Zinc finger transcription factor Slug is a novel target gene of aryl hydrocarbon receptor

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / y e x c r

Research Article

Zinc finger transcription factor Slug is a novel target gene of aryl hydrocarbon receptor Togo Ikuta a , Kaname Kawajiri a,b,⁎ a

Research Institute for Clinical Oncology, Saitama Cancer Center, 818 Komuro, Ina-machi, Kitaadachi-gun, Saitama 362-0806, Japan Solution Oriented Research for Science and Technology (SORST), Japan Science and Technology Agency, 4-1-8 Honmachi, Kawaguchi, Saitama 331-0012, Japan

b

ARTICLE INFORMATION

ABS T R AC T

Article Chronology:

The aryl hydrocarbon receptor (AhR) is a ligand-activated transcription factor. We previously

Received 30 May 2006

showed that AhR localizes predominantly in the cytoplasm under high cell densities of a

Revised version received

keratinocytes cell line, HaCaT, but accumulates in the nucleus at low cell densities. In the

1 August 2006

current report, we show that the Slug, which is a member of the snail/slug family of zinc finger

Accepted 2 August 2006

transcriptional repressors critical for induction of epithelial–mesenchymal transitions (EMT),

Available online 9 August 2006

is activated transcriptionally in accordance with nuclear accumulation of AhR. By reporter

Keywords:

analyses showed AhR directly binds to xenobiotic responsive element 5 at − 0.7 kb of the gene.

Ah receptor

AhR-targeted gene silencing by small interfering RNA duplexes led to the abolishment of not

Keratinocyte

only CYP1A1 but also Slug induction by 3-methycholanthrene. The Slug was co-localized to

Slug

the AhR at the wound margins of HaCaT cells, where apparent nuclear distribution of AhR and

EMT

Slug was observed. The induced Slug was associated with reduction of an epithelial marker of

Wound healing

cytokeratin-18 and with an increase in the mesenchymal marker, fibronectin. Taken

assay of the promoter of the Slug gene, gel shift and chromatin immunoprecipitation

together, these findings suggest that AhR participated in Slug induction, which, in turn, regulates cellular physiology including cell adhesion and migration. © 2006 Elsevier Inc. All rights reserved.

Introduction The aryl hydrocarbon receptor is a ligand-activated transcription factor belonging to the basic-helix–loop–helix (bHLH)/PERARNT-SIM homology region (PAS) family, and was initially identified as an intracellular mediator of the xenobiotic signaling pathways including 2,3,7,8-tetrachlorodibenzo-pdioxin (TCDD). In the absence of exogenous ligand, it remains predominantly in the cytoplasm as a complex with Hsp90 [1], p23 [2], and ARA9 [3]. When environmental contaminants such as TCDD and 3-methylcholanthrene (MC) bind to AhR, the receptor translocates into the nucleus to bind to the heterodimer partner, ARNT (AhR nuclear translocator) [4]. In the

nucleus, AhR/ARNT binds to the XRE (xenobiotic responsive element) [5], which is an enhancer DNA element located in the 5′-flanking region of the target genes such as CYP1A1, and several other proteins involved in xenobiotic metabolism. Since these enzymes are responsible for the metabolic activation of aromatic hydrocarbons to form active genotoxic metabolites, AhR plays an important role not only in chemical carcinogenesis [6] caused by these compounds, but also in teratogenesis [7] and thymic atrophy [8]. On the other hand, the combination of high degree of conservation and wide expression of AhR across different species [9] strongly suggests the existence of the physiological functions of AhR not related to xenobiotic metabolism. In this regard, recent studies have

⁎ Corresponding author. Fax: +81 48 722 1739. E-mail address: [email protected] (K. Kawajiri). 0014-4827/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2006.08.002

3586

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

shown that AhR participates in the regulation of such entity as p27Kip1 [10], Bax [11], and DNA polymerase κ [12]. Thus, it appears plausible that AhR may regulate the expression of genes involved in maintenance for normal physiological functioning. AhR is a nucleo-cytoplasmic shuttling protein that is mediated by both the nuclear localization signal (NLS) and the nuclear export signal (NES). The intracellular localization of AhR is regulated by the balance of nuclear import and export activity [13,14]. Molecular modulations of phosphorylation or dephosphorylation of NLS or NES also regulate intracellular localization of AhR [15]. Moreover, cell density regulates AhR intracellular localization through modulation of nuclear export activity, in which the p38 MAPK-mediated phosphorylation of NES by MC or medium change from DMEM to S-MEM and its dephosphorylation, regulated by cell–cell contact signals, plays a pivotal role in the novel AhR relocalization [16]. Because the regulated intracellular localization of transcription factors can serve as a biological switch in response to various physiological signals, a novel relocalization of AhR caused by cell density appears to be an essential event for the biological functioning of AhR. It has been generally accepted that the change of cell density from dense to sparsely distributed cells participates in both epithelial to mesenchymal transition (EMT) in embryonic development, when epithelial cells differentiate into various types of cells, and in acquisition of invasive and migratory properties during tumor progression [17]. E-cadherin is one of the most important molecules involved in tissue morphogenesis, wound healing, and cell–cell contact [18,19]. Downregulation of E-cadherin is often found during tumor cell invasion [20] and in EMT in embryonic development [21]. Although it has been suggested that proteolytic degradation of E-cadherin results in rapid changes in cell adhesion [22], E-cadherin expression is regulated by members of a zinc finger transcription factor snail/slug superfamily [21], which has been characterized as transcriptional repressors of E-cadherin acting through interaction with Eboxes of the proximal promoter. Disruption of E-cadherinmediated cell–cell contacts, mimicking a low cell density condition, led to nuclear localization of β-catenin, resulting in induction of Slug followed by inhibition of E-cadherin transcription [23]. In this study, we investigated whether AhR can activate the Slug gene as a novel downstream target of AhR, because several XREs exist in the human Slug promoter [24]. Our data suggest that AhR participates in Slug induction, which regulates cellular physiology such as cell adhesion and tumor cell invasion.

