|Home | About | Journals | Submit | Contact Us | Français|
Arsenic exposure has been reported to cause neoplastic transformation through the activation of PcG proteins. In the present study, we show that activation of p38α mitogen-activated protein kinase (MAPK) is required for arsenic-induced neoplastic transformation. Exposure of cells to 0.5 µM arsenic increased CRE and c-Fos promoter activities that were accompanied by increases in p38α MAPK and CREB phosphorylation and expression levels concurrently with AP-1 activation. Introduction of short hairpin (sh) RNA-p38α into BALB/c 3T3 cells markedly suppressed arsenic-induced colony formation compared with wildtype cells. CREB phosphorylation and AP-1 activation were decreased in p38α knockdown cells after arsenic treatment. Arsenic-induced AP-1 activation, measured as c-Fos and CRE promoter activities, and CREB phosphorylation were attenuated by p38 inhibition in BALB/c 3T3 cells. Thus, p38α MAPK activation is required for arsenic-induced neoplastic transformation mediated through CREB phosphorylation and AP-1 activation.
Arsenic is a widespread and well-known human carcinogen. Chronic arsenic exposure is associated with skin, lung, bladder, liver, kidney and prostate cancers [1,2]. Accumulating evidence implicates chronic exposure to arsenic with the malignant transformation of human bronchial epithelial cells , human HaCaT keratinocytes [4,5], human lung epithelial BEAS-2B cells , human small airway epithelial cells , rat liver epithelial TRL1215 cells , and human prostate epithelial RWPE-1 cells with estrogen-induced epigenetic changes . Even though chronic exposure to arsenic is known to be involved in carcinogenesis, the mechanism explaining the exact relationship of chronic arsenic exposure and tumor development is unclear.
The p38 mitogen-activated protein kinase (MAPK) family regulates various cellular stresses and inflammation. Many reports have suggested that p38 MAPK is an important regulator in bladder , prostate , liver  and lung  cancers. Overexpression or overactivation of p38 MAPK plays an important role in thyroid neoplasms  and transformed follicular lymphomas . However, p38 MAPK also plays an inhibitory role in Ras-induced neoplastic transformation of cells , hepatocellular carcinoma  and human ovarian cancer metastatic colonization . In addition, p38 MAPK specifically modulates malignant transformation induced by oncogenes that produce reactive oxygen species (ROS) . It also stimulates migration in melanoma  and promotes the invasive and proliferative phenotype of head and neck squamous cell carcinoma . Although p38 MAPK exhibits some antitumor activities, it is also associated with cellular neoplastic transformation that is dependent on cancer stage or tumor type. However, the relationship between p38 MAPK and arsenic-induced neoplastic transformation is not clear.
In a previous study, we reported that polycomb (PcG) proteins, including BMI1 and SUZ12, were required for arsenic-induced cell transformation through the inhibition of tumor suppressor expression . We sought to determine if p38α MAPK is involved in arsenic-induced fibroblast transformation and found that p38α MAPK plays a role in fibroblast proliferation induced by 0.5 µM arsenic possibly through AP-1 activation.
Chemical reagents, including arsenic trioxide (As2O3) and basal medium eagle (BME) were purchased from Sigma-Aldrich (St. Louis, MO). Calf serum (CS) was from GIBCO (Grand Island, NY). The prestained protein marker and protease inhibitor cocktail were from GenDEPOT (Barker, TX). Antibodies against phospho-p38 MAPK, total p38α MAPK, phospho-MSK1, total MSK1, phospho-CREB, total CREB, phospho-ATF2 and total ATF2 were from Cell Signaling Technology, Inc. (Beverly, MA).
