Fibronectins are cell-secreted glycoproteins that modulate cell attachment, spreading, migration, morphology, differentiation and oncogenic transformation. Fibronectin expression is activated during EMT (epithelial–mesenchymal transition) and is a hallmark of mesenchymal cells. It is shown in the present study that a transcription factor previously unrelated with EMT, TFCP2c/LSF/LBP-1c, was translocated to the nucleus and bound to the fibronectin promoter upon EMT induction by Snail1. Consequently, the interference of TFCP2c/LSF/LBP-1c's activity prevented fibronectin expression. Moreover, TFCP2c/LSF/LBP-1c was detected in nuclei of embryonic dermal mesenchymal cells adjacent to the hair bud, a cell population that expresses endogenous nuclear Snail1 and fibronectin. Therefore we indicate a new molecular role for TFCP2c/LSF/LBP-1c in fibronectin expression.
- epithelial–mesenchymal transition (EMT)
Advanced tumour cells frequently show a down-regulation of epithelial markers accompanied by increased expression of FN1 (fibronectin 1) and other mesenchymal genes. This morphogenetic process, referred to as EMT (epithelial–mesenchymal transition), was first described to be essential during embryogenesis  and it provides epithelial tumour cells with properties that support cancer progression, such as migration, invasion and extravasation , as well as a stem cell pathology associated with cancer . Fibronectins are a class of high molecular mass cell-secreted glycoproteins that modulate cell-substrate adhesion as well as a set of cellular processes, among which include cell attachment and spreading, migration, morphology, differentiation and oncogenic transformation . Fibronectin expression is a feature of mesenchymal cells and correlates with stimuli [such as TGF-β (transforming growth factor β)] that induce Snail1 expression and EMT . Snail1 is a member of the SNAI genes that is expressed in the EMT processes that have been analysed previously . This factor directly represses epithelial markers such as E-cadherin (the product of the CDH1 gene) by binding to E-boxes in their promoters [7,8].
The present study provides evidence that Snail1-induced FN1 gene expression requires the transcription factor TFCP2c (TFCP2c/LSF/LBP-1c) , and that SNAI1 and TFCP2c are co-expressed along with FN1 during embryogenesis. Although this transcription factor is known to interact with cellular and viral promoters, any interaction with FN1 or other mesenchymal genes has not been previously described. Therefore our results from the present study indicate a new role for TFCP2c in fibronectin regulation in Snail1-induced EMT.
Cell lines, transfection and infection
RWP1 (pancreatic carcinoma), HT-29 M6 (colorectal cancer), MEFs (mouse embryonic fibroblasts) and NMuMG (normal mouse mammary epithelial) cells were grown in DMEM (Dulbecco's modified Eagle's medium; Invitrogen) plus 10% fetal bovine serum under standard conditions. For NMuMG cells, medium was supplemented with 10 μg/ml of insulin. DNA transfection was performed with Lipofectamine™-PLUS (Invitrogen) reagent according to the manufacturer's instructions and DNA transduction was performed with lentivirus. These experimental procedures are detailed in the Supplementary online data (at http://www.BiochemJ.org/bj/435/bj4350563add.htm).
Transcription activity and DNA constructs
Luciferase-based reporter assays were performed as described previously [10,11]. DNA vectors used in the experiments are detailed in the Supplementary online data and in Supplementary Table S1 (at http://www.BiochemJ.org/bj/435/bj4350563add.htm).
Analysis of RNA
RNA was extracted with the Gene Elute™ Mammalian Total RNA Miniprep Kit (Sigma–Aldrich). For quantitative analysis, 1 μg of RNA was reverse-transcribed with Transcriptor First Strand cDNA Synthesis Kit (Roche) and 100 ng of the obtained cDNA was used as a template for quantitative SYBR-Green-based PCR. Correct product size was confirmed in agarose gels. The calculated RNA amount was systematically normalized by the Pumilio RNA levels. For the semi-quantitative analysis, 50 ng of RNA was analysed with SuperScript One-Step RT (reverse transcription)–PCR with Platinum Taq (Invitrogen) and agarose gels. See Supplementary Table S2 (at http://www.BiochemJ.org/bj/435/bj4350563add.htm) for the oligonucleotide sequences.