Materials and methods Cell culture The cell lines used for this study were human keratinocyte HaCaT cells provided by Dr. N. Fusenig [25], and COS cells. Cells were grown in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum at 37°C under 5% CO2 atmosphere. To reduce Ca2+ concentration, HaCaT cells were also cultured in S-MEM (Invitrogen, Carlsbad, CA) supplemented with 10% fetal calf serum dialyzed against phosphate-buffered saline.

Plasmid construction A 3.4 kb genomic fragment (GenBank, AF084243) containing the human Slug promoter was amplified by PCR with LA Taq DNA polymerase (Takara, Tokyo, Japan). The primers used were 5′ggatccctccatttcctaaacctggtagtc-3′ and 5′-ggatccgtacttgccagcgggtctggc), employing human genomic DNA as the template. The DNA sequence was confirmed. The PCR fragment was subcloned into pDrive (QIAGEN, Valencia, CA) followed by digestion with BamHI to isolate the BamHI–BamHI fragment of the Slug DNA. The fragment was subcloned into the pGL3 basic (Promega, Madison, WI), digested with BglII to construct the Slug-luciferase expression plasmid. To construct the Slug XREwt-Luc including two copies of Slug XRE5 sequence, two synthesized oligonucleotides, namely 5′-ccgcagcgccctcgccgcacgcaaggctggtac-3′ and 5′-cagccttgcgtgcggcgagggcgctgcgggtac-3′, were annealed and phosphorylated. For construction of the Slug XREmut-Luc, 5′-ccgcagcgccctcgccgcaaaaaaggctggtac-3′ and 5′-cagccttttttgcggcgagggcgctgcgggtac-3′ were prepared, then each double-stranded fragment was ligated to the KpnI site of the pX4TK-Luc digested by KpnI in order to remove four repeats of the XRE sequence. Human Slug cDNA was prepared by PCR amplification of reverse-transcribed products of total RNA from HaCaT cells, using the specific primers, (5′ggatccccatgccgcgctccttcctggtc-3′ and 5′-ggatccccgtgtgtacacagcagccaga-3′) and LA Taq DNA polymerase (Takara). The human slug cDNA was then inserted into the pDrive vector (Qiagen). For subsequent cloning into the pEGFP C2 (Clontech, Mountain View, CA) expression vector, pDrive-Slug was digested by BamHI in order to isolate the BamHI–BamHI fragment of the Slug cDNA, and then ligated into BglII site of pEGFP C2.

Immunoblot and immunofluorescence Samples for analysis by immunoblot were prepared as described previously [16]. For immunofluorescence, cells were incubated on a coverslip [16]. The following antibodies were used as primary antibodies: mouse monoclonal anti-cytokeratin 18 (Lab Vision, Fremont, CA), anti-fibronectin (SIGMAAldrich, St. Louis, MO), anti-E-cadherin (BD Bioscience, San Jose, CA), polyclonal rabbit anti-AhR (BIOMOL), and anti-Slug (Abgent, San Diego, CA). The samples were visualized with either fluorescein isothiocyanate (FITC) or rhodamine B isothiocyanate (RITC)-conjugated species-specific secondary antibodies.

Luciferase assay Luciferase activity expressed in stable HaCaT transformants containing pX4TK-Luc [16] was determined using a Luciferase Assay System (Promega) according to the manufacturer's instructions, and was adjusted to the protein content of each sample. Transiently expressed luciferase activities were normalized by β-galactosidase activity.

Reverse transcription and real-time quantitative PCR Total RNA was extracted from cultured cells with Isogen (Nippon Gene, Tokyo, Japan) according to the manufacturer's instructions. Random-primed, first strand cDNA was generated

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

from 0.5 μg total RNA with a Takara RNA PCR kit (Takara), and the PCR analysis was performed using gene-specific primers and first strand cDNA as the template. The sequence of primers was as follows: Slug, 5′-gacacacatacagtgattatt-3′ and 5′aaacttttcagcttcaatggc-3′; AhR, 5′-atactgaagcagagctgtgc-3′ and 5′-aaagcaggcgtgcattagac-3′; CYP1A1, 5′-tccctattcttcgctaccta-3′ and 5′-tctctgtaccctggggtt-3′; GAPDH, 5′-acatcgctcagacaccatgg3′ and 5′-gtagttgaggtcaatgaaggg-3′. The PCR products were fractionated on a 2% agarose gel and visualized after ethidium bromide staining. Quantitative real time PCR was performed on a LightCycler (Roche Diagnostics, Indianapolis, IN) using LightCycler FastStart DNA MasterPLUS SYBR Green I (Roche). The PCR was evaluated by a melting curve analysis following the manufacturer's instructions.