Wildtype BALB/c 3T3 murine embryonic fibroblasts (MEFs) were from ATCC® (Manassas, VA). BALB/c 3T3 cells are an immortalized transformed fibroblast cell line, which have not undergone neoplastic transformation. Cells stably expressing knockdown p38α MAPK were grown in 10% CS/DMEM supplemented with penicillin/streptomycin (100 units/mL; Invitrogen™, Carlsbad, CA) at 37°C in a humidified 5% CO2 incubator. To construct arsenic-induced transformed BALB/c 3T3 cells, As2O3 (0.5 µM in 0.1 M NaHCO3) or 0.1 M NaHCO3 only as a control was used to treat cells every two days with media change over 2 or 4 weeks. No additional arsenic was included in subsequent assays to assess neoplastic transformation or protein expression. To construct knockdown of p38α in BALB/c 3T3 cells, the shRNA-p38α (sh-p38α) or shRNA-GFP (sh-GFP) control vector (RNAi Core Facility, BioMedical Genomic Center, University of Minnesota, Minneapolis, MN) based on the pLKO.1 lentiviral vector, were infected into BALB/c 3T3 cells following the recommended protocols. Infected cells were selected in medium containing 2 µg/mL puromycin and the expression level of the p38α MAPK protein was confirmed by Western blot analysis.
To determine the effect of arsenic on protein phosphorylation, we analyzed phosphorylated proteins in untreated or arsenic-treated BALB/c 3T3 cell lysates using the Proteome Profiler™ Array/Human Phospho-Kinase Array Kit (R&D Systems, Minneapolis, MN) according to the instructions provided by the manufacturer.
In brief, cells (8 × 103/mL) were cultured in 1 mL of 0.3% Basal Medium Eagle (BME) agar containing 10% CS. The cultures were maintained in a 37°C, 5% CO2 incubator for 7 days and cell colonies were scored using a microscope and the Image-Pro PLUS (v.6) computer software program (Media Cybernetics, Silver Spring, MD).
For reporter gene assays, transient transfections were conducted using jetPEI (QBIOgene, Irvine, CA) and cells were co-transfected with 200 ng of the AP1-, c-Fos- or CRE-reporter plasmid and 50 ng of a β-galactosidase (β-Gal)-expressing plasmid in arsenic-induced transformed or wildtype cells without arsenic treatment. The β-Gal-expressing plasmid was co-transfected for normalizing the transfection efficiency. Routinely, at 36 h after transfection, cells were washed and then cell lysates were prepared. Luciferase and β-galactosidase activities were measured using the Luminoskan Ascent (Thermo Electron Corp., Marietta, OH) and Multiskan MCC (Thermo Electron Corp.), respectively.
We previously reported that exposure to arsenic caused neoplastic transformation of BALB/c 3T3 cells and these transformed cells could form tumors in nude mice . To begin to determine a mechanism of arsenic-induced neoplastic transformation, we used a phospho-kinase assay to examine the phosphorylated proteins from BALB/c 3T3 cells treated or not treated with 0.5 µM As2O3 for 2 or 4 weeks. We found that 0.5 µM exposure to arsenic significantly increased the phosphorylation of p38α MAPK, CREB, MSK1/2 and c-Jun proteins (Fig 1A, B). To determine which p38 MAPK isoforms were involved in arsenic-induced transformation, we detected p38α (Fig. 1), p38β, p38γ and p38δ (Fig. S1) isoform expression in BALB/c 3T3 cells treated or not treated with arsenic and found that only p38α MAPK was expressed. We then examined the phosphorylation and expression levels of p38α MAPK and confirmed that levels were increased up to almost 30-fold in arsenic treated BALB/c 3T3 cells (Figure 2A). In addition, the phosphorylation and expression levels of the transcription factors, CREB, c-Jun, and ATF2, were increased from about 2-fold up to 48-fold (Figure 2B). These findings demonstrated that the activation of p38α MAPK signaling is involved in arsenic-induced neoplastic transformation.
