Upregulation of CYP1B1 Expression by Inflammatory Cytokines Is Mediated by the P38 MAP Kinase Signal Transduction Pathway

Carcinogenesis vol.35 no.11 pp.2534–2543, 2014 doi:10.1093/carcin/bgu190 Advance Access publication September 18, 2014 Upregulation of CYP1B1 expression by inflammatory cytokines is mediated by the p38 MAP kinase signal transduction pathway

Lenka Šmerdová1, Jana Svobodová1,2, Markéta Kabátková1,2, which, despite sharing ~40% homology with both CYP1A1 and Jiří Kohoutek3, Dalibor Blažek4, Miroslav Machala3 and CYP1A2, has several distinct properties, including its tissue-specific Jan Vondráček1,* pattern of expression. CYP1B1 metabolizes a number of procarcinogens, including polycyclic aromatic hydrocarbons, N-heterocyclic 1Department of Cytokinetics, Institute of Biophysics, Academy of Sciences of amines, arylamines, amino azo dyes and several other compounds (2). the Czech Republic, Brno 62165, Czech Republic, 2Institute of Experimental CYP1B1 messenger RNA (mRNA) has been found at significant lev- 3 Biology, Faculty of Science, Brno 61137, Czech Republic, Department of els in a number of extrahepatic human tissues, including kidney and Chemistry and Toxicology, Veterinary Research Institute, Brno 62100, Czech colon, and also in hormonal tissues (including prostate, ovary, uterus Republic and 4Central European Institute of Technology (CEITEC), Masaryk University, Brno 62500, Czech Republic and mammary gland) (2). The expression of CYP1B1 in hormonal tissues could be related to the important role of CYP1B1 in the catab- *To whom correspondence should be addressed. Tel: +420 541517168; olism of 17β-estradiol, whose product, 4-hydroxyestradiol, may con- Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 Fax: +420 41211293; tribute to estrogen-mediated carcinogenesis (3,4). CYP1B1 has been Email: [email protected] also proposed to have an impact on course of the tamoxifen therapy Cytochrome P450 1B1 (CYP1B1) is an enzyme that has a unique of hormone-dependent cancers since it catalyzes the conversion of tumor-specific pattern of expression and is capable of bioactivat- trans-4-hydroxytamoxifen to weakly estrogenic cis-4-hydroxytamox- ing a wide range of carcinogenic compounds. We have reported ifen (5). The presence of CYP1B1 in various fetal tissues points to its previously that coordinated upregulation of CYP1B1 by inflam- role in normal fetal development and it has been demonstrated that a matory cytokines, such as tumor necrosis factor-α (TNF-α) and mutation in the CYP1B1 is linked with the development of primary the aryl hydrocarbon receptor ligands, may increase bioactiva- congenital glaucoma (6). tion of promutagens, such as benzo[a]pyrene (BaP) in epithelial CYP1B1 has been found to be overexpressed in a wide range of cells. Here, we extend those studies by describing a novel mech- malignant tumors of different tissue origin, including breast, colon, anism participating in the regulation of CYP1B1 expression, lung, brain, skin, prostate, ovary and liver cancers (7–9). Based on which involves activation of the p38 mitogen-activated protein this phenomenon, it has been suggested that CYP1B1 may have an kinase (p38) and mitogen- and stress-activated protein kinase endogenous function in tumor cells and contribute to drug resistance 1 (MSK1). Using inhibitors of p38 and MSKs, as well as mouse (10). Upregulation of CYP1B1 often correlates with upregulation of embryonic cells derived from p38α-deficient and MSK1/2 double the aryl hydrocarbon receptor (AhR), its major transcriptional regula- knockout mice, we show here that TNF-α potentiates CYP1B1 tor, suggesting that AhR may participate in the maintenance of high upregulation via the p38/MSK1 kinase cascade. Effects of this baseline CYP1B1 levels in tumors (11,12). The AhR ligands, such inflammatory cytokine on CYP1B1 expression further involve the as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) or benzo[a]pyrene positive transcription elongation factor b (P-TEFb). The inhibi- (BaP), are potent CYP1B1 inducers and CYP1B1 plays an impor- tion of the P-TEFb subunit, cyclin-dependent kinase 9 (CDK9), tant role in bioactivation of BaP and formation of the reactive anti- which phosphorylates RNA polymerase II (RNAPII), prevented 7α,8α-dihydroxy-9α,10α-epoxy-7,8,9,10-tetrahydro-BaP (BPDE), the enhanced CYP1B1 induction by a combination of BaP and which binds covalently to cellular macromolecules, including DNA, inflammatory cytokine. Furthermore, using chromatin immuno- and causes mutations (13). Although activation of the AhR is a major precipitation assays, we found that cotreatment of epithelial cells mechanism responsible for CYP1B1 induction, additional signal- with TNF-α and BaP resulted in enhanced recruitment of both ing pathways have been implicated in either inducing CYP1B1 or in CDK9 and RNAPII to the Cyp1b1 gene promoter. Overall, these maintaining its constitutive level of expression (reviewed in ref. 14). results have implications concerning the contribution of inflam- However, the mechanisms of transcriptional or post-transcriptional matory factors to carcinogenesis, since enhanced CYP1B1 induc- regulation of CYP1B1 are still not fully clear. tion during inflammation may alter metabolism of exogenous Recently, we have observed that CYP1B1 is significantly upregu- carcinogens, as well as endogenous CYP1B1 substrates playing lated in epithelial cells cotreated by AhR ligands and inflammatory role in tumor development. cytokine tumor necrosis factor-α (TNF-α) (15–17). We have also found that via CYP1B1 upregulation, inflammatory mediators may accelerate BaP metabolism and production of its genotoxic metabolites (18). This type of transcriptional regulation is in sharp contrast Introduction with a majority of other CYPs involved in the metabolism of carcinogens or in drug clearance, including CYP1A1 and CYP1A2 enzymes, Cytochrome P450 family 1 (CYP1) heme-thiolate monooxygenases whose activity/expression is generally downregulated by infection participate in the bioactivation of a variety of potent hydrophobic pro- or inflammation (19). It is presently unclear what mechanisms are carcinogens in the liver, lungs and other organs with a high metabolic responsible for the opposite CYP1B1 regulation under inflammatory capacity (1). CYP1B1 is a unique member of this enzyme family, conditions. Chronic inflammation, accompanied with deregulated and sustained Abbreviations: AhR, aryl hydrocarbon receptor; BaP, benzo[a]pyrene; CDK9, production of TNF-α, has been proposed to contribute to carcinogen- cyclin-dependent kinase 9; ChIP, chromatin immunoprecipitation; CREB, esis via mechanisms participating in tumor promotion, progression cyclic adenosine 3′,5′-monophosphate response element-binding protein; and, in some cases, also tumor initiation (20). Deregulated expres- CYP1, cytochrome P450 family 1; CYP1B1, cytochrome P450 1B1; DRB, sion/activity of enzymes, which bioactivate procarcinogens, may 5,6-dichlorobenzimidazole-1-β-d-ribofuranoside; EDTA, ethylenediaminetet- provide one of the mechanisms potentially contributing to tumor ini- raacetic acid; EGF, epidermal growth factor; MAPKAPK-2, MAP kinase-acti- tiation/development. In the present study, we identified the signaling vated protein kinase 2; MEF, mouse embryonic fibroblast; mRNA, messenger pathway responsible for the upregulation of CYP1B1 expression by RNA; MSK, mitogen- and stress-activated protein kinase; NF-κB, nuclear factor κ-light-chain-enhancer of activated B cells; PBS, phosphate-buffered saline; inflammatory cytokine(s). Our results suggest that activation of the P-TEFb, positive transcription elongation factor b; RNAPII, RNA polymer- p38 mitogen-activated protein (MAP) kinase p38/mitogen- and stress- ase II; SDS, sodium dodecyl sulfate; siRNA, short interfering RNA; TCDD, activated protein kinase (MSK) kinase cascade is responsible for 2,3,7,8-tetrachlorodibenzo-p-dioxin; TNF-α, tumor necrosis factor-α. the observed CYP1B1 upregulation by inflammatory cytokines. We