Gel shift assay Human AhR and murine ARNT cDNAs in pRc/CMV were transcribed and translated in vitro from the T7 promotor, using TNT-coupled rabbit reticulocyte lysate (Promega). These gene products were used for gel shift analysis as described previously [14,26].

Chromatin immunoprecipitation assay HaCaT cells cultured on 10-cm plates were fixed with 1% of formaldehyde to crosslink protein and DNA, harvested, and the chromatin was then extracted. Chromatin was sonicated to about 500-bp fragments in 0.2 ml of lysis buffer containing 50 mM Tris–HCl (pH 8.0), 10 mM EDTA, and 1% SDS. A solution containing chromatin fragments was diluted with 1.8 ml of buffer (50 mM Tris–HCl (pH 8.0), 167 mM NaCl, 1.1% Triton X100, and 0.11% sodium deoxycholate), precleared with protein G-Sepharose (Amersham, Piscataway, NJ), and then incubated with anti-AhR antibodies (Santa Cruz, Santa Cruz, CA) with overnight rotation, at 4°C. The chromatin bound by antibodies was then precipitated with protein G-Sepharose, washed and eluted with a buffer containing 10 mM Tris–HCl (pH 8.0), 300 mM NaCl, 5 mM EDTA, and 0.5% SDS. Protein–DNA crosslinking was reversed by incubation at 65°C for 4 h, after which the DNA fragments were purified and dissolved in 20 μl of Tris–EDTA buffer and used as a template for PCR amplification. A 256-bp sequence in the human Slug promoter region was amplified by PCR (primers: 5′-ttcaagccaccatagctaacacgg-3′ and 5′-tggcatctggagaggtttgccttg-3′). As a positive control, human CYP1A1 promoter region was amplified as described previously [27].

siRNA The siRNA, Silencer™, (pre-designated siRNA, Cat #16706) for AhR was purchased from Ambion (Austin, TX). The sequence of oligonucleotides (sense strand) was as follows: AhR, ccgacuuaauacagaguugtt, and control cgcgcuuuguaggauucgtt, which is derived from a message transcribed from the chloroplast genome of Euglena gracilis [28]. A total of 1 × 105 HaCaT cells was seeded in each well of a 24-well multiplate on the day before transfection. The siRNA was introduced into the cells by TransIT-TKO (Mirus, Madison, WI), at a concentration of 50 nM in cultured medium. The cells were collected 24 h after transfection.

3587

Results Ligand-independent activation of AhR in calcium-deficient medium We had previously reported that the subcellular localization and functioning of AhR are regulated by cell density [16]. When confluent HaCaT cells were cultured in DMEM containing 1.8 mM CaCl2, AhR was mainly localized in cytoplasm and the cells were characterized by immunohistochemical staining of E-cadherin at the site of cell–cell contact (Fig. 1A). Then, the medium was removed from the culture and followed by incubation in calcium-deficient S-MEM for 2 h, which resulted in both relocalization of E-cadherin and nuclear accumulation of AhR, implicating activation of AhR. In order to determine whether or not this culture condition increased AhR activity, stable transformant HaCaT cells, harboring pX4TK-Luc, were incubated in S-MEM. As shown in Fig. 1B, incubation for 10 h after the culture medium was replaced with S-MEM, the reporter activity was increased about 7-fold compared to incubation in DMEM. The addition of MC to DMEM caused only a 2-fold increase. Treatment of cytochalasin D, an inhibitor of actin polymerization, did not affect reporter activity, suggesting that signals other than cytoskeletal changes are required (data not shown). These results suggest that some downstream of target gene(s), regulated by AhR/ARNT, is upregulated in the process of the medium change.

Expression of Slug is AhR-dependent manner It is well known that one of the important effects of calcium on epidermal cells is a regulation of cell–cell contact. Among a number of factors regulating cell–cell contact, we were interested in Slug, harboring several XRE sequences in the 5′ flanking region of the gene. As shown in Fig. 2A, five XREs are located in 3.4 kb upstream of the initiation codon of the human Slug gene [24]. In order to examine whether transcription of Slug is regulated by AhR, the 3.4-kb region was amplified by PCR and cloned into pGL3 promoterless vector upstream of the luciferase gene. When the 5′-flanking sequence of the Slug gene was fused to the luciferase gene (Fig. 2A; Slug-Luc-a: −3371 ∼ +1 fused with luciferase and SlugLuc-b: −829 ∼ +1 fused with luciferase gene) and transfected into COS7 cells along with the expression plasmid of AhR, the Slug-luciferase activity was dependent on AhR as well as XREluciferase activity (Fig. 2B). In contrast, mutant AhR, which is unable to activate the XRE-dependent reporter gene transcription [14], failed to increase luciferase activity (Fig. 2C). The Slug-luciferase activity was induced by the AhR ligand MC (Fig. 2D). Thus, the XRE in the Slug is functional as the inducible enhancer mediated by AhR. Furthermore, we transfected the Slug-Luc-a into HaCaT cells incubated in DMEM, followed by replacement of the culture medium to S-MEM (Fig. 2E). After 10 h of incubation in S-MEM, the reporter activity was increased approximately 3-fold. These results led us to examine the induction of the Slug mRNA, followed by activation of AhR in HaCaT cells. As shown in Fig. 2F, both CYP1A1 and Slug mRNAs were induced by treatment with

3588

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

MC or S-MEM, and in a similar time course. The mRNA for Slug was also induced in a human breast cancer cell line, MCF-7, by treatment with MC or S-MEM (data not shown). These data suggest that below transcription level, expression of Slug is regulated by AhR.

