T cell factor-4 functions as a co-activator to promote NF-κB-dependent MMP-15 expression in lung carcinoma cells

Both TCF-4 and MMP-15 are closely linked to the development of lung cancer, while the regulatory role of TCF-4 in MMP-15 expression is still obscure. Here we found that expression of TCF-4 and MMP-15 was increased in lung cancer cells or tissues versus the normal ones. With gain-or loss-of -function studies, we demonstrated that TCF-4 positively regulated MMP-15 expression in mRNA and protein levels. With reporter gene assay, we found that TCF-4 regulated MMP-15 expression via a potential NF-κB binding element locating at -2833/-2824 in the mouse MMP-15 promoter. With ChIP and immunoblotting assays, we identified that TCF-4 functioned as a co-activator to potentiate the binding between p65 and MMP-15 promoter. Functionally, TCF-4 silence attenuated the migration activity of LLC cells, while additional overexpression of MMP-15 rescued this effect in cell scratch test and transwell migration assay. In xenograft model, TCF-4 silence-improved tumor lesions in lungs and survival time of LLC-tumor bearing mice were abolished by MMP-15 overexpression. In conclusion, we are the first to identify TCF-4 as a co-activator of NF-κB p65 to promote MMP-15 transcription and potentiate the migration activity of the lung cancer cells. Our findings shed light on the therapeutic strategies of this malignancy.


TCF-4 and MMP-15 are highly expressed in lung cancer cells versus the normal ones.
To observe the association between TCF-4 and MMP-15 in lung cancer, we firstly demonstrated that both TCF-4 and MMP-15 mRNA levels were notably elevated in lung cancer cells (H460, A549 and LLC) versus the normal epithelial cells SAEC (Fig. 1a). In addition, we also found that mRNA levels of TCF-4 and MMP-15 were significantly increased in LLC-tumors comparing with the adjacent normal lung tissues in the mouse xenograft models (Fig. 1b). Furthermore, identical results were obtained in the clinical samples (Fig. 1c,d). Therefore, we concluded that TCF-4 expression might positively correlate with MMP-15 levels in lung cancer. We presumed that TCF-4 might be a regulator of MMP-15 expression, since TCF-4 is a transcription factor. To

TCF-4 upregulates MMP-15 expression via the promoter region in lung cells.
We further explored the regulatory effect of TCF-4 on the activity of MMP-15 promoter, since TCF-4 was a well-documented transcription factor. As expected, we found that silence of TCF-4 largely attenuated the promoter activity of mouse (Fig. 3a) and human (Fig. 3b) MMP-15. We further focused on the regulatory mechanisms in mouse promoter by constructing a series of truncated reporter genes (Fig. 3c). The reporter gene assays indicated that TCF-4 regulated the promoter activity of mouse MMP-15 through the promoter region -3000/-2500 (Fig. 3d).  the region -2496/-2288 in human MMP-15 promoter) was observed according to the online alignment software (http://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE_TYPE= BlastSearch&BLAST_SPEC= blast2seq&LINK_ LOC= align2seq) (Fig. 4a). Generally, the conserved promoter regions contain important regulatory elements for gene expression. In line with our aforementioned results, the conserved region -2998/-2773 in mouse MMP-15 promoter was located in the functional region of TCF-4 (-3000/-2500). We further predicted the potential binding elements in the region -2998/-2773 by using an online software (http://alggen.lsi.upc.es/). Unexpectedly, we did not find any potential binding elements for TCF-4. Noteworthily, we found that two potential binding sites for NF-κ B (-2958/-2949 and -2833/-2824 in mouse MMP-15 promoter; -2456/-2447 and -2348/-2339 in human MMP-15 promoter) got high scores. We presumed that TCF-4 might regulate MMP-15 promoter activity through interacting with NF-κ B, because the protein-protein interaction between TCF-4 and NF-κ B was reported in a previous study 25 . Therefore, we further mutated those potential NF-κ B binding elements (Fig. 4a) and found that the NF-κ B binding elements locating at -2833/-2824 in mouse MMP-15 promoter (Fig. 4b) and -2348/-2339 in human MMP-15 promoter (Fig. 4c) were responsible for TCF-4-induced MMP-15 promoter activity.