© The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 2534 P38 MAP kinase regulates CYP1B1 expression also demonstrate that TNF-α contributes to the increased CYP1B1 was confirmed by performing the qPCR experiments under same conditions expression via enhanced recruitment of the positive elongation fac- with the respective pCR4-ZeroBlunt-TOPO vectors with inserted PCR prod- tor b (P-TEFb) and RNA polymerase II (RNAPII) to the CYP1B1 ucts. The conditions of assay were based strictly on the protocol provided by promoter. This mechanism may help to explain a unique pattern of the supplier of the one-step qRT-PCR kit. The amplifications were run on the Rotor-Gene 6000 (Corbett Research, Sydney, Australia) using the following CYP1B1 regulation during inflammation that alters the metabolic program: reverse transcription at 50°C for 30 min and initial activation step at activation of procarcinogens, such as BaP. 95°C for 15 min, followed by 35 cycles of 95°C for 10 s and 60°C for 40 s. The absolute quantification was performed using pCR4-ZeroBlunt-TOPO vector Materials and methods with inserted mouse Cyp1a1 and Cyp1b1 PCR products (Generi-Biotech, Hradec Králové, Czech Republic) as standards. The total RNA content was Chemicals and reagents determined using NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA). BaP (purity 99.9%) was from Ehrenstorfer (Augsburg, Germany). TCDD (purity 99%) was purchased from Cambridge Isotope Laboratories (Andover, Short interfering RNA transfections MA). Stock solutions were prepared in dimethyl sulfoxide (Merck, Darmstadt, Cells were plated at a density of 20,000 cells/cm2 in 24-well plates in culture Germany) and stored in the dark. Recombinant rat and human TNF-α, medium without antibiotics. After 24 h cultivation, transfections were per- purified murine epidermal growth factor (EGF) and recombinant EGF formed, using rat p65/RelA short interfering RNA (siRNA), rat AhR siRNA (expressed in Escherichia coli) were obtained from Sigma–Aldrich (Prague, or control siRNA, which was directed against mRNA encoding the red fluores- Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 Czech Republic). Anisomycin was purchased from Sigma–Aldrich and dis- cence protein DsRed from the coral Discosoma). All siRNAs were provided solved in dimethyl sulfoxide. Inhibitors of p38 MAP kinase (SB202190) by Ambion (Foster City, CA), based on the sequences reported previously and cyclin-dependent kinase 9 (CDK9) (5,6-dichlorobenzimidazole-1-β-d- (17,24). The transfections were performed in a total volume of 600 μl contain- ribofuranoside; DRB), both from Sigma–Aldrich, and inhibitor of MSK activ- ing 100 pmol siRNA and 1 μl of Lipofectamine 2000 (Invitrogen) according to ity, H89 (Enzo Life Sciences, Farmingdale, NY), were dissolved in dimethyl the manufacturer’s instructions. Transfection mix was removed 24 h later and sulfoxide and stored at −80°C prior to their use. cells were cultivated for another 24 h in media with antibiotics, followed by exposure to test compounds. Whole cell lysates were prepared and subjected Cell culture to western blot analysis. RLE-6TN rat lung epithelial cells (American Type Culture Collection, Manassas, VA) were cultured in Ham´s F-12 medium (Invitrogen, Carlsbad, Chromatin immunoprecipitation CA) with 2 mM l-glutamine, supplemented with bovine pituitary extract The chromatin immunoprecipitation (ChIP) assays were performed as (10 μg/ml), insulin (5 μg/ml), insulin-like growth factor (2.5 ng/ml), trans- described previously (25). WB-F344 cells were incubated with given regi- ferrin (1.25 μg/ml), EGF (2.5 ng/ml) and 5% heat-inactivated fetal bovine mens for 2 h, then washed with PBS and cross-linked with 1% formalde- serum. WB-F344 rat liver epithelial cells, kindly provided by James E.Trosko hyde/PBS solution for 10 min at room temperature. The cross-linking was (Michigan State University, East Lansing, MI), were cultured in Dulbecco's quenched for 5 min with glycine (final concentration 125 mM). After two modified Eagle's medium/F-12 medium (Invitrogen), supplemented with extensive washes with ice-cold PBS, cells were lysed in sonication buffer 15 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid, l-glutamine, [0.5% SDS, 20 mM Tris–HCl pH 8.0, 2 mM ethylenediaminetetraacetic acid penicillin/streptomycin (100 000 U/l and 100 mg/l, respectively) and 5% heat- (EDTA), 0.5 mM ethyleneglycol-bis(aminoethylether)-tetraacetic acid and inactivated fetal bovine serum. Only the cells at passage levels 15–22 were protease inhibitor—0.5 mM phenylmethylsulfonyl fluoride] and DNA was used for the study. HepaRG human hepatoma cells (Biopredic International, sonicated to fragments shorter than 500 bp. For immunoprecipitation, pro- Rennes, France) were cultured in original basal hepatic cell medium with tein extract was precleared with Protein A/G PLUS sepharose (Santa Cruz additives for 710 growth medium (Biopredic). Immortalized p38α knockout Biotechnology, Santa Cruz, CA) and then incubated overnight with control mouse embryonic fibroblasts (p38−/− MEFs) (21) were kindly provided by (immunoglobulin G), CDK9 or RNAPII antibodies, followed by 2 h incuba- Ángel R.Nebreda (Institute for Research in Biomedicine, Barcelona, Spain). tion with Protein A/G PLUS sepharose preblocked with 0.3 mg/ml of sheared MSK1/2 double knockout MEFs (MSK1/2−/− MEFs) (22) were kindly pro- salmon sperm DNA (Ambion) and 1 mg/ml of ultrapure bovine serum albu- vided by Simon J.Arthur (University of Dundee, Dundee, UK). All knockout min (Ambion). Rabbit polyclonal antibodies against CDK9 (# sc-484) and MEFs and respective wild-type MEFs were maintained in Dulbecco's modified RNAPII (# sc-899) were purchased from Santa Cruz Biotechnology. The Eagle's medium supplemented with 1% l-glutamine, 1% penicillin/streptomy- antibodies used in the ChIP experiments are well characterized, and they cin and 10% heat-inactivated fetal bovine serum. All cell lines were incubated have been routinely used in our previous work (25) as well as in studies in a humidified atmosphere of 5% CO2 at 37°C. Other than indicated, all tissue from other labs (26,27). The sepharose was washed once with low salt culture reagents were obtained from Sigma–Aldrich. buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris–HCl pH 8.0, 150 mM NaCl) followed by a triple wash with the same buffer containing Western blot analysis 500 mM NaCl and one wash with lithium buffer (2 mM EDTA, 20 mM Tris Cells were washed with phosphate-buffered saline (PBS) and lysed using 1% pH 8.0, 250 mM LiCl, 1% NP-40 and 1% sodium deoxycholate) and two sodium dodecyl sulfate (SDS) lysis buffer (10% glycerol, 100 mM Tris, pH washes with Tris–EDTA buffer. The immunoprecipitates were eluted with 7.4). Protein concentration was estimated using Bio-Rad DC Protein Assay 1% SDS and 100 mM sodium bicarbonate at room temperature for 15 min (Bio-Rad Laboratories, Hercules, CA). Equal amounts of protein (20 μg) were and cross-linking was reversed with 200 mM NaCl for 5 h at 65°C. Proteins separated on 10% SDS–polyacrylamide gel and transferred onto a polyvi- were digested with Proteinase K (Sigma–Aldrich) and DNA extracted with nylidene difluoride membrane. After incubation with primary and secondary phenol/chloroform/isoamyl alcohol (25:24:1) and precipitated by ethanol antibodies, detection was performed using ECLPlus western blotting detection overnight. A fraction of precipitated DNA was used in qPCR reactions with system (GE Healthcare, Little Chalfont, UK). Further details on the primary the SYBR Green mix (Sigma–Aldrich). The primers covering rat Cyp1b1 antibodies, their vendors, antibody dilutions, incubation and blocking con- gene promoter region (Supplementary Table 2, available at Carcinogenesis ditions are provided in Supplementary Table 1, available at Carcinogenesis Online) were designed and selected according to the previously reported Online. primers used for ChIP analysis of RNAPII binding to human CYP1B1 gene promoter (28). The primers for control intergenic region (Supplementary RNA isolation and real-time quantitative reverse transcription–PCR Table 2, available at Carcinogenesis Online) were designed upstream of Total RNA was isolated using the NucleoSpin® RNA II purification kit Cyp1b1 gene in region without any known genes. The signals from the con- (Macherey-Nagel, Düren, Germany). Quantification of gene expression was trol primers were considered a background. The qPCR reactions were run on performed using the SuperScript® III Platinum® One-Step Quantitative the Roche LightCycler 480II using the following conditions: initial activa- RT-PCR kit (Life Technologies, Carlsbad, CA). The amplification of the sam- tion step at 94°C for 2 min, followed by 42 cycles at 95°C for 30 s, 58 or ples was carried out in a final volume of 20 μl in a reaction mixture contain- 61°C for 30 s and 72°C for 30 s. ing 10 μl of 2× SuperScript Reaction Mix, 0.4 μl of SuperScript III Platinum Taq Mix, 1 μl of a solution of primers and probe, 6.6 μl of water and 2 μl Data analysis of RNA sample. The final concentration of each primer was 0.2 μM and the Data were expressed as means ± SD of at least three independent repeats. concentration of probe was 0.1 μM. The sequences of primers and respective Comparisons between treatments were made by one-way analysis of vari- TaqMan hydrolysis probes are listed in Supplementary Table 2, available at ance, with post hoc Tukey test, or Student’s t-test. If the variances were Carcinogenesis Online. For each cellular model, we have tested previously non-homogeneous, non-parametric Mann–Whitney U-test or Kruskal– several housekeeping genes and selected the most suitable one, or we selected Wallis analysis of variance was used. A P value of <0.05 was considered them based on available literature data (23). The specificity of target genes significant. 2535 L.Šmerdová et al.