AhR/ARNT complex binds to XRE5 in Slug promoter region In order to test whether the AhR/ARNT complex was able to bind to these putative XREs (Fig. 2A), a gel shift assay was performed using in vitro transcribed and translated AhR and ARNT proteins, and each oligonucleotide containing XRE1– XRE5. A ligand-dependent band shift was observed in the presence of both AhR and ARNT when the XRE probe derived from CYP1A1 was used (Fig. 3A; arrow). In the case of Slug, a similar band shift was observed when oligonucleotides containing XRE5 (–CACGCAA–) were used as the probe. The band lost in competition due to an excess amount of unlabeled probe derived from either CYP1A1 or Slug. Treatment with either anti-AhR or anti-ARNT before addition of labeled XRE5 probes resulted in reduction of the band

corresponding to the AhR/ARNT/XRE5 complex (data not shown) XRE1–XRE4 probes did not bind to AhR/ARNT (data not shown). In order to show if AhR/ARNT complexes bind to XRE5 in vivo, we conducted a chromatin immunoprecipitation (ChIP) assay using anti-AhR antibodies. The immunoprecipitates collected by protein G Sepharose were used as templates for amplification of the DNA fragments containing XRE5 by PCR. As shown in Fig. 3B, treatment with MC enhanced the binding of AhR to XRE from both CYP1A1 and Slug. To evaluate whether the XRE5 sequence is functional in transactivation of Slug, two copies of DNA fragments of XRE5 were subcloned in a pGL3 vector connected to TK promoter in order to generate XRE5-Luc. For comparison, we also prepared another reporter construct which interrupted the integrity of the XRE5 by introducing mutations. HaCaT cells, transfected with either wild or mutant luciferase construct, were incubated in S-MEM, and reporter activity was measured. Although the reporter activity from wild-type construct was efficiently activated, its activity in mutant construct remained at basal level (Fig. 3C). The significance of XRE5 in

Fig. 1 – AhR is activated by incubation in S-MEM. (A) Relocalization of AhR. HaCaT cells on coverslips were cultured in DMEM until growing to confluence. The culture medium was replaced with S-MEM and the cells were incubated for 2 h before fixation. The distribution of AhR and E-cadherin was detected by immunofluorescence with the antibodies indicated in the text. Nuclear staining was performed by Hoechst 33258. (B) XRE-mediated transcriptional activation. The luciferase activity was determined using HaCaT cells stably transformed with pX4TK-Luc. Cells were treated as indicated at time 0. 3-methylcolanthrene (MC) was used as a ligand at the concentration of 1 μM in culture medium. The activities are indicated as values normalized by the protein content of each sample. The results are given as the mean ± S.D. in three experiments.

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

Slug promoter (Slug-Luc-a) was confirmed as shown in Fig. 3D. Taken together, these observations show that AhR directly binds to XRE5 at −0.7 kb upstream of the initiation codon ATG, activating the Slug gene.

3589

Inhibition of AhR abolished Slug expression In order to confirm the function of AhR in regulating Slug transcription, we transiently transfected HaCaT cells with

Fig. 2 – Expression of Slug is upregulated by AhR. (A) Luciferase reporter constructs with the Slug promoter. Two regions of the slug promoter were connected to luciferase cDNA to generate constructs Slug-Luc-a and Slug-Luc-b. Boxes labeled 1 to 5 indicate the site of XRE motif. (B) Luciferase activity in COS cells. Cells were cotransfected with the reporter construct indicated in the text, and expression plasmid encoding AhR or empty vector (SRHis). Two days after transfection, cells were lysed and the reporter activities were determined. The results are given as the mean ± S.D. in three experiments. (C) Luciferase activity in COS cells cotransfected with Slug-Luc-a and either wild-type or mutant AhR (Leu70 and Leu72 were replaced to Ala) which was unable to bind to XRE. Luciferase activities expressed in COS cells were determined in the absence of exogenous ligand. (D) Luciferase activity in Hepa 1 cells treated with 1 μM MC for 18 h before harvest. Slug-Luc-a was used. (E) Luciferase activity in HaCaT cells incubated in S-MEM. Two days after transfection with Slug-Luc-a construct, culture medium was replaced with S-MEM, and cells were further incubated for the indicated time before harvest. (F) Induction of Slug and CYP1A1 mRNA in HaCaT was detected by RT-PCR. After reaction, samples were loaded onto 2% agarose gel and detected via staining with ethidium bromide. The amount of PCR products was quantified using the ImageJ software (National Institute of Health).

3590

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

either control siRNA or siRNA specific for AhR, and conducted RT-PCR to evaluate the amount of Slug mRNA. As shown in Fig. 4A, transfection of siRNA for AhR reduced the AhR mRNA level by 50%. As expected, expression of AhR was also inhibited in terms of protein level, while the amount of actin was not affected (Fig. 4B). The efficacy of the siRNA for AhR was determined by quantitative RT-PCR to detect CYP1A1 (Fig. 4C; left panel) and Slug mRNA (Fig. 4C; right panel). In the presence of control siRNA, treatment of MC for 1 h increased CYP1A1 mRNA by 2.3-fold. In contrast, the presence of siRNA specific for AhR completely inhibited the induction of CYP1A1 mRNA in response to MC. On the other hand, treatment with MC either for 1 h or for 2 h caused a 4.5fold or 2.7-fold increase in Slug mRNA in the presence of control siRNA, whereas induction of Slug mRNA in response to MC was abolished in the presence of siRNA specific for AhR.

