Hexavalent Chromium Cr(VI) Up-Regulates COX-2 Expression through an NFκB/c-Jun/AP-1–Dependent Pathway

Background: Hexavalent chromium [Cr(VI)] is recognized as a human carcinogen via inhalation. However, the molecular mechanisms by which Cr(VI) causes cancers are not well understood. Objectives: We evaluated cyclooxygenase-2 (COX-2) expression and the signaling pathway leading to this induction due to Cr(VI) exposure in cultured cells. Methods: We used the luciferase reporter assay and Western blotting to determine COX-2 induction by Cr(VI). We used dominant negative mutant, genetic knockout, gene knockdown, and chromatin immunoprecipitation approaches to elucidate the signaling pathway leading to COX-2 induction. Results: We found that Cr(VI) exposure induced COX-2 expression in both normal human bronchial epithelial cells and mouse embryonic fibroblasts in a concentration- and time-dependent manner. Deletion of IKKβ [inhibitor of transcription factor NFκB (IκB) kinase β; an upstream kinase responsible for nuclear factor κB (NFκB) activation] or overexpression of TAM67 (a dominant-negative mutant of c-Jun) dramatically inhibited the COX-2 induction due to Cr(VI), suggesting that both NFκB and c-Jun/AP-1 pathways were required for Cr(VI)-induced COX-2 expression. Our results show that p65 and c-Jun are two major components involved in NFκB and AP-1 activation, respectively. Moreover, our studies suggest crosstalk between NFκB and c-Jun/AP-1 pathways in cellular response to Cr(VI) exposure for COX-2 induction. Conclusion: We demonstrate for the first time that Cr(VI) is able to induce COX-2 expression via an NFκB/c-Jun/AP-1–dependent pathway. Our results provide novel insight into the molecular mechanisms linking Cr(VI) exposure to lung inflammation and carcinogenesis.


Research
Chromium (Cr) is a ubiquitous metal found in animals, plants, rocks, soil, and air (Hill et al. 2008). Exposure to hexa valent chromium [Cr(VI)] occurs in multiple occupational environments, and the approxi mate daily absorbed dose of Cr(VI) is 83-1,700 μg/kg/day (Beveridge 2010). The International Agency for Research on Cancer (1980) has classified Cr(VI) as a known human carcinogen. Previous in vivo studies strongly indicated that there is an association between Cr(VI) exposure and airway inflam mation and lung carcino genesis (Beaver et al. 2009a(Beaver et al. , 2009bZeidlerErdely et al. 2008). However, the molecular mechanisms by which Cr(VI) induces lung inflammation and cancers are not yet well understood.
Prostaglandin (PG) is an important mediator at all stages of cancer development (Menter 2002). Cyclooxygenase (COX) is the ratelimiting enzyme in the synthesis of PGs (Rao et al. 2004). The COX enzyme system is composed of two isoenzymes: COX1, the constitutive isoform, and COX2, the induc ible protein (Davies et al. 2002). COX2 can undergo rapid induction in response to many factors, such as growth factors and cytokines (Kirschenbaum et al. 2001), and is highly expressed in a variety of human cancers and cancer cell lines (Liao and Milas 2004). COX2 over expression is associated with more aggressive biological tumor behaviors (Liao and Milas 2004), and the inhibition of COX2 has been regarded as an effective anti cancer strategy (Davies et al. 2002). Thus, identification of the potential involvement of COX2 and molecular mecha nisms respon sible for COX2 induction due to Cr(VI) exposure will provide significant insight into understanding Cr(VI) lung inflamma tory and carcino genic effects. In the present study, we investigated the potential effects of Cr(VI) on COX2 expression and molecular mechanisms leading to this induction in cell culture models.

Reverse-transcription polymerase chain reaction (RT-PCR).