Results cells and in rat lung epithelial RLE-6TN cells, with the effects on the human liver HepaRG cell line, which is a frequently used model Inflammatory cytokine enhances CYP1B1 induction in both rodent for analysis of expression of human liver xenobiotic-metabolizing and human cells enzymes (29). We found that TNF-α enhanced induction of CYP1B1 We wanted to demonstrate that differential modulation of in all three of the studied cell lines treated with BaP (Figure 1). CYP1A1/1B1 is a more general phenomenon that is common for both Similar to our previous findings in rodent cells (15,16), induction human and rodent cells. Therefore, we compared the effects of TNF-α of CYP1A1 was suppressed both at mRNA and protein levels upon on BaP-induced CYP1B1 expression in rat liver epithelial WB-F344 TNF-α treatment in HepaRG cell line (Figure 1). Moreover, in this Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021

Fig. 1. TNF-α potentiates induction of CYP1B1 by BaP in rodent and human cellular models. (A) Rat liver epithelial WB-F344 and lung epithelial RLE-6TN cells were treated for 6 h with dimethyl sulfoxide (DMSO) (0.1%, solvent control), recombinant rat TNF-α (20 ng/ml), BaP (1 μM) or BaP + TNF-α. CYP1B1 protein and mRNA levels were determined by western blot analysis and quantitative reverse transcription–PCR (qRT–PCR), respectively. All western blotting results shown here are representative of at least three independent experiments; β-actin was used as a loading control. The qRT–PCR results of CYP1B1 mRNA detection were normalized to maximum CYP1B1 mRNA induction in the BaP + TNF-α group [the respective mean fold change relative to control (±SD) was 25.5 (±5.5) for WB-F344 cells and 18.9 (±6.7) for RLE-6TN cells, respectively] and then expressed as means ± SD of three independent experiments. A significant difference between BaP- and BaP + TNF-α-treated samples (*P < 0.05). A significant difference between BaP- and BaP + TNF-α-treated samples (**P < 0.01). (B) Human liver epithelial HepaRG cells were treated for 6 h by DMSO (0.1%, solvent control), recombinant human TNF-α (20 ng/ml), BaP (10 μM) or BaP + TNF-α. CYP1B1 and CYP1A1 protein and mRNA levels were determined by western blot analysis and qRT–PCR, respectively. The qRT–PCR results of CYP1B1 mRNA detection were normalized to maximum CYP1B1 mRNA induction in the BaP + TNF-α group [the respective mean fold change relative to control (±SD) was 82.9 (±45.2)] and then expressed as means ± SD of three independent experiments. The qRT–PCR results of CYP1A1 mRNA detection were normalized to maximum CYP1A1 mRNA induction in the BaP group [the respective mean fold change relative to control (±SD) was 162.7 (±57.8)] and then expressed as means ± SD of three independent experiments. A significant difference between BaP- and BaP + TNF-α-treated samples (*P < 0.05). A significant difference between BaP- and BaP + TNF-α-treated samples (**P < 0.01). 2536 P38 MAP kinase regulates CYP1B1 expression cell line, CYP1A2 mRNA induction was also repressed by TNF-α κ-light-chain-enhancer of activated B cells (NF-κB), which plays a treatment (data not shown). Finally, we used human breast carcinoma prominent role in TNF-α-dependent effects (30) and which has been MCF-7 cells to demonstrate that in this frequently used model for suggested to contribute to the deregulation of the AhR transcriptional analysis of CYP1B1 transcriptional regulation, BaP and TNF-α coor- targets through direct physical interaction with the AhR (31). However, dinately increase CYP1B1 mRNA levels significantly higher than BaP siRNA targeting the p65/RelA subunit did not prevent the synergistic alone (Supplementary Figure 1, available at Carcinogenesis Online). effect of TCDD and TNF-α on CYP1B1 induction (Supplementary Together, these results indicate that the inflammatory signaling does Figure 2B, available at Carcinogenesis Online). These results sug- not repress CYP1B1 expression in human cells and that an inflamma- gested that the yet unidentified, signaling pathway(s) that modulate tory cytokine may even potentiate CYP1B1 expression induced by the CYP1B1 induction would be independent of NF-κB activation. AhR ligand (a prototypical polycyclic aromatic hydrocarbon), at least Therefore, we next focused on the identification of a branch of in some human experimental models in vitro, which is in a marked TNF-α signaling that might be responsible for the observed CYP1B1 contrast with other CYP1 family members. For further work, we then upregulation. The MAP kinase signaling cascades are activated by employed the rat epithelial cell lines since in these cells we have dem- TNF-α, and they may both modulate the NF-κB-dependent activ- onstrated previously a functional role for CYP1B1 upregulation in ity and mediate further, NF-κB-independent, effects of this cytokine deregulation of metabolism and genotoxicity of a model genotoxic (30). In order to confirm that TNF-α activates MAP kinase signaling AhR ligand, BaP (15,16,18). in WB-F344 cells, we detected phosphorylated forms of both ERK1/2 Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 and p38 kinases that were induced by TNF-α and/or BaP at different Activation of MAP kinase signaling pathways potentiates the BaP- time points (Figure 2A). We found that TNF-α was a potent activa- mediated CYP1B1 induction tor of p38 pathway while it simultaneously induced a much weaker AhR is the principal transcription factor responsible for induction of ERK1/2 activation. The phospho-p38 levels increased particularly CYP1 family enzymes expression (1). As shown in Supplementary between 5 and 30 min after the treatment, and this effect was slightly Figure 2, available at Carcinogenesis Online, the AhR knockdown amplified by coadministration of BaP. by siRNA fully prevented induction of the CYP1A1 protein by the In order to elucidate whether the activation of MAP kinase cas- prototypical AhR ligand TCDD in WB-F344 liver epithelial cells. cades would be sufficient to mimic the effects of TNF-α, we then used Nevertheless, the AhR knockdown was not complete and when we EGF as a model MAP kinase activator. We have observed previously evaluated the effects of TCDD or TCDD in combination with TNF-α in WB-F344 cells that high doses of EGF induce cell proliferation of (Supplementary Figure 2A, available at Carcinogenesis Online), we WB-F344 cells in a dose-dependent manner (32) and that EGF acti- observed that the combined treatment still induced some CYP1B1 vates both p38 and ERK1/2 MAP kinases in the same cell line (33). protein, in contrast to TCDD alone. This indicated that although Based on these data and further preliminary experiments, we used AhR plays a major role in the regulation of CYP1B1 expression, EGF (50 ng/ml of mouse EGF purified from submaxillary glands) to there exists an additional signaling pathway(s) that contributes to study its effects on CYP1B1 induction. The EGF alone had no effect the transcriptional regulation of CYP1B1 expression during inflam- on induction of CYP1B1 transcription; however, in combination with mation. A potential candidate was transcription factor nuclear factor BaP, it significantly increased induction of both protein and mRNA