Analysis of proteins
A standard immunofluorescence protocol on cells grown for at least 48 h on ethanol-sterilized glass coverslips was applied. All steps were performed at room temperature (20°C). Cells were fixed with 4% PFA (paraformaldehyde) for 10 min. PFA autofluorescence was quenched with 50 mM NH4Cl/PBS for 5 min and cells were permeabilized with 0.2% Triton X-100 for 5 min. Blocking was carried out for 1 h with PBS containing 3% BSA. Coverslips were then incubated in a humid atmosphere for 1 h with a primary antibody (20 μg/ml for TFCP2c antibody) and, after extensive washing, for 45 min with the secondary antibody (in the dark). A complete list of antibodies used and their sources is described in Supplementary Table S3 (at http://www.BiochemJ.org/bj/435/bj4350563add.htm). Nuclear counterstaining was performed by incubating the coverslips for 1 min with 5 μg/ml propidium iodide and 100 μg/ml RNase. Coverslips were mounted with Fluoromount-G and fluorescence was viewed and captured through a Leica TCS-SP2 confocal microscope.
Analysis of proteins in the mouse embryo
All mice involved in the present study were maintained in a rodent barrier facility in order to guarantee the specific pathogen-free health status of the animals and the protocol was previously approved by the Animal Research Ethical Committee at the PRBB (Parc de Recerca Biomèdica de Barcelona). The protocol followed the one previously reported by Francí et al. . CD-1 pregnant mice (Harlan) were killed by cervical dislocation; the embryos were removed by Caesarean section, fixed in formalin and embedded in paraffin. Sections (4 μm) were dewaxed and rehydrated. Antigens were retrieved by boiling the samples in Tris/EDTA (50 mM Tris/HCl, 1 mM EDTA and 10 mM NaCl, pH 9.0) for 15 min. Endogenous peroxidase activity was quenched with 4% hydrogen peroxide in PBS containing 0.1% sodium azide for 15 min. After several rinses with PBS, sections were incubated with PBS containing 1% BSA to block non-specific binding and were washed with PBS. Sections were incubated with either 10 μg/ml of purified anti-Snail1 monoclonal antibody EC3, 5 μg/ml anti-fibronectin polyclonal antibody or 20 μg/ml anti-TFCP2 polyclonal antibody overnight at 4°C. After several further rinses with PBS, bound antibody was detected using anti-mouse or anti-rabbit Envision (Envision System Peroxidase; DAKO). Sections were counterstained with haematoxylin.
ChIP (chromatin immunoprecipitation)
Cells seeded in 150-mm-diameter plates were washed twice with pre-warmed PBS (37°C) and cross-linked with 1% formaldehyde for 10 min at 37°C in DMEM. The reaction was stopped by adding 250 μl of 2.5 M glycine (0.125 M final concentration) and incubating for an additional 2 min. Cells were washed twice with ice-cold PBS and 1 ml of Soft Lysis Buffer [50 mM Tris/HCl, pH 8.0, 2 mM EDTA, 0.1% Nonidet P40, 10% glycerol and protease inhibitors (10 μg/ml aprotinin, 1 mM leupeptin, 2 mM pefablock and 10 μg/ml pepstatin) and phosphatase inhibitors (1 mM β-glycerol phosphate, 10 mM sodium fluoride phosphatase and 2 mM sodium orthovanadate)] was added to the plates on ice. After scraping, lysates were transferred to Eppendorf tubes, incubated for 10 min on ice and centrifuged for 15 min at 800 g. The supernatant was discarded and the pellet was resuspended in SDS lysis buffer (1% SDS, 10 mM EDTA and 50 mM Tris/HCl, pH 8.0) and sonicated to generate fragments of DNA from 200 to 1000 bp (40% amplitude in Branson DIGITAL Sonifier® UNIT Model S-450D sonicator, 10 pulses of 10 s each). Lysates were incubated for 20 min on ice and subsequently centrifuged at maximum speed for 10 min.
