H1299, H460 and HEK293 cells in the present study were purchased from the American Type Culture Collection (ATCC; Gaithersburg, MD, USA) and propagated in RPMI-1640 medium or Dulbecco's modified Eagle's medium (DMEM) (Gibco) supplemented with 10% fetal bovine serum (FBS) (Invitrogen) and 1% penicillin and streptomycin in a cell incubator (37°C, 5% CO₂).

The antibodies for MMP12 (cat. #ab52897; Abcam^®), brachyury (cat. #sc-17743 and cat. #sc-20109), GAPDH (cat. #sc-32233), Hsp90 (cat. #sc-7949) (all from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) were purchased and used in the present study. HRP-goat anti-rabbit IgG and HRP-goat anti-mouse IgG were purchased from Jackson ImmunoResearch Laboratories. Transfection of DNA and RNAi was performed using Lipofectamine 2000 from Life Technologies (cat. #11668019).

A lentivirus vector expressing human brachyury and lentivirus particles were obtained from Shanghai R&S Biotechnology (Shanghai, China). In brief, human brachyury cDNA amplified from lung-derived cDNA samples was inserted into the Pac1/Nhe1 restriction enzyme sites within the multiple cloning site of the lentivirus vector pLenti6.3-IRES2-EGFP/V5 DEST, resulting in the vector pLenti-Brachyury-EGFP. Three siRNAs used to selectively silence the expression of brachyury with the given sequences (CCUCGAAUCCACAGUGATT, GCAAAUCCUCAUCCUCAGUTT and CUCCAACCUAUUCUGACAATT) were synthesized by Shanghai GenePharma Co., Ltd., while the sequence for the negative control was (UUCUCCGAACGUGUCACGUTT), as previously reported (22). The most efficient siRNA special for Brachyury in the H460 cells line was selected for the following shRNA construction (23), which targets ORF of Brachyury (+1021/+1041 bp). Plenti6.3-IRES-EGFP/V5 DEST was used to construct Bra-shRNA and con-shRNA according to shRNA construction protocol from Shanghai R&S Biotechnology (Shanghai, China).

Transfection of the PGL3-Basic and pcDNA3.1-Flag-Brachyury plasmids was performed using Lipofectamine 2000 (cat. #11668019) following the manufacturer's instructions. Basically, 5×10⁵ H1299 cells were seeded in a 6-cm culture dish overnight before transfection. Plasmid DNA (2 µg) and 6 µl Lipofectamine 2000 reagent were gently mixed in 500 µl RPMI-1640 medium without serum separately at room temperature for 5 min. The diluted DNA was combined with the diluted Lipofectamine 2000 and incubated for 20 min. The mixture was added into a culture dish drop by drop. Medium was replaced after 24 h.

The constructed plenti 6.3-Brachyury/shRNA plasmids and packaging mix (pLP1-Gag-pol, pLP-VSVG and pLP2-Rev) were co-transfected into H293T cells using POLOdeliverer™ 3000 transfection reagent (cat. #CT001-1; Shanghai R&S Biotechnology). The virus supernatant medium was collected after H293T cell infection for 48 h. The mixture was centrifuged at 3,000 rpm × 15 min at 4°C. Then, the liquid was filtered using a 0.45-µm filter membrane. The acquired virus was stored at −80°C until use. H1299-pBra (brachyury-overexpressing) and H1299-pcon (negative control), H460-pshT (Brachyury-knockdown) and H460-pshN stable cells were infected, and were then selected after the processes of fluorescence activated cell sorting (GFP selection) and Blasticidin treatment (cat. #60218ES50; InvivoGen).

Reverse-transcription PCR was performed using the SuperScript III One-Step RT-PCR system from Life Technologies (cat. #1257418) following the protocol provided in the manual. Total RNA (2 µg) was used for each reverse transcript reaction. The mRNA expression of 23 MMP members were semi-quantitatively measured via 2% agarose gel for H1299-pCon, H1299-pBra, H460-pshN and H460-pshT cells. Each experiment was performed in duplicate and was repeated 3 times. The MMP members were subjected to real-time PCR using SYBR-Green Master Mix (Bio-Rad) and the Mx3005P-quantitative RT-PCR system (Stratagene) according to the semi-quantitative PCR results. Mean Ct values for the target genes were normalized to the mean Ct values for the endogenous control 18s [−ΔCt = Ct (18s) − Ct (target gene)]. The ratio of mRNA expression of the target gene vs. 18s was defined as 2^−ΔCt.