Fig. 2. Activation of MAP kinase signaling contributes to upregulation of CYP1B1 expression. (A) WB-F344 cells were treated with dimethyl sulfoxide (DMSO) (0.1%, solvent control), TNF-α (20 ng/ml), BaP (1 μM), BaP + TNF-α, anisomycin (5 μM, a positive control for p38 activation) or EGF (50 ng/ml), for the times indicated, and levels of phosphorylated p38 (pp38) and ERK1/2 (pERK1/2) kinases were determined by western blot analysis, using the respective phospho-specific antibodies. Western blotting results shown here are representative of at least three independent experiments;β -actin was used as a loading control. (B and C) WB-F344 cells were incubated for 6 h with DMSO (0.1%, solvent control), BaP (1 μM), EGF (50 ng/ml) or BaP + EGF, and CYP1B1 protein (B) and mRNA (C) levels were determined by western blot analysis and quantitative reverse transcription–PCR (qRT–PCR), respectively. The western blotting results shown here are representative of at least three independent experiments; β-actin was used as a loading control. The qRT–PCR results were normalized to maximum CYP1B1 mRNA induction in the BaP + EGF group [the respective mean fold change relative to control (±SD) was 8.6 (±3.1)] and then expressed as means ± SD of three independent experiments. A significant difference between BaP- and BaP + EGF-treated samples (*P < 0.05). 2537 L.Šmerdová et al. levels of CYP1B1 compared with BaP treatment alone (Figure 2B and residues that get phosphorylated by p38 kinase in vitro and is essen- C). A similar effect on CYP1B1 expression was observed also when tial for the MAPKAPK-2 function (34). As shown in Supplementary using recombinant EGF expressed in E.coli, even at lower concentra- Figure 4, available at Carcinogenesis Online, SB202190 fully pre- tions (Supplementary Figure 3, available at Carcinogenesis Online). vented phosphorylation of MAPKAPK-2 induced by TNF-α (both These results indicated that strengthening of the BaP-induced when applied alone and in combination with BaP). In order to further CYP1B1 expression by TNF-α and EGF may involve common exclude any non-specific effects of this inhibitor, including a direct downstream signaling modules, presumably MAP kinases. Since p38 modulation of AhR action (35), we further compared induction of kinase activation induced by TNF-α was more pronounced than phos- CYP1B1 gene expression in wild-type MEFs and in p38−/− MEFs. phorylation of ERK1/2 kinases, we hypothesized that the p38 MAP As expected, we found that the induction of expression of CYP1B1 kinase cascade would contribute to CYP1B1 upregulation by TNF-α. induced by BaP + TNF-α was significantly reduced in p38−/− cells (Figure 3B). Moreover, the constitutive level of CYP1B1 mRNA was p38 MAP kinase plays a role both in the CYP1B1 upregulation significantly lower in p38−/− MEFs than in their wild-type counter- by inflammatory cytokine and in the control of basal CYP1B1 parts (Figure 3C). Importantly, no such effect was observed in the expression case of CYP1A1 (Figure 3C), thus confirming that CYP1B1 is a In order to confirm the causal role of p38 in CYP1B1 upregulation, unique CYP1 family member, which can be upregulated through p38 we next used the SB202190 inhibitor in WB-F344 and RLE-6TN MAP kinase signaling. Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 cells. We found that SB202190 suppressed the synergistic effects of BaP and TNF-α on CYP1B1 induction in both cell lines (Figure 3A). MSK activation mediates effects of TNF-α on CYP1B1 expression In order to confirm that the SB202190 inhibitor at the given concen- Next, we attempted to identify further downstream mechanisms con- tration prevented p38 activity in our cell model, we further performed tributing to the effects of p38 MAP kinase on CYP1B1 expression. the western blotting detection of phosphorylation of MAP kinase-acti- Since p38 kinase has been implicated in mRNA stabilization of a wide vated protein kinase 2 (MAPKAPK-2) at T384. The MAPKAPK-2 is range of transcriptional targets of TNF-α, including various inflam- a direct target of p38 kinase and T384 is one of the three amino acid matory mediators (36), we first hypothesized that the stabilization

Fig. 3. The p38 MAP kinase plays a role in both TNF-α-mediated potentiation of CYP1B1 induction and in maintaining the constitutive level of CYP1B1 expression. (A) WB-F344 and RLE-6TN cells were pretreated for 1 h with SB202190 inhibitor (10 μM) and then treated for 6 h with dimethyl sulfoxide (DMSO) (0.1%, solvent control), TNF-α (20 ng/ml), BaP (1 μM) or BaP + TNF-α. CYP1B1 protein levels were determined by western blot analysis; β-actin was used as a loading control. (B) Wild-type and p38−/− MEFs cells were incubated for 6 h with DMSO (0.1%, solvent control), TNF-α (20 ng/ml), BaP (1 μM) or BaP + TNF-α. CYP1B1 mRNA levels were determined by quantitative reverse transcription–PCR (qRT–PCR). The qRT–PCR results were normalized to maximum CYP1B1 mRNA induction in the BaP + TNF-α group [the respective mean fold change relative to control (±SD) was 5.4 (±2.3)] and then expressed as means ± SD of three independent experiments. (C) Absolute quantification of CYP1B1 and CYP1A1 mRNA copies in wild-type and p38−/− MEFs. Absolute numbers were calculated from calibration curves plotted using the respective standards. The results of qRT–PCR were expressed as means ± SD of three independent experiments. A significant difference between a wild-type and the respective p38−/− MEFs sample (**P < 0.01). 2538 P38 MAP kinase regulates CYP1B1 expression of CYP1B1 mRNA could be involved in the observed effects, as MSK1/2 double knockout mice. We found that induction of CYP1B1 CYP1B1 has been shown previously to contain AU-rich element mRNA by BaP + TNF-α was significantly reduced in MSK1/2−/− cells motifs located within its 3′ untranslated region (37). However, the in comparison with wild-type MEFs (Figure 4C). Similar to p38−/− SB202190 inhibitor had no effect on the rate of CYP1B1 mRNA MEFs, constitutive CYP1B1 mRNA level in MSK1/2−/− cells was sig- decay (Supplementary Figure 5, available at Carcinogenesis Online). nificantly lower than in wild-type cells (Figure 4D). Again, no such Among the wide range of p38 downstream substrates, MSKs effect was observed, when we compared the CYP1A1 mRNA level in represent protein kinases implicated in transcriptional regulation wild-type MEFs with MSK1/2−/− cells (Figure 4D). Collectively, these of various inflammatory genes (38). To study the potential involve- data suggested that MSKs mediate the effects of TNF-α on CYP1B1 ment of MSKs in CYP1B1 regulation, we first used H89 inhibitor in expression. WB-F344 cells (Figure 4A), and we found that H89 partly prevented synergistic effects of BaP and TNF-α on CYP1B1 induction. In order Involvement of P-TEFb in regulation of CYP1B1 gene expression to confirm that MSKs are indeed activated by TNF-α through p38 The positive transcription elongation factor b (P-TEFb), which is MAP kinase activity in our experimental settings, we next detected composed of CDK9 and either cyclin T1 or T2 (39), activates tran- phosphorylation of MSK1 (Figure 4B). We found that TNF-α induced scriptional elongation and RNA processing of numerous genes by phosphorylation of MSK1 after 15 min incubation and that this effect RNAPII (40,41). Recently, activation of p38/MSK1 kinase cascade was abolished by the SB202190 pretreatment. Nevertheless, due to has been implicated in recruitment of CDK9 to the cyclooxygenase-2 Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 the off-target effects of H89 (such as inhibition of protein kinase A), gene (PTGS2) promoter (42). In order to investigate the possible func- its use cannot unequivocally prove the involvement of MSKs in the tion of P-TEFb in CYP1B1 gene upregulation in WB-F344 cells, we observed effects. Complementary data from MSK-knockout experi- used DRB, an inhibitor of CDK9, which blocks RNAPII-dependent ments have been suggested to be necessary to confirm the role of transcription (43,44). We found that inhibition of CDK9 prevented MSKs (38). Therefore, in order to confirm the functional role of enhanced induction of CYP1B1 mRNA (Figure 5A) and protein MSKs in CYP1B1 regulation, we next used MEFs derived from (Figure 5B) by combined BaP + TNF-α treatment. No such effect was

