Molecular Mechanism of the Regulation of Glutathione Synthesis by Tumor Necrosis Factor- a and Dexamethasone in Human Alveolar Epithelial Cells*

Glutathione (GSH) is an important physiological antioxidant in lung epithelial cells and lung lining fluid. We studied the regulation of GSH synthesis in response to the pro-inflammatory cytokine tumor necrosis factor- a (TNF- a ) and the anti-inflammatory agent dexamethasone in human alveolar epithelial cells (A549). TNF- a (10 ng/ml) exposure increased GSH levels, concomitant with a significant increase in g -glutamylcysteine synthetase ( g -GCS) activity and the expression of g -GCS heavy subunit ( g -GCS-HS) mRNA at 24 h. Treatment with TNF- a also increased chloramphenicol acetyltransferase (CAT) activity of a g -GCS-HS 5 * -flanking region reporter construct, transfected into alveolar epithelial cells. Mutation of the putative proximal AP-1-binding site ( 2 269 to 2 263 base pairs), abolished TNF- a -mediated activation of the promoter. Gel shift and supershift analysis showed that TNF- a increased AP-1 DNA binding which was predominantly formed by dimers of c-Jun. Dexamethasone (3 m M ) produced a significant decrease in the levels of GSH, decreased g -GCS activity and g -GCS-HS mRNA expression at 24 h. The increase in GSH levels, g -GCS-HS mRNA, g -GCS-HS promoter activity, and AP-1 DNA binding produced by TNF- a were abrogated by co-treating the cells with dexamethasone. Thus these data demonstrate

The tripeptide, L-␥-glutamyl-L-cysteinylglycine, or glutathione (GSH), is a ubiquitous cellular non-protein sulfhydryl, which plays an important role in maintaining intracellular redox balance and in cellular defenses against oxidative stress (1,2). GSH is present in high concentrations in lung epithelial lining fluid (3). It also has an important role in maintaining the integrity of the airspace epithelium, in both type II alveolar cells in vitro, and in lungs in vivo (4,5). A physiological role for GSH as an antioxidant has been described in numerous inflammatory disorders (6). Depletion of lung epithelial lining fluid GSH has been described in conditions such as human immunodeficiency virus infection (7), idiopathic pulmonary fibrosis (8), adult respiratory distress syndrome (9), and cystic fibrosis (10). Elevated levels of GSH occur in the epithelial lining fluid of chronic smokers and in patients with lung cancer (11).
Inflammation results in the release of inflammatory mediators, such as cytokines, which affect the local tissue oxidant/ antioxidant balance. Anti-inflammatory drugs might also influence this balance. Tumor necrosis factor-␣ (TNF-␣) is a ubiquitous pro-inflammatory cytokine and is recognized as an important mediator of multiple inflammatory events in the lungs. It induces chronic inflammatory changes associated with the increase in a variety of defense mechanisms, through the up-regulation of mRNA for various inflammatory mediators (16). Regulation of mRNA expression by TNF-␣ is mediated by activation of transcription factors c-Fos/c-Jun (activator protein-1, AP-1) and nuclear factor-kappa B (NF-B) (17). It has been reported that prior exposure to a sublethal dose of TNF-␣ renders cells resistant to a subsequent TNF-␣ challenge. The mechanism by which this TNF-␣-mediated cellular resistance occurs is related to the enhancement of the intracellular antioxidant capacity (18). Examples of this are increased activity of mitochondrial manganese superoxide dismutase (19), and other protective proteins, including plasminogen activator inhibitor type 2, the zinc finger protein A20, and the Bcl-2-related family member A1 (20 -22). However, the molecular basis of the TNF-␣-induced cellular tolerance is not fully understood at present. One possible mechanism may involve TNF-␣-induced generation of reactive oxygen species (ROS) by leakage from the mitochondrial electron transport chain (23,24). Such intracellular ROS could induce glutathione synthesis in response to this oxidant stress by a mechanism involving the up-regulation of the ␥-GCS-HS mRNA.
