Bronchial epithelium and its associated tissues act as a

Transcription

1 The Journal of Immunology A JNK-Independent Signaling Pathway Regulates TNF -Stimulated, c-jun-driven FRA-1 Protooncogene Transcription in Pulmonary Epithelial Cells 1 Pavan Adiseshaiah,* Dhananjaya V. Kalvakolanu, and Sekhar P. Reddy 2 * mong the several effectors that mediate TNF- action is AP-1, which consists of transcription factors belonging to the JUN and FOS families. Although the effects of TNF- in immune cells, such as the induction of NF-, are well known, the mechanisms by which it induces transcriptional activation of AP-1 in pulmonary epithelial cells are not well defined. In this study, we report that TNF- stimulates the expression of the FRA-1 protooncogene in human pulmonary epithelial cells using c-jun, acting via a 12-O-tetradecanoylphorbol-13 acetate response element located at 318. Although TNF- stimulates phosphorylation of c-jun, the inhibition of JNK activity had no significant effect on FRA-1 induction. Consistent with this result, ectopic expression of a c-jun mutant lacking JNK phosphorylation sites had no effect on the TNF- -induced expression of the promoter. In contrast, inhibition of the ERK pathway or ectopic expression of an ERK1 mutant strikingly reduced FRA-1 transcription. ERK inhibition not only blocked phosphorylation of Elk1, CREB, and ATF1, which constitutively bind to the FRA-1 promoter, but also suppressed the recruitment of c-jun to the promoter. We found that short interfering RNA-mediated silencing of FRA-1 enhances TNF- -induced IL-8 expression, whereas overexpression causes an opposite effect. Our findings collectively indicate that ERK signaling plays key roles in both Elk1, CREB, and ATF-1 activation and the subsequent recruitment of c-jun to the FRA-1 promoter in response to TNF- in pulmonary epithelial cells. The Journal of Immunology, 2006, 177: Bronchial epithelium and its associated tissues act as a primary interface for the interaction of a plethora of environmental stressors in the vertebrates. Acute lung injury caused by pathogenic and toxic products activates the synthesis of proinflammatory cytokines. These cytokines not only act directly on the lung cells themselves during the early phase of the response but also help recruit the cells of the immune system to alleviate the effects of the injury. Human pulmonary epithelial cells are known (1) to secrete many proinflammatory cytokines. High-level expression of these cytokines has been (2) causally linked to the development of pulmonary diseases, such as chronic obstructive pulmonary disease and asthma. These proinflammatory signals initiate intracellular signaling cascades, leading to an activation of various immediate early transcription factors, which then bind to target sequences commonly found in the regulatory regions of various cytokine and cytokine receptor genes and activate their transcription (3). One of the proinflammatory cytokines, TNF-, plays a critical role in diverse physiologic events and contributes to the development of air pollutant-induced lung pathogenesis and airway remodeling (4). Apart from NF- B, activation of immediate transcription factors such as AP-1 has been reported (3, 5, 6) to occur *Department of Environmental Health Sciences, Johns Hopkins University, Baltimore, MD 21205; and Greenbaum Cancer Center and Department of Microbiology and Immunology, University of Maryland School of Medicine, Baltimore, MD Received for publication April 17, Accepted for publication August 7, The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This study was supported by National Institutes of Health Grants ES11863 and HL66109 and (to S.P.R.) and by National Cancer Institute Grants CA and CA (to D.V.K.). 2 Address correspondence and reprint requests to Dr. Sekhar P. Reddy, Department of Environmental Health Sciences, Johns Hopkins University, Bloomberg School of Public Health, 615 North Wolfe Street, Room E7610, Baltimore, MD address: sreddy@jhsph.edu in other cell types in response to TNF-. AP-1 is a dimeric complex composed mainly of Jun (c-jun, JunB, and JunD), Fos (c-fos, FosB, Fra-1, and Fra-2), and ATF family proteins. Fos/Jun dimers bind to 12-O-tetradecanoylphorbol-13-acetate (TPA) 3 response elements (TREs, also known as AP-1 sites) and regulate the expression of genes involved in cell proliferation, inflammation, and pulmonary defenses (7, 8). A combinatorial interaction among the Jun, Fos, and ATF families of proteins has been shown (9) to regulate gene expression in a signal, cell-type-, and stressor-specific manner. The abundance and regulated autoinduction of certain members of the AP-1 family in response to specific stimuli control the duration and magnitude of a stress-related or mitogenic response (10). Consistent with this observation, overexpression of some AP-1 proteins results in various diseases associated with inflammation. For example, targeted expression of JunB in T lymphocytes promotes high levels of Th2 cytokines (11). Abrogation of JunB in keratinocytes triggers chemokine/cytokine expression, leading to the development of psoriasis, whereas abrogation of c-jun has the opposite effect (12). A role for JunD in T lymphocyte proliferation and cell differentiation has been reported (13). Given that specific members of this family are rapidly induced and the composition of AP-1 protein complex distinctly regulates gene expression, an understanding of the mechanisms of activation of Jun and Fos members is critical to our understanding of the molecular pathogenesis promoted by inflammatory stimuli. The mechanisms by which TNF- induces effector functions in immune cells are well recognized. However, it is unclear how this cytokine stimulates the activation of immediate response genes, such as transcription factors, that regulate subsequent expression of a variety of inflammatory mediators in pulmonary epithelial 3 Abbreviations used in this paper: TPA, 12-O-tetradecanoylphorbol-13-acetate; TRE, TPA response element; MMP, matrix metalloproteinases; WT, wild type; MEF, mouse embryonic fibroblast; rrna, rrna-encoding DNA; ChIP, chromatin immunoprecipitation; sirna, small interfering RNA; SRE, serum response element. Copyright 2006 by The American Association of Immunologists, Inc /06/$02.00

2 7194 INTERPLAY BETWEEN c-jun AND Elk1 AT THE FRA-1 PROMOTER cells. The FRA-1 was isolated as a TPA-inducible gene from monocytes, suggesting a role for this transcription factor in cell differentiation (14). Human T cell leukemia virus type 1 Tax1 activates the transcription of FRA-1 (15). Recently, we and others have shown that respiratory toxins that promote airway inflammation, such as cigarette smoke (16), asbestos (17), and diesel exhaust particles (18), strongly up-regulate the expression of FRA-1 in lung epithelial cells, suggesting a key role for this transcription factor in airway inflammation, injury, and repair processes. FRA-1 up-regulates the expression of several matrix metalloproteinases (MMPs), such as MMP-12 (19) and MMP-9 (18, 20 22), which are known to promote airway inflammation. Although the activation of c-fos by cytokines has been investigated in great detail (23 25) in cells of the immune system, the induction of FRA-1 by cytokines and its role in inflammatory responses in pulmonary epithelial cells are poorly understood. In this study, we report that JNK activation is not required for TNF- -induced, c-jun-mediated FRA-1 transcription in pulmonary epithelial cells. The induction occurs instead via an ERK signaling pathway through the activation of