Total protein extracts of each group were resolved by 11% SDS-PAGE and transferred onto polyvinylidene difluoride (PVDF) (Millipore) membranes. After blocking with 5% not-fat milk for 1 h at room temperature, the PVDF membranes were washed 3 times for 10 min with phosphate-buffered saline (PBS) followed by incubation with the primary antibodies [goat anti-human brachyury (1:8,000; 3% BSA dilution), rabbit anti-human MMP12 (1:250; 3% no-fat milk dilution) and mouse anti-human HSP90 (1:4,000; 3% BSA dilution)] overnight at 4°C. Following extensive washing, the membranes were incubated with secondary antibodies for 2 h. After washing 3 times for 10 min with Tris-buffered saline/Tween-20 (TBST) at room temperature once more, the immunoreactivity was visualized using an ECL kit (cat. #SK6668-100; Sangon Biotech) and the membranes were exposed to Kodak XAR-5 film (F5263; Sigma-Aldrich, USA).

Blind well chambers (Neuro Probe Inc.) with 8-µm pore polycarbonate filters coated with Matrigel (BD Biosciences) were used. RPMI-1640 medium supplemented with 10% FBS was added to the lower chambers; cells (1×10⁵) were added in serum-free medium to the upper chambers. After a 24-h incubation at 37°C, the filters were fixed and stained with 0.1% solution of crystal violet. Cells on the bottom side of the filters were counted in five random ×100 objective microscopic fields. Experiments were conducted in triplicate for each cell line. Two MMP12 siRNAs were synthesized (Shanghai GenePharma, Co., Ltd.) as previously reported (24). The sequence for MMP12-siRNA1 was 5′-GCUGUUUUU AACCCACGUUTT-3′; and the sequence for MMP12-siRNA2 was 5′-CCGUGAGGAUGUUGACUACTT-3′. Negative control siRNA was also used.

Immunohistochemical staining (IHC) was carried out following a previously published protocol (25). Specimens were deparaffinized with xylene, subjected to antigen repair, and were blocked for 15 min with 5% goat serum. The slides were incubated with the anti-brachyury (1:400 dilution) and anti-MMP12 (1;200 dilution) antibodies at 4°C overnight, followed by incubation with anti-rabbit second antibodies at room temperature for 30 min. Positive expression of IHC was reflected as brown staining estimated by averaging the numbers of positively stained cells under ×200 high power vision fields. The demographic information of the NSCLC cases used for IHC is provided in Table I. The present study was approved by the Institutional Review Board of Shanghai Changzheng Hospital. Informed consent was obtained before tissue collection. Additionally, the Ethics Committee of Shanghai Changzheng Hospital approved the use of the obtained patient specimens for this study.

Brachyury transcription factor cDNA was directionally cloned between XhoI and HindIII sites of the pCDNA vector which contained an N-terminal flag tag. Sequencing analysis was performed to verify the construct. H1299 cells were transfected with constructed pcDNA3.1-flag-Brachyury, followed by chromatin immunoprecipitation assay (ChIP). Briefly, nuclear proteins of the transfected H1299 cells were cross-linked and sonicated to obtain DNA fragments of 200–1000 bp. The lysates were diluted in ChIP dilution buffer and then incubated with salmon sperm DNA/protein A agarose slurry (Invitrogen). The chromatin solutions were incubated with anti-flag tag (Sigma) and RNA pol2 (cat. #sc-899; Santa Cruz Biotechnology) antibodies overnight at 4°C with rotation. Beads were pelleted by centrifugation. The IgG was used as a negative control. PCR amplification of the genomic fragments was performed with 10 specific primers flanking putative binding sites on the MMP12 promoter. The PCR products were separated by 2.0% agarose gel electrophoresis.

The brachyury promoter sequence (−1/−1800 bp) was directionally cloned between the XhoI and HindIII sites of the pGL3-basic vector with the following sense primer: 5′-CCGCTCGAGTGTGGCTTTGGATAAGTTAAGTTCA-3′ and antisense primer, 5′-CCCAAGCTTTGTAAACTTCTAAACGGATCAATTC-3′. Sequencing was performed to verify the construct. Three subcloning vectors of point mutation were performed using Site-Directed Mutagenesis kit (cat. #D0206; Shanghai Beyotime, China) according to the manufacturer's protocol. Primers of pGL3-mmp12-MT1 (mutation at −959/−953) and pGL3-mmp12-MT2 (mutation at −923/−917) were: 5′-GAGGCAGCGAACTTATGGCTCGATTTGAAGGGGTGTTGA-3′ (forward) and 5′-TCAACACCCCTTCAAATCGAGCCATAAGTTCGCTGCCTC-3′ (reverse); 5′-TTGAGTAGAATCTATCCTCGTGCCGCAG CACCACTGCTT-3′ (forward), and 5′-AAGCAGTGGTGCTGCGGCACGAGGATAGATTCTACTCAA-3′ (reverse), respectively. PGL3-mmp12-MT1+2 (mutation at −959/−953 and −923/−917) was also constructed. These constructs were verified by sequencing.