Corticosteriods are widely used as anti-inflammatory agents in various inflammatory lung diseases (17). Corticosteroids act to reduce inflammation and cellular damage by two different mechanisms as follows: by direct binding of glucocorticoids to their consensus glucocorticoid response element site, which activates transcription processes, and by an indirect mechanism involving nuclear protein interactions (25). These latter effects may be mediated by an interaction between the glucocorticoid-receptor complex and transcription factors such as NF-B and AP-1, which regulate transcription for inflammatory mediators such as cytokines (17,25). We have shown that AP-1 plays a critical role in enhancing ␥-GCS-HS mRNA expression, and therefore regulation of GSH synthesis (15,26,27). Thus corticosteroids may have an effect on intracellular GSH through a mechanism involving AP-1 (28). However, the effects of corticosteroids on GSH synthesis at the molecular level have not been studied so far.
Thus the aim of this study was to elucidate the molecular regulatory mechanisms for GSH synthesis in human alveolar epithelial cells (A549) in response to TNF-␣ and dexamethasone, important pro-and anti-inflammatory agents, respectively.
When required for the assays, confluent monolayers of A549 epithelial cells were washed twice with phosphate-buffered saline (PBS). Thereafter, trypsin EDTA solution was added to detach the cells. The cells were then washed with DMEM, containing 10% FBS at 250 ϫ g for 10 min, to neutralize the trypsin, and were resuspended in DMEM with 10% FBS and maintained in 162-cm 2 cell culture flasks (Corning Costar, High Wycombe, UK).
Epithelial Cell Exposure to TNF-␣ and Dexamethasone-Monolayers of confluent A549 epithelial cells were prepared by seeding 0.8 ϫ 10 6 cells/well in a 6-well or 3 ϫ 10 6 cells/100-mm plates in DMEM with 10% FBS at 37°C, 5% CO 2 , until 70 -80% confluency was reached. Confluent monolayers were rinsed twice with DMEM and were treated with TNF-␣ (10 ng/ml) and/or dexamethasone (3 M) for time intervals between 4 and 24 h in 2 ml or 5 ml of full media, incubated at 37°C, 5% CO 2 . Thereafter, the monolayers were washed twice with cold PBS, and the cells were harvested with 0.05% trypsin/EDTA solution and resuspended in cold PBS (pH 7.4). An aliquot of cells was diluted with trypan blue, counted, and the cell viability determined. This cell suspension was thereafter used in the GSH and GSSG assays. For ␥-GCS activity, mRNA assays, and transfection experiments, monolayers were scraped using a Teflon scraper (Corning Costar, High Wycombe, UK).
GSH, GSSG, and ␥-GCS Assays-The epithelial cells were centrifuged at 250 ϫ g for 5 min at 4°C, and the cell pellets were suspended in 1 ml of cold 0.6% sulfosalicylic acid, sonicated on ice, homogenized with a Teflon pestle, and vortexed vigorously. The cell homogenates were then centrifuged at 4,000 ϫ g for 5 min at 4°C. The supernatant was immediately used in the GSH assay by the 5,5Ј-dithiobis-(2-nitrobenzoic acid)-glutathione reductase recycling method described by Tietze (29). For the GSSG assay, supernatant was treated with 2-vinylpyridine and triethanolamine as described previously (30) and thereafter was used in the assay for GSH as described above.
Cell extracts for ␥-GCS activity were prepared 4 or 24 h after each treatment. The cells were washed with ice-cold PBS, scraped into PBS, and collected by centrifugation at 250 ϫ g for 5 min at 4°C. After two additional washes with PBS, the cells were resuspended in 100 mM potassium phosphate buffer (pH 7.4), sonicated, and homogenized using a Teflon pestle on ice with Triton X-100 to a final concentration of 0.1% (v/v). The extracts were spun at 13,000 ϫ g for 15 min at 4°C. The supernatants were recovered and dialyzed at 4°C against four changes of 500 ml each of the Tris-HCl buffer (50 mM, pH 8.0) and then stored at 0°C. The protein concentrations were determined using the bicin-choninic acid reagent assay (Pierce, Chester, UK) (31).