Protein concentration in the supernatants was determined by the Lowry method  and 500 μg of protein per immunoprecipitation was diluted in Dilution Buffer (0.001% SDS, 1.1% Triton X-100, 16.7 mM Tris/HCl, pH 8.0, 2 mM EDTA, 2 mM EGTA and 167 mM NaCl). Pre-clearing in order to reduce background was performed for 3 h at 4°C with agitation using mouse IgG (Dako) and salmon sperm/BSA-blocked agarose–Protein G (Upstate). Samples were then centrifuged at 400 g to remove the beads, 10% of the sample was stored as the input and the rest of the sample was divided for immunoprecipitation and incubated with either specific anti-TFCP2c antibody (ab42973, Abcam) or an irrelevant antibody of the same species overnight at 4°C with agitation.
Blocked beads were added to each sample and incubated for an additional 1 h at 4°C. Five washes were performed in MoBiTec columns with each of the buffers: Low Salt Buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris/HCl, pH 8.0, and 150 mM NaCl), High Salt Buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris/HCl, pH 8.0, and 500 mM NaCl), LiCl Buffer (250 mM LiCl, 1% Nonidet P40, 1% sodium deoxycholate, 1 mM EDTA and 10 mM Tris/HCl, pH 8.0) and TE Buffer (10 mM Tris/HCl and 1 mM EDTA, pH 8.0). The columns were centrifuged to eliminate traces of buffer (2 min, 400 g) and the remaining beads were incubated with Elution Buffer (100 mM Na2CO3 and 1% SDS) at 37°C for 30 min. DNA was recovered by centrifugation (5 min, 400 g) to separate the eluate from the beads.
NaCl (4 μl of 5 M; 250 mM final concentration) was added to each eluate and de-cross-linking was performed by incubating samples at 65°C overnight. After 2–4 h of digestion with Proteinase K (55°C), DNA was purified by the GFX PCR DNA and Gel Band Purification kit (Amersham Biosciences) and finally analysed by quantitative SYBR-Green PCR (see Supplementary Table S4 at http://www.BiochemJ.org/bj/435/bj4350563add.htm for the oligonucleotides used).
For ChIP of exogenous promoters , 3.5×106 RWP1 cells were plated in 150 mm dishes and 5 μg of the indicated FN1 promoter was transfected with Lipofectamine™-PLUS reagent (Invitrogen) according to the manufacturer's instructions. At 48 h after transfection, cells were treated following the protocol described above. The amount of protein used per immunoprecipitation was 100 μg.
TFCP2c binds to the FN1 promoter in Snail1 expressing cells
Up-regulation of mesenchymal genes, such as FN1, is a characteristic molecular event related to EMT [10,11]. Accordingly, fibronectin RNA and protein levels increased when RWP1 and HT-29 M6 epithelial cells were forced to transit towards a mesenchymal phenotype by ectopic Snail1–HA (where HA is haemagglutinin) expression (Figure 1a). In line with this increment, the transcriptional activity of the −341/+265 proximal FN1 promoter was higher in Snail1 than in control cells (Figure 1b). Whereas fibronectin basal levels in HT-29 M6 were lower than in RWP1 cells, the increase induced by Snail1 was higher. As expected for EMT, we detected reduced expression of the epithelial gene E-cadherin in Snail1-expressing cells (Figures 1a and 1b).
To find the transcription factors mediating FN1 activation by Snail1, we looked for transcription factor-binding sites in the FN1 proximal promoter. We narrowed our search to the initial 19 bp (−341/−323) sequence of the proximal promoter, since the exclusion of this region was sufficient to interfere with FN1 transcription by Snail1, as observed through reporter assays of 5′-deleted proximal FN1 promoters (Figure 1c). The transcription factor-binding algorithm [TESS (Transcription Element Search Software)] detected three consensus binding motifs for the transcription factor TFCP2c, which was conserved among species (Figure 1d); therefore, we decided to evaluate the possibility that TFCP2c was mediating fibronectin activation in Snail1 cells.
