroxidase kit and Apoptag Direct Fluorescein kit were obtained from Oncor, Inc. (Gaithersburg, MD). The compounds SB203580, AA861, FPL55712, and ONO1078 were a gift of SmithKline Beecham Pharmaceuticals (King of Prussia, PA), Takeda Chemical Industries (Osaka, Japan), Fisons Pharmaceuticals (Loughborough, UK), and Ono Pharmaceuticals (Osaka, Japan), respectively. The fluorescence probe 6-carboxy-2',7'-dichlorodihydrofluorescein diacetate, di(acetoxymethyl ester) (CDCFH) and monochlorobimane (mBCI) came from Molecular Probes, Inc. (Eugene, OR). Polyclonal antibodies to p44/42-MAPK, p38-MAPK, c-JNK, ATF2, and c-Jun, specific for the phosphorylated forms of these proteins, were obtained from New England Biolabs, Inc. (Beverly, MA).

Cell Culture and Thiol Depletion

Normal human fetal lung fibroblasts (IMR-90; Clonetics Corp., Palo Alto, CA) were maintained in growth medium that consisted of Dulbecco's modified Eagle's medium (DMEM) containing 48 µg/ml cystine, 10% fetal calf serum (FCS), 100 U/ml penicillin, and 100 µg/ml streptomycin. Cells were passaged weekly and cells from passages 8-20 were used for experiments. Before experiments, confluent cells were trypsinized and plated into 60-mm culture dishes, 24-well culture plates, or 96-well culture plates at a density of 3.8 × 10⁴ cells/cm² in growth medium. After 24 h, unless indicated otherwise, cells were subjected to thiol depletion protocols described later. At this time point, cells were 50 to 70% confluent.

For thiol depletion, cells were rinsed twice with phosphate-buffered saline (PBS) and replenished with the following media: serum-free DMEM containing or lacking cystine, or serum-free DMEM containing cystine enriched with 5 × 10⁴ M diethyl maleate or 2 × 10⁴ M CDNB. Medium lacking cystine was used to deplete cellular GSH and other thiols, through depletion of their limiting amino-acid precursor, cysteine (25). Diethyl maleate and CDNB were used to deplete thiols by forming adducts with GSH and other thiols (11, 22).

Determination of Total Cellular GSH and Free Thiol Levels

To assay for total GSH, cells in 96-well plates were lysed by the addition of 5% trichloroacetic acid containing 20 mM ethylenediaminetetraacetic acid and a cycle of freeze- thawing. The acid-precipitated protein was pelleted by centrifugation at 4°C for 30 min at 10,000 × g. The total protein content in each sample was determined by Bradford's method using a Bio-Rad protein assay (Bio-Rad, Hercules, CA). The aliquot was assayed for total GSH content (the sum of the oxidized and reduced forms) using the enzymatic recycling procedure in which GSH or oxidized GSH and GSH reductase reduce 5,5'-dithiobis (2-nitrobenzoic acid) (DTNB) to form 5-thio-2-nitrobenzoate (TNB) (28). The formation of TNB was followed spectrophotometrically at 412 nm. Total GSH content was determined in comparison with reference curves generated with known amounts of GSH. The content of intracellular free thiols was determined using mBCI, which passively diffuses across the plasma membrane into the cytoplasm where it forms blue fluorescent adducts with the reduced form of GSH and other thiol-containing proteins (29). Briefly, cells in 96-well plates were incubated with 50 µM mBCI for 40 min and then plates were read on a Cytofluor II multiplate fluorimeter (Perseptive Biosystems, Inc. Framingham, MA) using excitation and emission wavelengths of 395 and 460 nm, respectively.

Detection of Intracellular ROS Accumulation

Intracellular ROS accumulation was monitored using CDCFH, which is trapped inside cells and forms yellow fluorescent adducts with ROS (29). The CDCFH mainly measures peroxide (e.g., H₂O₂ and lipid peroxide), which can be further used to generate other ROS. Cells in 96-well plates were incubated with 10⁵ M CDCFH for 30 min. The cells were rinsed with PBS and fed with indicated medium. Fluorescence was monitored using excitation and emission wavelengths of 485 and 530 nm, respectively.