Fig. 4. Signaling through MSK1 contributes to CYP1B1 upregulation. (A) WB-F344 cells were pretreated for 1 h with H89 inhibitor (10 μM) and then treated for 6 h with dimethyl sulfoxide (DMSO) (0.1%, solvent control), TNF-α (20 ng/ml), BaP (1 μM) or BaP + TNF-α. CYP1B1 protein levels were determined by western blot analysis; β-actin was used as a loading control. (B) WB-F344 cells were pretreated for 1 h with SB202190 inhibitor (10 μM) and then treated for 15 min with DMSO (0.1%, solvent control), TNF-α (20 ng/ml), BaP (1 μM) or BaP + TNF-α. The levels of Thr581-phosphorylated MSK1 (pMSK1) were determined by western blot analysis. All western blotting results shown in (A) and (B) are representative of three independent experiments; β-actin was used as a loading control. (C) Wild-type and MSK1/2−/− MEFs were incubated for 6 h with DMSO (0.1%, solvent control), TNF-α (20 ng/ml), BaP (1 μM) or BaP + TNF- α. CYP1B1 mRNA levels were determined by quantitative reverse transcription–PCR (qRT–PCR). The qRT–PCR results were normalized to maximum CYP1B1 mRNA induction in the BaP + TNF-α group [the respective mean fold change relative to control (±SD) was 3.6 (±0.3)] and then expressed as means ± SD of three independent experiments. (D) Absolute quantification of CYP1B1 and CYP1A1 mRNA copies in wild-type and MSK1/2−/− MEFs cells. Absolute numbers were calculated from calibration curves plotted using the respective standards. The results of qRT–PCR were expressed as means ± SD of three independent experiments. A significant difference between a wild-type and the respective MSK1/2−/− MEFs sample (**P < 0.01). 2539 L.Šmerdová et al. Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021

Fig. 5. CDK9 activity is required for enhanced CYP1B1 induction. WB-F344 cells were treated for 6 h with dimethyl sulfoxide (DMSO) (0.1%, solvent control), TNF-α (20 ng/ml), DRB inhibitor (30 μM), TNF-α + DRB, BaP (1 μM), BaP + DRB, BaP + TNF-α or BaP + TNF-α + DRB. CYP1B1 mRNA (A) and protein (B) levels were determined by quantitative reverse transcription–PCR (qRT–PCR) and western blot analysis, respectively. The western blotting results shown here are representative of least three independent experiments; β-actin was used as a loading control. The qRT–PCR results were normalized to maximum CYP1B1 mRNA induction in the BaP + TNF-α group [the respective mean fold change relative to control (±SD) was 11.6 (±6.9)] and then expressed as means ± SD of three independent experiments. A significant difference between the BaP + TNF-α-treated sample and the respective DRB-treated sample (**P < 0.01). (C) WB-F344 cells were treated for 2 h with DMSO (0.1%, solvent control), TNF-α (20 ng/ml), BaP (1 μM) or BaP + TNF-α. ChIP assays were performed with antibodies recognizing either CDK9 or RNAPII, and qPCR was carried out to quantify the associated coding regions of the rat Cyp1b1 gene in the immunoprecipitated complexes with two primer sets corresponding to promoter sequences and intergenic region upstream of Cyp1b1 promoter. Immunoglobulin G corresponds to the empty beads control. A representative experiment out of two independent repeats is presented. observed for CYP1A1 mRNA (Supplementary Figure 6, available at specific xenobiotic response elements within the enhancer region and Carcinogenesis Online), which pointed to the specific role of CDK9 in leading to initiation of transcription, whereas p38/MSK signaling may CYP1B1 regulation during inflammation. Overall, these results sug- then provide an additional stimulus for transcriptional elongation by gested that P-TEFb activity is required for enhanced CYP1B1 induc- enhancing recruitment of both CDK9 and RNAPII to the Cyp1b1 gene tion by BaP + TNF-α. To further analyze the association between promoter. P-TEFb and CYP1B1 upregulation, we next performed ChIP assays in order to examine recruitment of CDK9 and RNAPII to the rat Cyp1b1 Discussion gene promoter region. As shown in Figure 5C, cotreatment of cells with BaP + TNF-α resulted in enhanced/prominent recruitment of The tumor- and cell-specific pattern of expression of CYP1B1, both CDK9 and RNAPII to the rat Cyp1b1 gene promoter, compared together with its ability to bioactivate a wide range of carcinogens, with individual treatments. These experiments suggested that P-TEFb has previously led several authors to advocate its use as a potential activity may contribute to an increased expression of CYP1B1 during target for anticancer drugs/chemoprevention or as a possible tumor inflammation. Together, as summarized inFigure 6 , the results of the biomarker (45). Although its expression is mainly controlled by AhR, present study indicate that inflammatory cytokines, through activating other transcription factors participate in tissue- and stimulus-specific the p38/MSK cascade, could play a costimulatory role in the regula- regulation of CYP1B1. In a majority of normal tissues, CYP1B1 tion of CYP1B1 expression. Activation of AhR results in binding to is regulated through AhR/AhR nuclear translocator activity. This 2540 P38 MAP kinase regulates CYP1B1 expression

experiments. Previously, it has been postulated that AhR and NF-κB compete for transcriptional coactivators, such as p300/CBP or SRC- 1, and that this competition may lead to the activation of one pathway while simultaneously repressing the other pathway, which may account for the inhibition of CYP1A1 induction during inflammation (31,54). However, the siRNA targeting the p65/RelA subunit of NF-κB had no effect on CYP1B1 induction. These results indicated that the synergistic effect of inflammatory mediator and AhR ligands on CYP1B1 induction is controlled by additional signaling pathways that do not involve NF-κB activity. Activation of TNF-α receptors has been shown to activate a wide range of signaling pathways in a cell-dependent manner (30). Among them, MAP kinase signaling has been implicated both in the TNF-α-activated pathways and in medi- ating the toxic effects of polycyclic aromatic hydrocarbons (33,55). Interestingly, apart from p38 kinase contributing to the regulation of the AhR transcriptional targets (35), the activity of CYP1B1 itself Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 (presumably through generation of reactive oxygen species) has been recently shown to contribute to MAP kinase activation (56). In the present study, we observed that p38 MAP kinase is rapidly activated by both TNF-α and TNF-α + BaP cotreatment. Importantly, activation of MAP kinases by EGF