After the cells were treated with Na 2 CrO 4 , total RNA was extracted using TRIZOL reagent (Invitrogen) following the manufacturer's instructions. Firststrand cDNA was synthesized with oligo(dT) 20 Background: Hexavalent chromium [Cr(VI)] is recognized as a human carcinogen via inhalation. However, the molecular mechanisms by which Cr(VI) causes cancers are not well understood. oBjectives: We evaluated cyclooxygenase-2 (COX-2) expression and the signaling pathway leading to this induction due to Cr(VI) exposure in cultured cells. Methods: We used the luciferase reporter assay and Western blotting to determine COX-2 induction by Cr(VI). We used dominant negative mutant, genetic knockout, gene knockdown, and chromatin immuno precipitation approaches to elucidate the signaling pathway leading to COX-2 induction. results: We found that Cr(VI) exposure induced COX-2 expression in both normal human bronchial epithelial cells and mouse embryonic fibroblasts in a concentration-and time-dependent manner. Deletion of IKKβ [inhibitor of transcription factor NFκB (IκB) kinase β; an upstream kinase responsible for nuclear factor κB (NFκB) activation] or over expression of TAM67 (a dominant-negative mutant of c-Jun) dramatically inhibited the COX-2 induction due to Cr(VI), suggesting that both NFκB and c-Jun/AP-1 pathways were required for Cr(VI)-induced COX-2 expression. Our results show that p65 and c-Jun are two major components involved in NFκB and AP-1 activation, respectively. Moreover, our studies suggest crosstalk between NFκB and c-Jun/AP-1 pathways in cellular response to Cr(VI) exposure for COX-2 induction. conclusion: We demonstrate for the first time that Cr(VI) is able to induce COX-2 expression via an NFκB/c-Jun/AP-1-dependent pathway. Our results provide novel insight into the molecu lar mechanisms linking Cr(VI) exposure to lung inflammation and carcinogenesis. volume 120 | number 4 | April 2012 • Environmental Health Perspectives primers using the SuperScript III FirstStrand Synthesis System for RTPCR; Invitrogen), and 1 μg of total RNA was used to perform reverse transcription. Specific primer pairs were designed for amplifying murine cox-2 (forward, 5´tca ccc gag gac tcc gcc3´; reverse, 5´tcc tgc ccc aca gca aac tgc3´) and βactin (forward, 5´gac gat gat att gcc gca ct3´; reverse, 5´gat acc acg ctt gct ctg ag3´). For specific amplifications, 50 ng of cDNA tem plates was used.
Luciferase reporter assay. MEFs trans fected with the luciferase reporter constructs were seeded into 96well plates (8 × 10 3 /well) and subjected to various treatments when cul tures reached 80-90% confluence. For ultra violet B (UVB) radiation, culture plates were covered with a thin layer of fresh medium (0.1% FBSDMEM) and exposed to UVB light for 1 min, corresponding to a dose of 1 kJ/m 2 , as reported previously (Song et al. 2007). The UVB light source (UVP Inc., Upland, CA, USA) emitted > 95% 302nm UVB light. Luciferase activity was determined using a luminometer (Wallac 1420 Victor 2 multi label counter system; PerkinElmer, Waltham, MA, USA) as described previously (Huang et al. 2002). The results are expressed as relative activity normalized to the luciferase activity in the control cells without treatment.
Western blotting assay. Cells (2 × 10 5 ) were seeded and cultured in each well of sixwell plates until 70-80% confluence. The cells were exposed to Cr(VI) at varying doses and time points and then extracted with sodium dodecyl sulfate sample buffer as previously described (Ouyang et al. 2007b). The cell extracts were used for Western blotting with specific anti bodies. The protein band, specifically bound to the primary antibody, was detected using an antirabbit IgGalkaline phosphatase (AP)-linked antibody and an electro chemi fluorescence (ECF) Western blotting system (Amersham Biosciences, Piscataway, NJ, USA). The images were obtained by scanning using the Storm 860 phospho imager (Molecular Dynamics, Sunnyvale, CA, USA) Electrophoretic mobility shift assay (EMSA) and super gel shift. We performed the EMSA using the LightShift Chemiluminescent EMSA Kit (Pierce, Rockford, IL, USA) according to the manufacturer's instructions. Nuclear extracts were isolated with a Nuclear/ Cytosol Fractionation Kit (BioVision, Mountain View, CA, USA). The specific probe pair designed for activated NFκB was 5´agt tga ggg gac ttt ccc agg c3´ and 5´gcc tgg gaa agt ccc ctc aac t3´. The specific probe pair designed for activated AP1 was 5´cgc ttg atg agt cag ccg gaa3´ and 5´ttc cgg ctg act cat caa gcg3´. The probes were conjugated with biotin by a Biotin 3´ End DNA Labeling Kit (Pierce) following the manufacturer's instructions. Nuclear protein (4 μg) was subjected to the gel shift assay by incubation with 1 μg poly(dIdC) DNA carrier in DNA binding buffer [10 mM Tris (pH 8.0), 150 mM potassium chloride, 2 mM EDTA, 10 mM magnesium chloride, 10 mM dithiothreitol, 0.1% bovine serum albumin, 20% glycerol]. The biotinlabeled doublestranded oligo nucleo tide (1 μL) was then added, and the reaction mixture was incubated at room temperature for 50 min.