Slug expression at wound edge is associated with AhR activation Expression of the GFP reporter gene mediated by XRE was activated at wound edge in an in vitro wound-healing model using HaCaT cells [16]. Also, expression of Slug was stimulated at the wound margin, as analyzed by in situ hybridization [29]. We then investigated the expression of AhR and Slug by immunostaining in an in vitro wound-healing model. Anti-Slug antibodies used here reacted specifically to Slug when the GFP-Slug fused protein expressed in COS7 cells was examined for antigen (Fig. 5A). Wound margins were produced by incubation for 4 days after scraping confluent serum-deprived culture of HaCaT cells as shown in dotted line (Fig. 5B). In each panel, area on the right side is denuded region of the cells, while area on the left is more confluent region of the culture. The cells with tightly association, where

Fig. 3 – AhR/ARNT complex binds to XRE5. (A) Gel shift analysis with AhR and ARNT protein produced in reticulocyte lysate and with oligonucleotides including XRE derived from CYP1A1 or Slug. (B) ChIP assays of CYP1A1 and Slug genes using a control IgG or an AhR antibody. Confluent HaCaT culture was treated with 1 μM MC for the indicated times before harvest. A 289-bp sequence located in the promoter region of CYP1A1 (positive control) and a 256-bp sequence containing XRE5 in Slug were amplified. (C) XRE5 present in the Slug promoter is required for transactivation. HaCaT cells were transfected with either a wild-type XRE5-Luc (XRE5wt) or a mutated XRE5-Luc (XRE5mut) construct. Two days after transfection, the culture medium was replaced with S-MEM and the cells were further incubated for the indicated time before harvest. The results are given as the mean ± S.D. in three experiments. (D) XRE5 is essential for Slug transcription by AhR. COS cells were transfected with indicated luciferase construct along with either control vector or AhR expression plasmid. Mutations were introduced in the XRE5 (CACGCAA to CAAAAAA) of the Slug promoter (Slug-Luc-a) by site-directed mutagenesis. Cells were incubated for two days before harvest. The results were given as the mean ± S.D. in three experiments.

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

3591

with TGF-β1, an interaction known to trigger EMT [30], also increased the Slug and decreased CK18 mRNA in a similar fashion compared to that of S-MEM treatment, whereas the expression of Vim was slightly upregulated. Furthermore, by using immunofluorescence, we examined the effect of longer incubation times in S-MEM on the expression of EMT markers (Fig. 5D). Expression of CK18 was observed after 8 days of incubation in serum-deprived DMEM, indicating an epithelial character of HaCaT cells. Fibronectin (FN), a marker of mesenchymal cells, was only weakly detected in these cells. Incubation in S-MEM resulted in decreased expression of CK18 and increased FN in a significant number of cells. Cells demonstrating intense expression of FN did not show CK18 staining(arrowhead). Most of the cells treated with TGF-β1 showed clear staining with FN and reduction in CK18. These data suggest that incubation with S-MEM is able to induce epithelial–mesenchymal transition.

Discussion Fig. 4 – Inhibition of AhR expression abolished Slug induction. (A) 1 × 105 HaCaT cells per well were seeded onto a 24-well plate, cultured overnight to reach 40–50% confluence, and were transfected with AhR or control siRNA at a final concentration of 50 nM. At 24 h post transfection, total RNA was extracted to detect AhR by real-time RT-PCR. Values were normalized by expression of GAPDH, and are indicated by mean ± S.D. (B) Cellular extract was subjected to 8% SDS-PAGE in order to detect AhR by immunoblotting. The results shown are from two independent experiments. Actin was used as the loading control. (C) Detection of CYP1A1 (left panel) and Slug (right panel) mRNA by real-time RT-PCR. Cells were treated with 1 μM methylcolanthrene for 1 h or 2 h before harvest.

E-cadherin was detected at cell–cell contact, revealed both cytoplasmic and nuclear localization of AhR (Fig. 5B, arrow). In contrast, cells at wound margin, where the cells seem to be loosely associated, showed exclusive nuclear AhR (arrowhead). Immunostaining with anti-Slug antibodies revealed that expression of Slug was enhanced and mainly localized in the nuclei in cells at wound margin (arrowhead). These results show the beneficial association between activation of AhR and induction of Slug at wound edge.