After transfection and/or treatment, the cells were washed with PBS 3 times. The cells were then lysed in luciferase cell culture lysis buffer provided with the Dual-Luciferase Assay kit (Promega). After −80°C cryopreservation for 1 h and rotation at room temperature for 20 min, the whole cell lysates were collected into EP tubes and centrifuged at 4°C at 12,000 × g for 10 min. The supernatant was collected in a fresh tube and 20 µl of the supernatant was added to the luciferase assay substrate (20 µl). Luminescence was measured as relative light units, twice for each lysate. Readings for the luciferase assay were carried out using LUMIstar OPTIMA (BMG Labtech). Each assay was repeated 3 times. Fold repression values were represented as mean of three experiments.

All of the measurements were collected in triplicate for each independent operation. The data were statistically analyzed using the Student's t-test. Statistical analysis was carried out using SPSS 16.0 statistical software (SPSS, Inc., Chicago, IL, USA). Statistical significance was set as a p-value <0.05.

T-box transcription factor brachyury has been recently reported as a poor prognostic factor in various types of carcinomas and is mainly characterized as a driver of EMT (5,6). MMP family members play an important role in tumor cell metastasis promoted by the brachyury transcription factor in NSCLC. However the correlation between brachyury and MMPs remains unclear. We constructed and selected two stable cell lines (H1299-pBra and H460-pshT) and two negative control stable cell lines (H1299-pCon and H460-pshN). mRNA expression of brachyury in the H1299-pBra cells was significantly increased by 64.8-fold (p<0.001), compared with that in the H1299-pCon cells (Fig. 1A). mRNA expression of brachyury in the H460-pshT cells was significantly reduced by 73% (p<0.001), compared with that in the H1299-pCon cells (Fig. 1B). Protein expression of brachyury in the H1299-pBra and H460-pshT cell lines also showed consistent results (Fig. 2A and B).

After extraction of total RNA and reverse transcription of cDNA from the 4 NSCLC stable cell lines (H1299-pBra, H1299-pCon, H460-pshT and H460-pshN) we investigated the mRNA expression levels of 23 MMP members by semi-quantitative PCR. Furthermore, we selected 11 MMP members (MMP2, MMP7, MMP9, MMP10, MMP12, MMP14, MMP15, MMP19, MMP24, MMP25 and MMP28) for quantitative real-time PCR according to the agarose gel results (data not shown), and reported MMP members with increased expression in NSCLC tissues as reported in previous literature (10–20). mRNA expression levels of MMPs in the H1299-pBra cells were as follows. MMP2 exhibited a 47% decrease (p=0.07), MMP7 exhibited a 32% increase (p=0.61), MMP9 exhibited a 18% increase (p=0.11), MMP10 exhibited a 66% increase (p=0.34), MMP12 exhibited a 323% increase (p=0.02), MMP14 exhibited a 41% increase (p=0.88), MMP15 exhibited a 28% increase (p=0.94), MMP19 exhibited a 62% increase (p=0.59), MMP24 exhibited a 183% increase (p=0.40), MMP25 exhibited a 46% increase (p=0.86), MMP28 exhibited a 8% increase (p=0.60), compared with the relative levels in the negative control H1299-pCon cells (Fig. 1C). mRNA expression levels of MMPs in the H460-pBra cells were as follows. MMP2 exhibited an 80% increase (p=0.13), MMP7 exhibited a 34% decrease (p=0.09), MMP9 exhibited a 43% decrease (p=0.08), MMP10 exhibited a 56% decrease (p=0.07), MMP12 exhibited a 77% decrease (p<0.001), MMP14 exhibited a 19% decrease (p=0.68), MMP15 exhibited a 28% decrease (p=0.77), MMP19 exhibited a 23% increase (p=0.04), MMP24 exhibited a 35% decrease (p=0.29), MMP25 exhibited a 12% decrease (p=0.08), MMP28 exhibited a 110% increase (p=0.06), compared with the levels in the negative control H1299-pCon cells (Fig. 1D). We concluded that the expression of MMP12 was the most significantly correlated with the expression of brachyury. Consistent with the mRNA expression of MMP12 and brachyury, we observed that the protein level of MMP12 (54 kDa) and the active protein level of MMP12 (45 kDa) were increased in the H1299-pBra cells as compared to the H1299-pCon cells. The protein level of MMP12 and the active protein level of MMP12 were decreased in the H460-pshT cells as compared to the H460-pshN cells (Fig. 2C and D).

