Down-regulation of miR-129-5p via the Twist1-Snail feedback loop stimulates the epithelial-mesenchymal transition and is associated with poor prognosis in breast cancer.

The epithelial to mesenchymal transition (EMT) plays a pivotal role in breast cancer progression. We found that overexpression of miR-129-5p reversed EMT, whereas depletion of miR-129-5p induced EMT in breast cancer cells. We demonstrated that Twist1 is a direct target of miR-129-5p. Both Twist1 and Snail transcriptionally suppressed miR-129-5p expression. Levels of miR-129-5p were low in breast cancer tissues. miR-129-5p down-regulation correlated with advanced clinical stage and poor prognosis in patients with breast cancer. miR-129-5p expression negatively correlated with Twist1 and Snail expression. Thus, miR-129-5p down-regulation fosters EMT in breast cancer by increasing Twist1-Snail and activating a negative feedback loop.

While miR-129-5p has been studied in a variety of human cancers, its role in breast cancer is unclear as are the molecular mechanisms underlying miR-129-5p regulation. Breast cancer is one of the most common malignant tumors in females [22], and it is a heterogeneous disease with respect to molecular alteration, cellular composition, and clinical outcome [23]. The incomplete understanding of its carcinogenic mechanisms has lead to difficulties in selecting targeted treatments, contributing to low survival rates. In this study, we investigated the mechanisms of miR-129-5p regulation in breast cancer cells. We find that miR-129-5p serves as a tumor-suppressor in the development and progression of breast cancer.

miR-129-5p is a tumor suppressor in breast cancer
Using RT-qPCR, we examined miR-129-5p expression in six breast cancer cell lines (MCF7, T47D, BT474, BT549, MDA-MB-468, and MDA-MB-231) and one breast epithelial cell line (MCF10A). miR-129-5p expression was down-regulated in all breast cancer cell lines as compared to MCF10A cells ( Figure 1A). Next, we investigated the influence of miR-129-5p on cell proliferation by transfecting miR-129-5p inhibitor or mimic in MCF7 and MDA-MB-231 cell lines ( Figure 1B). miR-129-5p inhibition resulted in a higher proliferation rate in MCF7 cells as assessed by MTT and plate colony formation assays ( Figure 1C left and 1D left). In contrast, miR-129-5p over-expression reduced cell growth and produced smaller plate colonies in MDA-MB-231 cells ( Figure 1C right and Figure 1D right). To investigate the role of miR-129-5p in cell migration and invasion we used a trans-well assay. miR-129-5p over-expression reduced both cell migration and invasion in MDA-MB-231 cells, while its inhibition enhanced cell migration and invasion as compared to control cells ( Figure 1E). The number of apoptotic cells was significantly higher in miR-129-5poverexpressed MDA-MB-231 cells compared with the control cells ( Figure 1F). These results indicate that miR-129-5p inhibits breast cancer cell proliferation, migration, and invasion and promotes cell apoptosis, suggesting that miR-129-5p is a tumor suppressor in breast cancer.

miR-129-5p suppresses EMT in human breast cancer cells
To further examine the effect of miR-129-5p expression in breast cancer EMT, we measured expression of epithelial and mesenchymal markers by RT-qPCR and immunoblotting. miR-129-5p-depleted MCF7 cells exhibited a significant up-regulation of vimentin and N-cadherin, while epithelial markers E-cadherin and ZO-1 were dramatically down-regulated ( Figure 3A, left and Figure 3B, left). In contrast, vimentin and N-cadherin were down-regulated, while E-cadherin and ZO-1 were up-regulated in miR-129-5p-overexpressing MDA-MB-231 cells ( Figure 3A, right and 3B, right). E-cadherin protein expression was decreased in miR-129-5p-depleted MCF7 cells whereas vimentin protein expression was decreased in miR-129-5p-overexpressed MDA-MB-231 cells compared to controls ( Figure 3C). To probe the interactions between miR-129-5p and EMTinducing transcription factors, we examined mRNA levels of known EMT inducers Slug, Snail, Twist1, ZEB1, and ZEB2. Expression of all five transcription factors increased in response to miR-129-5p depletion in MCF7 cells ( Figure 3D, left), while all but ZEB1 had decreased expression in response to miR-129-5p over-expression,in MDA-MB-231 cells (Figure 3D right). These results suggest that miR-129-5p is a suppressor of breast cancer EMT.