AhR-associated EMT due to Slug induction It is well known that one of the important roles of Slug is to serve as a trigger of EMT, leading to the conversion of epithelial cells into mesenchymal cells [21]. We examined whether the expression of either epithelial or mesenchymal marker was altered by changing the medium from DMEM to S-MEM (Fig. 5C). Slug induction was initiated after 1 or 2 h of incubation with the HaCaT cells. We observed that the expression of the epithelial marker, cytokeratin 18 (CK18), was gradually decreased, but the mesenchymal marker, vimentin (Vim), was not detected. On the other hand, treatment of HaCaT cells

We have previously shown that intracellular localization and function of AhR are regulated by cell density in the absence of exogenous ligand [16]. In the current study, we demonstrated that a predominant cytoplasmic localization of AhR and low transcription activity with high cell density was altered by medium change with Ca2+-deficient S-MEM towards a more nuclear distribution having high transcription activity (Fig. 1). This suggests that E-cadherin-mediated signals derived from cell–cell contact may have pivotal roles in the events described above. In this context, we identified the Slug gene, which is a member of the snail/slug family of zinc finger transcriptional repressors that are critical for induction of EMT [21], as a novel downstream target for AhR. The Slug-luciferase reporter activity was increased by cotransfection with expression plasmid coding AhR (Fig. 2B) and its ligand, MC (Fig. 2D), while a mutant AhR, which was deficient in DNA binding [14], was unable to activate the reporter gene (Fig. 2C). The reporter was activated in HaCaT cells when the cells were incubated in S-MEM after transfection (Fig. 2E). The RT-PCR analysis revealed that the expression of Slug mRNA was induced by treatment with either MC or SMEM (Fig. 2F). The AhR-targeted gene silencing by small interfering RNA duplexes led to the abolishment of not only CYP1A1 via medium change with Ca2+-deficient S-MEM, but also Slug induction by MC (Fig. 4). We also showed that the AhR/ARNT complex binds XRE5 in the promoter of the Slug gene (Figs. 3A, 3C), as well as endogenous AhR binding to the Slug promoter as shown by ChIP assay in HaCaT cells (Fig. 3B), indicating that AhR did bind XRE5 in vivo. Taken together, these observations indicate that transcription of the Slug gene is regulated by an AhR/ARNT system in response to the signals derived from cell–cell contact. As shown in this study, AhR appeared to be activated when cell–cell contact was disrupted either by incubation in S-MEM or in scattering cells in a wound-healing model. In both cases, the cells were characterized by loss of organized arrangement of E-cadherin at the site of cell adhesion (Figs. 1A, 5B). Some investigators also have reported that AhR activity is associated

3592

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

Fig. 5 – (A) Specificity of anti-Slug IgG. COS7 cells were transfected with expression plasmid encoding for either GFP or GFP-Slug fused protein. Slug was visualized by RITC-conjugated secondary antibodies. (B) Slug expression associated with AhR activation. Confluent cultures of HaCaT cells were scraped and further incubated for 4 days. After fixation, cells were incubated with primary antibodies indicated in the text. AhR and Slug were visualized by RITC-conjugation. E-Cadherin was visualized by FITC-conjugated secondary antibodies. Cells at wound margin are indicated by arrowhead, and cells surrounded by E-cadherin at cell contact are indicated by arrow. Wounded edge is indicated by dotted line. (C) Cells were treated with either 2.5 ng/ml TGF-β1 or S-MEM for the indicated times and were collected to extract total RNA. RT-PCR was performed to detect Slug, cytokeratin 18 (CK18), vimentin (Vim,) and GAPDH. (D) HaCaT cells were cultured for 8 days in DMEM (control), S-MEM, and DMEM supplemented with 2.5 ng/ml TGF-β1. After fixation, cells were incubated with primary antibodies indicated in the text, and then visualized by appropriate fluorescent-conjugated secondary antibodies. Nuclear staining was performed with Hoechst 33258. Cells were observed under microscope at 400× magnification.

with the state of cell adhesion [31,32]. It is fascinating to postulate that the signaling cascade leading to AhR activation links to cell adhesion molecules. EMT and tumor cell metastasis are known to share common properties, including destruction of E-cadherinmediated cell–cell adhesion, acquisition of a mesenchymal phenotype, invasion into the extracellular environment, and movement to distant sites [17,29]. Disruption of E-cadherinmediated cell–cell contact is considered to be a pivotal first step during EMT, both in embryonic development and pathological situations [18–20], and the snail/slug superfamily of zinc-finger transcription factors has a central role in triggering EMT [21]. Different signaling molecules have been implicated in the activation of snail/slug genes in several processes that subsequently lead to the conversion of epithelial cells into mesenchymal cells. Among these signal-

ing molecules, we have particular interest in TGF-β-directed EMT compared with that of AhR. The gene expression profiling study showed complex context-dependent signaling pathways and transcriptional events that determine epithelial plasticity mediated by TGF-β pathway [30]. As shown in Fig. 5, AhR induced Slug in the nuclei of HaCaT cells and resulted in a decrease of the epithelial marker CK18, but not any increase in the mesenchymal marker, Vim. Immunofluorescent analysis revealed that activation of AhR led to an increase of FN, one of the mesenchymal markers (Fig. 5D). On the other hand, TGFβ1 increased Slug, followed by a decrease of CK18 and increase of Vim and FN. Although the effect of AhR on EMT is not the same to TGF-β, AhR is considered to be included in the signaling molecules associated with EMT. Accumulated evidence has shown that AhR induces p27Kip1 [10], which elicits cell migration by interaction with Rac-1 [33],

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

a member of the small GTPases of Rho family that regulates structural modifications to actin cytoskeleton. MuleroNavarro et al. also reported that AhR−/− mouse mammary fibroblasts showed a significant decrease in their migratory potential in collagen I matrix and caused a marked reduction in active Rac-1 [34]. Moreover, AhR upregulates matrix metalloproteinases [35,36] a family of proteolytic enzymes that degrade the protein components of the ECM, and ECM remodeling is a key process in cell migration and tumor invasions [37,38]. Overall, our results suggest that AhR plays a pivotal role in regulation of cell motility not only by promoting EMT due to direct induction of Slug but also ECM remodeling via indirect induction of genes such as MMPs, due to cross-talk between AhR and a variety of signaling pathways.