For competition experiments, a 50fold molar or NHBECs (C,D) were exposed to Cr(VI) as indicated. The cells were extracted and COX-2 expression was determined by Western blotting (A,C,D) or by RT-PCR (B). β-Actin was used as a loading control. (E) WT MEFs were treated with Cr(VI) at indicated doses for 6 and 12 hr and then allowed to recover in normal culture medium for 24 hr; cytotoxicity was determined by colony survival assay. (F,G) COX-2 promoter-driven luciferase transcription relative to control (relative COX-2 transcription) was determined in MEFs treated with 20 μM Cr(VI) for various times (F) or at different Cr(VI) doses for 12 hr (F). Data are mean ± SD of triplicates. *p < 0.05, compared with control cells (medium only).

Time (hr) Time (hr) Cr(VI) (µM)
24 36 Cr(VI) 0 µM Cr(VI) 5 µM Cr(VI) 20 µM excess of the unlabeled doublestranded oligo nucleo tide was added before the addition of the labeled probe. For the super gel shift assay, nuclear extracts were incubated with 2 μg antibody for 30 min at 4°C before addition of the probe. DNA-protein complexes were resolved by electrophoresis on 5% non denaturing glycerolpolyacrylamide gels. The luminescent signal was developed by a LightShift® Chemiluminescent EMSA Kit and detected by an automatic developing machine. Chromatin immunoprecipitation (ChIP) assay. The ChIP assay was performed using the EZ ChIP kit (Upstate, Billerica, MA, USA) according to the manufacturer's instructions. Briefly, cells were either untreated or treated with Cr (20 μM) for 12 hr, and then genomic DNA and the proteins were crosslinked with 1% formaldehyde. The crosslinked cells were pelleted, resuspended in lysis buffer, and soni cated to generate 200 to 500bp chroma tin DNA fragments. After centrifugation, the supernatants were diluted 10fold and then incubated with antip65 or anticJun anti bodies, respectively, or the control rabbit IgG at 4°C over night. The immune complex was captured by protein G agarose saturated with salmon sperm DNA and then eluted with elu tion buffer. DNA-protein crosslinking was reversed by heating at 65°C for 4 hr. DNA was purified and subjected to PCR analysis.
To specifically amplify the region con taining the putative NFκBresponsive ele ments on the mouse COX-2 promoter, we performed PCR using the following prim ers: 5´ctg acg agc gag cac gtc3´ (forward) and 5´ttt ggc ctc tgg ggt ttc3´ (reverse). To specifically amplify the region containing the putative AP1-responsive elements on the mouse COX-2 promoter, PCR was per formed with the following primers: 5´ttc cca taa gac tcc g3´ (forward) and 5´gct tca tgt gca agc t3´ (reverse). Primers targeting the region 1 kb upstream of the NFκB and AP1 binding sites on the COX-2 promoter were also used in the PCR analysis to support the specificity of the ChIP assay: 5´tga ttt ggt ttg gga ca3´ (forward) and 5´ctg gag gac aag agc agt3´ (reverse).
Clonogenic survival assay. MEFs were treated with Cr(VI) at 5 μM and 20 μM for 6 and 12 hr and recovered for 24 hr in nor mal culture medium. Cells were then plated at 500 cells/dish in 100mm cell culture dishes and cultured for 2 weeks. Cells were stained with Giemsa solution, and the num ber of colonies was counted and presented as mean ± SD (n = 3).
Statistical analysis. We used the Student's ttest to determine the significance of differ ence in COX2 induction and AP1, NFAT, or NFκB activation in luciferase reporter assays among various groups. The statistical significance level was set at p < 0.05.

Cr(VI) exposure induced COX-2 expression.