Twist1 is a direct target of miR-129-5p
Loss of E-cadherin is a critical event in EMT. We next used a luciferase reporter assay to investigate whether miR-129-5p transcriptionally regulates E-cadherin expression. Relative luciferase activity of E-cadherin was decreased in MCF7 cells transiently co-transfected with the E-cadherin promoter (pE-CAD-wt) and miR-129-5p inhibitor or control ( Figure 4A, left). Conversely, luciferase activity was increased in MDA-MB-231 cells transiently co-transfected with pE-CAD-wt and miR-129-5p mimic or control ( Figure 4A, right). To further explore the interaction of miR-129-5p with the E-cadherin promoter, we co-transfected an E-box mutated E-cadherin promoter (pE-CAD-mu) with miR-129-5p inhibitor/mimic or control. Relative luciferase activity was not changed in either MCF7 and MCA-MB-231 cells ( Figure 4A), indicating that miR-129-5p does not directly regulate E-cadherin transcription.
Twist1 was predicted as a potential target gene of miR-129-5p based on miR-target analysis (http:// www. microrna.org). The predicted binding of miR-129-5p with the Twist1 3′UTR is shown in Figure 4B. To demonstrate that miR-129-5p regulates the expression of Twist1, miR-129-5p inhibitor and mimic were transfected into MCF7 and MDA-MB-231 cells, respectively. Twist1 protein expression increased when MCF7 cells were treated with miR-129-5p inhibitor and decreased when MDA-MB-231 cells were treated with miR-129-5p mimic ( Figure 4C). To examine whether miR-129-5p directly interacts with the predicted 3′UTR of Twist1, the 3′UTR of human Twist1 was cloned downstream of the LUC coding gene of pGL3-control (Tw-3′UTR-wt) and co-transfected with miR-129-5p inhibitor or mimic www.impactjournals.com/oncotarget To ascertain whether miR-129-5p regulates breast cancer progression through its interaction with Twist1, we performed a rescue experiment. Knockdown or overexpression of Twist1 rescued the effects of miR-129-5p depletion or overexpression on breast cancer proliferation ( Figure 4E and 4F), migration, and invasion ( Figure 4G) in MCF7 and MDA-MB-231 cells. These data indicate that miR-129-5p targets Twist1, resulting in the negative regulation of breast cancer proliferation, migration, and invasion.

miR-129-5p is a transcriptional target of Twist1 and Snail
Analysis of the miR-129 promoter revealed three E-box motifs (™300 to +1 bp; Figure 5A). To investigate the binding of Twist1 and Snail to the miR-129 promoter, we performed a ChIP assay in MDA-MB-231 cells. The miR-129 promoter fragment was enriched with both anti-Twist1 and anti-Snail antibodies but not by IgG control ( Figure 5B), suggesting that both Twist1 and Snail can bind to the miR-129 promoter region. Next, we investigated whether Twist1 and Snail regulate the transcriptional activation of miR-129 using a dualluciferase reporter assay. The miR129-Ebox-wt plasmid, which contained the fragment of miR-129 promoter, was transiently co-transfected with Twist1 and/or Snail plasmids into MCF7 cells, or with Twist1 and/or Snail siRNAs and their controls into MDA-MB-231 cells. miR-129 promoter activity was decreased in Twist1and/ or Snail-overexpressed MCF7 cells, whereas activity was increased in Twist1 and/or Snail-depleted MDA-MB-231 cells compared to those of controls ( Figure 5C). We next co-transfected miR-129-5p, an Ebox-mutated luciferase report plasmid (miR129-Ebox-mu), with Twist1 and/ or Snail plasmids into MCF7 cells, or with Twist1 and/ or Snail siRNAs into MDA-MB-231 cells. The miR129-Ebox-mu was no longer responsive to the presence of Twist1 or Snail ( Figure 5C).

miR-129-5p is down-regulated in breast cancer and correlates with patients' prognosis
To examine whether the biological effects of miR-129-5p on breast cancer progression were clinically relevant, we assessed its expression in 30 cases of paired normal and breast cancer tissues. miR-129-5p expression was down-regulated in 26/30 collected breast cancer tissues as compared with paired adjacent normal breast tissues ( Figure 6A). Furthermore, miR-129-5p expression levels were lower in advanced tumor stage III than in tumor stages I/II ( Figure 6B). We then subdivided the tumors into two groups according to the miR-129-5p median expression and analyzed the relationship between miR-129-5p expression and clinicopathologic features known to affect patient outcome, including patient age, menopausal status, tumor size, clinical stage, histological grade, lymph node status, and ER/PR/Her2 status. As shown in Table 1, miR-129-5p expression levels were correlated with clinical stage (P < 0.001), histological grade (P < 0.001), lymph node status (P < 0.001) and PR status (P = 0.016). A Kaplan-Meier survival analysis revealed that patients in miR-129-5p high group had a more favorable outcome than those in the miR-129-5p low group ( Figure 6C).
Multivariate analysis was performed to determine the prognostic value of miR-129-5p expression using the Cox proportional hazard model. The risk variables examined included miR-129-5p expression level, as well as the factors examined in correlation analysis, above.
In the univariate analysis, low miR-129-5p expression level (P < 0.001), lymph node status (P = 0.005), Clinical stage (P < 0.001), and histological grade (P = 0.022) were  (Table 2). In the final multivariate Cox regression model, low miR-129-5p expression (P = 0.012) and advanced clinical stage (P = 0.008) were associated with a poor survival prognosis independent of other clinical covariates (Table 2). To address whether the expression of miR-129-5p is associated with its targets or regulators, both Twist1 and Snail mRNA expression were examined in 30 cases of primary breast cancer tissues by RT-qPCR. There was a significant inverse correlation between miR-129-5p and Twist1 ( Figure 6E) or Snail ( Figure 6F) expression inav breast cancer tissues. These results indicate that lower  miR-129-5p expression is associated with breast cancer progression and a worse prognosis.

