Cell lines and breast tumor specimens

The cell lines MDA-MB-231 (HTB-26™), MDA-MB-436 (HTB-130™), MDA-MB-453 (HTB-131™), MDA-MB-468 (HTB-132™), BT-549 (HTB-122™), ZR-75-1 (CRL-1500™), ZR-75-30 (CRL-1504™), T47D (HTB-133 ™), MCF-7 (HTB-22™), SK-BR-3 (HTB-30™) and MCF-10A (CRL-10317™) were obtained from American Type Culture Collection (ATCC) and maintained in complete growth medium according to the ATCC-recommended culture conditions. HBL-100 was obtained from the Shanghai Cell Bank Type Culture Collection Committee (CBTCCC, Shanghai, China) and cultured in DMEM (Gibco, Gaithersburg, MD, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco). Mycoplasma testing was routinely conducted by HD Biosciences. All cells were stored in liquid nitrogen and used for no more than 6 months after being thawed.

Patients with locally advanced breast cancer (>5 cm in diameter or with clinically palpable axillary adenopathy) received preoperative chemotherapy, weekly paclitaxel (80 mg/m2) and carboplatin (AUC = 2). All human breast cancer specimens were obtained from patients who were diagnosed with breast cancer and underwent surgery at the Shanghai Cancer Center during 2003–2009 (n = 110). All specimens contained over 90% tumor cells and were stored in liquid nitrogen until analysis. The study was approved by the Ethics Committee of the Cancer Hosipital, Fudan University and each patient gave written informed consent to participation in this research.

siRNA, miRNA, plasmid construction, transfection, and luciferase assays

Specific siRNAs and scrambled siRNA were purchased from Genepharma, and miR-200a mimics, antagomiRNA, and negative controls were obtained from Ribobio. MicroRNA-200a represents miR-200a-5p in the current study. Transfection of siRNA were same as previously described [10]. The negative controls had no detectable effects in human cell lines or tissues (Additional file 1: Table S1).

The 3’-UTR of human TP53INP1 mRNA was cloned into the XhoI/NotI sites of the psiCHECK2 vector (Promega) using the In-Fusion® Advantage PCR Cloning Kit (Clontech). Luciferase assays were conducted as we previously reported [10]. The primers are listed in Additional file 1: Table S1.

Immunoblotting

Immunoblotting was performed using a standard method [10]. Primary antibodies used were: anti-p73, anti-cleaved caspase-3, and anti-Bim (Epitomics); anti-Noxa (AbD Serotec); anti-YAP1, anti-Bax, and anti-GAPDH (Proteintech); and anti-TP53INP1 (Prosci). Secondary antibodies used were horseradish peroxidase-conjugated anti-rabbit and anti-mouse IgG (Proteintech). Quantity One software (BioRad) was used to quantify protein expression.

RNA extraction and quantification of miRNAs

The tissues were homogenized with Polytron PT100. Total RNA was extracted using the Trizol reagent (Invitrogen), according to the manufacturer’s instructions. cDNA was synthesized from total RNA using a specific stem-loop RT primer (50 nM) with the ReverTra Ace®qPCR RT Kit (Toyobo). The primers for miR-200a and U6 detection assays were purchased from Ribobio. Real-time PCR was performed using the SYBR® Green Real-time PCR Master Mix (Toyobo) and conducted on the 7900HT Fast Real-Time PCR System (Applied Biosystems). The relative amount of miRNA or mRNA in each sample was calculated using the comparative CT method as described in our previous study [10].

Establishment of gemcitabine resistant cell line

MDA-MB-231 cells were selected using 1 μM gemcitabine for 24 h, and then they were washed with PBS and recultured in drug-free medium. The treated cells were pulsed with 1 μM gemcitabine during periods of exponential growth. This cycle of treatment was repeated six times. The cell lines were cultured in 7 μM gemcitabine to generate gemcitabine-resistant sublines. Prior to the start of the experiments, the MDA-MB-231 GR cells were cultured in a gemcitabine-free medium.