mimicked the effects of TNF-α on CYP1B1 upregulation. Together, these results indicated that p38 MAP kinase could mediate the synergistic effects of TNF-α and BaP on CYP1B1 induction. This hypothesis was further supported by our observation that inhibition of p38 MAP kinase by its specific inhibitor prevented synergistic effects of TNF-α and BaP. Finally, using p38−/− MEFs, we confirmed the significant role of p38 MAP kinase in regulation of CYP1B1 expression. Not only was the enhanced induction of Fig. 6. Inflammatory signaling, through activating the p38/MSK cascade, CYP1B1 prevented in p38−/− MEFs but the basal level of CYP1B1 could play a costimulatory role in regulation of CYP1B1 expression. mRNA was also significantly lower in p38−/− cells compared with Activation of AhR (left) results in binding to specific xenobiotic response elements within enhancer region and initiation of transcription, whereas p38/ wild-type MEFs. The p38 kinase has been suggested to be directly MSK signaling (right) may provide an additional stimulus for transcriptional involved in regulation of nuclear/cytoplasmic localization of the AhR elongation by enhancing recruitment of both CDK9 and RNAPII to Cyp1b1 (57). However, the effect of p38 on CYP1B1 expression was unlikely gene promoter. ARNT, AhR nuclear translocator; Hsp90, heat shock protein to be mediated via its direct impact on the AhR because we did not 90; p23, p23 co-chaperone; p38, p38 mitogen-activated protein kinase; RNA observe any significant differences in AhR levels, or its intracellu- Pol II, RNA polymerase II; XAP2, X-associated protein 2 (AIP/ARA9); lar localization, between p38−/− cells and wild-type MEFs (data not XRE, xenobiotic responsive element. shown). The substrates of p38 MAP kinases comprise downstream pro- heterodimer activates transcription by specifically binding to xeno- tein kinases, nuclear proteins (including both transcription factors biotic/dioxin response elements that are present in CYP1B1 enhancer and regulators of chromatin remodeling) and cytosolic proteins, region across different species (46). In contrast, CYP1B1 levels in which regulate processes such as protein degradation and localiza- steroidogenic tissues are primarily determined by hormones elevat- tion, mRNA stability, endocytosis, apoptosis, cytoskeleton dynamics ing intracellular cyclic adenosine 3′,5′-monophosphate (47,48), via or cell migration (58). The regulation of mRNA stability has been activation of the cyclic adenosine 3′,5′-monophosphate response ele- shown to play an important role in the modulation of mRNA levels of ment-binding protein (CREB) or activation protein 1 (49). In target various target genes by p38. The degradation of mRNA is controlled tissues of estrogens, the regulation of CYP1B1 is mainly controlled through AU-rich element motifs, located within the 3′-untranslated through the activity of estrogen receptor-α (50). Epigenetic mecha- region, which interact with specific AU-rich element-binding proteins nisms, post-transcriptional modifications and degradation pathways that can stabilize various transcripts, such as HuR (36). Importantly, further contribute to CYP1B1 regulation (reviewed in ref. 14). The p38 MAP kinase has been implicated in the regulation of the mRNA activities and expression of a majority of CYPs are downregulated by half-life of a number of inflammatory mediators, including cyclooxy- infection and/or inflammation 19( ), which, in turn, has a major impact genase-2, TNF-α, interleukin-8 and others (59–62), and we have on drug efficiency/toxicity (51). Contrary to that, it has been observed recently observed that p38 activity contributes to synergistic upregula- in a number of cell models that inflammatory mediators may even tion of proinflammatory genes by TNF-α and BaP (16). Therefore, we enhance CYP1B1 expression (reviewed in ref. 52) and this may lead first hypothesized that p38 activation may increase CYP1B1 mRNA to stimulation of metabolism and bioactivation of carcinogens, such stability. However, we found that unlike cyclooxygenase-2 mRNA, as BaP (18). It has been demonstrated that CYP1B1 is upregulated CYP1B1 mRNA decay is not enhanced upon inhibition of p38 MAP in human end-stage liver disease (which is characterized by dereg- kinase. This indicated that increased levels of CYP1B1 mRNA and ulated inflammation and predisposes for liver cancer development) protein, mediated by p38 kinase activation, are not due to the stabili- (7) or that upregulation of CYP1B1 is associated with inflammation zation of CYP1B1 mRNA. during rat esophageal tumorigenesis (53). Thus, both in vitro and in MSKs are p38 MAP kinase targets that regulate transcription at vivo data indicate that CYP1B1 can be upregulated during inflamma- multiple levels (38). We found in the present study that TNF-α induces tion. Since such an effect may potentially contribute to production of p38-dependent phosphorylation of MSK1. The chemical inhibition of carcinogenic metabolites, the main aim of the present study was to MSK activity led to decreased CYP1B1 induction, and the same effect identify signaling pathways possibly regulating enhanced expression was confirmed by using MSK1/2 double knockout MEFs. Moreover, of CYP1B1 by inflammatory mediators. we found that MSK1/2−/− cells have significantly lower basal levels of The AhR plays a major role in the regulation of CYP1 family CYP1B1 mRNA similar to p38−/− cells. These results indicated that enzymes in a variety of cells, and we confirmed, by using the siRNA- MSKs could be the molecular target of p38 MAP kinase, which medi- mediated AhR knockdown, that induction of CYP1B1 is primarily ate the effects of inflammatory cytokines on CYP1B1 upregulation. under the control of AhR also in the cell models that we used in our MSKs are nuclear kinases that can directly phosphorylate a number 2541 L.Šmerdová et al. of transcription factors, including CREB, NF-κB, signal transducer providing p38−/− MEFs, MSK1/2−/− MEFs and WB-F344 cells, respectively. and activator of transcription 3 and others (38), and they can also The expert technical assistance of I.Lišková and M.Urbánková is gratefully phosphorylate the nucleosomal proteins, such as histone H3 and high- acknowledged. mobility group 14 (63). Therefore, they may both directly activate Conflict of Interest Statement: None declared. transcription and induce chromatin remodeling or recruitment of the transcription machinery (64). Interestingly, quite recently, it has been also reported that activated AhR might facilitate recruitment of sev- References eral nuclear kinases, including MSKs, to enhancer/promoter of AhR target genes, such as Cyp1a1, and this has been implicated in the epi- 1. Nebert,D.W. et al. (2006) The role of cytochrome P450 enzymes in endogenous signalling pathways and environmental carcinogenesis. Nat. Rev. genetic regulation of their transcription (65). Cancer, 6, 947–960. The transcription of eukaryotic genes by RNAPII is a highly 2. Shimada,T. et al. (1996) Activation of chemically diverse procarcinogens regulated process, in which regulation of transcriptional elongation by human cytochrome P-450 1B1. Cancer Res., 56, 2979–2984. by P-TEFb plays an important role. Shortly after the initiation of 3. Hayes,C.L. et al. (1996) 17 beta-estradiol hydroxylation catalyzed by transcription, RNAPII is paused by the action of two negative elon- human cytochrome P450 1B1. Proc. Natl Acad. Sci. USA, 93, 9776–9781. gation factors (NELF—negative elongation factor and DSIF—DRB- 4. Cavalieri,E. et al. (2000) Estrogens as endogenous genotoxic agents— sensitivity inducing factor) downstream (50–200 nucleotides) of DNA adducts and mutations. J. Natl Cancer Inst. Monogr. 