Evaluation of Cell Survival

Cell survival was determined in 96-well plates by a colorimetric MTT assay as described elsewhere (30). This assay is based on the conversion of the tetrazolium salt MTT by mitochondrial dehydrogenase to a formazan product, as measured at an absorbance of 570 nm.

Evaluation of Apoptosis

Apoptotic cells were detected in two different ways. For nuclear staining with Hoechst33342, cells were incubated for 24 h in indicated medium in 24-well plates. Then both detached and attached cells released with trypsin were combined, fixed in 3% paraformaldehyde, and stained with 5 µg/ml Hoechst33342. Cells containing condensed or fragmented nuclei were identified as apoptotic cells (31) on fluorescence microscopy at ×200 magnification. At least 300 cells were counted and the percentage of apoptotic cells was determined. For DNA nick-end labeling, cells were seeded on glass coverslips and placed in 24-well plates. After a 24-h incubation in indicated medium, detached cells were removed and attached cells were fixed in 4% neutral buffered formalin. DNA strand breaks in apoptotic cells were detected in situ by terminal deoxynucleotidyl transferase-mediated nucleotide nick-end labeling (TUNEL) using the ApopTag Plus Peroxidase kit or the ApopTag Direct Fluorescein kit.

Immunoblot Analysis of Protein Phosphorylation

Cell lysates were solubilized in RIPA buffer (0.15 M NaCl, 50 mM Tris-Cl [pH 7.4], 0.5% NP40, and 0.1% sodium dodecyl sulfate [SDS]) containing 10 µg/ml leupeptin, 1 mM PMSF, 10 µg/ml aprotinin, and 1 mM sodium vanadate; fractionated by SDS-polyacrylamide gel electrophoresis (PAGE); transferred to polyvinylidene difluoride membrane; and probed with antibodies, all of which were used at a dilution of 1:1,000. Primary antibody was detected by horseradish peroxidase-conjugated antibody (1:2,500), which, in turn, was visualized using enhanced chemiluminescence (SuperSignal; Pierce, Rockford, IL).

Immunocytochemistry

Cells incubated on glass coverslips were fixed in 3% paraformaldehyde, permeabilized with 0.5% Triton X-100, and stained with primary antibodies (1:200). Primary antibody was detected by fluorescein isothiocyanate-conjugated antibody (1:100). Stained cells were observed under fluorescence microscopy.

Determination of LTC4, LTD4, and LTE4 Release

Cells were incubated in 150 µl of indicated medium in 96-well plates. After 16 h, the conditioned medium was recovered and immediately assayed for LTC4, LTD4, and LTE4 contents using a Biotrak Leukotriene C4, D4, E4 Enzyme Immunoassay System (Amersham Life Science, Buckinghamshire, UK).

Wounding of Fibroblast Cultures

Fibroblasts were grown in growth medium containing 10% FCS on coverslips placed in 24-well plates. After reaching confluence, the monolayer was scratched with a sterile plastic pipette tip. The wounded monolayers were then rinsed twice with PBS, replenished with the growth medium containing serum, and allowed to commence repopulation of the denuded area. After 24 h, the growth medium was replaced with serum-free DMEM with or without cystine and incubation was continued for another 24 h. For in situ detection of intracellular ROS accumulation, cells were loaded with 10⁵ M CDCFH for 30 min and rinsed with PBS before the growth medium was replaced with indicated medium. ROS accumulation was qualitatively monitored by fluorescence microscopy. For TUNEL, cells were fixed with 4% neutral buffered formalin and stained using the ApopTag Direct Fluorescein kit, followed by counterstaining with a mixture of 5 µg/ml propidium iodide and 50 µg/ml RNAse A. For in situ detection of p38-MAPK phosphorylation, cells were immunostained with an antibody specific for the phosphorylated form of p38-MAPK as described previously.