TUNEL assays

Synthetic miR-200a- and miR-control-transfected cells were plated onto poly-lysine-treated cover clips in six-well plates. Then, cells were treated with 5 μM cis-platin for 24 h. Apoptosis was detected using the in situ cell death detection kit with TMR red (Roche). Phalloidin-TRITC (10 μg/ml) (Sigma) was used to stain for actin, and DAPI was used to stain the nucleus. Confocal microscopy was then used to analyze the results.

Proliferation assays

Cells were plated in 96-well plates. Following a 12-h culture, drugs were added. Each condition was repeated in triplicate, and untreated cells served as controls. After treatment for 72 h, 10 μl per well of WST-8 (Dojindo) was added to analyze cell viability. After culturing for 3 h, the absorbance at 450 nm was recorded using a 96-well plate reader. The growth inhibition ratio was calculated as follows: (%) = (OD. control well – OD. treated well) / OD. control well *100%.

miRNA targets prediction

The following online software programs were used: TargetScan 5.2, and BioGRID 3.1 [11, 12]. The predicted targets of miR-200a, including the p53 family members as well as binding proteins that may regulate the p53 pathway were exported into Cytoscape and analyzed for evidence of intersection [13].

Statistical analysis

The results of at least 3 experiments are expressed as the mean ± SD. An ANOVA and the Student’s t test were used to compare values between the test and control samples. Fisher’s exact χ2 test was used for categorical patient variables. Statistical significance was represented as P values <0.05. The optimal cut-off value for miR-200a expression was calculated by a X-tile program [14]. All statistical calculations were performed using STATA software.

miR-200a plays a role in chemoresistance of breast cancer cells

In our previous studies, it has been shown that MDA-MB-231 and ZR-75-30 had low levels of miR-200a. To further examine the role of miR-200a in chemosensitivity, MDA-MB-231 and ZR-75-30 cells were transfected with miR-200a mimics, respectively. The transfected cells were treated with 5 μM cis-platin for apoptosis assays. First, the confocal TUNEL analysis showed that the miR-200a-transfected MDA-MB-231 cells had lower levels of apoptosis than the miR-Ctrl-transfected MDA-MB-231 cells (Fig. 1a, b, and p < 0.05). Second, Annexin V/PI staining showed that ZR-75-30-miR-200a mimic cells were resistant to cis-platin-induced apoptosis. In contrast, the ZR-75-30-miR-Ctrl cells were sensitive to cis-platin-induced apoptosis (Fig. 1c, d, and p = 0.009). Third, cleavage of Caspase-3 was suppressed in both miR-200a mimic transfected cells compared with the miR-Ctrl transfected cells (Fig. 1e). Together, these data suggest that miR-200a plays a role in chemoresistance.

miR-200a expression promotes chemoresistance in breast cancer cell lines

To examine the effects of miR-200a on chemosensitivity, we stably expressed miR-200a in the human breast cancer cell lines MDA-MB-231 and ZR-75-30. After treating with paclitaxel, cis-platin and gemcitabine, cells viability and Caspase 3/7 activity of stably miR-200a and scramble transfected cells were tested. In the cells viability assay, miR-200a overexpression significantly decreases the inhibitory effect of these chemotherapeutic agents. Furthermore, caspase3/7 activity assays showed that expression of miR-200a renders MDA-MB-231 and ZR-75-30 cells resistant to paclitaxel, cis-platin and gemcitabine induced apoptosis (Fig. 2a and b). On the other hand, scramble transfected cells were sensitive to apoptosis induced by these chemotherapy drugs (Fig. 2c and d). These results demonstrate a critical role for miR-200a in the chemoresistance of breast cancer cells.