The AP-1 transcription factor is a dimeric complex that can include combinations of Fos, Jun or ATF proteins . AP-1 activation is known to play an important role in multistage development of tumors [24,25]. To determine whether arsenic affects AP-1 activation, we co-transfected the AP-1-, c-Fos-, c-Jun- or CRE-luciferase reporter plasmid and the pCMV-β-Gal gene into arsenic-treated or -untreated BALB/c 3T3 cells. Our results showed that arsenic significantly increased AP-1 activation (Figure 3A) and c-Fos (Figure 3B) and c-Jun (Figure 3C) promoter activity in transformed BALB/c 3T3 cells. We also found a significant enhancement of the CRE promoter activity in arsenic-treated BALB/c 3T3 cells (Figure 3D). CRE is a cAMP response promoter element that regulates RNA polymerase activity and binding of activated CREB in cAMP-response genes . These data indicated that AP-1 transactivation activity and CRE promoter activity are induced in cells transformed by arsenic (0.5 µM) exposure.
To verify that the p38α MAPK protein is associated with neoplastic transformation induced by 0.5 µM arsenic exposure, we established knockdown of p38α in BALB/c 3T3 cells with sh-p38α to compare with cells transfected with a control vector (sh-GFP). Results indicated that endogenous p38α protein expression was suppressed by about 90% in p38α knockdown BALB/c 3T3 cells compared with cells expressing sh-GFP (Figure 4A). To examine the effect of suppressing p38α expression on arsenic-induced cell transformation, we stimulated cells with 0.5 µM As2O3 for 2 weeks in a 5% CO2 incubator. Arsenic exposure for 2 weeks had no effect on p38α expression in p38α knockdown cells compared with sh-GFP control cells (Figure 4B). Arsenic induction of neoplastic transformation in sh-p38α cells was significantly inhibited compared with cells expressing sh-GFP (Figure 4C). These results showed that p38α knockdown suppressed neoplastic transformation induced by 0.5 µM arsenic exposure. Overall, this result further demonstrated that p38α activation plays a key role in arsenic-induced cell transformation.
We showed that arsenic-induced phosphorylation levels of CREB were decreased in p38α knockdown cells (Figure 5A). To determine whether knockdown of p38α inhibits AP-1 activation in arsenic-induced neoplastic transformation, we introduced the AP-1, c-Fos- or CRE-luciferase reporter plasmid and the pCMV-β-Gal gene into BALB/c 3T3 cells treated with arsenic for 2 weeks. Results demonstrated that induction of AP-1 activation was significantly decreased in p38α knockdown cells treated with arsenic for 2 weeks compared with sh-GFP control cells (Figure 5B, left). Similar results were obtained for arsenic-induced c-Fos promoter activity (Figure 5B, middle) and arsenic-induced CRE promoter activity (Figure 5B, right), confirming that the p38α protein plays a key role in AP-1 activation and c-Fos or CRE promoter activity induced in arsenic-treated transformed cells. These findings suggest that the p38α protein is involved in arsenic-induced cell transformation mediated through AP-1 activation and regulation of p38α downstream molecules, including the CREB transcription factor.
Taken together, our results provide evidence showing that p38α MAPK signaling and AP-1 activation are required for neoplastic transformation induced by arsenic exposure. Finally, our findings suggest that p38α MAPK activation is essential for cell transformation induced by 0.5 µM exposure to arsenic.
Several studies indicated that an arsenic exposure was associated with malignant transformation in cancer cell lines and cancer patients [1,2,6,9,13,17,22]. Chronic arsenic exposure was reported to cause malignant transformation of lung cells and induce the expression of the mdig oncogene through JNK and STAT3 activation . Others reported that the Ras oncogene is activated in malignant transformation of human prostate epithelial cells by arsenic . HIF-2a-mediated inflammation is involved in arsenite-induced transformation of human bronchial epithelial cells  and c-Myc and c-Fos protein expression is increased in long-term arsenic-treated human small airway epithelial cells . Epigenetic silencing of let7-c by Ras/NF-κB is reportedly involved in neoplastic transformation of human keratinocytes  and hypermethylation-mediated silencing of MLH1 is associated with chronic exposure of human epithelial cells to arsenic . Herein, we found that BALB/c 3T3 cells is a cell line that can be useful for examining neoplastic transformation induced by arsenic treatment . The BALB/c 3T3 cell model is widely used to study cancer development induced by chemicals, physical trauma and biological agents. The transformation mechanism is very likely somewhat different between epithelial and fibroblast cells but some similarities could also exist. However, caution should be used in comparing results between transformation of epithelial cells and fibroblasts. To determine the mechanism of arsenic-induced neoplastic transformation in BALB/c 3T3 cells, we examined the signaling pathways that were activated by arsenic treatment and found that 0.5 µM arsenic exposure significantly increased the phosphorylation and expression levels of p38α MAPK (Figures 1 and and22).