Statistics

Results are presented as means ± standard error of the mean (SEM). Comparisons were made by Student's t test or analysis of variance with Scheffe's correction as appropriate. A value of P < 0.05 was accepted as significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Culturing Fibroblasts in Cystine-Free Medium Leads to Depletion of GSH and Other Thiol Levels

In normal tissue culture medium stored aerobically, the primary source of cysteine available to most cell types is cystine, due to oxidation of all cysteine to cystine during storage. Cysteine, a limiting sulfur-amino acid required for synthesis of thiols such as GSH, is the most abundant cellular thiol (10). Thus, the availability of cyst(e)ine is usually rate-limiting for synthesis of thiols in culture (25- 27). When fibroblasts were cultured in medium lacking cystine, total GSH and free thiol levels in cells declined to undetectable levels by 10 to 14 h (Figure 1). In contrast, the culture of cells in medium lacking the other sulfur-containing amino acid methionine did not affect cellular thiol levels (data not shown). These results confirm that limited cyst(e)ine availability depletes thiol levels in fibroblasts.

View larger version (13K): [in this window] [in a new window]   Figure 1.   Culture of fibroblasts in cystine-free medium causes depletion of total GSH (A) and total free thiol (B) levels in cells. Fibroblasts grown for 24 h were rinsed with PBS and cultured in medium containing (filled circles) or lacking (open circles) cystine. Total GSH (n = 4) and free thiol (n = 6) contents were monitored using the DTNB- and mBCI-based assays, respectively.

Depletion of Cellular Thiols Induces Accumulation of ROS

Because cellular thiols such as GSH represent the first line in the cellular antioxidant defense mechanisms, we examined whether depletion of cellular thiols induces intracellular accumulation of ROS. Intracellular ROS was measured using a fluorometric assay with CDCFH that mainly reflects intracellular peroxides (e.g., H₂O₂ and lipid peroxides), which can be further used to generate other ROS. When fibroblasts were cultured in cystine-free medium for 24 h, the intensities of CDCFH fluorescence increased significantly to about 3-fold greater than those in cystine-containing medium (Figure 2). Similarly, more than 2-fold increases in CDCFH fluorescence were observed when cells were cultured with diethyl maleate (5 × 10⁴ M) and CDNB (2 × 10⁴ M) that deplete cellular GSH and other thiols by forming adducts with them. Compared with the thiol-depleted cells, CDCFH fluorescence was increased by only 1.4-fold in cells cultured with aminotriazole (5 × 10² M) that was used to inhibit catalase (32). These results indicate that cellular thiols are a critical regulator of peroxide accumulation in fibroblasts. The major source of ROS accumulation in nonphagocytic cells is known to be a leak from mitochondrial electron transport (33), but some enzymatic mechanisms can also generate ROS. Several inhibitors of these enzymes were used to test for their ability to prevent ROS accumulation in response to thiol depletion. However, increased intensities of CDCFH fluorescence in cells in cystine-free culture were not blocked by any of the inhibitors of cyclooxygenase (5 × 10⁴ M indomethacin; 107 ± 1.5% of control CDCFH in cystine-free culture), 5-LO (2 µg/ml AA861; 110 ± 2.5%), monoamine oxidase (10⁶ M cloglyline; 120 ± 8.1%), xanthine oxidase (3 × 10⁴ M allopurinol; 108 ± 2.4%), nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (10⁴ M neopterin; 127 ± 5.5%), and nitric oxide synthase (10⁶ M L-NAME; 192 ± 5.2%). Thus, these results indicate that depletion of cellular thiols induces accumulation of ROS, although the exact source(s) of ROS in thiol-depleted fibroblasts remain to be determined.

View larger version (21K): [in this window] [in a new window]   Figure 2.   Thiol depletion allows accumulation of ROS in fibroblasts. Fibroblasts grown for 24 h were rinsed with PBS and cultured in medium containing cystine (open circles), medium lacking cystine (filled circles), or medium containing cystine supplemented with 5 × 104 M diethyl maleate (filled triangles), 2 × 104 M CDNB (filled squares), or 5 × 102 M aminotriazole (open squares). Intracellular ROS accumulation was monitored using CDCFH. Results are expressed as means ± SEM of n = 5 experiments. *P < 0.05, **P < 0.01 versus cells in cystine-containing medium.