Elk1, CREB, ATF, and the subsequent recruitment of c-jun to the promoter. Materials and Methods Reagents and plasmids Abs specific for c-jun (SC-45X), FRA-1 (SC-605X), JNK1 (SC-474), ERK2 (SC-154), Elk1 (SC-355), and p-elk1 (Ser 383, SC-8406) were obtained from Santa Cruz Biotechnology. The -actin Ab and phosphospecific Abs for JNK (T 183 /Y 185 ), c-jun (Ser 73 ), and ERK (T 202 /Y 204 ) were all obtained from Cell Signaling Technology. Phosphospecific-CREB (Ser 133 ; catalog no ) and nonphosphorylated CREB (catalog no ) were purchased from Upstate Cell Signaling. Details concerning the various deletion and mutant FRA-1 promoter reporter luciferase constructs used in this study have been published elsewhere (26). Expression vectors coding for the c-jun mutant (c-jun TAM), Elk1 mutant (dn-elk1), SRF mutant (SRF-mt), ATF1 mutant (ATF1-mt), and CREB mutants (CREBmt) used in this study are detailed in our earlier publication (26). Plasmid constructs of wild-type (WT) c-jun and mutant c-jun lacking JNK phosphorylation sites, serines 63 and 73 and threonines 91 and 93 (27), were gifts from W. G. Kaelin, Jr. (Harvard Medical School, Boston, MA). The 165- to 19-bp promoter of human IL-8 (IL-8-Luc), which contains the functional motifs, such as AP-1 and NF- B sites (21), fused to luciferase (Luc) gene was described elsewhere (28). Cell culture A549, a human alveolar type II-like epithelial cell line, was maintained in RPMI 1640 medium supplemented with 5% FBS and antibiotics (Invitrogen Life Technologies). The primary human bronchial epithelial cells were cultured in MEM supplemented with growth factors according to the supplier s recommendation (Cambrex). Mouse embryonic fibroblasts (MEFs) lacking the erk1 gene (erk1 / ) and their isogenic WT cells (29) were cultured as previously described (30). To generate stable cell lines that overexpress FRA-1, A549 cells were transfected with FRA-1 wild-type cdna (gift from E. Tulchinsky, University of Leicester, U.K.) or an empty pcmv mammalian expression vector containing a selection marker neomycin. Stable cell clones overexpressing FRA-1 (referred to as A549-F1 cells) or a control empty vector (referred to as A549-C) were isolated following selection with 600 g/ml G418 (Invitrogen Life Technologies), pooled, and used for subsequent gene expression and functional studies. Northern and Western blot analyses For Northern blot analysis, cells were serum-starved for 24 h and subsequently treated with TNF- (10 ng/ml) for various times as indicated. Total RNA (15 g/lane) was separated on a 1.2% agarose gel, blotted onto a nylon membrane, and hybridized with 32 P-labeled cdnas of FRA-1 and 18S RNA as previously described (31). For Western blot analysis, total protein was extracted using a lysis buffer consisting of 20 mm Tris (ph 7.5), 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 1% Triton X-100, 2.5 mm sodium pyrophosphate, 1 mm Na 3 VO 4,5mM -glycerophosphate, and 1 g/ml leupeptin. A comparable quantity of protein from each sample was separated on a 10% SDS-PAGE, and the membranes were probed with specific Abs (Santa Cruz Biotechnology). Real-time RT-PCR TaqMan gene expression assays for mouse and human FRA-1, c-jun, and GAPDH were purchased from Applied Biosystems, and mrna levels were quantified in triplicate according to the supplier s recommendations. The absolute values for FRA-1 and c-jun were normalized to that of GAPDH. The relative value from the vehicle-treated control group was considered equal to one arbitrary unit. IL-8 and IL-6 expression was analyzed by a LightCycler (Roche) using the SYBR Green QuantiTech RT-PCR kit (Qiagen). Primersequenceswere: IL-8sense, GTTTTTGAAGAGGGCTGA GAATTC; IL-8 antisense, CATGAAGTGTTGAAGTAGATTTGC T; IL-6 sense, GGCAGAAAACAACCTGAACCT TC; IL-6 antisense, ACCTCAA ACTCCAAAAGACCAGTG; and 18S rrna-encoding DNA (rrna) sense, GTAACCCGTTGAACCCCATT; 18S rrna antisense, CCATCCAATC GGTAGTAGCG. The reaction was performed in a 20- l final volume consisting of 25 ng of total RNA (IL-8 and IL-6) or 2.5 ng of total RNA (for 18S rrna), 10 l of QuantiTech SYBR Green PCR mastermix (Qiagen), and 1.5 mm primers. Negative controls without template were included in all of the RT-PCR. Quantification of IL-8 and IL-6 mrna in each sample was normalized to the abundance of its corresponding 18S rrna in each sample. Transient transfection assays Cells were transfected with 100 ng of IL-8-Luc or FRA-1 promoter reporter construct, 1 ng of Renilla luciferase (prl-tk) vector (Promega), and ng of empty or expression plasmids. At h posttransfection, cells were serum-starved for 24 h and then treated with vehicle or TNF-. Cell extracts were assayed for firefly and Renilla luciferase activities using a commercially available kit (Promega). Luciferase activity of individual samples was normalized to that of Renilla luciferase activity (31). EMSAs Serum-starved cells were treated with TNF- for 60 min, nuclear extracts were prepared, and the EMSAs were performed using 2 3 g of nuclear extract and 32 P-labeled double-stranded 318 TRE oligonucleotide as a probe, as described previously (26). For supershift assays, nuclear extracts were incubated with 1 2 g of specific Abs or nonimmune IgG on ice for 2 h before the addition of labeled probe. Chromatin immunoprecipitation (ChIP) assays ChIP assays were conducted as described earlier (32): Cells ( ) were exposed to TNF- for 60 min, and ChIP was performed using a commercially available kit (Upstate Biotechnology). Chromatin was crosslinked by adding formaldehyde (1%) to the tissue culture medium for 10 min at 37 C. A fraction of the soluble chromatin (1%) was saved for measurement of total chromatin input. Precleared chromatin was incubated with specific Abs for 18 h at 4 C. DNA recovered from the immunoprecipitated products was used as a template for PCR with FRA-1 promoterspecific primers (32). After cross-linking and immunoprecipitation, purified DNA isolated from MEFs was subjected to PCR amplification for 40 cycles using primers specific for fra-1 promoter (GenBank accession no. AF017128): forward primer ( 208/ 185), 5 -GCGGAGCTCGGCCACA GGATTTTGTTTCGCCCT-3 and reverse primer ( 44/ 64), 5 -GGC GCTAGCCCTCTGACGCAGCTGCCCAT-3. PCR was performed at 95 C for 5 min, followed by 40 cycles of 95 C for 30 s, 55 C for 45 s, and 72 C for 1 min, with a final extension at 72 C for 10 min. The amplified 165-bp DNA fragment was separated on gel electrophoresis. Small interfering RNA (sirna) and gene expression analysis SMARTpool sirna duplexes for c-jun (catalog no. M ) and a scrambled sirna (catalog no. D ) were purchased from Dharmacon. To silence the endogenous FRA-1 expression, a plasmid-based expression vector, prnatin-h1.2/neo (GenScript) containing FRA-1 sirna sequence, GGATCCCGCTGACTGCCACTCATGGTGCCACACCCACCA TGAGTGGCAGTCAGTTTTTTCCAAAAGCTT, was used. Empty vector was used as a control. A549 cells at 30 40% confluence were transfected with sirnas at 20 nm concentrations and were harvested at h to determine the effect of sirna on the expression of endogenous c-jun and FRA-1 using Western blotting. For reporter assays, A549 cells were transfected with 100 ng of the 379-Luc promoter-reporter construct along with c-jun or scrambled sirnas for 36 h. Cells were serum-starved overnight before stimulation with TNF-, and luciferase activity was measured as described above. Statistical analysis Data are expressed as means SE. Statistical significance was determined using t tests and accepted at p All assays were performed using two or three (n 2 3) independent samples, and each experiment was repeated at least two times.