DISCUSSION
miRNAs are emerging as a new class of tumor suppressor genes or oncogenes in cancer [25,26] and, in the future, they may serve as new biomarkers to predict clinical outcomes [27]. In this study, we provide the first evidence that miR-129-5p plays a role in breast carcinogenesis [7]. Its expression is down-regulated in breast cancer tissues, with lower levels in advanced tumor stages. miR-129-5p down-regulation is associated with a poor prognosis. The negative relationship of miR-129-5p expression levels with proliferation and motility supports the notion that it acts as a tumor suppressor miRNA. miR-129-5p was first found to be downregulated in undifferentiated gastric cancer tissues [28] where it inhibits gastric cancer proliferation [13]. Later it was found to inhibit growth and induce cell death in hepatocellular carcinoma, bladder carcinoma, colorectal carcinoma, esophageal carcinoma, and laryngeal squamous cell carcinoma [14,18,21,29]. Conversely, it is up-regulated in esophageal squamous cell carcinoma [20]. These conflicting data suggest that miR-129-5p may have different functions depending on the cell context and cancer type. Our data suggest that its down-regulation is associated with more aggressive tumors and a poorer EMT enables epithelial cells to acquire an invasive mesenchymal phenotype and is a key mechanism for the initial step of metastasis [30,31]. Recently, several miRNAs, such as the miR-200 [32] and miR-34 [33] families, have been found to participate in the regulation of EMT. Our data demonstrates that over-expression of miR-129-5p prevents TGF-β-induced EMT and downregulation of miR-129-5p is required for EMT initiation and maintenance in breast epithelial cells. In addition, it is sufficient to reverse the mesenchymal characteristics and decrease the invasiveness of mesenchymal cells. In contrast, inhibition of miR-129-5p in epithelial cells triggered EMT, resulting in a more invasive phenotype. miR-129-5p suppressed EMT in human breast cancer by directly targeting Twist1. Twist1 promotes the EMT process [34,35], and elevated Twist1 expression is also correlated with tumor proliferation, invasion, and metastasis in a variety of solid tumors, including breast cancer [36], lung cancer [37], prostate cancer [38], and oral cancer [39]. miR-300 and miR-720 inhibit EMT and metastasis by targeting Twist1 expression [40,41]. Similarly, we show that Twist1 is a direct target of miR-129-5p. Moreover, there is a significant inverse correlation between miR-129-5p and Twist1 expression levels in breast cancer tissues, suggesting a functional interaction of miR-129-5p and Twist1.
Both Twist1 and Snail are over-expressed in various human solid tumors and enhance proliferation, invasion, and metastasis by induction of EMT [34,42,43]. They bind to the E-box motif (canonical, 5′-CANNTG-3′ and non-canonical, 5′-CANNTT-3′) in target gene promoters and suppress the expression of target genes [7,43]. Our work indicates that both Twist1 and Snail bind to the miR-129-5p promoter region and suppress its expression. As Twist1 and Snail play an essential role in EMT as well as breast cancer metastasis, our data establish a mechanistic link between miR-129-5p, Twist1, Snail, in EMT and breast cancer metastasis.  with 10% FBS without CO 2 at 37°C. All cell lines were cultured with 100 μg/mL streptomycin and 100 units/ mL penicillin. The cells were expanded immediately and multiple aliquots were cryopreserved. Cells were used within 6 months of resuscitation.

Clinical samples
Breast cancer specimens were obtained from Tianjin Medical University Cancer Institute and Hospital. A total of 200 paraffin-embedded specimens and 30 primary breast cancer tissue and the paired adjacent normal breast tissue specimens were included in this study. All tumors were from patients with a newly diagnosed breast cancer who had received no therapy before sample collection. After radical prostatectomy, the primary breast cancer tissue and the adjacent normal tissues were flash-frozen in liquid nitrogen and stored at −80°C. This study was approved by the Institutional Review Board of the Tianjin Medical University Cancer Institute and Hospital and written consent was obtained from all participants.
Luciferase assays were carried out on extracts from different breast cancer cells co-transfected for 24/48 hours with the corresponding plasmids, miRNAs, or siRNAs by using a dual luciferase assay kit (Promega) according to the manufacturer's recommendations. The results were normalized against Renilla luciferase activity. All transfections were performed in triplicate.