27, 75–93. Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 transcription start site. P-TEFb then phosphorylates NELF and DSIF 5. Crewe,H.K. et al. (2002) Metabolism of tamoxifen by recombinant human to liberate RNAPII from their inhibitory impact. Simultaneously, cytochrome P450 enzymes: formation of the 4-hydroxy, 4′-hydroxy and N-desmethyl metabolites and isomerization of trans-4-hydroxytamoxifen. P-TEFb phosphorylates the C-terminal domain of RNAPII, mani- Drug Metab. Dispos., 30, 869–874. fested by synthesis of nascent transcripts (40,66). Importantly, it 6. Stoilov,I. et al. (1997) Identification of three different truncating mutations has been shown recently that MSK1, activated by p38 MAP kinase, in cytochrome P4501B1 (CYP1B1) as the principal cause of primary con- may facilitate recruitment of P-TEFb to the promoter of the PTGS2 genital glaucoma (Buphthalmos) in families linked to the GLC3A locus on gene (encoding cyclooxygenase-2), thus enhancing RNAPII activity chromosome 2p21. Hum. Mol. Genet., 6, 641–647. and, consequently, the transcription of PTGS2 (42). Also, it has been 7. Kurzawski,M. et al. (2012) Expression of genes involved in xenobiotic reported that TNF-α triggers recruitment of P-TEFb to promoters of metabolism and transport in end-stage liver disease: up-regulation of various genes, such as interleukin-8 (67) or matrix metalloprotein- ABCC4 and CYP1B1. Pharmacol. Rep., 64, 927–939. ase-9 (68). This has led us to hypothesize that CDK9 might contrib- 8. McFadyen,M.C. et al. (2001) Cytochrome P450 CYP1B1 protein expression: a novel mechanism of anticancer drug resistance. Biochem. ute to the CYP1B1 upregulation. When blocking CDK9 activity with Pharmacol., 62, 207–212. DRB inhibitor, which inhibits C-terminal domain phosphorylation 9. Tokizane,T. et al. (2005) Cytochrome P450 1B1 is overexpressed and of RNAPII (43,44), we observed a significant reduction of CYP1B1 regulated by hypomethylation in prostate cancer. Clin. Cancer Res., 11, mRNA induction after TNF-α and BaP cotreatment. When we exam- 5793–5801. ined recruitment of CDK9 and RNAPII to the rat Cyp1b1 gene pro- 10. Murray,G.I. et al. (1997) Tumor-specific expression of cytochrome P450 moter region, we found that cotreatment of cells with TNF-α and BaP CYP1B1. Cancer Res., 57, 3026–3031. resulted in enhanced recruitment of both CDK9 and RNAPII to the 11. Chang,J.T. et al. (2007) Requirement of aryl hydrocarbon receptor overex- rat Cyp1b1 gene promoter. These experiments indicated that P-TEFb pression for CYP1B1 up-regulation and cell growth in human lung adeno- contributes to an enhanced induction of CYP1B1 by inflammatory carcinomas. Clin. Cancer Res., 13, 38–45. 12. Yang,X. et al. (2008) Constitutive regulation of CYP1B1 by the aryl hydro- mediators. carbon receptor (AhR) in pre-malignant and malignant mammary tissue. J. In conclusion, we report here a novel type of regulation of CYP1B1 Cell. Biochem., 104, 402–417. that is mediated via the activation of the p38 MAP kinase/MSK1 13. Baird,W.M. et al. (2005) Carcinogenic polycyclic aromatic hydrocarbon- signaling cascade and involves enhanced recruitment of CDK9 and DNA adducts and mechanism of action. Environ. Mol. Mutagen., 45, RNAPII to the promoter region of CYP1B1, thus supporting its tran- 106–114. scriptional elongation. Activation of the p38 pathway by inflammatory 14. Sissung,T.M. et al. (2006) Pharmacogenetics and regulation of human cytokines, or its deregulation observed in various types of neoplastic cytochrome P450 1B1: implications in hormone-mediated tumor metabo- cells (69), may employ this mechanism to upregulate CYP1B1, which lism and a novel target for therapeutic intervention. Mol. Cancer Res., 4, may in turn alter metabolism of endogenous CYP1B1 substrates, such 135–150. 15. Umannová,L. et al. (2008) Tumor necrosis factor-alpha potentiates as estrogens, or contribute to bioactivation of procarcinogens. These genotoxic effects of benzo[a]pyrene in rat liver epithelial cells through results have implications concerning the role of inflammation in can- upregulation of cytochrome P450 1B1 expression. Mutat. Res., 640, cer development and progression. 162–169. 16. Umannová,L. et al. (2011) Benzo[a]pyrene and tumor necrosis factor-α coordinately increase genotoxic damage and the production of proinflam- Supplementary material matory mediators in alveolar epithelial type II cells. Toxicol. Lett., 206, 121–129. Supplementary Tables 1 and 2 and Figures 1–6 can be found at http:// 17. Umannová,L. et al. (2007) Tumor necrosis factor-alpha modulates effects carcin.oxfordjournals.org/ of aryl hydrocarbon receptor ligands on cell proliferation and expression of cytochrome P450 enzymes in rat liver “stem-like” cells. Toxicol. Sci., 99, 79–89. Funding 18. Šmerdová,L. et al. (2013) Inflammatory mediators accelerate metabolism of benzo[a]pyrene in rat alveolar type II cells: the role of enhanced Czech Science Foundation (P503/11/1227 to J.V., 14-09979S to D.B.); cytochrome P450 1B1 expression. Toxicology, 314, 30–38. Internal Grant Agency of the Ministry of Health of the Czech Republic 19. Morgan,E.T. (2001) Regulation of cytochrome p450 by inflammatory (NT14599-3/2013 to J.K.); Czech Ministry of Education, Youth and mediators: why and how? Drug Metab. Dispos., 29, 207–212. Sports (CZ.1.07/2.3.00/30.0030 to L.Š.); Czech Ministry of Agriculture 20. Balkwill,F. (2009) Tumour necrosis factor and cancer. Nat. Rev. Cancer, 9, (MZE0002716202 to M.M. and J.K.); project ‘CEITEC-Central 361–371. European Institute of Technology’ (CZ.1.05/1.1.00/02.0068 to D.B.). 21. Dolado,I. et al. (2007) p38alpha MAP kinase as a sensor of reactive oxygen species in tumorigenesis. Cancer Cell, 11, 191–205. 22. Wiggin,G.R. et al. (2002) MSK1 and MSK2 are required for the mitogen- Acknowledgements and stress-induced phosphorylation of CREB and ATF1 in fibroblasts. Mol. Cell. Biol., 22, 2871–2881. The authors thank Á.R.Nebreda (Institute for Research in Biomedicine, 23. Ceelen,L. et al. (2011) Critical selection of reliable reference genes for Barcelona, Spain), S.J.Arthur (University of Dundee, UK) and gene expression study in the HepaRG cell line. Biochem. Pharmacol., 81, J.E.Trosko (Michigan State University, East Lansing, MI) for kindly 1255–1261. 2542 P38 MAP kinase regulates CYP1B1 expression