[11]

[12]

[13]

[14]

Acknowledgment

[15]

This study was supported by the Grants-in-Aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science and Technology.

[16]

[17] REFERENCES

[18] [19]

[1] G.H. Perdew, Association of the Ah receptor with the 90-kDa heat shock protein, J. Biol. Chem. 263 (1998) 13802–13805. [2] A. Kazlauskas, L. Poellinger, I. Pongratz, Evidence that the co-chaperone p23 regulates ligand responsiveness of the dioxin (aryl hydrocarbon) receptor, J. Biol. Chem. 274 (1999) 13519–13524. [3] L.A. Carver, C.A. Bradfield, Ligand-dependent interaction of the aryl hydrocarbon receptor with a novel immunophilin homolog in vivo, J. Biol. Chem. 272 (1997) 11452–11456. [4] H. Eguchi, T. Ikuta, T. Tachibana, Y. Yoneda, K. Kawajiri, A nuclear localization signal of human aryl hydrocarbon receptor nuclear translocator/hypoxia-inducible factor 1β is a novel bipartite type recognized by the two components of nuclear pore-targeting complex, J. Biol. Chem. 272 (1997) 17640–17647. [5] A. Fujisawa-Sehara, K. Sogawa, C. Nishi, Y. Fujii-Kuriyama, Regulatory DNA elements localized remotely upstream from the drug-metabolizing cytochrome P-450c gene, Nucleic Acids. Res. 14 (1986) 1465–1477. [6] Y. Shimizu, Y. Nakatsuru, M. Ichinose, Y. Takahashi, H. Kume, J. Mimura, Fujii-Kuriyama, T. Ishikawa, Benzo[a]pyrene carcinogenicity is lost in mice lacking the aryl hydrocarbon receptor, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 779–782. [7] J. Mimura, K. Yamashita, K. Nakamura, M. Morita, T.N. Takagi, K. Nakao, M. Ema, K. Sogawa, M. Yasuda, M. Katsuki, Y. Fujii-Kuriyama, Loss of teratogenic response to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in mice lacking the Ah (dioxin) receptor, Genes Cells 2 (1997) 645–654. [8] P.M. Fernandez-Salguero, D.M. Hilbert, S. Rudikoff, J.M. Ward, F.J. Gonzalez, Aryl-hydrocarbon receptor-deficient mice are resistant to 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced toxicity, Toxicol. Appl. Pharmacol. 140 (1996) 173–179. [9] M.E. Hahn, S.I. Karchner, M.A. Shapiro, S.A. Perera, Molecular evolution of two vertebrate aryl hydrocarbon (dioxin) receptors (AHR1 and AHR2) and the PAS family, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 13743–13748. [10] S.K. Kolluri, C. Weiss, A. Koff, M. Gottlicher, p27(Kip1) induction and inhibition of proliferation by the intracellular

[20]

[21] [22]

[23]

[24]

[25]

[26]

[27]

[28]