TCF-4 upregulates
TCF-4 interacts with NF-κB p65 to promote the DNA binding activity. To further verify the binding between NF-κ B p65 and mouse MMP-15 promoter DNA in live cells, ChIP assay was performed. The results indicated that p65 could bind with the sites (-2833/-2824) in mouse MMP-15 promoter, and this effect was attenuated by TCF-4 silence (Fig. 5a). In addition, we found that the promoter DNA could also be immunoprecipitated by TCF-4 antibody (Fig. 5b), indicating the interaction between TCF-4 and p65. Furthermore, we performed immunoprecipitation and immunoblotting assays, and verified the direct protein-protein interaction between mouse TCF-4 and p65 (Fig. 5c). Finally, we also verified this finding in human SAEC cells (Fig. 5d).
TCF-4 promotes NF-κB p65 translocation from cytosol to nucleus. To investigate the reciprocal regulation between TCF-4 and p65, we operated the expression of TCF-4 and p65 in LLC cells. We demonstrated that overexpressing TCF-4 upregulated MMP-15 mRNA levels, with the p65 mRNA levels being not altered significantly (Fig. 6a). Likewise, silence of TCF-4 attenuated MMP-15 mRNA expression, with the p65 mRNA levels not altered notably (Fig. 6b). However, TCF-4 regulated MMP-15 expression in a p65 dependent manner ( Fig. 6a,b). Those results indicated that TCF-4 might regulate MMP-15 expression via modulating protein activity of p65. Indeed, as shown in Fig. 6c, silence of TCF-4 decreased the levels of p65 in the nucleus, but increased p65 levels in the cytosol (Fig. 6c). Consistent with the p65 translocation, silence of TCF-4 could suppress the activity of a reporter gene harboring the promoter of CCL20, which is a target of NF-κ B p65 26 . Those results indicated that TCF-4 promoted NF-κ B p65 translocation from cytosol to nucleus in lung cells.

TCF-4/MMP-15 pathway aggravates the progression of LLC-tumor.
To observe the function of TCF-4/MMP-15 pathway in lung cancer cells, the scratch test and transwell assay were performed, since MMP-15 was a metastasis-associated factor in previous studies 22 . As expected, silence of TCF-4 in LLC cells notably inhibited the cell migration rate in the scratch test, while additionally overexpressing MMP-15 in LLC cells rescued this effect (Fig. 7a,b). In addition, identical results were obtained in the transwell assay ( Fig. 7c,d). Finally, the xenograft model was carried out to verify the findings in vivo. As shown in Fig. 8a, silence of TCF-4 in LLC cells significantly decreased the tumor lesions in the lungs, while additional overexpression of MMP-15 could rescued this effect (Fig. 8b). Consistently, the TCF-4 silence-improved survival time was abolished by additional overexpression of MMP-15 in LLC cells (Fig. 8c).

Discussion
Both TCF-4 and MMP-15 were closely linked to the development of lung cancer 18,22 , while the regulatory role of TCF-4 in MMP-15 expression was still not elucidated. Here in this study, we identified MMP-15 as a target gene of TCF-4 in lung cancer cells, providing a mechanism linking Wnt signaling and TCF-4 to the metastasis and progression of lung cancer.
We were the first to identify TCF-4 as a regulator of MMP-15 expression in normal and malignant lung cells in mouse and human species. This regulatory mechanism might be universal, because both TCF-4 and MMP-15 were expressed ubiquitously in other normal and malignant cells. Therefore, the roles of TCF-4/MMP-15 pathway in other types of tumors should be explored in future.
MMP-15 played a critical role in promoting tumor growth and invasion in previous studies 22,27 . In the present study, we identified MMP-15 as a target gene of TCF-4 downstream of Wnt signaling. Those findings indicated that MMP-15 might mediate Wnt signaling-regulated tumor progression. Actually, we demonstrated that the TCF-4/MMP-15 pathway was causally linked to the invasion and metastasis of lung cancer cells. Blockage of the TCF-4/MMP-15 pathway might represent an effective strategy to manage the therapy of lung cancer in clinic. We and others found that MMP-15 expression was significantly elevated in lung cancer cells or tissues related to the normal ones [27][28][29][30] , indicating that MMP-15 might also be a good marker for the diagnosis and therapy of lung cancer. As a transcription factor, TCF-4 had a series of target genes, such as c-myc 14 , cyclin D1 15 , c-jun 16 and MMP7 17 . Those target genes exerted powerful function in the development of various types of tumors. However, whether those targets were still regulated by TCF-4 and which one was the most important in lung cancer were still obscure. In the present study, we identified that at least MMP-15 was rather important in TCF-4-midiated metastasis of lung cancer. Thus, MMP-15 might be an effective target for metastatic lung cancer. Future studies in clinic need to be performed to verify this presumption.
Noteworthily, we identified TCF-4 was a co-activator of NF-κ B p65, which directly targeted to a potential NF-κ B binding element in the distant promoter region of MMP-15. To our knowledge, TCF-4 was a downstream effector of Wnt signaling and was co-activated by β -catenin 3 . Meanwhile, NF-κ B controlled the transcriptional specificity via the assembly of homodimers or heterodimers of 5 different NF-κ B proteins (p65, p50, c-Rel, p105 and p100) 31 . According to our data, we concluded that the protein complex binding to the promoter DNA of MMP-15 comprised with at least two components: TCF-4 and p65. Other potential components like β -catenin, p50, c-Rel, p105 and p100 need to be identified in the future study.
In conclusion, we are the first to identify TCF-4 as a co-activator of NF-κ B p65 to promote MMP-15 transcription and potentiate the migration activity of lung cancer cells. Our findings shed light on the therapeutic strategies of this malignancy.