We analysed whether TFCP2c was binding to the FN1 promoter by means of ChIP assays. We detected binding of TFCP2c to the proximal sequence of the FN1 promoter in HT-29 M6 Snail1–HA, but not in control cells. As expected, Snail1 expression did not stimulate the binding of an irrelevant transcription factor (Figure 2a). Taking advantage of the fact that the transfection efficiency of RWP1 is higher than that of HT-29 M6, we tested whether the TFCP2c is binding to the exogenous proximal FN1 promoter transfected into RWP1 control and Snail1 cells. Similarly to the endogenous promoter, we found higher binding of TFCP2c to the transfected proximal promoter (−341/+265) in RWP1 Snail1–HA compared with control cells; however, this was not the case for a promoter missing the sequence (−341/−332) including the TFCP2c binding motifs (Figure 2b). Next, we analysed whether the increased TFCP2c binding to the FN1 promoter in Snail1 cells was accompanied by an increase in the nuclear amount of TFCP2c using immunofluorescence. We detected higher nuclear TFCP2c staining in HT-29 M6 Snail1 than in control cells (Figure 2c). Since total levels of TFCP2c analysed by Western blotting were not increased in Snail1 cells, altogether our data indicate that Snail1 expression triggers the nuclear localization of TFCP2c and its binding to the FN1 proximal promoter.
TFCP2c is required for Snail1-induced fibronectin expression
The binding of TFCP2c to the FN1 promoter in Snail1 cells suggests that TFCP2c activity is required for FN1 transcription in these cells. Therefore we analysed the effect of a dominant-negative mutant for TFCP2c (TFCP2c Q234L/K236E) on fibronectin expression. This mutant interferes with TFCP2c transcriptional activity by preventing the binding of the endogenous wild-type TFCP2c to DNA . Although expression of the dominant-negative did not affect the low fibronectin protein levels in RWP1 and HT-29 M6 control cells, it clearly decreased the fibronectin levels detected in Snail1 cells (Figure 3a), especially in the case of HT-29 M6 cells. A similar decrease was detected at the mRNA level, quantified as 50 or 60% in RWP1 Snail1–HA or HT-29 M6 Snail1–HA cells respectively (see Supplementary Figure S1 at http://www.BiochemJ.org/bj/435/bj4350563add.htm). In accordance with the dominant-negative effects, we found that HT-29 M6 Snail1 cells depleted of TFCP2c by infection of a shRNA (small-hairpin RNA) specific for TFCP2c produced less fibronectin than Snail1 cells infected with an irrelevant shRNA (see Supplementary Figure S1). The involvement of TFCP2c in fibronectin activation was also tested in epithelial cells forced to transit to the mesenchymal phenotype by TGF-β, which induces endogenous Snail1 levels. As shown in Figure 3(c), fibronectin mRNA was increased after 8 h of TGF-β treatment in control NMuMG-sensitive cells, but remained at basal levels in cells transduced with the dominant-negative TFCP2c.
To further analyse the relevance of TFCP2c on cells expressing endogenous Snail1, we transduced the dominant-negative form of TFCP2c into MEFs in which fibronectin RNA expression depends on endogenous Snail1 (see the Supplementary online data and Supplementary Table S5 at http://www.BiochemJ.org/bj/435/bj4350563add.htm). In the same way as epithelial cells expressing exogenous Snail1, the expression of TFCP2c Q234L/K236E also decreased FN1 expression (Figure 3b). Other mesenchymal genes analysed such as LEF1, Snail1 or ZEB1 were not altered by the dominant-negative form. Since fibronectin modulates the cell–substrate interaction, we analysed whether the dominant-negative form also affected the migration of MEFs. Wound healing (Figure 3d) and single-cell tracking assays (Figure 3e) showed that the dominant-negative form of TFCP2c reduced the migratory capacity of fibroblasts, and this effect was rescued by providing exogenous fibronectin to the cells (see Supplementary Figure S2 at http://www.BiochemJ.org/bj/435/bj4350563add.htm).