3593

Ah receptor in developing thymus and hepatoma cells, Genes Dev. 13 (1999) 1742–1753. T. Matikainen, G.I. Perez, A. Jurisicova, J.K. Pru, J.J. Schlezinger, H.Y. Ryu, J. Laine, T. Sakai, S.J. Korsmeyer, R.F. Casper, D.H. Sherr, J.L. Tilly, Aromatic hydrocarbon receptor-driven Bax gene expression is required for premature ovarian failure caused by biohazardous environmental chemicals, Nat. Genet. 28 (2001) 355–360. T. Ogi, J. Mimura, M. Hikida, H. Fujimoto, Y. Fujii-Kuriyama, H. Ohmori, Expression of human and mouse genes encoding polkappa: testis-specific developmental regulation and AhR-dependent inducible transcription, Genes Cells 6 (2001) 943–953. T. Ikuta, H. Eguchi, T. Tachibana, Y. Yoneda, K. Kawajiri, Nuclear localization and export signals of the human aryl hydrocarbon receptor, J. Biol. Chem. 273 (1998) 2895–2904. T. Ikuta, T. Tachibana, J. Watanabe, M. Yoshida, Y. Yoneda, K. Kawajiri, Nucleocytoplasmic shuttling of the aryl hydrocarbon receptor, J. Biochem. (Tokyo) 127 (2000) 503–509. K. Kawajiri, T. Ikuta, Regulation of nucleo-cytoplasmic transport of the aryl hydrocarbon receptor, J. Health Sci. 50 (2004) 215–219. T. Ikuta, Y. Kobayashi, K. Kawajiri, Cell density regulates intracellular localization of aryl hydrocarbon receptor, J. Biol. Chem. 279 (2004) 19209–19216. J.P. Thierry, Epithelial–mesenchymal transitions in tumor progression, Nat. Rev., Cancer 2 (2002) 442–454. M. Takeichi, Cadherin cell adhesion receptor as a morphogenetic regulator, Science 251 (1991) 1451–1455. B.M. Gumbier, The molecular basis of tissue architecture and morphogenesis, Cell 84 (1996) 345–357. A.K. Perl, P. Wilgenbus, U. Dahl, H. Semb, G. Christofori, A causal role for E-cadherin in the transition from adenoma to carcinoma, Nature 392 (1998) 190–193. M.A. Nieto, The snail superfamily of zinc-finger transcription factors, Nat. Rev., Mol. Cell. Biol. 3 (2002) 155–166. Y. Fujita, G. Krause, M. Schffner, D. Zechner, H.E. Leddy, J. Behrens, T. Sommer, W. Birchmeier, Hakai, a c-Cbl-like protein, ubiquitinates and induces endocytosis of the E-cadherin complex, Nat. Cell Biol. 4 (2002) 222–231. M. Conacci-Sorrell, I. Simcha, T. Ben-Yedidia, J. Blechman, P. Savagner, A. Ben-Ze'ev, Autoregulation of E-cadherin expression by cadherin–cadherin interactions: the roles of β-catenin signaling, Slug, and MAPK, J. Cell Biol. 163 (2003) 847–857. K. Stegmann, J. Boecker, C. Kosan, A. Ermert, J. Kunz, M.C. Koch, Human transcription factor SLUG: mutation analysis in patients with neural tube defects and identification of a missense mutation (D119E) in the Slug subfamily-defining region, Mutat. Res. 406 (1999) 63–69. P. Boukamp, R.T. Petrussevska, D. Breitkreutz, J. Hornung, A. Markham, N.E. Fusenig, Normal keratinization in a spontaneously immortalized aneuploid human keratinocyte cell line, J. Cell Biol. 106 (1998) 761–771. N. Matsushita, K. Sogawa, M. Ema, A. Yoshida, Y. Fujii-Kuriyama, A factor binding to the xenobiotic responsive element (XRE) of P-4501A1 gene consists of at least two helix–loop–helix proteins, Ah receptor and ARNT, J. Biol. Chem. 268 (1993) 21002–21006. F. Ohtake, K. Takeyama, T. Matsumoto, H. Kitagawa, Y. Yamamoto, K. Nohara, C. Tohyama, A. Krust, J. Mimura, P. Chambon, J. Yanagisawa, Y. Fujii-Kuriyama, S. Kato, Modulation of oestrogen receptor signaling by association with activated dioxin receptor, Nature 423 (2003) 545–550. M. Abdelrahim, R. III, S. Safe, Aryl hydrocarbon receptor gene silencing with small inhibitory RNA differentially modulates Ah-responsiveness in MCF-7 and HepG2 cancer cells, Mol. Pharmacol. 63 (2003) 1373–1381.

3594

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 3 5 8 5 –35 9 4

[29] P. Savagner, D.F. Kusewitt, E.A. Carver, F. Magnino, C. Choi, T. Gridley, L.G. Hudson, Developmental transcription factor slug is required for effective re-epithelialization by adult keratinocytes, J. Cell. Physiol. 202 (2005) 858–866. [30] J. Zavadil, M. Bitzer, D. Liang, A. Y.C.Yang, S. Massimi, E. Kneitz, E.P. Piek, Genetic programs of epithelial cell plasticity directed by transforming growth factor-β, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 6686–6691. [31] C.M. Sadek, B.L. Allen-Hoffmann, Suspension-mediated induction of Hepa1c1c7 Cyp1a-1 expression is dependent on the Ah receptor signal transduction pathway, J. Biol. Chem. 269 (1994) 31505–31509. [32] Y.C. Cho, W. Zheng, C.R. Jefcoate, Disruption of cell–cell contact maximally but transiently activates AhR-mediated transcription in 10T1/2 fibroblasts, Toxicol. Appl. Phamacol. 199 (2004) 220–238. [33] S.S. McAllister, M. Becker-Hapak, G. Pintucci, M. Pagano, S.F. Dowdy, Novel p27Kip1 C-terminal scatter domain mediates rac-dependent cell migration independent of cell cycle arrest functions, Mol. Cell. Biol. 23 (2003) 216–228. [34] S. Mulero-Navarro, E. Pozo-Guisado, P.A. Perez-Mancera, A. Alvarez-Barrientos, I. Catalina-Fernandez,

[35]

[36]

[37]

[38]

E. Hernandez-Nieto, J. Saenz-Santamaria, N. Martinez, J.M. Rojas, I. Sanchez-Garcia, P.M. Fernandez-Salguero, Immortalized mouse mammary fibroblasts lacking dioxin receptor have impaired tumorigenicity in a subcutaneous mouse xenograft model, J. Biol. Chem. 280 (2005) 28731–28741. K. Murphy, C.M. Villano, R. Dorn, L.A. White, Interaction between the aryl hydrocarbon receptor and retinoic acid pathways increases matrix metalloproteinase-1 expression in keratinocytes, J. Biol. Chem. 279 (2004) 25284–25293. M. Haque, J. Francis, I. Sehgal, Aryl hydrocarbon exposure induces expression of MMP-9 in human prostate cancer cell lines, Cancer Lett. 225 (2005) 159–166. M. Luca, S. Huang, J.E. Gershenwal, R.K. Singh, R. Reich, M. Bar-Eli, Expression of interleukin-8 by human melanoma cells up-regulates MMP-2 activity and increases tumor growth and metastasis, Am. J. Pathol. 151 (1997) 1105–1113. M. Krampert, W. Bloch, T. Sasaki, Activities of the matrix metalloproteinase Stromelysin-2 (MMP-10) in matrix degradation and keratinocyte organization in wounded skin, Mol. Cell. Biol. 15 (2004) 5242–5254.