␥-GCS activity was assayed by the method described by Seelig and Meister (12) using a coupled assay with pyruvate kinase and lactate dehydrogenase. The rate of decrease in absorbance at 340 nm was followed at 37°C. Enzyme-specific activity was measured as micromoles of NADH oxidized per min/mg protein, which is equal to 1 international unit (IU). For each experimental assay, BSO (50 M) was added to the cell homogenate and incubated for 1 h at 37°C to check the specificity of the ␥-GCS assay.
Isolation of RNA and Reverse Transcription-RNA was isolated from A549 cells using the TRIZOL reagent (Life Technologies, Inc., Paisley, UK). Total RNA was reverse-transcribed according to the manufacturer's instructions (Life Technologies, Inc., catalog number 8025SA). The resultant cDNA was stored at Ϫ20°C until required.
Assessment of ␥-GCS-HS mRNA by Polymerase Chain Reaction (PCR)-Oligonucleotide primers were chosen using the published sequence of human ␥-GCS-HS cDNA (32) and human glyceraldehyde-3phosphate dehydrogenase (GAPDH) (33) (Stratagene, Cambridge, UK). The primers for ␥-GCS-HS were synthesized by Oswel DNA Services, University of Southampton, UK (25,34). The sequences of the primers used in the PCR were as follows: ␥-GCS-HS (sense 5Ј-GTG GTA CTG CTC ACC AGA GTG ATC CT-3Ј) and (antisense 5Ј-TGA TCC AAG TAA CTC TGG ACA TTC ACA-3Ј); GAPDH (sense 5Ј-CC ACC CAT GGC AAA TTC CAT GGC A-3Ј) and (antisense 5Ј-TC TAG ACG GCA GGT CAG GTC AAC C-3Ј). Five microliters of the reverse-transcribed mRNA mixture was added directly to the PCR mixture and used for the PCR reactions, which we have previously described (26,27,34). Bands were visualized by a UV transilluminator, and photograph negatives were scanned using a white/ultraviolet transilluminator, UVP (Orme Technologies, Cambridge, UK). The intensity of the ␥-GCS-HS mRNA (531 bp) bands were expressed as a percentage of the intensity of the GAPDH bands (600 bp). The pKS-hGCS plasmid (American Type Culture Collection, Rockville, MD, catalog number 79023) was used as a positive control for ␥-GCS-HS. Fifty femtograms was used for each experiment to check the specificity of the PCR (data not shown).
Generation of the Mutant ␥-GCS-HS Promoter Construct-A selective mutation in the proximal AP-1-binding site (Ϫ269 to Ϫ263 bp) was introduced by PCR amplification with the mutated upstream primers. The sequences of the oligonucleotides used in the PCR were as follows: upstream 5Ј-ATGGTGAGTTCGTCATGTTATCAA-3Ј and downstream 5Ј-GCAGGCATGCCCAGTCTTTGCG-3Ј and were synthesized by MWG-BIOTECH GmbH, (Ebersberg, Germany). The mutated consensus sequence of AP-1 is denoted by the underlined letters, and the italic letters on the downstream oligonucleotide indicates additional bases from the cloning site of the pCAT Basic vector. The PCR conditions were 94°C for 10 min and then 35 cycles at 94°C for 30 s, 58°C for 30 s, 72°C for 120 s, and a final extension at 72°C for 10 min with 1 unit of Taq DNA polymerase (Life Technologies, Inc., Paisley, UK). The resulting PCR-amplified DNA fragment (Ϫ285 to ϩ47 bp) was confirmed by DNA sequencing. The mutated PCR fragment was isolated and subcloned into pCBGCS using HpaII and SphI restriction sites. This construct was denoted pCBmGCS.