Depletion of Cellular Thiols Causes Fibroblast Death through ROS Accumulation

To explore the contribution of thiols and ROS to fibroblast survival, an MTT assay was performed to quantify the number of viable cells. When fibroblasts were cultured in medium lacking cystine or medium containing diethyl maleate (5 × 10⁴ M) and CDNB (2 × 10⁴ M), the survival was markedly decreased (Figure 3). In contrast, the survival of cells cultured with the catalase inhibitor aminotriazole (5 × 10² M) remained unchanged (Figure 3B). The decrease in cell survival in cystine-free medium was inhibited by addition of AA (10³ M) that acts as antioxidant and catalase (400 U/ml) that decomposes H₂O₂. Because H₂O₂ is freely permeant in the cell membrane, and therefore readily diffuses out of cells if generated intracellularly, the reduction of extracellular H₂O₂ by catalase is expected to decrease intracellular H₂O₂ (34). In the presence of Fe²⁺, H₂O₂ can decompose to the hydroxyl radical, an extremely reactive species responsible for most of the covalent modification and damage to macromolecules, including DNA, proteins, and lipid membranes (35). However, the hydroxyl radical is unlikely to be involved in our systems because the iron chelator deferoxamine mesylate (10⁴ M) did not rescue cell death in cystine-free culture (Figure 3A). These results indicate that thiol depletion causes fibroblast death at least partly through accumulation of ROS including H₂O₂, although the possible involvement of other reactive species such as peroxynitrite cannot be excluded.

View larger version (20K): [in this window] [in a new window]   Figure 3.   Effect of thiol depletion on fibroblast survival in the presence or absence of antioxidants. Fibroblasts were cultured for 24 h in medium containing or lacking cystine with or without 103 M AA, 400 U/ml catalase, and 104 M deferoxamine (A); or in medium containing cystine with or without 5 × 104 M diethyl maleate, 2 × 104 M CDNB, or 5 × 103 M aminotriazole (B). Survival of fibroblasts was determined by MTT assay. Results are expressed as means ± SEM of n = 6 experiments. **P < 0.01 versus control cells in medium containing cystine. P < 0.01 versus cells in medium lacking cystine.

Depletion of Cellular Thiols Induces Fibroblast Apoptosis through ROS Accumulation

To determine whether cell death was due to the induction of apoptosis, the morphology of cell nuclei was evaluated by fluorescence microscopy of Hoechst33342-stained nuclei. When fibroblasts were cultured in medium lacking cystine for 24 h, 43% of cells had condensed and fragmented nuclei typical of apoptosis (Figures 4B and 5). DNA strand breaks, also characteristic of apoptosis, were detected in these cells by TUNEL (Figures 4D and 4F). Apoptosis was also induced in cells cultured with diethyl maleate (5 × 10⁴ M) or CDNB (2 × 10⁴ M) (Figure 5). Induction of apoptosis in cystine-free culture was blocked by addition of AA (Figure 5), consistent with the MTT assay data. Together, these results suggest that depletion of cellular thiols induces cell death via apoptosis by a mechanism dependent on ROS accumulation.

View larger version (43K): [in this window] [in a new window]   Figure 4.   Induction of fibroblast apoptosis in cystine-free culture. Fibroblasts were cultured for 24 h in medium containing (A, C, and E) or lacking (B, D, and F ) cystine. Both detached and attached cells were then collected, stained with Hoechst33342, and observed by fluorescence microscopy (A and B; original magnification: ×200). Alternatively, only attached cells were stained with the ApopTag Plus Peroxidase kit and observed by phase-contrast (C and D; original magnification: ×400) or brightfield (E and F; original magnification: ×1,000) microscopy. When cultured in medium lacking cystine, many cells have condensed and fragmented nuclei (B) and shrunken cytoplasm (D) with TUNEL-positive nuclei (F).

View larger version (14K): [in this window] [in a new window]   Figure 5.   Effect of thiol depletion on fibroblast apoptosis in the presence or absence of antioxidants. Fibroblasts were cultured for 24 h in medium containing or lacking cystine with or without 103 M AA, 5 × 104 M diethyl maleate, or 2 × 104 M CDNB. Percent apoptosis was determined by nuclear morphology after Hoechst33342 nuclear staining. Results are expressed as means ± SEM of n = 3 experiments. **P < 0.01 versus control cells in medium containing cystine. P < 0.01 versus cells in medium lacking cystine.

p38-MAPK Phosphorylation by ROS Is Involved in Signaling Events Mediating Thiol Depletion-Induced Apoptosis