3 The Journal of Immunology 7195 FIGURE 1. FRA-1 induction by TNF- in pulmonary epithelial cells. A, A549 cells were serum-starved for 24 h and then treated with TNF- (10 ng/ml) for various times as indicated. Northern blot analysis was conducted using a 32 P-labeled FRA-1 cdna probe. Arrows, Positions of the 3.3- and 1.7-kb transcripts of FRA-1. The membrane was stripped and probed with 28S RNA cdna to monitor equal RNA loading of all samples. B, Cells were treated with actinomycin D (Act-D, 10 g/ml) or vehicle (DMSO) for 30 min before stimulation without (Con) or with TNF- for 90 min. FRA-1 and 28S RNA expression was analyzed as detailed above. A representative blot of two independent experiments is shown for A and B. C, Cells were transfected with various FRA-1 promoter reporter constructs as indicated along with a reference plasmid, prl-tk. After overnight incubation, cells were serum-starved for 24 h and then stimulated without ( ) or with TNF- (f) for 5 h. The promoter activity of the constructs was expressed using the basal activity of the 861-Luc as one unit. The fold activation of the individual reporters was calculated with the basal values of the respective construct set to one. The data represent the values of three independent samples of a representative experiment, which was repeated at least twice to obtain reproducible results. Results TNF- -induced FRA-1 expression in pulmonary epithelial cells To understand the regulation of FRA-1 expression by cytokines, A549 cells were treated with TNF- for min, RNA was isolated, and Northern blot analysis was performed using a 32 P- labeled human FRA-1 cdna as probe. As shown in Fig. 1A, TNF- significantly stimulated FRA-1 mrna expression as early as 30 min; the levels reached a maximum by 90 min and remained elevated through 360 min. The induction of FRA-1 mrna expression by TNF- was also correlated with a corresponding increase in its protein levels and was determined by Western blotting (data not shown). The two alternatively spliced mrna transcripts of FRA-1 were induced similarly, as previously reported (14). However, pretreatment of cells with actinomycin D, an inhibitor of transcription, blocked TNF- -stimulated FRA-1 expression (Fig. 1B), indicating that the induction was mainly regulated at the transcriptional level. To map the promoter region required for TNF- -inducible transcription, promoter-reporter constructs bearing various lengths of the 5 -flanking region of FRA-1 were transfected into A549 cells, and reporter gene expression was monitored (Fig. 1C). Consistent with our previous results, the 283-Luc yielded an 4-fold higher basal activity when compared with the 105-Luc and 68-Luc constructs. However, the levels of reporter expression following TNF- stimulation were unchanged, suggesting these constructs lack the cis-elements required for the induction. In contrast, the 328-Luc construct bearing the serum response element (SRE) had a 2-fold higher luciferase activity in response to TNF- (Fig. 2B). However, nearly a 5- to 7-fold rise in promoter activity was noticed with the 379-Luc, 570-Luc, and 861-Luc constructs, suggesting that the DNA sequences spanning 379 and 283 regulate the induction by TNF-. The 318 TRE mediates c-jun-dependent, TNF- -inducible FRA-1 promoter activity The 379/ 283 region harbors functional elements such as 318 TRE (26) and SRE (32). Because AP-1 acts as a major downstream effector of TNF- -induced signaling, we first examined the role of 318 TRE in mediating cytokine-inducible FRA-1 transcription. Disruption of 318 TRE site markedly reduced TNF- -inducible promoter activity (Fig. 2A). Consistent with this result, ectopic expression of a c-jun mutant lacking the transactivation domain greatly reduced ( 80%) TNF- -stimulated promoter activity (Fig. 2B). Furthermore, the knockdown of c-jun expression by sirna strongly reduced TNF- -stimulated luciferase activity (Fig. 2C, bar 2), when compared with control scrambled sirna (Fig. 2C, FIGURE 2. TRE mediates TNF- -stimulated FRA-1 transcription. A, The position of the 318 TRE of the 379-Luc construct is shown. Mutations were introduced into the 318 TRE of the 379-Luc as previously described (32). Cells were transfected with 100 ng of 379-Luc and 318 TRE mutant construct (379-TRE mt) in the presence of the prl-tk plasmid. The TNF- -inducible promoter activity was expressed as fold change over the activity of the respective constructs in unstimulated cells. B, Cells were transfected with 100 ng of empty or c-jun mutant expression vectors to determine the role of c-jun in TNF- -stimulated FRA-1 transcription. Fold induction was calculated with the values for empty parental vector-transfected cells set to one. C, Cells were transfected with the 379-Luc construct and prl-tk plasmid in the presence of scrambled (Scr) or c-jun sirna (20 nm) sequences as previously described (32). After 36 h of transfection, cells were serum-starved and then treated without or with TNF- for 5 h. The promoter activity (-fold activation) was calculated with the value for unstimulated cells set to one unit. Values are mean SE from two independent experiments conducted in duplicate (n 4). D, To confirm the knockdown of endogenous c-jun expression, sirna transfected cells were lysed in parallel experiments and immunoblotted with antic-jun and tubulin Abs. Lane 1, Scr-SiRNA and lane 2, c-jun sirna.