As shown in Figure 1A, treatment of MEFs with Cr(VI) resulted in an increase in COX2 protein expression in a dose and timedependent manner. We observed marked induction at 12 hr and 24 hr after exposure. Cr(VI) exposure was previously reported to induce either cell growth arrest and/or apoptosis in a dose, time and, cell type-dependent manner (Wang et al. 2004). To evaluate the cyto toxicity of Cr(VI) in our experimental system, we subjected Cr(VI) treated MEFs to a colonysurvival assay. Results showed only marginal toxicity on MEFs exposed to 20 μM Cr(VI) after 12 hr of exposure, whereas there was no observable cyto toxicity at 5 μM ( Figure 1E). These results are consisted with a previous report showing that the viability of HaCaT (human keratino cyte) cells is not affected at Cr(VI) concentrations as high as 30 μM (Wang et al. 2010). Consistent with protein induction, marked induction of COX-2 mRNA by 20 μM Cr(VI) was present as early as 6 hr after exposure, suggesting that Cr(VI) might induce COX-2 expression at a transcriptional level ( Figure 1B). To test this notion, we investigated the effects of Cr(VI) on COX-2 promoter activity in the stable transfectant of COX-2 promoter-driven luciferase reporter. As shown in Figure 1F and 1G, treatment with Cr(VI) resulted in a marked increase in COX-2 promoter activity. This induction was also observed with 20 μM Cr(VI) as early as at 6 hr after exposure ( Figure 1F), which is consistent with the results of the RTPCR assay. The respiratory tract is the primary target organ of Cr(VI) (Goldoni et al. 2008). Thus, we used NHBECs to test the effect of Cr(VI) on COX2 expression. Cr(VI) exposure did cause COX2 expression in NHBECs ( Figure 1C,D). Collectively, these results indicate that Cr(VI) is able to induce COX2 expression in both MEFs and NHBECs.
Cr(VI) exposure induced the activation of NFκB and AP-1 but not NFAT. Cr(VI) treatment did not result in observable NFAT activation (Figure 2A), whereas UVB expo sure, the positive control, resulted in signifi cant NFAT activation ( Figure 2B) in the same stable NFATLuc reporter transfectant.  A and B), NFκB-dependent (C), and AP-1-dependent (D) transactivation in MEFs was determined by specific luciferase reporter assay after exposure to different concentrations of Cr(VI) for 12 hr or UVB (1 kJ/m 2 ) for 6 hr. Values are mean ± SD of triplicates. (E) MEFs were exposed to Cr(VI) for 1 hr and then extracted and subjected to Western blotting analysis. β-Actin was used as a loading control. (F,G) MEFs were exposed to 20 μM Cr(VI), and the nuclear extracts were subjected to the gel shift assay with NFκB (F) or AP-1 (G) probe. For competition experiments, a 50-fold molar excess of unlabeled NFκB or AP-1 cold probe was added to the binding reaction mixtures to determine the specific binding. *p < 0.05, compared with control. In contrast to NFAT, NFκB activation was signifi cantly increased by Cr(VI) treatment in the NFκBLuc reporter assay ( Figure 2C). The activation of the NFκB pathway by Cr(VI) was further verified by the observation of increased IκBα phosphoryla tion and degradation in the Western blotting assay ( Figure 2E) and NFκB DNA binding activity analyzed by an EMSA assay ( Figure 2F). We further determined the involvement of the AP1 pathway in cells exposed to Cr(VI). As shown in Figure 2D and 2G, treatment of cells with Cr(VI) for 6 hr also led to marked AP1 induction in the AP1Luc reporter assay ( Figure 2D) and the AP1 EMSA assay ( Figure 2G). These results demonstrate that Cr(VI) exposure induced activation of NFκB and AP1 but not NFAT.
DNA binding activity of NFκB induced by Cr(VI) reached to peak at 3 hr ( Figure 2F), whereas the maxi mum AP-1 DNA binding activity was achieved at 9 hr after exposure ( Figure 2G). The difference could be due to the differential pathways responsible for activation of NFκB and AP1. NFκB activation is fully dependent on IKKβ/IκB phosphorylation/degradation , whereas AP1 activation is dependent on both cJun phosphorylation and increased cJun protein expression (Huang et al. 1999a(Huang et al. , 1999b. The induction of cJun protein expression may lead to the delay of maximum AP1 activation compared with the peak of NFκB activation. Cr(VI) has been reported to inhibit tumor necrosis factorα-induced NFκB transcriptional competence through inhibiting inter actions with coactivators of transcription rather than DNA binding (Shumilla et al. 1999). Another study found that Cr(VI) prevented the benzo[a]pyrene dependent release of histone deacetylase1 from cytochrome P450 1a1 chromatin and blocked p300 recruitment (Wei et al. 2004).