24. Andrysík,Z. et al. (2007) The aryl hydrocarbon receptor-dependent deregu- 48. Otto,S. et al. (1991) A novel adrenocorticotropin-inducible cytochrome lation of cell cycle control induced by polycyclic aromatic hydrocarbons in P450 from rat adrenal microsomes catalyzes polycyclic aromatic hydrocar- rat liver epithelial cells. Mutat. Res., 615, 87–97. bon metabolism. Endocrinology, 129, 970–982. 25. Blazek,D. et al. (2011) The Cyclin K/Cdk12 complex maintains genomic 49. Zheng,W. et al. (2005) Steroidogenic factor-1 interacts with cAMP response stability via regulation of expression of DNA damage response genes. element-binding protein to mediate cAMP stimulation of CYP1B1 via a far Genes Dev., 25, 2158–2172. upstream enhancer. Mol. Pharmacol., 67, 499–512. 26. Czudnochowski,N. et al. (2012) Serine-7 but not serine-5 phosphorylation 50. Tsuchiya,Y. et al. (2004) Human CYP1B1 is regulated by estradiol via primes RNA polymerase II CTD for P-TEFb recognition. Nat. Commun., estrogen receptor. Cancer Res., 64, 3119–3125. 3, 842. 51. Renton,K.W. (2005) Regulation of drug metabolism and disposition dur- 27. Galbraith,M.D. et al. (2013) HIF1A employs CDK8-mediator to stimulate ing inflammation and infection.Expert Opin. Drug Metab. Toxicol., 1, RNAPII elongation in response to hypoxia. Cell, 153, 1327–1339. 629–640. 28. Beedanagari,S.R. et al. (2009) Resveratrol inhibits dioxin-induced expres- 52. Vondráček,J. et al. (2011) Interactions of the aryl hydrocarbon receptor sion of human CYP1A1 and CYP1B1 by inhibiting recruitment of the aryl with inflammatory mediators: beyond CYP1A regulation. Curr. Drug hydrocarbon receptor complex and RNA polymerase II to the regulatory Metab., 12, 89–103. regions of the corresponding genes. Toxicol. Sci., 110, 61–67. 53. Devlin,A.H. et al. (2009) Cytochrome P450 1B1 expression in rat 29. Andersson,T.B. et al. (2012) The HepaRG cell line: a unique in vitro tool esophageal tumorigenesis promoted by gastric and duodenal reflux. Mol. for understanding drug metabolism and toxicology in human. Expert Opin. Carcinog., 48, 110–117. Drug Metab. Toxicol., 8, 909–920. 54. Ke,S. et al. (2001) Mechanism of suppression of cytochrome P-450 1A1 Downloaded from https://academic.oup.com/carcin/article/35/11/2534/418774 by guest on 29 September 2021 30. Wajant,H. et al. (2003) Tumor necrosis factor signaling. Cell Death Differ., expression by tumor necrosis factor-alpha and lipopolysaccharide. J. Biol. 10, 45–65. Chem., 276, 39638–39644. 31. Tian,Y. et al. (1999) Ah receptor and NF-kappaB interactions, a potential 55. Henklová,P. et al. (2008) Role of mitogen-activated protein kinases mechanism for dioxin toxicity. J. Biol. Chem., 274, 510–515. in aryl hydrocarbon receptor signaling. Chem. Biol. Interact., 172, 32. Chramostová,K. et al. (2004) Polycyclic aromatic hydrocarbons modulate 93–104. cell proliferation in rat hepatic epithelial stem-like WB-F344 cells. Toxicol. 56. Yaghini,F.A. et al. (2010) Angiotensin II-induced vascular smooth muscle Appl. Pharmacol., 196, 136–148. cell migration and growth are mediated by cytochrome P450 1B1-dependent 33. Andrysík,Z. et al. (2006) Activation of ERK1/2 and p38 kinases by poly- superoxide generation. Hypertension, 55, 1461–1467. cyclic aromatic hydrocarbons in rat liver epithelial cells is associated with 57. Ikuta,T. et al. (2004) Cell density regulates intracellular localization of aryl induction of apoptosis. Toxicol. Appl. Pharmacol., 211, 198–208. hydrocarbon receptor. J. Biol. Chem., 279, 19209–19216. 34. Ben-Levy,R. et al. (1995) Identification of novel phosphorylation sites 58. Cuadrado,A. et al. (2010) Mechanisms and functions of p38 MAPK signal- required for activation of MAPKAP kinase-2. EMBO J., 14, 5920–5930. ling. Biochem. J., 429, 403–417. 35. Shibazaki,M. et al. (2004) Suppression by p38 MAP kinase inhibitors 59. Brook,M. et al. (2000) Regulation of tumour necrosis factor alpha mRNA (pyridinyl imidazole compounds) of Ah receptor target gene activation by stability by the mitogen-activated protein kinase p38 signalling cascade. 2,3,7,8-tetrachlorodibenzo-p-dioxin and the possible mechanism. J. Biol. FEBS Lett., 483, 57–61. Chem., 279, 3869–3876. 60. Fernau,N.S. et al. (2010) Role of HuR and p38MAPK in ultraviolet 36. Clark,A.R. et al. (2003) Post-transcriptional regulation of gene expression B-induced post-transcriptional regulation of COX-2 expression in the by mitogen-activated protein kinase p38. FEBS Lett., 546, 37–44. human keratinocyte cell line HaCaT. J. Biol. Chem., 285, 3896–3904. 37. Graham,J.R. et al. (2010) mRNA degradation plays a significant role in 61. Frevel,M.A. et al. (2003) p38 mitogen-activated protein kinase-dependent the program of gene expression regulated by phosphatidylinositol 3-kinase and -independent signaling of mRNA stability of AU-rich element-contain- signaling. Mol. Cell. Biol., 30, 5295–5305. ing transcripts. Mol. Cell. Biol., 23, 425–436. 38. Vermeulen,L. et al. (2009) The versatile role of MSKs in transcriptional 62. Holtmann,H. et al. (1999) Induction of interleukin-8 synthesis integrates regulation. Trends Biochem. Sci., 34, 311–318. effects on transcription and mRNA degradation from at least three different 39. Price,D.H. (2000) P-TEFb, a cyclin-dependent kinase controlling elonga- cytokine- or stress-activated signal transduction pathways. Mol. Cell. Biol., tion by RNA polymerase II. Mol. Cell. Biol., 20, 2629–2634. 19, 6742–6753. 40. Kohoutek,J. (2009) P-TEFb—the final frontier. Cell Div., 4, 19. 63. Soloaga,A. et al. (2003) MSK2 and MSK1 mediate the mitogen- and 41. Nechaev,S. et al. (2008) Promoter-proximal Pol II: when stalling speeds stress-induced phosphorylation of histone H3 and HMG-14. EMBO J., 22, things up. Cell Cycle, 7, 1539–1544. 2788–2797. 42. Keum,Y.S. et al. (2013) UVB-induced COX-2 expression requires histone 64. Sundar,I.K. et al. (2012) Mitogen- and stress-activated kinase 1 (MSK1) H3 phosphorylation at Ser10 and Ser28. Oncogene, 32, 444–452. regulates cigarette smoke-induced histone modifications on NF-κB- 43. Biglione,S. et al. (2007) Inhibition of HIV-1 replication by P-TEFb inhibi- dependent genes. PLoS One, 7, e31378. tors DRB, seliciclib and flavopiridol correlates with release of free P-TEFb 65. Kurita,H. et al. (2014) The Ah receptor recruits IKKα to its target binding from the large, inactive form of the complex. Retrovirology, 4, 47. motifs to phosphorylate serine-10 in histone H3 required for transcriptional 44. Tian,Y. et al. (2003) Interactions between the aryl hydrocarbon receptor activation. Toxicol. Sci., 139, 121–132. and P-TEFb. Sequential recruitment of transcription factors and differential 66. Garriga,J. et al. (2004) Cellular control of gene expression by T-type cyc- phosphorylation of C-terminal domain of RNA polymerase II at cyp1a1 lin/CDK9 complexes. Gene, 337, 15–23. promoter. J. Biol. Chem., 278, 44041–44048. 67. Barboric,M. et al. (2001) NF-kappaB binds P-TEFb to stimulate transcrip- 45. Guengerich,P.F. et al. (2003) Cytochrome P450 1B1: a target for inhibition tional elongation by RNA polymerase II. Mol. Cell, 8, 327–337. in anticarcinogenesis strategies. Mutat. Res., 523–524, 173–182. 68. Shan,B. et al. (2005) Cyclin-dependent kinase 9 is required for tumor 46. Zheng,W. et al. (2003) Cell selective cAMP induction of rat CYP1B1 necrosis factor-alpha-stimulated matrix metalloproteinase-9 expression in in adrenal and testis cells. Identification of a novel cAMP-responsive far human lung adenocarcinoma cells. J. Biol. Chem., 280, 1103–1111. upstream enhancer and a second Ah receptor-dependent mechanism. Arch. 69. Wagner,E.F. et al. (2009) Signal integration by JNK and p38 MAPK path- Biochem. Biophys., 416, 53–67. ways in cancer development. Nat. Rev. Cancer, 9, 537–549. 47. Brake,P.B. et al. (1995) Regulation of cytochrome P4501B1 in cultured rat adrenocortical cells by cyclic adenosine 3′,5′-monophosphate and 2,3,7,8- Received October 3, 2013; revised August 12, 2014; accepted August 15, tetrachlorodibenzo-p-dioxin. Endocrinology, 136, 5034–5041. 2014

2543