4 7196 INTERPLAY BETWEEN c-jun AND Elk1 AT THE FRA-1 PROMOTER FIGURE 3. Analysis of c-jun binding at the FRA-1 promoter. A, ChIP analysis of c-jun binding to the endogenous FRA-1 promoter. The arrows indicate the positions of the primers flanking 318 TRE that were used in the ChIP assays. Cells were treated with TNF- for 0, 30, or 60 min, and then chromatin protein-dna complexes were cross-linked using formaldehyde. The purified nucleoprotein complexes were immunoprecipitated with c-jun Abs or nonimmune IgG and amplified by PCR as detailed in Materials and Methods. Experiments were repeated at least twice to obtain reproducible results. B, The quantification of the amplified band with c-jun Ab normalized against the input reference band. bar 1). The c-jun-specific sirna markedly suppressed endogenous c-jun protein levels by 80% (Fig. 2D, lane 2), when compared with scrambled sirna (Fig. 2D, lane 1). The expression level of tubulin was comparable between these two samples, confirming a specific inhibitory effect of c-jun sirna. The level of expression of c-jun was similar for a scrambled sirna and reagent control (data not shown). These results collectively indicate a requirement for c-jun in TNF- -stimulated FRA-1 expression in pulmonary epithelial cells. c-jun is recruited to the FRA-1 promoter following TNF- stimulation We performed ChIP assays to examine the binding of c-jun to the 318 TRE of the FRA-1 promoter in vivo following TNF- stimulation (Fig. 3A). In the unstimulated state, c-jun bound only minimally to the FRA-1 promoter (Fig. 3A, lane 1). However, TNF- induced the binding of c-jun to the promoter as early as 30 min (Fig. 3A, lane 2), and it remained high through 60 min (Fig. 3A, lane 3). We chose these time points because FRA-1 message levels were maximal at min after TNF- stimulation (Fig. 1A). In contrast, ChIP assays of nonimmune IgG showed no amplification of the FRA-1 promoter. Quantification of c-jun binding revealed a nearly 8- to 12-fold increase in the induced binding of c-jun to the 318 TRE after TNF- stimulation (Fig. 3B). Collectively, these results support a critical role for c-jun in controlling TNF- induced FRA-1 transcription. TNF- -stimulated c-jun expression precedes and is essential for subsequent FRA-1 induction We next examined the role of c-jun in this process. We measured the message levels of c-jun and FRA-1 by real-time PCR following TNF- stimulation at 30 and 90 min. TNF- treatment increased the expression levels of c-jun after as little as 30 min, and the levels remained elevated up to 90 min (Fig. 4A, left panel). In contrast, FRA-1 induction by TNF- peaked between 30 and 90 min (Fig. 4A, right panel). The increase in c-jun and FRA-1 mrna expression was confirmed at the protein level by Western blot analysis using -actin as a loading control (Fig. 4B). We next examined the role of c-jun in controlling FRA-1 expression using sirnas. Cell cultures were transfected with c-jun or a controlscrambled sirna and then stimulated with TNF-. Total RNA was isolated, and FRA-1 expression was measured by real-time PCR (Fig. 4C). Transfection of c-jun sirna significantly diminished TNF- -stimulated FRA-1 expression (Fig. 4C, bar 4), when compared with the scrambled sirna control (Fig. 4C, bar 2). These results (Figs. 3 and 4) demonstrate a requirement for c-jun for TNF- -stimulated FRA-1 induction in pulmonary epithelial cells. JNK signaling is not essential for TNF- -inducible FRA-1 expression TNF- stimulates the activation of the JNK pathway, and c-jun acts as a major downstream effector of JNK kinases in other cell types. We therefore asked whether the JNK pathway was necessary for TNF- -induced expression of FRA-1. Cells were serumstarved for 24 h, then treated with TNF-, and JNK1/2 activation was assessed using phosphospecific Abs (Fig. 5A). As anticipated, TNF- strongly stimulated the phosphorylation of JNK1/2 (Fig. 5A). However, pretreatment of cells with the JNK inhibitor SP suppressed TNF- -stimulated JNK1/2 activation (Fig. 5B, compare lane 4 and lane 1). In contrast, SP did not inhibit ERK1/2 phosphorylation. In contrast, treatment of cells with the ERK1/2 and p38 MAPK pathway inhibitors PD98059 (Fig. 5B, lane 2) and SB (Fig. 5B, lane 3), respectively, had no effect on TNF- -stimulated JNK1/2 activation (Fig. 5B, lane 4). To examine the role of JNK signaling in TNF- -stimulated FRA-1 expression, RNA was isolated from cells stimulated with TNF- in the presence or absence of SP600125, and a Northern blot analysis was performed. As shown in Fig. 5C, JNK inhibition had no effect on the TNF- induced expression of FRA-1. Similar results were obtained with primary cultures of human bronchial epithelial cells (Fig. 5D). Above results suggest that the activation of the JNK pathway is not essential for TNF- -stimulated FRA-1 expression. To further FIGURE 4. c-jun is required for TNF- -stimulated FRA-1 expression. A, Cells were stimulated with TNF- for 0 90 min, total RNA was isolated, and c-jun and FRA-1 mrna expression was analyzed by real-time PCR. Bars, Mean SE of triplicates. B, Cell extracts (40 g) isolated from cells treated with TNF- as detailed above were immunoblotted using the c-jun, FRA-1, and tubulin Abs. A representative blot of two independent experiments is shown. C, Cells were transfected with scrambled (Scr) or c-jun (20 nm) sirna sequences as detailed in Fig. 2C, and then stimulated without ( ) or with TNF- (f) for 90 min. Total RNA was isolated and FRA-1 mrna expression quantified by real-time PCR. Bars, Mean SE (n 4). *, p 0.05.

5 The Journal of Immunology 7197 FIGURE 5. Effect of JNK pathway inhibition on TNF- -induced FRA-1 expression. A, Cells were stimulated with TNF- for 0 60 min, cell extracts were isolated, and JNK1/2 activation was analyzed using phosphospecific Abs. Membranes were stripped and probed with JNK2 Abs. B, Cells were incubated with ERK inhibitor PD98059 (PD, 30 M), p38 inhibitor SB (SB, 10 M), or JNK inhibitor SP (SP6, 10 M) for 30 min and then treated with TNF- for 30 min. DMSO was used as vehicle control. Cell extracts (40 g) were analyzed by Western blotting using the phosphospecific JNK1/2 Abs. Membranes were stripped and subsequently probed with phosphospecific ERK1/2 and total ERK2 Abs. A representative blot of two independent experiments is shown. C, Cells were incubated with SP (SP6, 10 M) for 30 min and then treated with TNF- for 90 min, and FRA-1 mrna expression was analyzed by realtime PCR. Bars, Mean SE (n 6). D, Primary cultured human bronchial epithelial (PHBE) cells were serum-starved for 2 h and then treated with SP (SP6) before stimulation with TNF- for 90 min. FRA-1 mrna expression was analyzed by real-time PCR. Bars, Mean SE (n 4). confirm this hypothesis, we examined the JNK-mediated phosphorylation of c-jun at Ser 73 in response to TNF- under our experimental conditions. As anticipated, TNF- markedly stimulated c-jun expression (Fig. 6A) and its phosphorylation (Fig. 6B). Pretreatment of cells with SP inhibited TNF- -stimulated JNK MAPK activation and the subsequent c-jun phosphorylation. To rule out a role for JNK phosphorylation in c-jun-dependent FRA-1 transcription, we transiently transfected cells with a c-jun mutant ( JNK c-jun) lacking JNK phosphorylation sites, Ser 63 and Ser 73 and Thr 91 and Thr 93 (27) (Fig. 6C, bar 2), then compared FRA-1 promoter activation to that of the WT construct (Fig. 6C, bar 3). Ectopic expression of the c-jun mutant robustly stimulated FRA-1 promoter activity to a level equivalent to that of the WT protein. Moreover, the c-jun mutant had no effect on TNF- -induced reporter expression (Fig. 6D, bar 3). Collectively, these results indicate that JNK1/2 signaling and c-jun phosphorylation do not contribute to TNF- -stimulated FRA-1 expression. The ERK1/2 pathway is essential for TNF- -induced FRA-1 expression A critical role for ERK1/2-dependent control of toxin- and mitogen-stimulated FRA-1 expression has