Methods
Cell lines and culture conditions. The human lung cancer cell lines A549 and H460 as well as mouse lung cancer cell line LLC were grown in Dubelco's Minimum Essential Medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Gibco), 100 unites/ml penicillin, and 100 μg/ml streptomycin. The primary small airway epithelial cell (SAEC) was maintained in Airway Epithelia Cell Basal Medium (ATCC, PCS-300-030). All the cell lines were purchased from American Type Culture Collection (ATCC, USA) and cultured in a humidified incubator of 5% CO2 at 37 °C. Animal studies. All the animal experiments were approved by the Institutional Animal Care and Use of Committee at Chongqing Medical University and performed in accordance with the "Guide for the care and use of laboratory animals" published by the US National Institutes of Health (publication no. 85-23, revised 1996). All the mice were housed in a pathogen-free facility with a 12-h light, 12-h dark cycle and were fed with food and water ad libitum. Each cages contained no more than 5 mice.
In tail vein injection of xenograft model, the 8-week old female C57BL/6 mice received a tail vein injection of 5 × 10 5 LLC tumor cells (PBS as control) for 14 days. Then, the mice were sacrificed and the lungs were dissected for subsequent pathological observation (H&E staining). These studies were approved by the Institutional Animal Care and Use Committee of Chongqing Medical University.