Many extracellular stimuli can induce the phosphorylation of p38 MAPK resulting in the activation of a wide range of substrates such as protein kinases, transcription factors, and other cytosolic and nuclear proteins . Previous studies suggested that p38 MAPK exhibits tumor-suppressing functions [16,17,31]. However, p38 MAPK has also been associated with malignant transformation in many cancers [10,12,13,20,32]. Activated p38 MAPK contributed to cell migration and in vivo growth of melanoma  and TGF-β-mediated epithelial-mesenchymal transition (EMT) in mammary epithelial cells . p38 MAPK promoted the malignant phenotype of squamous carcinoma cells by regulating survival, proliferation and invasion . p38 MAPK plays a critical role in the H-Ras-induced invasive phenotype and migration in human breast epithelial cells . Short-term exposure to arsenic reportedly activated the p38 MAPK signaling pathway in non-transformed human mammary epithelial MCF 10A cells  and human uroepithelial SV-HUC-1 cells . Our study results suggested that p38α MAPK activation induced by arsenic exposure is heavily involved in neoplastic transformation (Figure 4). This finding supports the idea that p38α MAPK exhibits an oncogenic function and thus, p38α MAPK activation is linked with arsenic-induced cell transformation.
Based on an accumulation of evidence, AP-1 activation is closely associated with malignant transformation [24,36–39]. AP-1-mediated cyclin D1 expression plays an important role in arsenic-induced transformation of mouse epidermal cells  and c-Fos overexpression is correlated with cell transformation [39,40], late stage tumorigenesis , and tumor invasion . Our data showed that the activation of AP-1 and c-Fos promoter activity were closely associated with cell transformation induced by exposure to arsenic (Figure 3A, B, and Figure 5B). Moreover, phosphorylation of CREB (Figure 1, ,2B)2B) and CRE promoter activity (Figure 3D) were significantly increased with arsenic treatment but decreased in cells when p38α MAPK expression was blocked (Figure 5A, B). Previous reports indicated that CREB might be linked with liver cancer development , hepatocellular carcinomas , tumorigenesis of endocrine tissues  and acute myeloid cell transformation . Also, CREB phosphorylation by MSK1 reportedly mediated AP-1 activation or c-Fos gene expression in keratinocytes and promoted the growth of human epidermoid carcinoma cells [47,48]. We also found that phosphorylation and expression of MSK1, which is downstream of p38α MAPK, were increased in arsenic-treated transformed cells (Figure 1, ,2A).2A). The anchorage-independent colony formation ability induced by arsenic treatment was significantly decreased in BALB/c 3T3 cells deficient in MSK1 (Supplementary Figure 2A, B, C). Both AP-1 activation and c-Fos promoter activity were decreased in MSK1 knockdown cells (Supplementary Figure S2D). Taken together, AP-1 activation and increased c-Fos or CRE promoter activity and CREB expression are likely involved in arsenic-induced neoplastic transformation.
Overall, we demonstrated that total and phosphorylated p38α MAPK is required for cell transformation induced by 0.5 µM arsenic exposure. Arsenic exposure significantly increased AP-1 transcription activity and CREB activation, resulting in increased cell transformation. Taken together, our findings indicate that p38α MAPK could be a crucial target for cancer prevention or chemotherapy in arsenic-caused carcinogenesis.
This work was supported by The Hormel Foundation and National Institutes of Health grants ES106548 and CA172457.
NOTE: whole blots are presented in Supplementary Figures 3–9.