To determine the molecular mechanism by which thiol depletion induces apoptosis, immunoblot analysis was performed using phosphospecific antibodies to detect phosphorylation of MAPKs and their substrates (Figure 6). When fibroblasts were cultured in medium lacking cystine or containing diethyl maleate (5 × 10⁴ M), phosphorylation of p38-MAPK (Tyr-182) was observed at 4 h and later. Simultaneous phosphorylation was observed in ATF2 (Thr-71), which is a nuclear substrate of p38-MAPK (36, 37). Addition of AA completely blocked the phosphorylation of p38-MAPK in cells in cystine-free culture, suggesting the involvement of ROS in thiol depletion-induced phosphorylation of p38-MAPK. The phosphorylation of p38-MAPK was selective because no significant phosphorylation was detected in the other MAPK members p44/42-MAPK (Tyr-204) and c-JNK (Thr-185) and its substrate c-Jun (Ser-63/73) under conditions in which p38 was phosphorylated. To determine whether p38-MAPK activation is required for induction of apoptosis, the effect of the selective p38-MAPK inhibitor that has no inhibitory action on p44/42-MAPK and c-JNK (38) was tested. As shown in Figure 7, application of 10⁶ M SB203580 almost completely blocked cell death and apoptosis in cystine-free culture. These results suggest that selective p38-MAPK phosphorylation by ROS is involved in the mechanism of thiol depletion-induced apoptosis.

View larger version (26K): [in this window] [in a new window]   View larger version (35K): [in this window] [in a new window]   Figure 6.   Effect of thiol depletion on phosphorylation of MAPKs and their substrates. Fibroblasts were cultured in medium lacking cystine for 0, 2, 4, and 8 h (A), or with or without 103 M AA and 5 × 104 M diethyl maleate (DM) for 4 h (B). Cell lysates were prepared and subjected to SDS-PAGE (45 µg protein/lane) and immunoblotting with control antibodies (Con) to p38-MAPK, ATF2, p44/42-MAPK, c-JNK, c-Jun, and phosphospecific antibodies (Phos) to each protein. The control antibodies react with both the phosphorylated and nonphosphorylated forms of these proteins. This figure represents one of three separate experiments performed.

View larger version (12K): [in this window] [in a new window]   Figure 7.   Protection of thiol depletion-induced cell death and apoptosis by the selective p38-MAPK inhibitor SB203580. Fibroblasts were cultured in medium containing or lacking cystine with or without the indicated concentrations of SB203580. After 24 h, cells were examined for cell survival (A: n = 6) by MTT assay, and for percent apoptosis (B: n = 3) by morphology of Hoechst33342-stained nuclei. **P < 0.01 versus control cells in medium containing cystine. P < 0.01 versus cells in medium lacking cystine.

LT Production Is an Upstream Event Mediating p38-MAPK Phosphorylation and Apoptosis in Thiol-Depleted Cells

Since previous studies showed that arachidonic acid metabolism is involved in certain apoptosis induction protocols (18, 19), we tested the abilities of inhibitors of cyclooxygenase and lipoxygenase to protect against apoptosis (Figures 8A and 8B). When the 5-LO inhibitor AA861 (2 µg/ml) was added into cystine-free medium, both cell death and apoptosis were almost completely blocked. Similar protective effects were observed with FPL55712 (5 × 10⁵ M) and ONO1078 (1 × 10⁵ M), receptor antagonists of LTC4, LTD4, and LTE4 that are arachidonic acid products via the 5-LO. In contrast, the cyclooxygenase inhibitor indomethacin had no protective effect on cell death or apoptosis (data not shown). The protective effect of AA861, FPL55712, and ONO1078 was unlikely to be due to inhibition of ROS accumulation because these drugs failed to reduce the increased intensity of CDCFH fluorescence in cells in cystine-free culture (data described previously or not shown). These results suggest that LTs produced via the 5-LO are involved in the signaling mechanism of apoptosis induction.

View larger version (13K): [in this window] [in a new window]   View larger version (50K): [in this window] [in a new window]   Figure 8.   Effect of inhibition of the lipoxygenase pathway on thiol depletion-induced cell death (A), apoptosis (B), and p38-MAPK phosphorylation (C). Fibroblasts were cultured in medium containing or lacking cystine with or without the 5-LO inhibitor AA861 (2 µg/ml) or the LTC4, LTD4, LTE4 receptor antagonists FPL55712 (5 × 105 M) and ONO1078 (105 M). After 24 h, cells were examined for survival (A: n = 6) and percent apoptosis (B: n = 3) as described in Figure 7. After 8 h, phosphorylation of p38-MAPK (C) was analyzed by immunoblotting. AA and FPL represent AA861 and FPL55712, respectively. **P < 0.01 versus control cells in medium containing cystine. P < 0.01 versus cells in medium lacking cystine.

To strengthen these results, we evaluated LT release from cells in cystine-free culture. When fibroblasts were cultured in cystine-free medium for 16 h, the release of LTC4, LTD4, and LTE4 was increased by 1.8-fold compared with cystine-containing culture (1.28 ± 0.36 to 2.27 ± 0.16 pg/well, P < 0.05). This increase was effectively blocked by the addition of 10³ M AA (1.73 ± 0.10 pg/ well, P < 0.05), suggesting that accumulation of ROS is involved upstream of LT production in thiol-depleted cells. To further determine the position of LT production in signaling events in thiol-depleted cells, we explored the effect of AA861 and FPL55712 on p38-MAPK phosphorylation. As shown in Figure 8C, both reagents inhibited p38-MAPK phosphorylation in cells in cystine-free culture, suggesting that LT production lies at a point upstream of p38-MAPK. Together, these results suggest that thiol depletion triggers an ordered cell-death pathway in which ROS accumulation activates LT production via the 5-LO pathway which, in turn, stimulates the p38-MAPK route, thereby inducing apoptosis.

Subconfluent Cells but Not Confluent Cells Undergo Apoptosis in Response to Thiol Depletion

All the results described previously were obtained using cells that were 50 to 70% confluent when thiol-depletion procedures were applied. It was also necessary to examine whether confluent cells display similar behavior with regard to apoptosis. As shown in Figure 9A, when cells at 50 to 70% confluence (Day 1 culture) were exposed to cystine-free medium, 43% of cells underwent apoptosis after 24 h. Percentage of apoptosis was decreased to 12% when cells at 70 to 90% confluence (Day 2 culture) were exposed to cystine-free medium. Only 1% of cells became apoptotic when fully confluent cells (Day 4 culture) were exposed to cystine-free medium. Consistent with these data, phosphorylation of p38-MAPK was exclusively detectable in subconfluent cells in cystine-free culture (Figure 9B). The resistance of confluent cells to apoptosis was unlikely to be due to incomplete thiol depletion because (1) after 8 h of cystine-free culture, total GSH was reduced to similar levels in subconfluent (1.89 ± 0.97 µg/mg protein) and confluent cells (2.15 ± 0.13 µg/mg protein) from the baseline level (9.82 ± 0.1 µg/mg protein); and (2) confluent cells also failed to undergo apoptosis induced by thiol depletion with diethyl maleate (5 × 10⁴ M) (33.3 ± 2.0 versus 3.3 ± 0.7% apoptosis for subconfluent and confluent cells, respectively; P < 0.01). Further, the resistance of confluent cells to apoptosis was not due to the production of any soluble or insoluble survival factor(s) because conditioned medium and immobilized cell-derived matrix (biosynthesized matrix) (39) obtained from confluent cell cultures did not protect against apoptosis in subconfluent cells exposed to cystine-free medium (data not shown).

View larger version (66K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 9.   Induction of apoptosis and phosphorylation of p38-MAPK by thiol depletion are dependent on cell density. Fibroblasts were plated at a density of 3.8 × 104 cells/cm2 and allowed to grow in growth medium containing 10% FCS. After the indicated days, cells were rinsed with PBS and refed with serum-free medium containing (open bars) or lacking (filled bars) cystine. After 24 h, cells were examined for percent apoptosis by morphology of Hoechst33342-stained nuclei (A: n = 3). Alternatively, after 8 h, p38-MAPK phosphorylation was analyzed by immunoblotting (B). Fibroblasts were found to be 50 to 70% confluent on Day 1 and 70 to 90% confluent on Day 2, and to have reached confluence on Day 3 after cell plating.

Repopulating Fibroblasts at the Wound Margin but Not Quiescent Cells in the Intact Site Undergo Apoptosis in Response to Thiol Depletion

The most distinct feature of subconfluent versus confluent cells is their high activities for proliferation and migration, which are believed to be important for normal wound healing. Thus, we attempted to evaluate the potential role of thiols in fibroblastic wound repopulation in vitro. Our hypothesis was that fibroblasts that are actively repopulating a wound are susceptible to induction of apoptosis by thiol depletion. This hypothesis was tested using an in vitro scratch wound model in which fibroblasts were first allowed to repopulate the denuded area for 24 h in serum-containing growth medium, followed by exposure to serum-free medium containing or lacking cystine. When repopulating fibroblast cultures were exposed to cystine-containing medium for 24 h, no apoptosis was detected by TUNEL at either the injured or intact sites (Figures 10A and 10B). In contrast, when repopulating cell cultures were exposed to medium lacking cystine for 24 h, many apoptotic cells were observed exclusively at the wound margin where fibroblasts were actively repopulating the denuded area (Figures 10C and 10D). Very few apoptotic cells were observed at the intact area (Figures 10C and 10D). Consistent with the selective induction of apoptosis, ROS accumulation and p38-MAPK phosphorylation were detected exclusively in fibroblasts at the wound margin but not in those at the intact area (Figures 10E and 10F). Apoptosis of fibroblasts at the wound margin was completely blocked by the addition of AA861, FPL55712, ONO1078 or SB203580 (Figures 10G-10N), suggesting that apoptosis was dependent on LT production and p38-MAPK activation. These observations suggest that actively repopulating fibroblasts at the wound margin, but not quiescent fibroblasts at the intact site, generate ROS, increase phosphorylation of p38-MAPK, and undergo apoptosis in response to thiol depletion.

View larger version (20K): [in this window] [in a new window]   Figure 10.   Repopulating fibroblasts at the wound margin, but not quiescent fibroblasts at the intact site, generate ROS, exhibit phosphorylation of p38-MAPK, and undergo apoptosis in response to thiol depletion. Confluent fibroblast cultures were scratched with a plastic pipette tip and allowed to repopulate the denuded area in growth medium containing 10% FCS for 24 h. The wounded cultures were then exposed to serum-free medium containing (A and B) or lacking (C-F ) cystine, or serum-free medium lacking cystine supplemented with 2 µg/ml AA861 (G and H), 5 × 105 M FPL55712 (I and J), 105 M ONO1078 (K and L), or 106 M SB203580 (M and N). For monitoring of intracellular ROS accumulation, cells were loaded with 105 M CDCFH and rinsed with PBS before exposure to the medium described above. After 24 h, cells were examined for apoptosis by TUNEL (B, D, H, J, L, and N) and by counterstaining with propidium iodide to identify cell nuclei (A, C, G, I, K, and M); and for intracellular ROS accumulation (E) and for the presence of phosphorylated p38-MAPK (F) by immunostaining with a phosphospecific p38-MAPK antibody. Photographs were taken under fluorescence microscopy (original magnification: ×100). This represents one of three experiments performed. W and I in each panel denote wounded and intact sites, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study we have demonstrated that thiol depletion caused by limiting cyst(e)ine availability or sulfhydryl adduction with diethyl maleate and CDNB induces apoptosis of lung fibroblasts through a cell-death pathway composed of ROS accumulation, LT production, and selective p38-MAPK phosphorylation. Using an in vitro scratch wound model we also demonstrated that repopulating fibroblasts at the wound margin, but not quiescent cells at the intact site, undergo thiol depletion-induced apoptosis; and that this apoptosis is completely protected against by inhibition of the 5-LO and p38-MAPK pathways. Although the scratch wound model we used does not necessarily reflect a complex process of wound healing that involves the participation of multiple cell types, it is widely used to analyze the process of wound repair in vitro (40, 41).