Generation of a ␥-GCS-HS AP-1 Construct-Multiple restriction sites within the ␥-GCS-HS promoter were utilized to create a ␥-GCS-HS construct containing a putative proximal AP-1 construct. Digestion of the ␥-GCS-HS promoter using KpnI produced a short (Ϫ1050 to Ϫ818 bp) and a large fragment (Ϫ817 to ϩ82 bp). Further digestion of the large fragment by BalI and DraII generated a proximal fragment (Ϫ305 to ϩ82 bp). This fragment contains a putative AP-1 at Ϫ269 to Ϫ263 bp, CAAT and TATA boxes, which was subcloned into the pCAT Basic vector (pCBGC⌬D). Restriction enzyme analyses were performed to confirm the orientation and validity of all constructs.
Transient Transfection and CAT Assay-A549 cells (0.8 ϫ 10 6 per well) were seeded into 6-well tissue culture plates and cultured at 37°C until they were 70 -80% confluent. Plasmid DNA transfections were performed using the LipofectAMINE reagent (Life Technologies, Inc.). Following treatment with TNF-␣ (10 ng/ml) and/or dexamethasone (3 M), cell extracts were prepared and assayed for protein content using the BCA reagent (Pierce, Chester, UK) (31). Chloramphenicol acetyltransferase (CAT) activity was quantified by a CAT enzyme-linked immunosorbent assay. A ␤-galactosidase expression plasmid (PSVgal, Promega) was co-transfected as an internal control to normalize transfection efficiency. In all of the transfection experiments, pCAT-Basic and pCAT-Control were used as negative and positive controls, respectively.
Supershift Assay-The nuclear extracts were preincubated with 3 l of antiserum (1 mg/ml), at 4°C for overnight, before analysis by EMSA as described above. Human anti-c-Jun and anti-c-Fos sera were obtained from Santa Cruz Biotechnology, Inc. These sera specifically detect the presence of the corresponding transcription factor and do not interfere with nuclear factor binding. Rabbit preimmune sera (SAPU, Edinburgh, Scotland) were incubated with the nuclear extracts as described above and used as a control.
Statistical Analysis-Results were expressed as means Ϯ S.E. Differences between values were compared by Duncan's multiple range test.

Effect of TNF-␣ and Dexamethasone on GSH Levels in Alveolar Epithelial
Cells-TNF-␣ (10 ng/ml) significantly decreased GSH levels after 4 h treatment concomitant with an increase in GSSG levels in A549 epithelial cells. This was associated with a significant increase in GSH at 24 h, without any change in GSSG levels, compared with control values ( Fig. 1 and Table I). By contrast, dexamethasone (3 M) depleted intracellular GSH levels significantly at both 4 and 24 h. At 4 h dexamethasone produced no change in GSSG, but at 24 h GSSG was decreased compared with control levels ( Fig. 1 and Table I). However, exposure to dexamethasone produced a 31% decrease in the GSH/GSSG ratio at 4 h and an 24% increase at 24 h in A549 cells, compared with control values. Co-incubation of TNF-␣ and dexamethasone produced further depletion of GSH at 24 h in epithelial cells, compared with TNF-␣ or dexamethasone alone, without any significant change in the GSH/GSSG ratio ( Fig. 1 and Table I). Co-incubation of TNF-␣ and dexamethasone produced a similar decrease in GSH levels to TNF-␣ alone at 4 h, but with an 50% decrease in the GSH/GSSG ratio. Cell viability remained Ͼ95% after all of the above treatments.
Effect of TNF-␣ and Dexamethasone on ␥-GCS Activity in Alveolar Epithelial Cells-␥-GCS activity was not affected by TNF-␣ and/or dexamethasone treatment at 4 h, compared with control values (Fig. 2). However, TNF-␣ produced an increase in ␥-GCS activity at 24 h (Fig. 2). By contrast, dexamethasone significantly decreased ␥-GCS activity at 24 h (Fig. 2). Coincubation of TNF-␣ with dexamethasone produced a further decrease in ␥-GCS activity at 24 h, compared with dexamethasone alone. Addition of BSO (50 M) only inhibited 65-72% of the total ␥-GCS activity. The results were corrected for the BSO non-inhibitable enzyme activity after each treatment. The mean BSO-inhibitable ␥-GCS activity in the A549 cell homogenate was 0.065 Ϯ 0.01 units/mg protein.