Tissue collection. This study was approved by the Ethics Committee of the First Affiliated Hospital of
Chongqing Medical University and carried out in accordance with the approved guidelines. All patient-derived tissues were obtained with their written informed consent. We randomly collected 10 cases of primary lung carcinoma tissues and the corresponding adjacent normal tissue from the patients undergoing surgery. All the patients were pathologically diagnosed at the First Affiliated Hospital of of Chongqing Medical University. The age range of the patients was 36 to 58 years (median, 50 years). The patients did not receive any preoperative chemotherapy. The detailed clinical information of the patients was displayed in Table 1. Two to three pieces (around 5 mm in diameters) of the specimens were obtained from one patient after tumorectomy. All the specimens were stored in liquid nitrogen.
Realtime PCR. Total RNAs were extracted from cells with Trizol reagent (#15596-026, Invitrogen, USA) according to the manufacture's protocol. RNAs were transcribed into cDNAs using Omniscript (Qiagen, Hilden, Germany). Quantitative Real-Time PCR was performed using the 7900HT Fast Real-Time PCR system (Applied Biosystems, Darmstadt, Germany). The mRNA expression levels were normalized to the expression of GAPDH. Reactions were done in duplicate using Applied Biosystems Taqman Gene Expression Assays and Universal PCR Master Mix. The relative expression was calculated by the 2( −DDCt ) method. All the primers used for PCR are available upon request. Cell scratch tests. The LLC cells were transfected with si-NC (20 nmol/ml) or si-TCF4-1 (20 nmol/ml) plus PCMV (0.4 μg/ml) or PCMV-MMP-15 (0.4 μg/ml) for 24 hours. A scratch was made across the monolayer of LLC cells seeded in 6-well plates. Images of the scratches were captured at 0 and 18 hours using a digital camera (C5060, Olympus, Tokyo, Japan) mounted on an inverted microscope (CKX41, Olympus). Four different fields from each sample were considered for quantitative estimation of the distance between the borderlines and in each image four different equidistant points were measured in order to better estimate the real width of the wounded area. The migration rate is expressed as percentage of the control, and it was calculated as the proportion of the mean distance between both borderlines caused by scratching, to the distance which remained cell-free after re-growing.
Transwell migration assays. The LLC cells were transfected with si-NC (20 nmol/ml) or si-TCF4-1 (20 nmol/ml) plus PCMV (0.4 μg/ml) or PCMV-MMP-15 (0.4 μg/ml) for 36 h, and then the cells were digested and resuspended at 1 × 10 5 /ml. In migration assay, 200 μl of cell suspension was sucked into each insert of the Transwell (PC membrane with 8.0-μm pore size; No. 3422, BD, USA). After culture for 12 h, the upper inserts were air-dried, fixed with paraformaldehyde for 15 min, and stained with 0.1% crystal violet, and five fields of view were randomly selected to count cells under a microscope (× 200). The cell migration activity was described as the relative cell numbers of the transmitted cells.
Chromatin immunoprecipitation (ChIP) assays. This experiment was to measure the binding activity between p65 protein and the promoter DNA of MMP-15. In brief, the cultured LLC cells (transfected with si-NC, si-TCF4-1 or si-TCF4-2 for 36 hours, respectively) were cross-linked with 1% formaldehyde, followed by sonication. The supernatant with equal amounts of protein were immunoprecipitated with 1 μg of mouse p65 (sc-109, Santa Cruz) or TCF-4 antibody (#2569, Cell Signaling) or rabbit IgG as control using the ChIP Kit (#17-10460, Millipore Corp.) according to the manufacture's protocol. The immunoprecipitates were analyzed by PCR for detecting the co-immunoprecipitated DNA containing the functional p65 binding site (GGGAAAGTAC,-2833/-2824). The ChIP primers were designed as: forward: 5′ -ttccatctggcaccgaagcctt-3′ , reverse: 5′ -gaaggaagaaataaacaaacca-3′ . The length of the desired product was 100 base pairs.  Gain or loss of function studies. P65 overexpression was performed by transfecting the LLC cells with a constitutive expression plasmid for mouse p65 (empty vector pCDNA3.1 as control). Mouse endogenous p65 was knocked down by using the two mixed commercial siRNAs (sc-29411 and sc-44213, Santa Cruz). The control siRNA was purchased from Santa Cruz (sc-37007). TCF-4 overexpression was carried out by transfecting the cells with commercial constructs (MR207325 for mouse cells and SC109746 for human cells, OriGene). The endogenous TCF-4 was silenced by OriGene commercial siRNAs (SR414923: si-TCF4-1, RefSeq NM_001142918; si-TCF4-2, RefSeq NM_001142919) for mouse cells and SR304745 (si-TCF4, RefSeq NM_001146274) for human cells. MMP-15 overexpression was performed by transfecting the cells with a commercial construct (MC202385, OriGene). The transfection concentration for the siRNAs was 20 nmol/ml and for the expression plasmids was 0.4 μg/ml.
Cloning of reporter gene and site-directed mutation. The DNA fragments for the human or mouse MMP-15 promoter fusion reporter constructs shown in Fig. 3a-c were generated from SAEC cell or LLC cell genomic DNA by PCR amplification using KOD Plus(Toyobo), respectively. The DNA fragments for the mouse CCL20 promoter (-3000/-1) fusion reporter constructs shown in Fig. 6d was generated from LLC cell genomic DNA. The Amplified DNA fragments were subcloned into pGL4-Basic vector by blunting, kination, and ligation reaction. Finally, the mutant plasmid was transformed into Escherichia coli DH5α , and the positive clone was selected and confirmed by DNA sequencing. The NF-κ B specific binding sites in the human or mouse MMP-15 promoter region were mutated using MutanBEST Kit (Takara Bio, Inc., Shiga, Japan) according to the manufacturer's protocol. The primers used for the DNA fragments were designed upon request and displayed in Table 2.
Cell transfection and reporter gene assays. The transfections were performed by using Lipofectamine-2000 according to the protocol from the manufacturer. Briefly, the plasmids (0.4 μg/ml) or the siRNAs (20 nmol/ml) were transfected in the serum-free and antibiotic-free media. After transfection for 6 h, the media was removed and replaced with complete growth media with different treatment. Then, the luciferase activity of the cell lysate was evaluated according to the manufacture's instructions (Promega Corp., Madison, WI), and the total protein concentration in each well was measured as an internal control. The relative luciferase activity was displayed to compare the difference between each group.
Statistical analysis. All the data are displayed as mean ± S.E.M and were analyze by either one-way ANOVA or two-tailed unpaired Student's test. For each parameter of all data, *P < 0.05, **P < 0.01, and ***P < 0.005.