Next, we investigated the role of MMP12 in the promotion of NSCLC cell invasion by brachyury. We synthesized two MMP12 siRNAs for transfection into the H1299-pBra and H1299-pCon stable cells.

As shown in Fig. 3C, the MMP12 protein levels were reduced in the H1299-pCon and H1299-pBra cell lines following transfection of the two synthesized siRNAs targeting MMP12, compared to the levels in the cells transfected with a control siRNA. Importantly, the H1299-pBra cells displayed high protein expression of MMP12 compared with the level in the H1299-pCon cells, when we performed the same MMP12-siRNA treatment suggesting that brachyury promotes the expression of MMP12.

Consistent with the protein expression between brachyury and MMP12, the Matrigel invasion assay revealed higher invasive ability in the H1299-pBra cells than that in the H1299-pCon cells following treatment with MMP12-NC or MMP12-siRNA1 and MMP12-siRNA2 (Fig. 3). Statistical analysis between the H1299-pCon-NC and H1299-pBra-NC cells; between H1299-pCon-siRNA1 and H1299-pBra-siRNA1 cells; between H1299-pCon-siRNA2 and H1299-pBra-siRNA2 cells, showed a significant reduction in invasive ability (p<0.05) (Fig. 3B).

Expression levels of brachyury and MMP12 protein were assessed by IHC in a series of 30 primary NSCLC carcinomas and 22 NSCLC metastatic tissues. Fig. 4A and B shows representative results of the intensity scores observed for brachyury and MMP12 expression, suggesting a positive correlation between brachyury and MMP12 in the primary NSCLC carcinomas and metastatic tissues. Brachyury expression exhibited a high level in 4/30 cases, a medium level in 7/30 cases, a low level in 10/30 cases and a negative level in 9/30 cases in the primary NSCLC carcinoma tissues. MMP12 expression exhibited a high level in 4/30 cases, a medium level in 8/30 cases, a low level in 10/30 cases and a negative level in 8/30 cases. Furthermore, brachyury expression exhibited a high level in 4/22 cases, a medium level in 6/22 cases, a low level in 5/22 cases and a negative level in 5/22 cases in the lung cancer metastatic tissues. MMP12 expression exhibited a high level in 5/22 cases, a medium level in 5/22 cases, a low level in 7/22 cases and a negative level in 3/22 cases. We analyzed the correlation between brachyury and MMP12 in the same sets of tumors by scatter and agreement plots. These plots indicated that these two proteins were highly correlated in the same sets of tumor tissues. Pearson's correlation coefficient was used as a measure of correlation between brachyury and MMP12, which was 0.59 for primary lung cancer tissues (Fig. 4C) and 0.665 for lung cancer metastatic tissues (Fig. 4D).

Since the mRNA and protein expression levels of MMP12 were correlated with brachyury, we sought to explore whether brachyury regulates MMP12 in the invasion and metastasis of NSCLC cells. We designed a CHIP-PCR for the MMP12 promoter in the H1299 cells transfected with the pcDNA3.1-Flag-Brachyury vector. The 5′ MMP12 promoter sequence (−3049 bp to −1 bp) was divided into 10 parts, and 10 pairs of primers were designed to amplify each part of the promoter. The anti-Flag Tag antibody specifically pulled down DNA fragments corresponding to the −1443 bp to −266 bp region (Fig. 5A). This effect was not observed when an irrelevant control IgG was used.

Brachyury and other T-box transcription family members preferentially bind to the consensus elements AATTTCACAC CTAGGTGTGAAATT (6). Bioinformatic analysis revealed two putative half-site 5′-GGTGTGA-3′ (−959/−953 bp) and 5′-TCACACC-3′ (−923/−917 bp) within the region (Fig. 5B). Then, we performed 4 cloning plasmids including pGL3-MMP12-WT (−1/−1,843 bp) and other three point mutation pGL3-basic plasmids (pGL3-MMP12-MT1, pGL3-MMP12-MT2 and pGL3-MMP12-MT1+2) (Fig. 5B). MMP12 promoter luciferase reporter assay showed that the point mutations at each or both potential half-site of the consensus elements reduced the luciferase units, and the second position (−923/−917 bp) presented a more marked and significant reduction by 39.1% (p<0.05) (Fig. 5C). These samples of the luciferase reporter assay were subjected to western blot assay using GAPDH and brachyury antibodies (Fig. 5D).