Effect of TNF-␣ and Dexamethasone on ␥-GCS-HS mRNA Expression-We investigated the mechanism of the above effects on GSH and ␥-GCS activity following TNF-␣ and dexamethasone alone or in combination. Neither TNF-␣ nor dexa-   methasone alone or in combination produced any significant change in the ␥-GCS-HS mRNA level after 4 h, compared with control values (Fig. 3A). However, TNF-␣ significantly increased ␥-GCS-HS mRNA expression, whereas dexamethasone treatment significantly depleted ␥-GCS-HS mRNA, after 24 h, compared with GAPDH mRNA expression (Fig. 3B). The increased ␥-GCS-HS mRNA expression produced by TNF-␣ at 24 h was completely abolished when the cells were co-treated with dexamethasone.
Role of AP-1 in TNF-␣ and Dexamethasone-mediated Regulation of ␥-GCS-HS-To determine if AP-1 plays an important role in TNF-␣ and dexamethasone-mediated ␥-GCS-HS gene regulation in A549 epithelial cells, we used a DNA fragment of the 5Ј-flanking region of the ␥-GCS-HS containing a proximal putative AP-1 site and a commercially available DNA fragment containing an AP-1 consensus in the EMSA. Nuclear proteins were isolated at 4 and 24 h after TNF-␣ and/or dexamethasone treatment and were incubated with the DNA probes containing the AP-1 site. We found that TNF-␣ exposure increased AP-1 DNA binding activities in A549 cells using both the ␥-GCS-HS AP-1 probe and commercial AP-1 probe at 4 h exposure (Figs. 4 and 5), in A549 cells compared with untreated cells. Dexamethasone treatment alone did not produce any changes in the nuclear binding of ␥-GCS-HS AP-1 probe at 4 or 24 h, compared with control values. However, dexamethasone treatment produced a significant decrease in the nuclear binding using commercial AP-1 at 24 h, without any change at 4 h, in A549 cells. TNF-␣ exposure increased ␥-GCS-HS AP-1 DNA binding activity at 24 h, without any change in AP-1 binding of the commercial probe at 24 h in A549 epithelial cells. TNF-␣-mediated increase in AP-1 DNA binding was abolished when cells were co-treated with dexamethasone at 4 and 24 h (Figs. 4 and 5). The specificity of the binding was checked using 100-fold excess unlabeled ␥-GCS-HS AP-1 and AP-1 oligonucleotides and nonspecific oligonucleotides for NF-B (data not shown). addition of the c-Jun antibody shifted the AP-1 DNA band considerably (Fig. 6). Furthermore, both oligonucleotides were equally able to bind to c-Jun-purified protein (Promega) (data not shown).

Involvement of c-Jun in TNF-␣-induced AP-1 DNA Binding in A549 Epithelial
Role of NF-B in TNF-␣ and Dexamethasone-mediated Regulation of ␥-GCS-HS-We also determined whether the pleiotropic transcription factor NF-B was also affected by these treatments. Using a commercially available NF-B oligonucleotide, TNF-␣ treatment of A549 cells produced a significant increase in NF-B DNA binding both at 4 and 24 h (Fig. 7). By contrast, dexamethasone produced a significant decrease in NF-B DNA binding activity at 24 h, without any change at 4 h, compared with control levels. TNF-␣-mediated increased NF-B binding activity was not inhibited either at 4 or 24 h by co-treatment with dexamethasone.