Overexpression of miR-200a induced chemoresistance was mediated through TP53INP1 and YAP1

To elucidate the mechanism behind the chemoresistance induced by miR-200a, we used the BioGRID and TargetScan databases (http://www.targetscan.org/) to search for potential targets of miR-200a [12]. Cytoscape was used to find intersection between predicted miR-200a targets and p53 family binding proteins responsible for the regulation of p53, p63 and p73 (Fig. 3a, Additional file 1: Table S2) [13]. We found that TP53INP1 and YAP1 are involved in p73-mediated apoptosis pathway and have a high degree of complementarity with the seed region of miR-200a (Fig. 3b). To confirm whether miR-200a also directly regulates the expression of TP53INP1, the 3’ UTR region of the TP53INP1 was cloned downstream of the luciferase open reading frame to construct the reporter plasmid psiCHECK2–3’UTR-TP53INP1-luc (Fig. 3b). Transient transfection of 293 T cells with the reporters and a miR-200a mimic led to significantly decreased reporter activity compared to the transfection of control oligonucleotides. Importantly, the miR-200 mimic did not decrease the luciferase activity of a mutant construct that contained substitutions at three nucleotides within the miR-200a binding site (Fig. 3c). These results indicate that miR-200a downregulates TP53INP1 expression by directly targeting its 3’ UTR.

TP53INP1 and YAP1 are involved in the DNA damage-induced p73-mediated apoptosis. We further detected the expression proapoptotic proteins puma, Bax, bim and noxa in miR-200a transfected cell lines. Unexpectedly, in miR-200a transfected MDA-MB-231, the p73 pathway expression levels of YAP1 and TP53INP1 were low, which resulted in the transcriptional repression of the pro-apoptosis target genes puma, Bax, bim and noxa (Fig. 3d). No significant difference was detected in expression of p73 in mRNA level (Additional file 2: Fig. S1). These data suggest that miR-200a confer insensitivity to drug-induced apoptosis by antagonizing YAP1 and TP53INP1 expression.

In the next series of studies, we established the gemcitabine-resistant cell line (GR), MDA-MB-231 GR (Fig. 4a). qRT-PCR and western blot confirmed a significantly higher miR-200a expression in gemcitabine-resistant cell lines than in its parent cells (5.5-fold change of mRNA level in MDA-MB-231) (Fig. 4b). The treatment resistance of the MDA-MB-231 GR cell line was reversed by transfecting with the antagomiR-200a (Fig. 4c). Taken together, these data indicate that miR-200a is a determinant of chemosensitivity in breast cancer cells.

Overexpression of miR-200a was associated with poor response to preoperative chemotherapy and poor prognosis in patients with breast cancer

Significantly, the expression of miR-200a correlated inversely and significantly with the response to chemotherapy (Table 1; Fig. 5a; p = 0.0021). Furthermore, a survival analysis was performed to calculate the disease-free survival (DFS) for breast cancer patients (n = 110). The results showed that overexpression of miR-200a correlated significantly with shorter disease free survival duration of patients with breast cancer. (Fig. 5b). These results suggest the involvement of miR-200a overexpression in chemoresistance and poor prognosis.

Knockdown of TP53INP1 and YAP1 phenocopied effects of overexpression of miR-200a

To further identify the association between TP53INP1 and miR-200a in breast cancer, the expression of miR-200a and TP53INP1 were examined in various breast cancer cell lines (Fig. 6a and b). A reverse correlation was observed between miR-200a expression and TP53INP1 expression. The cells overexpressing miR-200a had a lower level of TP53INP1 expression (R = −0.780, p = 0.0279, n = 12) (Fig. 6c). This data further indicated the functional link between miR-200a and TP53INP1 in breast cancer. Moreover, silencing TP53INP1 and YAP1 can partly reproduce treatment resistance of miR-200a overexpression (Fig. 6d and e). These results implied that, enforced expression of miR-200a in breast cancer cell lines conferred chemoresistance via targeting YAP1 and TP53INP1.