been demonstrated (16) in lung epithelial cells. We, therefore, examined the role of this pathway in TNF- -stimulated FRA-1 expression. Cells were treated with TNF- for various times, and ERK1/2 kinase activation was determined by Western blot analysis with Abs specific for phosphorylated (active) forms of ERK1/2 (Fig. 7A). The TNF- -stimulated phosphorylation of ERK1/2 kinases was robust at 15 min FIGURE 6. Effect of a c-jun mutant lacking JNK phosphorylation sites on TNF- -induced FRA-1 promoter activity. A, The extracts isolated from cells stimulated with TNF- at various points were probed with phosphospecific c-jun (Ser 73 ) Abs. Membranes were stripped and probed with c-jun Abs. B, The extracts isolated from cells stimulated with TNF- in the presence of MAP kinase inhibitors, PD98059, SB202190, and SP600125, were analyzed by Western blotting using the phosphospecific c-jun (Ser 73 ) or JNK1/2 Abs. Membranes were stripped and probed with native c-jun and tubulin Abs. A representative blot of two independent experiments is shown. C and D, Cells were transfected with the 379-Luc promoter reporter construct (0.1 g) in the presence of the parental empty vector (vector), c-jun WT vector (c-jun), or a mutant form of c-jun lacking JNK phosphorylation sites ( JNK-c-Jun). Both basal (C) and inducible (D) promoter activities were determined after normalization to the value of the empty parental vector-transfected cells, which was set to 100%. Bars, Mean SE of triplicates of a representative experiment. (Fig. 7A, lane 2) but returned to basal levels thereafter. As shown in Fig. 7B, treatment of cells with the ERK inhibitor PD98059 completely blocked TNF- -stimulated ERK1/2 activation. Next, we analyzed the effect of the ERK inhibition on TNF- enhanced FRA-1 transcription. PD98059 markedly blocked TNF- -stimulated FRA-1 mrna expression (Fig. 7C). A similar result was obtained with another MEK-ERK pathway-specific inhibitor, U0126. These results were further confirmed at the transcription level using FRA-1 reporter constructs in transient transfection assays (Fig. 7D). TNF- strongly stimulated FRA-1 promoter activity (Fig. 7D, bar 1), but this stimulation did not occur in the presence of ERK inhibitor PD98059 (Fig. 7D, bar 2), supporting a role for ERK signaling in controlling FRA-1 induction by TNF-. To further assess the importance of ERK1 signaling in the regulation of FRA-1 induction in lung epithelial cells, A549 cells were transfected with the dominant-negative ERK1 (dn-erk1) plasmid, and TNF- -stimulated FRA-1 promoter activity was analyzed. A control transfection with empty expression vector was used for comparison. Overexpression of dn-erk1 significantly inhibited both basal and TNF- -stimulated FRA-1 promoter-driven reporter expression, as compared with empty vector-transfected, TNF- -treated cells (Fig. 7E). Taken together, these results strongly support a critical role for ERK signaling in controlling TNF- -induced FRA-1 transcription.

6 7198 INTERPLAY BETWEEN c-jun AND Elk1 AT THE FRA-1 PROMOTER FIGURE 7. The ERK pathway regulates TNF- -induced FRA-1 promoter activity. A, Cells were stimulated with TNF- for various times, and extracts were analyzed by Western blotting with ERK1/2 Abs. B, Western blot showing the effect of PD98059 (20 M) on TNF- -stimulated ERK1/2 activation. C, Cells were incubated with the ERK inhibitors PD98059 (30 M) and UO126 (10 M) for 30 min and then stimulated without ( ) or with TNF- (f) for 90 min. FRA-1 mrna expression was analyzed by real-time PCR. Bars, Mean SE (n 6). D, Effect of PD98059 on TNF- -stimulated FRA-1 promoter activity. Cells were transfected with the 379- Luc reporter (100 ng) and prl-tk (1 ng). The fold-activation of was calculated with the basal values of the respective DMSO or PD98059 treated samples as one unit. Bars, Mean SE (n 4). E, Cells were transfected with the 379-Luc reporter and prl-tk along with an equimolar amount of empty vector or dominant negative ERK1 (dn-erk1) plasmid, and the TNF- -stimulated FRA-1 promoter activity was analyzed with the value for empty vector-transfected cells set to one. *, p Inhibition of the ERK pathway suppresses TNF- -induced Elk1 and CREB phosphorylation To determine the downstream effector mechanisms by which ERK signaling controls FRA-1 induction by TNF-, we focused our studies on the Elk1 and CREB transcription factors that are targets of ERK signaling and are known (32) to regulate the induction of FRA-1 in response to tumor promoters and mitogens. As anticipated, TNF- stimulated the phosphorylation of Elk1, CREB, and ATF-1 after as little as 15 min (Fig. 8A, lane 2), and this stimulation was decreased in the presence of PD98059 (Fig. 8A, lanes 6 and 7). To examine the role of Elk1 and ATF/CREB proteins in the transcriptional up-regulation of FRA-1 by TNF-, cells were transfected with the reporter constructs bearing a mutation in the Elk1 binding site TCF and the ATF/CREB binding site of the FRA-1 promoter (Fig. 8B). The CArG element, flanking these sites, has been shown (33) to be critical for efficient binding of Elk1 to the SRE. We, therefore, examined the impact of mutations in the CArG element on TNF- -induced FRA-1 promoter activity. Mutation of the individual TCF site, the CArG box, or the ATF site significantly diminished ( 50%) TNF- -induced reporter expression, when compared with the results for the WT construct that lack these mutations (Fig. 8B). To further confirm the role of these transcription factors, we transfected cells with plasmids coding for dominant-negative mutants of the SRF, Elk1, ATF1, and CREB proteins. Coexpression of SRF mutant or an Elk1 mutant significantly repressed TNF- -induced FRA-1 promoter activity (Fig. 8C). Ectopic expression of the ATF1 and CREB mutants had a similar effect on reporter gene expression (Fig. 8D). These results collectively suggest that SRF and TCF proteins, such as Elk1, ATF1, and CREB, regulate TNF- -stimulated FRA-1 expression through the SRE (TCF and CArG) and the ATF sites located in the enhancer region. ERK1 kinase regulates FRA-1 induction by TNF- To examine the role of the ERK1 pathway in FRA-1 induction, we used MEFs lacking the erk1 gene (erk1 / ) and compared the magnitude of fra-1 induction by TNF- in these cells to that in isogenic WT cells. We examined the activation of the ERK1/2 pathway in these two cell types by Western blot analysis using phosphospecific Abs. As shown in Fig. 9A, TNF- strongly stimulated both ERK1 and ERK2 phosphorylation in WT cells (cf lanes 1 and 2). As expected, there was no ERK1 activation in FIGURE 8. The SRE is essential for TNF- -induced FRA-1 promoter activity. A, Cells stimulated with TNF- for 0, 15, and 30 min (lanes 1-3) and the extracts were analyzed by Western blotting using phosphospecific Elk1 (Ser 383 ) and CREB (Ser 133 ) Abs. Membranes were stripped and probed with total Elk1 Abs to monitor equal loading. A (right panel, lanes 4-7), cells were treated with PD98059 and then stimulated with TNF- for 30 min, and Elk1 and CREB/ATF1 activation was analyzed. DMSO was used as a vehicle control (lanes 4 and 5). Note that CREB (Ser 133 ) Abs also recognize the phosphospecific form of ATF1 (Upstate Cell Signaling). B, A549 cells were transfected with the FRA-1 promoter-reporter constructs bearing mutations within the TCF site, the CArG box, and the ATF site. prl-tk plasmid was used as an internal control. C and D, Cells were transfected with the 379-Luc along with the prl-tk plasmid in the presence of an equimolar amount of either empty vector, mutant SRF (dn-srf), or Elk1 (dn-elk1) plasmid (C). The effects of mutant ATF1 (dn-atf-1) and CREB (dn-creb) plasmids on TNF- -stimulated FRA-1 promoter is shown in D. TNF- -induced promoter activity (-fold activation) was calculated with the value for unstimulated cells set to one unit. Data shown are mean SE of triplicates from a typical experiment., Statistically significant difference at p 0.05.