Nuclear TFCP2c is expressed in embryonic dermal mesenchymal cells adjacent to the hair bud together with Snail1 and fibronectin
To verify that the effect of TFCP2c on FN1 expression was not restricted to cultured Snail1-expressing cells, we analysed the subcellular localization of TFCP2c in vivo. Expression of endogenous Snail1 is higher in embryonic than in adult tissues, but it is not constitutive and has a limited distribution in a subset of mesenchymal cells; for instance, Snail1 is expressed in embryonic mesenchymal cells adjacent to hair bud cells [12,16]. In line with this, by immunohistochemical analysis of whole mounted mice embryos at E14.5 (embryonic day 14.5), we found embryonic dermal mesenchymal cells with positive nuclear staining for Snail1 (Figure 4a). Interestingly, we found a similar pattern of nuclear expression of TFCP2c (Figure 4b), and fibronectin was detected in the same areas (Figure 4c). These results therefore suggest that the transcription factor TFCP2c is involved in the transcriptional activation of fibronectin by Snail1 in vivo.
Our results in the present study indicate for the first time a role for the transcription factor TFCP2c in FN1 transcription. LSF was known to regulate diverse cellular and viral promoters [9,17], but none of them have a direct role in EMT regulation. An indirect relationship between TFCP2c targets and EMT can be established considering the fact that EMT and Snail1 expression generate cells with stem properties that hold potential for tumour initiation, tumour recurrence and metastasis [3,18,19]. Thus it was described that TFCP2 activates the cENS-1 (chicken embryonic stem 1) gene expression early in embryonic development  and the osteopondin gene in the context of hepatocellular carcinoma . We have observed that overexpression of TFCP2c in epithelial HT-29 M6 or RWP1 cells did not increase fibronectin RNA levels (see Supplementary Figure S3 at http://www.BiochemJ.org/bj/435/bj4350563add.htm). Moreover, the epithelial phenotype was maintained and no changes in E-cadherin RNA levels (see Supplementary Figure S3) were observed in these cells. Although an increase of TFCP2c was not sufficient to induce a complete EMT in the epithelial cells tested, our results directly involve TFCP2c in the molecular events triggering the expression of the mesenchymal gene FN1. Thus we have unveiled a new molecular mechanism which could be targeted to prevent cancer cell properties associated with fibronectin expression.
Montserrat Porta-de-la-Riva performed most of the experimental work and generated most of the required reagents. Jelena Stanisavljevic participated in the work for Figures 2a, 2b and 4; Josue Curto for Figures 2a, 2b, 3c and Supplementary Figure S2; and Josep Baulida for Figures 3b–3e and Supplementary Figures S1 and S2. Clara Francí performed the animal work for Figure 4. Montserrat Porta-de-la-Riva and Josep Baulida designed the research. Victor Manuel Díaz and Antonio García de Herreros contributed in the analysis and interpretation of the data. Antonio García de Herreros and Josep Baulida provided financial support. Josep Baulida wrote the paper and all of the authors critically revised the paper.
This study was supported by the Fondo de Investigaciones Sanitarias of the ISCIII [grant number PI071054 (to J.B.)] and by the Ministerio de Ciencia y Tecnología [grant number SAF2006-SA00339] and the Fundació La Marató of TV3 to A.G.H. Partial support from the Instituto Carlos III [grant number RD06/0020/0040] and from the Generalitat de Catalunya [grant number 2005SGR00970] is also acknowledged.
We are grateful for the technical assistance of Dr Raúl Peña, Alicia Riquel and Marta Garrido.
Abbreviations: ChIP, chromatin immunoprecipitation; DMEM, Dulbecco's modified Eagle's medium; E14.5, embryonic day 14.5; EMT, epithelial–mesenchymal transition; FN1, fibronectin 1; HA, haemagglutinin; HPRT, hypoxanthine–guanine phosphoribosyltransferase; MEF, mouse embryonic fibroblast; PFA, paraformaldehyde; RT, reverse transcription; shRNA, small-hairpin RNA; TESS, Transcription Element Search Software; TGF-β, transforming growth factor β; TFCP2c, TFCP2c/LSF/LBP-1c
- © The Authors Journal compilation © 2011 Biochemical Society