IKKβ is required for CI(VI)-induced COX-2 expression.
To clarify the potential role of IKKβ in Cr(VI)induced COX2 expres sion, we used IKKβKM, an inactive mutant of IKKβ, and IKKβ -/-MEFs. As shown in Figure 3A, over expression of IKKβKM in MEFs inhibited Cr(VI)induced COX2 expression in the COX2Luc reporter assay. The knockout of IKKβ ( Figure 3B) impaired the phosphorylation and degradation of its downstream target IκBα after Cr(VI) treat ment ( Figure 3C), indicating the necessary role of IKKβ in Cr(VI)induced NFκB activa tion. Cr(VI)induced COX2 protein expres sion was consistently blocked in IKKβ -/cells ( Figure 3D). Moreover, reconstituted expres sion of IKKβ in IKKβ -/cells restored COX2 induction ( Figure 3E). Our results demonstrate that IKKβ was required for COX2 induc tion after Cr(VI) exposure. Overexpression of IKKβKM was not able to completely inhibit COX-2 promoter-driven luciferase transcrip tion ( Figure 3A), whereas IKKβ deletion (IKKβ -/-) was able to block COX2 expression completely ( Figure 3D). These results suggest that IKKβKM over expression was not able to completely impair endogenous IKKβ function.
The potential role of NFκB p65 in the regu lation of COX-2 expression due to Cr(VI) exposure. NFκB components are expressed in a variety of cell types (Karin and Greten 2005). In a previous study we showed that the NFκB p65 sub unit, but not the p50 sub unit, is required for nickelinduced COX2 expression in Beas2B cells (Ding et al. 2006b). In the present study, we determined the dif ferential involvement of p65 and p50 sub units in Cr(VI)induced COX2 expression. We per formed a super gel shift assay in the presence of the anti bodies specific for p65 or p50. As shown in Figure 3F, selective reduction of the p65 band was observed using antip65 anti body, whereas no reduction of DNA binding activity was observed with antip50 anti body. Incubation of cell nucleus extracts with anti p65 antibody reduced the extract protein bind ing to the NFκB probe but did not cause the supershift band. The explanation for this may be that binding of antip65 antibody to p65 protein changes the p65 protein conforma tion and in turn leads to p65 losing its bind ing activity to the NFκB probe. These results suggest that p65 might be the major compo nent involved in NFκB activation after Cr(VI) exposure. This notion is further supported by ChIP assay data. As shown in Figure 3G, Cr(VI) treatment markedly enhanced recruit ment of the p65 subunit to its binding site in COX-2 promoters, whereas control IgG and primers targeting the DNA sequence located at approximately 1 kb upstream of the NFκB binding site in the COX-2 promoter did not , and IKKβ -/-(IKKβ) MEFs were seeded into six-well plates, and Western blotting analysis was performed with anti-P-IκBα and anti-IκBα (C), anti-COX-2 (D), or anti-COX-2 and anti-IKKβ (E). β-Actin was used as a loading control. (F) MEFs were exposed to 20 μM Cr(VI) for 3 hr, and then the nuclear extracts were subjected to a super gel shift assay using anti-p65 and anti-p50. (G) MEFs were exposed to 20 μM Cr(VI) for 3 hr, and then the ChIP assay was performed. *p < 0.05, compared with WT (vector) cells. show detectable PCR products ( Figure 3G). Taken together, these results demonstrate that NFκB p65, rather than the p50 subunit, plays a key role in NFκB activation and COX2 induction after Cr(VI) exposure.

Involvement of c-Jun/AP-1 in Cr(VI)induced COX-2 expression. Different AP1
dimers play different roles in the regulation of cellular function and carcino genesis . Western blotting shows that Cr(VI) exposure resulted in cJun phosphorylation, but we observed no activation of other AP1 members Jun B, Jun D, cFos, or Fra1 ( Figure 4A). To determine the role of cJun in Cr(VI)induced AP1 activation, we performed a super gel shift assay using antibodies specific for cJun and cFos. As shown in Figure 4B, we observed a selective supershift band of cJun in cell extracts from Cr(VI)treated cells, but no cFos supershift band was observable, suggesting that cJun was the major component involved in AP1 activation due to Cr(VI) exposure.