7 The Journal of Immunology 7199 FIGURE 9. ERK1 is critical for TNF- -stimulated FRA-1 induction. A, The activation of ERK1/2 kinases by TNF- in the WT and erk1 / MEFs. The MEFs were serum-starved for 2 h and then stimulated with TNF- or TPA (10 ng/ml) for 15 min, and ERK activation was analyzed using phosphospecific Abs. B, The effects of ERK1 deficiency on endogenous fra-1 mrna expression. MEFs were stimulated without ( ) or with TNF- (f) for 90 min, and fra-1 mrna expression was analyzed by real-time PCR. Bars, Mean SE (n 6). C, After transfection with the 379-Luc FRA-1 promoter construct, WT and erk1 / MEFs were treated with TNF-, and luciferase activity was analyzed. erk1 / MEFs, which lack this gene (Fig. 9A, lane 4). In transient transfection assays, TNF- strongly stimulated FRA-1 promoter activity in WT MEFs as compared with the erk1 / MEFs (Fig. 9B). Furthermore, the ERK inhibitor PD98059 repressed TNF- stimulated FRA-1 promoter activity in both cell types. We further validated these results at the level of endogenous fra-1 expression levels by real-time PCR (Fig. 9C). Treatment of cells with TNF- stimulated fra-1 mrna expression in WT MEFs. However, the magnitude of the induction was greatly diminished in the erk1 / MEFs when compared to WT cells. To further validate these results, we treated WT cells with the MEK-ERK pathway-specific inhibitors PD98059 and U0126 and examined the endogenous fra-1 expression. Treatment of WT MEFs with either PD98059 or U0126 obliterated the TNF- -stimulated response. These results collectively indicate a prominent role for ERK signaling, especially ERK1, in regulating fra-1 induction by TNF-. ERK1/2 signaling is essential for the recruitment of c-jun to the FRA-1 promoter We have previously shown (32) that mutations in the 318 TRE or TCF and CArG sites of SRE as well as the flanking ATF site ablate the mitogen-induced FRA-1 promoter activity. Because inhibition of the JNK pathway did not block c-jun activation, and recruitment of c-jun following TNF- stimulation precedes FRA-1 induction, we wondered whether inhibition of the ERK pathway affects the recruitment of c-jun at the promoter. For this purpose, we exposed cells to TNF- for 60 min in the presence or absence of the ERK inhibitor UO126, which completely blocks FRA-1 induction, and analyzed the recruitment of c-jun using ChIP assays as detailed in Materials and Methods. As shown in Fig. 10, TNF- strongly enhanced the binding of c-jun at the promoter (lanes 2 and 3). However, pretreatment of cells with ERK inhibitor before TNF- stimulation markedly reduced ( 80%) the recruitment of c-jun at the FRA-1 promoter (Fig. 10, lanes 5 and 6). In a complementary experiment, we performed a ChIP analysis to determine whether the lack of erk1 / altered the recruitment of c-jun at the endogenous fra-1 promoter in MEFs. The WT and erk1 / MEFs were stimulated with or without TNF-, DNAprotein complexes were cross-linked, and ChIP assays were performed using mouse fra-1 promoter-specific primers as detailed in Materials and Methods. As expected, ChIP assays with the nonimmune IgG showed no amplification of the fra-1 promoter (data not shown). The binding of c-jun at the promoter is very low under steady-state conditions (Fig. 10B, lanes 1 and 2). However, the recruitment of c-jun was strongly enhanced following TNF- treatment (Fig. 10B, lanes 3 and 4). In contrast, the binding of c-jun to the fra-1 promoter was significantly diminished in MEFS lacking the erk1 / signaling (cf lanes 7 and 8 with lanes 3 and 4). Role of FRA-1 in TNF- -stimulated proinflammatory gene transcription To examine a role for FRA-1 in regulating TNF- -induced pulmonary epithelial responses, we used two complimentary approaches: 1) an RNAi-mediated knockdown of gene expression and 2) stable overexpression. To silence endogenous FRA-1 expression, A549 cells were transfected with FRA-1 shrna expression vector or empty vector; after a 48-h incubation, cell lysates FIGURE 10. Effect of ERK inhibition on the recruitment of c-jun to the FRA-1 promoter. A, A549 cells were serum-starved for 24 h and incubated with UO126 (10 M) or DMSO for 30 min before treatment without ( ) or with TNF- ( ). After a 60-min incubation, formaldehyde was added to the cells to cross-link chromatin. ChIP assays were performed using c-jun Abs and nonimmune IgG as detailed in Materials and Methods. Experiments were repeated at least twice to obtain reproducible results. C, Quantification of amplified DNA band intensity normalized against input DNA from two independent experiments is shown (n 4). C, The position of forward (F) and reverse (R) primers used to amplify the fra-1 promoter encompassing the TRE site located at nucleotide position 119/ 113 is shown. D, WTanderk1 / MEFs were serum-starved for 2 h and stimulated without ( ) or with TNF- ( ). ChIP assays (bottom) were performed using c-jun Abs or nonimmune IgG (data not shown) using fra-1 promoter-specific primers as detailed in Materials and Methods. A representative blot of two independent experiments performed in duplicate is shown. E, Quantification of amplified DNA band intensity normalized against input DNA from two independent experiments (n 4) is shown.