Proliferation, migration and invasion assays
Both MTT and plate colony formation assays were used to evaluate the ability of cell proliferation. For MTT assay, 24 h after transfection, 5 × 10 3 cells were seeded in 96-well plates per well. After incubation for indicated time, the cells were incubated with 10 μl MTT (0.5 mg/ml; Sigma-Aldrich) at 37°C for 4 h. Then, the medium was removed, and precipitated Formosan was dissolved in 150 μl DMSO. The absorbance at 570 nm was detected using a micro-plate auto-reader (Bio-Rad, Richmond, CA, USA). For the plate colony formation assay, 24 h after transfection, 500 cells were seeded in 6-well plates per well. After about 3 weeks, the colonies obtained were washed with PBS and fixed with 10% formalin for 15 min at room temperature and then washed with PBS followed by staining with hematoxylin. The colonies were counted and compared with control cells.
The invasion and migration abilities of breast cancer cells in vitro were evaluated by Matrigel coated Transwell and Transwell inserts (BD Biosciences, San Diego, CA, USA), respectively. Briefly, 5 × 10 4 cells in 500 μl serumfree medium were added to the upper chamber, and medium containing 20% FBS was added into the lower chamber. 24 h later, the migrant cells that had attached to the lower, surface were fixed with 20% methanol and stained for 20 min with crystal violet. The membranes were then carved and embedded under cover slips with the cells on the top. The number of migrating cells was counted under a microscope in five predetermined fields.

RNA extraction and reverse transcription quantitative PCR
Total RNA of cultured cells, surgically resected fresh breast tissues, and formalin-fixed paraffin-embedded clinical specimens were extracted using mirVana PARIS kit (Life Technologies) according to the manufacturer's instruction. For mRNA detection, reverse transcription was performed using a First Strand cDNA Synthesis kit (TakaRa, Dalian, China), according to the manufacturer's instruction. The real-time quantitative PCR was performed using GoTaq qPCR Master Mix (Promega) on a Bio-Rad iQ5 Optical System (Bio-Rad, Hercules, CA, USA). β-actin was used as an internal control. For miRNA detection, miRNA was reverse transcribed with TaqMan MicroRNA Reverse Transcription kit and real-time quantitative PCR was performed using TaqMan miR-129-5p and u6 RNA (used as a normalizer) assays (Life Technologies) following the manufacturer's instructions.
For immunofluorescence assays, cell were seeded onto glass coverslips in 24-well plates, washed with PBS, fixed in 4% formaldehyde solution for 30 min and then permeabilized with 0.2% Triton X-100/PBS for 15 min. Cells were blocked with 2% BSA in PBS for 30 min. Coverslips were incubated with primary antibodies overnight at 4°C, followed by incubation with FITC-/TRITC-conjugated secondary antibodies for 1 h at room temperature, and then stained with DAPI. Finally, coverslips were observed under a fluorescence microscope.

Chromatin immunoprecipitation
The Chromatin immunoprecipitation (ChIP) assay was carried out by using a ChIP assay kit (Upstate, Lake Placid, NY, USA) according to the manufacturer's recommendations. Briefly, approximately 1 × 10 7 cells were crosslinked with 1% formaldehyde and resuspended, lysed in lysis buffer, then sonicated to shear the genomic DNA to a range of 500 bp to 2000 bp. The chromatin fraction was immunoprecipitated by using an anti-Twist1 or anti-Snail antibody. The chromosomal DNA was purified by using a MinElute PCR purification Kit (Qiagen, Hilden, Germany) according to the manufacture's instruction. PCR was carried out using primers specific for the Twist1 or Snail binding region in the human mir-129 promoter (forward 5′-GGCGCAGGATGCAGGAGAT-3′ and reverse 5′-GAGGGAGAGCCAGGACCTA-3′). The PCR products were separated on a 2% agarose gel containing ethidium bromide.

Statistical analysis
Data are presented as mean ± standard deviation. Student's t-test was used to compare the differences between the experimental group and control group. The relationship between miR-129-5p expression and various clinicopathological variables was analyzed by the chi-square test. Kaplan-Meier survival curves and the log-rank test were used to evaluate the breast cancer patients with different miR-129-5p expression. Multivariate survival analysis was performed by a backward stepwise Cox proportional hazards regression model. All calculations were performed with the SPSS for Windows statistical software package (SPSS Inc., Chicago, IL, USA). The level of significance was set to P < 0.05.

CONCLUSION
In summary, we find that miR-129-5p suppresses breast cancer development and progression and that down-regulation of miR-129-5p, through a Twist1-Snail negative feedback loop, is required for EMT initiation and maintenance ( Figure 6G).