COX-2 has been shown to be a typical AP1regulated gene in several experimental systems (Zhang et al. 2010). Thus, we determined the recruitment of cJun to the COX-2 promoter region using the ChIP assay. The detection of the COX-2 promoter in the antibodycaptured genomic DNA fragments was performed by PCR amplification with primers designed to specifically recognize the region containing AP1-responsive elements. AnticJun antibody strongly coimmuno precipitated the target COX-2 promoter region DNA in Cr(VI) treated cell extract but not in the control cell extract ( Figure 4C), indicating the inducible recruitment of cJun to the endogenous COX-2 promoter after Cr(VI) exposure. This demonstrates Cr(VI)inducible recruitment of AP1 onto the endogenous COX-2 promoter region ( Figure 4C), suggesting that AP1 might play a role in the regulation of COX2 expression due to Cr(VI) exposure. To test this notion, we used TAM67, a dominant negative mutant of cJun. The ectopic expression of TAM67 in WT cells attenuated Cr(VI)induced cJun phosphorylation in MEFs ( Figure 4E). Unlike over expression of IKKβKM in MEFs ( Figure 3A), COX-2 promoter-driven luciferase transcription was impaired in WT/ TAM67 transfectant ( Figure 4D), suggesting that TAM67 over expression was able to block the endogenous cJun function. COX2 protein induction by Cr(VI) was also blocked ( Figure 4E). These results demonstrate that cJun activation is essential for COX2 induction after Cr(VI) exposure.
Crosstalk between AP-1 and NFκB pathways after Cr(VI) exposure. Crosstalk between AP1 and NFκB has been reported to be responsible for the synergistic increase in their activity in the regulation of target gene expression (Adcock 1997). Thus, we determined the potential relation ship of these two transcription factors in response to Cr(VI) exposure in cells. We used IKKβ -/-MEFs to examine whether the impairment of the NFκB pathway could affect cJun phosphoryla tion. Impairment of the NFκB pathway inhibited cJun phosphorylation ( Figure 5A), suggesting Figure 4. Requirement of c-Jun/AP-1 for Cr(VI)-induced COX-2 expression. (A) MEF cells were exposed to Cr(VI) for 12 hr, and cell extracts were subjected to Western blotting. (B) MEFs were exposed to 20 μM Cr(VI) for 6 hr, and nuclear extracts were subjected to a super gel shift assay for c-Jun and c-Fos. (C) MEFs were exposed to 20 μM Cr(VI) for 3 hr, before performing the ChIP assay. (D) MEFs transiently transfected with the COX-2-Luc reporter construct or COX-2-Luc reporter together with a c-Jun mutant construct (TAM67) were then exposed to Cr(VI), and the luciferase activities were determined 6 hr after treatment. Results are expressed as COX-2 induction relative to control. (E) WT (vector) or TAM67 MEFs cells were treated with Cr(VI) for 24 hr, and cell extracts were subjected to Western blotting. β-Actin was used as a loading control. that NFκB activation has a positive effect on cJun activation after Cr(VI) exposure. To further reveal the potential effects of cJun/AP1 on NFκB activation, we used a dominant negative mutant of cJun (TAM67). As shown in Figure 5B, ectopic expression of TAM67 had an inhibitory effect on IκBα phosphorylation, suggesting that the phosphorylation of cJun was also involved in the regulation of the NFκB pathway. Taken together, the AP1 and NFκB pathways did show crosstalk after Cr(VI) treatment, which might play a role in Cr(VI)induced COX2 induction and carcinogenesis ( Figure 5C).

Discussion
The data we present here indicate that Cr(VI) induced expression of COX2 and activation of AP1 and NFκB, and show that both AP1 and NFκB are required for Cr(VI)induced COX2 expression. Our data also indicate the presence of cross talk between the NFκB and AP1 pathways after Cr(VI) exposure, which mainly occurred via IKKβ/p65 dependent and cJun-dependent pathways. Considering the important role of COX2 in the mediation of chronic inflammation and lung carcino genesis, we anticipated that acti vation of NFκB and AP1 pathways and their crosstalk in the regulation of COX2 expres sion might be key factors in Cr(VI)induced lung carcino genesis. Further elucidating the relationship among chronic inflammation, COX2 induction, and lung carcinogenic effect after various doses of Cr(VI) exposure in vivo animal models will be a major focus for future investigations in our laboratory, which might help determine a threshold dose for lung carcinogenesis of Cr(VI) exposure.
Inflammation is implicated in Cr(VI) induced human lung cancer development. Repetitive exposure to Cr(VI) results in persis tent inflammation, and such an inflammatory micro environment can further promote lung carcino genesis (Beaver et al. 2009a(Beaver et al. , 2009b. COX2 plays an important role in the develop ment of various types of cancer, including lung cancer (Sahin et al. 2009), and drugs targeting this enzyme have achieved widespread clinical use (Bertagnolli 2007). Our previous studies have shown that COX2 induction is involved in several carcino genic responses (Ding et al. 2006a(Ding et al. , 2006bLi et al. 2006). In the pres ent study, we initially found that exposure to Cr(VI) induced COX2 expression in both NHBECs and MEFs. Considering the critical role of COX2 in the inflammatory processes of cancer and the importance of an inflamma tory micro environment during carcino genesis after Cr(VI) exposure, our results may shed light into the mechanisms of Cr(VI)induced carcinogenic effects.
The COX-2 promoter region contains the binding sites of three major transcription factors: NFκB (Crofford et al. 1997), AP1 (Subbaramaiah et al. 2002), and NFAT (Iniguez et al. 2000). These three factors have been reported to be major mediators for the regulation of cell proliferation, differentia tion, and transformation (Huang et al. 1999a(Huang et al. , 1999b. In the present study, we observed that Cr(VI) exposure resulted in the activa tion of NFκB and AP1, whereas there was no observable NFAT activation, which is consistent with published studies showing that Cr(VI) exposure leads to the activation of NFκB and AP1 in an oxidativestressdependent manner (Yao et al. 2008). NFκB activation has been reported to be involved in the development of several cancers (Biswas et al. 2004;Wang et al. 2003). Our published studies have shown that NFκB activation is involved in cellular responses to several envi ronmental carcinogens Ouyang et al. 2007b). In the present study, we found that IKKβ was critical for Cr(VI) induced NFκB activation and COX2 expres sion. In addition, we showed that p65, rather than p50, was required for Cr(VI)induced NFκB activation and COX2 expression. We observed that Cr(VI) exposure induces NFκB activation via an IKKβ/p65dependent path way, which further leads to COX2 induc tion. Cr(VI) increases formation of reactive oxygen species (ROS) in certain cell types (Wang et al. 2010), and the inductive COX2 expression of manganese is accompanied by generation of oxidative stress and increased NFκB and AP1 DNA binding activities (Chen et al. 2007). Thus, we anticipate that ROS generation may also be involved in the activation of NFκB and AP1, which further leads to COX2 expression.
The cJun/AP1 pathway is crucial for COX2 induction caused by some environ mental stresses (Ouyang et al. 2007a;Zhang et al. 2008). Because of the multiple functions of AP1 proteins, the selection of the different AP1 dimers is considered as another mechanism for the modulation of AP1 activity . The results of the present study indicate that AP1 activation due to Cr(VI) exposure mainly involves cJun phosphorylation. The predominant role of cJun in Cr(VI)induced AP1 transactivation and COX2 induction was further confirmed by super gel shift assay and ChIP assay. Furthermore, transfection with the dominant negative cJun mutant (TAM67) blocked Cr(VI)induced COX2 expression. In addition, the knockout of IKKβ impaired Cr(VI)induced cJun phosphorylation, whereas inhibition of the cJun/AP1 pathway by over expression of TAM67 also inhibited Cr(VI)induced IκBα phosphorylation, suggesting crosstalk between the cJun/AP1 pathway and the IKKβ/NFκB pathway in the Cr(VI) response. Because both the cJun/AP1 pathway and the IKKβ/NFκB pathway are crucial for COX2 induction, we anticipate that this crosstalk may play a key role in Cr(VI)induced COX2 expression, which provides a novel model of the inter action between NFκB and AP1 pathways for environmental responses. Considering that inhibition of NFκB, AP1, and COX2 has been proposed as potential anti cancer strategies, our results may lead to new targets for chemo prevention of Cr(VI)induced human carcinogenesis.