8 7200 INTERPLAY BETWEEN c-jun AND Elk1 AT THE FRA-1 PROMOTER 11D). An equal number of viable cells were plated on a 6-well plate, serum-starved, and then stimulated with or without TNF- for 6 h. Total RNA was isolated and IL-8 expression was analyzed. TNF- markedly (9-fold) stimulated IL-8 mrna expression (Fig. 11D, bar 2), as compared with untreated cells (Fig. 11D, bar 1). The basal level expression of IL-8 is significantly lower in A549-F1 as compared with A549-C cells. Moreover, FRA-1 overexpression completely suppressed the TNF- -stimulated expression of IL-8 (Fig. 11D, bar 4). Consistent with this result, the magnitude of TNF- -stimulated IL-8 promoter-driven reporter expression was remarkably lower in A549-F1 cells (Fig. 11E, bar 4) as compared with A549-C (Fig. 11E, bar 2). Collectively, these data indicate that the FRA-1 induction by TNF- may play a role in dampening a sustained IL-8 induction by TNF-. FIGURE 11. The effects of FRA-1 on TNF- -induced proinflammatory cytokine gene transcription. A, A549 cells were transfected with a plasmid construct containing a FRA-1 sirna sequence (FRA1-sh) and incubated for 48 h. Empty vector without sirna sequence was used as control. Whole-cell lysates were prepared and immunoblotted with anti-fra-1 and -actin Abs. A representative immunoblot of three independent experiments is shown. B, A549 cells transfected with the empty vector or plasmid-bearing FRA1-siRNA as in A and then treated without ( ) or with TNF- (f) for 6 h. IL-8 gene expression was analyzed by real-time RT- PCR., p 0.01 and, p C, An equal amount of whole-cell lysates (40 g) isolated from the stable FRA-1 overexpressing (A549-F1) or control empty vector (A549-C) cells were immunoblotted with anti- FRA-1 and -actin Abs. Results shown are from two independent experiments. D, A549-F1 and A549-C cells were treated without ( ) or with (f) TNF- for 6 h, and IL-8 mrna expression was analyzed. Data are representative of two independent experiments performed in duplicate., p E, Stable A549-F1 and A549-C were cotransfected with 100 ng of IL-8-Luc reporter along with 1 ng of a reference plasmid, prl-tk. TNF- -unstimulated ( ) and -stimulated (f) luciferase activity was determined as detailed in Fig. 1. The values obtained from the untreated ( ) A549-C cells were set as 1.0. Data represent mean SD from at least two three independent experiments done in triplicate., p were prepared, and the knockdown of endogenous FRA-1 expression was analyzed by Western blot analysis (Fig. 11A). FRA-1 shrna suppressed 60% of the total level of FRA-1 protein (Fig. 11A, lane 2), as compared with the empty vector-transfected control (Fig. 11A, lane 1). The sirna exerted no effect on the expression of -actin (Fig. 11A, bottom panel) or ERK2 (data not shown) protein. Given a critical role of IL-8 in mediating TNF- induced phenotypic effects, we have analyzed the effects of FRA-1 silencing on IL-8 expression. FRA-1-silencing caused a significant increase (3-fold) in the basal level expression of IL-8, as compared with vector-transfected control (Fig. 11B). As expected, TNF- markedly stimulated IL-8 expression (Fig. 11B, bar 2). However, knockdown of FRA-1 further significantly enhanced the TNF- enhanced IL-8 mrna expression (Fig. 11B, cf bars 2 and 4). These results indicate that FRA-1 induction by TNF- may play a role in attenuating IL-8 induction by TNF-. To confirm this notion, we have stably overexpressed FRA-1 in A549 cells (termed as A549-F1) and its expression was confirmed by immunoblot analysis (Fig. 11C, lane 2). As anticipated, empty vector-bearing A549 cells (A549-C) showed a little expression of FRA-1 (Fig. 11C, lane 1). The biological activity of ectopically expressed FRA-1 in A549-F1 cells was significantly high compared with A549-C cells, as assessed by EMSA using a TRE as probe and by transfection assays using TRE-Luc as a reporter, and was markedly (4-fold) higher in A549-F1 cells, as compared with A549-C cells (data not shown). We next examined the effects of ectopically expressed FRA-1 on TNF- stimulated IL-8 expression (Fig. Discussion In this study, we have shown that c-jun, a major effector of JNK1/2 signaling, is required for FRA-1 protooncogene induction by TNF- in pulmonary epithelial cells. Unexpectedly, inhibition of the JNK pathway, which is known to play a critical role in AP-1 activation in other cell types, failed to suppress TNF- -stimulated FRA-1 expression in both A549 and primary cultured human bronchial epithelial cells (Fig. 5, C and D). Furthermore, ectopic expression of a c-jun mutant lacking the N-terminal JNK phosphorylation sites did not suppress TNF- -inducible FRA-1 promoter activation (Fig. 6D). A requirement of c-jun, but not the JNK1/2 pathway, indicates that JNK1/2-dependent c-jun phosphorylation may not be essential for TNF- -stimulated FRA-1 promoter transactivation in pulmonary epithelial cells. The JNK pathway, in contrast, has been implicated (34) in FRA-1 induction by TNF- in MEF cells. JNK-deficient MEFs, lacking both Jnk1 and Jnk2 genes, showed diminished levels of fra-1 expression in response to TNF- (34). Thus, it seems that the JNK1/2 pathway regulates FRA-1 transcription in a cell type-specific manner. Consistent with this view and our results, Catani et al. (35) have shown that ascorbate, which blocks the activation of the JNK pathway, strongly up-regulates FRA-1 expression in the HACAT cell line of epithelial origin. On the contrary, we have shown (16) that JNK1/2 inhibition blocks cigarette smoke-stimulated FRA-1 expression in human bronchial epithelial cells, underscoring the notion that the JNK1/2 pathway regulates FRA-1 transcription in a context-dependent manner. The c-jun activity is primarily regulated at the level of its transcription and posttranslational modifications. c-jun expression is rapidly induced within the first 30 min by a variety of mitogenic and stress stimuli, as well as in response to cytokines in multiple cell types (36). The induction of the c-jun gene is controlled by multiple regulatory elements, including TRE, GC box, CAAT box, ATF, and MEF2 sites (37). Several transcription factors that bind to these elements are targets of ERK1/2, ERK5, JNK1/2, and p38 kinases, which have been shown (38) to regulate c-jun transcription in response to stress, cytokine, and mitogenic stimuli. For example, ATF and c-jun, which bind to the ATF/Jun site, are targets of ERK1/2, JNK, and p38 kinases (39). In contrast, MEK5- ERK5 signaling regulates c-jun transcription via a phosphorylation of the MEF family of transcriptions factors that bind to the functional MEF site located next to the TATA box (40, 41). In addition to transcriptional induction, posttranslational modifications, such as phosphorylation of c-jun play key roles in c-jundependent gene transcription. The N-terminal phosphorylation of c-jun protein by JNK kinases enhances both transactivation potential and the stability of c-jun protein (42), which otherwise undergoes ubiquitin-dependent proteolysis (43).