Effect of TNF-␣ and Dexamethasone on ␥-GCS-HS Promoter Construct-derived Chloramphenicol Acetyltransferase (CAT)
Activity-To confirm that AP-1 plays a key role in TNF-␣induced ␥-GCS-HS gene expression, we used a CAT reporter system to explore the mechanisms by which TNF-␣ and dexamethasone regulate ␥-GCS-HS at the transcriptional level. TNF-␣ treatment of A549 cells transfected with the full ␥-GCS-HS promoter linked to the CAT reporter system (pCBGCS, Fig.  8A) produced a significant increase in CAT activity at 24 h, compared with control values (Fig. 8B). By contrast, CAT activity was significantly decreased when the cells were treated with dexamethasone at 24 h. Dexamethasone also blocked the TNF-␣-induced CAT activity at 24 h.
Effect of TNF-␣ and Dexamethasone on ␥-GCS-HS AP-1 Construct-derived CAT Activity-To establish whether the TNF-␣induced ␥-GCS-HS AP-1 DNA binding activity was related to activation of the ␥-GCS-HS AP-1 fragment in the CAT reporter system (pCBGCS⌬D), we transfected epithelial cells with a plasmid containing the putative AP-1 transcription factor (pCBGCS⌬D). pCBGCS⌬D displayed significant TNF-␣-induced CAT activity in cells transfected with pCBGCS⌬D, whereas dexamethasone inhibited the CAT activity (Fig. 8B). Furthermore, TNF-␣-induced pCBGCS⌬D CAT activity was abolished when the transfected cells were co-treated with dexamethasone.
To confirm further whether the proximal putative AP-1 site is a key regulator of TNF-␣-induced ␥-GCS-HS promoter activation, a construct was generated from the parent pCBGCS plasmid where the wild-type proximal AP-1 site was mutated (Fig. 9A). This mutated pCBmGCS reporter construct, where the proximal AP-1 site was abrogated, showed a reduced response to TNF-␣, compared with the wild-type ␥-GCS-HS promoter (Fig. 9B). These data imply that the proximal AP-1binding site present in Ϫ269 to Ϫ263 bp is the major cisregulatory element responsible for the TNF-␣ induction of the ␥-GCS-HS gene. Incubation with anti-c-Fos, anti-c-Jun sera before the gel shift assay results in a super-retardation in band mobility (arrow with ss). Experiments were repeated three times, and a representative autograph is shown.

DISCUSSION
Rapid induction of intracellular GSH synthesis occurs in response to various oxidant stresses (36 -38). This may be a critical determinant of cellular tolerance to oxidant stress. In this study, we demonstrated that TNF-␣ caused a transient depletion of GSH, concomitant with an increase in GSSG levels, suggesting that TNF-␣ induced oxidative stress in epithelial cells. However, prolonged exposure (24 h) to TNF-␣ increased GSH levels, without changing GSSG levels in A549 epithelial cells. The increase in GSH was associated with increased ␥-GCS activity. The levels of GSH and ␥-GCS activity are regulated by the expression of the catalytic ␥-GCS-HS (26,27). Recently, we and others (27, 36 -43) have demonstrated that oxidants, radiation, heat shock, heavy metals, and chemotherapeutic agents can increase GSH concentrations by induction of ␥-GCS-HS expression in various cell types. In this study, we show, for the first time, that in human alveolar epithelial cells (A549), ␥-GCS-HS gene expression is induced by TNF-␣. Similar induction of ␥-GCS-HS mRNA and elevation of GSH levels have been shown in human hepatocytes and mouse endothelial cells treated with TNF-␣ (44,45). Elevation of glutathione levels has also been observed in cultured rat hepatocytes following treatment with TNF-␣, which protected the cells against the cytotoxic effects of further oxidant stresses (46). However, the mechanism of this induction was not studied.
To understand the molecular mechanism of the transcrip-tional induction of ␥-GCS-HS in response to TNF-␣, we cloned the 5Ј-flanking region of the ␥-GCS-HS into a CAT reporter system. Following transfection into A549 epithelial cells, we observed that TNF-␣ up-regulated the promoter region of the ␥-GCS-HS gene, measured as increased CAT activity. This suggests that TNF-␣ acts at the level of transcription to induce ␥-GCS-HS mRNA in epithelial cells. The signaling mechanism whereby TNF-␣ exerts its effect is currently not known. TNF-␣ is known to generate ROS, particularly the superoxide anion (O 2 . ) and H 2 O 2 , by leakage from the electron transport chain in mitochondria. This could trigger transcriptional up-regulation of ␥-GCS-HS possibly by activating signaling pathways, such as activation of the c-Jun N-terminal kinase/stress activated protein kinase (47). We next elucidated the transcriptional regulatory mechanism by which TNF-␣ exerted its effect on the induction of ␥-GCS-HS. By using deletion and mutation studies, we have recently shown the critical role of a putative AP-1 site in oxidant-mediated regulation of the ␥-GCS-HS promoter (15). However, other investigators have suggested that an NF-Bbinding site in the ␥-GCS-HS promoter may be important in regulating ␥-GCS-HS gene expression (40,42). Activation of NF-B is known to be regulated by a variety of pro-inflammatory (e.g. TNF-␣) and anti-inflammatory agents (e.g. dexamethasone) (17). We therefore used TNF-␣ and dexamethasone to assess the role of the AP-1 and NF-B transcription factors in the transcriptional up-regulation of ␥-GCS-HS. Exposure of alveolar epithelial cells to TNF-␣ produced a significant increase in the DNA binding activities of both nuclear proteins, using AP-1 and NF-B as consensus probes. Thus both AP-1 and NF-B transcription factors are activated by TNF-␣. AP-1 activation by TNF-␣ was confirmed by increased CAT activity in a CAT reporter system with a cloned ␥-GCS-HS-AP-1 fragment. These data indicate that NF-B binding to a consensus site in the ␥-GCS-HS promoter is not necessary for TNF-␣induced transcriptional activation of the ␥-GCS-HS gene and suggest the involvement of an AP-1 response element. Further confirmation of the key role of a putative proximal AP-1-binding site in the regulation of the ␥-GCS-HS gene comes from the mutational study of the putative AP-1 site (Ϫ269 to Ϫ263 bp). Mutation of the sequence 5Ј-TTGATTCAA-3Ј to 5Ј-TGTTAT-CAA-3Ј in the proximal AP-1 oligonucleotide (pCBmGCS) effectively eliminated TNF-␣-induced promoter activity. This provides strong evidence that this sequence, but not NF-B, is involved in the regulation of the expression of the endogenous ␥-GCS-HS subunit gene.
To identify which components of AP-1 are responsible for the up-regulation of ␥-GCS-HS, we used antibodies directed against c-Fos and c-Jun in an effort to demonstrate a supershift in the EMSA. Only an antibody that cross-reacted with c-Jun produced a supershift, supporting the view that the AP-1 EMSA data is likely to have resulted from c-Jun/c-Jun homodimer binding to ␥-GCS-HS AP-1 sites. In support of this, several investigators have recently suggested possible involvement of AP-1/Jun family members in the regulation of ␥-GCS-HS (40,42,43,48).
Corticosteroids, such as dexamethasone, are known to suppress both immune responses and inflammation by activation of the glucocorticoid receptor and by interaction with various transcription factors (17). The role of intracellular GSH redox status in the regulation of transcription factors is of considerable interest (17,49). However, the effect of dexamethasone on the regulation of GSH synthesis has not been studied. We showed that depletion of intracellular GSH by dexamethasone occurs concomitant with a decrease in ␥-GCS activity, without any change in GSSG levels in A549 alveolar epithelial cells. However, the ratio of GSH/GSSG levels was decreased by dexamethasone, suggesting that dexamethasone may impose oxidative stress in A549 epithelial cells. Depletion of liver GSH in mice and inhibition of GSH synthesis by dexamethasone have been observed in a rat hepatic cell line (50,51). Similarly, depletion of antioxidant enzyme activities have been shown in indicates an additional 50 bp from multiple cloning sites of the pCRII vector. The structure of the ␥-GCS-HS-CAT plasmid is shown below on the right. The complete 5Ј-flanking region (pCBGCS) and a proximal putative AP-1 construct (pCBGCS⌬D) were ligated to the CAT gene in a pCAT Basic vector (pCB) and transfected into A549 epithelial cells. Transfected A549 cells were exposed to TNF-␣ (10 ng/ml) or dexamethasone (Dex, 3 M) alone or in combination. After 24 h incubation, the cells were harvested and assayed for CAT activity by an enzyme-linked immunosorbent assay. B, transcriptional activity was standardized by the amount of CAT activity relative to ␤-galactosidase activity. The results are shown as percentages of the CAT concentration compared with that of pCBGCS. The histograms represent the means and the error bars the S.E. of the CAT-derived activities of pCBGCS and pCBGCS⌬D constructs of three transfection experiments, each performed in duplicate with the activity of pCBGCS set at 100%. **, p Ͻ 0.01; ***, p Ͻ 0.001, compared with pCBGCS. †, p Ͻ 0.001, compared with TNF-␣ alone. various rat tissues by glucocorticoid supplementation, and this depletion was more pronounced when rats were challenged with oxidative stress (52). In this study, we show that the depletion of GSH, decrease in GSH/GSSG ratio, and decreased ␥-GCS activity produced by dexamethasone is also associated with a decrease in the expression of ␥-GCS-HS mRNA. However, dexamethasone had no effect on the basal level of ␥-GCS-HS gene expression at 4 h. In support of these data, dexamethasone also had no effect on basal level of manganese superoxide dismutase mRNA expression, whereas TNF-␣-induced gene expression was completely abolished by dexamethasone treatment (53).
Protein-protein interactions between the glucocorticoid re-ceptor and the transcription factor AP-1 are thought to mediate a negative cross-talk between dexamethasone and TNF-␣induced gene regulation (17,25). In this study, dexamethasone significantly inhibited TNF-␣-mediated activation of ␥-GCS-HS AP-1 DNA binding activity and proximal AP-1 (pCBGCS⌬D)-derived CAT activity in alveolar epithelial cells. Dexamethasone also inhibited TNF-␣-induced changes in GSH levels, ␥-GCS activity, ␥-GCS-HS mRNA expression, and ␥-GCS-HS promoter activity. Dexamethasone did not inhibit TNF-␣-mediated activation of NF-B but did block the increase in GSH and ␥-GCS-HS mRNA expression induced by TNF-␣. These data support the concept that activation of NF-B does not have a role in mediating the transcriptional activation of ␥-GCS-HS in response to TNF-␣. The results of our mutation and deletion studies confirm that the 5Ј-flanking proximal sequence of the ␥-GCS-HS gene, containing a putative AP-1-binding site at Ϫ269 to Ϫ263 bp, plays an important role in the transcriptional up-regulation of the ␥-GCS-HS gene in TNF-␣-treated alveolar epithelial cells. The supershift assay showed that this AP-1 DNA-binding complex was predominantly formed by dimers of c-Jun. Furthermore, our data provide supportive evidence for negative interaction between glucocorticoid receptor and AP-1 (c-Jun) which prevents TNF-␣-induced ␥-GCS-HS mRNA expression in A549 epithelial cells.
In conclusion, these studies show that TNF-␣ causes an increase in intracellular GSH content, ␥-GCS activity, and transcriptional activation of ␥-GCS-HS in alveolar epithelial cells, whereas dexamethasone decreased GSH levels by downregulating the transcription of the ␥-GCS-HS gene. Electrophoretic mobility gel shift and CAT reporter assays revealed that the modulation of ␥-GCS-HS gene expression in alveolar epithelial cells by TNF-␣ and dexamethasone occurs by a mechanism involving AP-1 (c-Jun homodimer). These data may have implications for dexamethasone treatment in patients with inflammatory lung diseases, since such treatment may prevent synthesis of increased levels of the protective antioxidant GSH.