Our study is consistent with previous studies that demonstrate that depletion of cellular thiols such as GSH induces apoptosis in neutrophils, T cells, and neural cells (11, 27). As has been reported previously (8, 9, 42), our study also suggests the role of ROS as a mediator of thiol depletion-induced apoptosis because ROS scavengers including catalase and AA protected against cell death by thiol depletion. Although the type(s) of ROS causing apoptosis have remained speculative, several lines of evidence in our study suggest that peroxide plays an important role in mediating apoptosis in response to thiol depletion. First, CDCFH fluorescence that mainly reflects H₂O₂ and lipid peroxides increased markedly within cells after thiol depletion. Second, the cell death was effectively blocked by application of catalase that is expected to reduce intracellular H₂O₂ by decomposing H₂O₂ that has diffused out of cells (34). Our interpretation is in line with previous studies showing that direct exposure of cells to peroxide can induce apoptosis in a variety of cell types (43). Previous studies demonstrated that endogenous generation of H₂O₂ following GSH depletion caused mitochondrial damage (48, 49), which has been implicated in apoptosis (50). However, our results do not exclude the possibility that other reactive species, such as peroxynitrite, are also involved in apoptosis caused by thiol depletion.

How does thiol depletion induce apoptosis in fibroblasts? The present study has suggested the involvement of two signaling events: LT production and p38-MAPK phosphorylation. Our observations also permit an initial ordering of the cell-death pathway in which thiol depletion allows ROS accumulation that activates LT production via the 5-LO pathway, which, in turn, stimulates the p38-MAPK route, thereby inducing apoptosis. The MAPK isoforms p38-MAPK and c-JNK have been shown to be involved in apoptotic induction by environmental stresses such as radiation, DNA-damaging agents, heat shock, and inflammatory cytokines (17). However, its involvement in apoptosis by thiol depletion/ROS generation systems has not been reported previously.

On the other hand, our results regarding the involvement of LT production in apoptosis are consistent with previous studies showing that inhibition of the lipoxygenase pathway protects against radiation-induced thymocyte apoptosis (19) and tumor necrosis factor-mediated apoptosis of fibrosarcoma cells (18). Regarding the involvement of ROS in LT production, several reports have documented increased arachidonic acid release due to the synthesis and activation of phospholipase A₂ during oxidative stress (51). Further, some enzymes that produce eicosanoids appear to be regulated by cellular oxidative or redox levels. It has recently been shown that 5-LO, which produces LTs, contains a catalytically important iron atom at its active site and is regulated by agents that change cellular redox status (55). In addition, a recent study has shown that application of arachidonic acid activates the other MAPK c-JNK through NADPH oxidase stimulation in kidney epithelial cells (56).

Our study of incised fibroblast culture has demonstrated that fibroblasts that are stimulated to repopulate the wound, but not quiescent fibroblasts at the intact site, underwent ROS accumulation, p38-MAPK phosphorylation, and apoptosis in response to thiol depletion. In this regard, our study has demonstrated that subconfluent cells that are actively proliferating and migrating, but not confluent cells that are quiescent, underwent p38-MAPK phosphorylation and apoptosis after thiol depletion. Together, these observations suggest that the ability or inability of fibroblasts to apoptose in response to thiol depletion may depend on the ability of cells to proliferate and/or migrate. To date, little information is available to explain our observations. However, recent evidence suggests that apoptosis requires cell division processes or cell cycle-related proteins such as c-myc, p53, and cdc2 kinase (57). Further, intracellular ROS are closely related to growth-related signals and regulate cell proliferation (60). In fact, stimulation of cells by growth factors has been shown to cause H₂O₂ generation in cells (61), suggesting that intracellular ROS are generated when cells are stimulated to proliferate in response to external stimuli. Thus, cells may become oxidative and susceptible to thiol depletion-induced apoptosis when stimulated to proliferate and/or migrate in response to wound injury. Further studies will be needed to test the validity of this idea.

Recent studies have demonstrated the role of cellular thiols in wound healing (20). Wounds, particularly inflamed and ischemic wounds, are frequently exposed to enhanced oxidative stress and reduced thiol antioxidant levels, both of which could lead to tissue injury and impaired wound healing (20, 24). Taken together, our observation in the in vitro scratch wound model suggests that regulation of fibroblast apoptosis by altered cellular thiol levels may play a role in wound healing processes after injury to tissues, including the lung. However, it should be noted that the degree of thiol depletion observed in this study is unlikely to occur under physiologically relevant stress conditions such as exposure to some toxicants. Thus, extrapolation of our data to in vivo situations requires caution.