9 The Journal of Immunology c-jun has been shown to mediate various cellular responses in a JNK phosphorylation-dependent and -independent manner. Although phosphorylation of c-jun by JNK on Ser 63 and Ser 73 is required to protect cells from UV-induced cell death, it is not required for cell growth (41). Consistent with this, although c-jun deletion leads to embryonic lethality (44), the c-jun mice lacking the JNK phosphorylation sites, Ser 63 and Ser 73, are viable, fertile, and displayed no phenotypic defects (45). However, these mutant mice are less susceptible to kinate-induced neuronal apoptosis than the WT mice (42, 45). c-jun-regulated cell cycle progression (46) and Ras-induced cellular transformation (47, 48) do not require the JNK-induced phosphorylation at Ser 63 and Ser 73 of c-jun. Several recent studies have shown (49, 50) that JNK signaling is dispensable for transactivation of c-jun. For example, interaction of c-jun with CBP coactivator or RNA helicase requires the N-terminal region but not the JNK phosphorylation sites. Consistent with these findings, it has been shown (51) that JNK phosphorylation of c-jun, which is essential for disassociation of c-jun from the HDAC3 repressor, is not essential for subsequent transcriptional activation of c-jun. In contrast, JNK phosphorylation sites are required for an efficient interaction of c-jun with TCF4 and the subsequent recruitment of -catenin on the c-jun promoter, thereby resulting in an enhanced c-jun transcription (52). Thus, it appears that the status of c-jun JNK phosphorylation has a distinct effect on gene transcription. Our findings show a prominent role for ERK1/2 in controlling TNF- -induced FRA-1 transcription, despite the transient nature of the activation of this MAPK pathway by TNF- (Figs. 7 and 9). Our analysis revealed that the ERK pathway is required for TNF- -stimulated Elk1, CREB, and ATF1 phosphorylation (Fig. 8A) and is consistent with the previous suggestion (53) that these proteins are putative substrates for ERKs. We have previously shown (26, 32, 54) using ChIP assays that these proteins are constitutively bound to a critical SRE of the FRA-1 promoter in pulmonary epithelial cells. Similarly, a variety of external stimuli, such as epidermal growth factor, 12-O-tetradecanoylphorbol-13-acetate, and cigarette smoke, also did not enhance their binding to the promoter. A similar scenario exists for c-fos, whose promoter is occupied by these factors in vivo in the unstimulated state (55, 56). Based on these observations, we speculate that ERK inhibition likely affects the activation of the DNA-bound Elk1, CREB, and ATF proteins. Consistent with this notion, the translocation of ERKs from the cytoplasm to the nucleus following external stimuli has been firmly established (57). Finally, the phosphorylation of Elk1 by MAPK has been shown (26, 32, 54) to enhance its interaction with the coactivator p300, leading to gene transcription. Phosphorylation of CREB at Ser 133 is critical for CBP recruitment to the promoter in response to mitogenic and stress signals (60). sirna-mediated knockdown of endogenous c-jun expression profoundly inhibited FRA-1 induction by TNF-. Consistent with this result, MEFs lacking the c-jun gene showed a strong decrease in the level of fra-1 expression in response to mitogens (61, 62). Furthermore, we have recently found that overexpression of a c- Jun mutant or knockdown of endogenous c-jun expression significantly reduces the mitogen-inducible FRA-1 transcription in lung epithelial cells (32). Importantly, our findings indicate that inhibition of the ERK pathway decreases c-jun recruitment to the FRA-1 promoter in response to TNF- (Fig. 10A). A similar result was obtained in MEFs lacking the erk1 gene (Fig. 10B). However, we have noted that inhibition of the ERK1/2 pathway with PD98059 does not significantly reduce TNF- -stimulated c-jun mrna expression, which precedes FRA-1 transcription (Fig. 4A). These results collectively support the involvement of cross-talk between c-jun and ERK targets, such as Elk1, ATF, and CREB, binding at the respective 318 TRE, 274 TCF and the 248 ATF sites (detailed in Fig. 2 of Ref. 32) of the FRA-1 enhancer. Consistent with this notion, mutational inactivation of the 318 TRE, 274 TCF, or 248 ATF sites crippled TNF- inducibility of the FRA-1 promoter (Fig. 8B). Conversely, coexpression of mutant forms of c-jun, Elk1, SRF, ATF1, or CREB repressed FRA-1 induction (Fig. 8, C and D). The inability of c-jun to bind to the FRA-1 promoter in the absence of ERK signaling and the fact that Elk1, SRF, and CREB are bound to the promoter in the steady state suggest that the activation of Elk1, SRF, and CREB proteins by ERK signaling may facilitate, in some way, the recruitment of c-jun at the FRA-1 promoter in response to TNF-. This effect seems to occur independently of JNK signaling. Our findings indicate that FRA-1 may play a key role in regulating TNF- -induced proinflammatory cytokine gene expression (Fig. 11). Silencing of FRA-1 enhanced both basal and TNF- stimulated IL-8 expression. In contrast, FRA-1 overexpression caused a repression of IL-8 gene expression. The suppressive effect of FRA-1 on IL-8 gene expression appears to be regulated at the level of transcription (Fig. 11E). Our results are consistent with a recent report (63) that demonstrated a negative role for FRA-1 in attenuating or limiting the IL-1-induced IL-8 gene expression in non-pulmonary epithelial cells. In that study, the authors have shown that a delayed recruitment of FRA-1 to the IL-8 promoter counteracts c-fos and NF- B-mediated IL-1-induced IL-8 expression. Similarly, we have noticed that TNF- -stimulated c-fos expression precedes FRA-1 induction in pulmonary epithelial cells (data not shown). Thus, a repression of IL-8 induction by FRA-1 may probably be mediated by the displacement of c-fos from the IL-8 promoter. In summary, induction of the FRA-1 by TNF- occurs independently of the JNKs. Instead, ERKs seem to play a critical role in this process. Our findings also suggest that FRA-1 may attenuate the magnitude of the TNF- -induced activation of IL-8 expression in pulmonary epithelial cells. Acknowledgments We thank all of the scientists for providing us with the various expression vectors used in this study. We also thank Suneetha Peddakama and Won Kyung Lee for technical assistance on real-time PCR and Bill Spannhake for his help in the analysis of IL-6 and IL-8 expression. Disclosures The authors have no financial conflict of interest References 1. Bals, R., and P. S. Hiemstra Innate immunity in the lung: how epithelial cells fight against respiratory pathogens. Eur. Respir. J. 23: Sheppard, D Functions of pulmonary epithelial integrins: from development to disease. Physiol. Rev. 83: Baud, V., and M. Karin Signal transduction by tumor necrosis factor and its relatives. Trends Cell Biol. 11: Erzurum, S. C Inhibition of tumor necrosis factor for refractory asthma. N. Engl. J. Med. 354: Aggarwal, B. B Signalling pathways of the TNF superfamily: a doubleedged sword. Nat. Rev. Immunol. 3: Foletta, V. C., D. H. Segal, and D. R. Cohen Transcriptional regulation in the immune system: all roads lead to AP-1. J. Leukocyte Biol. 63: Karin, M., and E. Shaulian AP-1: linking hydrogen peroxide and oxidative stress to the control of cell proliferation and death. IUBMB Life 52: Reddy, S. P., and B. T. Mossman Role and regulation of activator protein-1 in toxicant-induced responses of the lung. Am. J. Physiol. 283: L1161 L Eferl, R., and E. F. Wagner AP-1: a double-edged sword in tumorigenesis. Nat. Rev. Cancer. 3: Karin, M., Z. Liu, and E. Zandi AP-1 function and regulation. Curr. Opin. Cell Biol. 9: Li, B., C. Tournier, R. J. Davis, and R. A. Flavell Regulation of IL-4 expression by the transcription factor JunB during T helper cell differentiation. EMBO J. 18: