The paraffin-embedded specimens of 150 breast cancer tissues (33 tissues of TNBC cases and 117 of non-TNBC cases) and 30 paracancerous tissues were selected from the First Affiliated Hospital of China Medical University and the Affiliated Hospital of Binzhou Medical University. The 16 pairs of primary breast cancer and the corresponding adjacent non-tumor tissues were acquired from the First Affiliated Hospital of China Medical University. The collected fresh samples were immediately stored at −80°C for protein and RNA extraction. None of the patients had undergone radiotherapy and chemotherapy prior to the sample collection. Informed consent was obtained prior to surgery, from all enrolled patients. The study was approved by the Medical Ethics Committee of China Medical University and Binzhou Medical University.

Human breast cancer cell lines MCF-7 and MDA-MB-231, and normal human breast epithelial cell line MCF-10A were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). MCF-7 and MCF-10A cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco, Grand Island, NY, USA) and MDA-MB-231 cells were grown in Leibovitz's L-15 medium (Gibco). Both media were supplemented with 10% fetal bovine serum (FBS; Biological Industries Israel Beit Haemek Ltd., Israel) and 1% Pen-Strep solution in a 5% CO₂ humidified incubator at 37°C.

Total RNA was extracted from the breast tissues and cells using RNAiso Plus (Takara, Dalian, China) and reverse-trancribed to synthesize cDNA using the PrimeScript TMRT reagent kit (Takara). The expression of Livin was determined with SYBR^® Premix Ex Taq™ II (Takara) using the 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The quantification of Livin was normalized to GAPDH using the ΔΔCt method. The primers for Livin were 5′-GACAGAGGAGGAAGAGGAGGA-3′/5′-TCAGCGGCCAGTCATAGAAG-3′ and for GAPDH were 5′-GCACCGTCAAGGCTGAGAAC3′/5′-TGGTGAAGACGCCAGTGGA-3′.

Total proteins from breast tissues and cells were extracted in RIPA buffer and quantified using the Bradford method. Lysates containing 60 µg of total proteins were subjected to 10% SDS-PAGE and subsequently transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with 5% non-fat milk powder or 5% bovine serum albumin (BSA) for 2 h at room temperature. Subsequently, the blots were probed overnight at 4°C with indicated antibodies against Livin, Snail, Slug, MMP-2 and MMP-7 (all 1:500 diluted; Santa Cruz Biotechnology, Santa Cruz, CA, USA), E-cadherin (1:500 diluted; Wuhan Boster Biological Technology, Ltd., Wuhan, China), vimentin (1:500 diluted; Bioss, Beijing, China), N-cadherin (1:500 diluted; Bioss), p38, p-p38, GSK3β, p-GSK3β, p-ATF2 (all 1:1,000 diluted; Cell Signaling Technology, Beverly, MA, USA), GAPDH (1:2,000 diluted; ZSGB-Bio, Beijing, China) and incubated with appropriate secondary antibodies at room temperature for 2 h. Immunolabeled proteins were detected with ECL (Thermo Fisher Scientific, Waltham, MA, USA). GAPDH was used as an internal control. To detect the expression of the relevant protein, the gray level of the indicated protein band was detected using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

All specimens were made into paraffin sections with 4 µm thickness. Immunohistochemistry staining was performed according to the Envision method (ZSGB-Bio, Beijing, China), following the manufacturer's protocol. The primary antibodies were rabbit polyclonal antibody against Livin (1:200), rabbit polyclonal antibody against E-cadherin, vimentin and N-cadherin (1:300, respectively). For the negative controls, the primary antibodies were replaced by phosphate-buffered saline (PBS). Previously identified strongly staining breast tissue sections were used as positive controls. By multiplying the staining intensity and the percentage of positive cells, we evaluated the expression of Livin. Livin was mainly located in the cytoplasm. The intensity of staining was graded as 0–3 (0, none; 1, weak; 2, moderately strong; 3, intense) and positive cell proportion was scored as 0–4 (none, 0; 1–25%, 1; 26–50%, 2; 51–75%, 3; 76–100%, 4). By multiplying these two factors, positive expression was indicated when the immunoreactive score was ≥2. E-cadherin mainly localized in the cytomembrane and ≥50% cells with deep yellow or brown granules was scored as positive expression. N-cadherin exhibited in the cytomembrane and (or) cytoplasm and ≥10% cells with deep yellow or brown granules was scored as positive expression. Vimentin mainly existing in the cytoplasm and ≥10% cells with deep yellow or brown granules was scored as positive expression (14,15).

The pCMV6-myc-Livin plasmid was purchased from Sino Biological Technology (Beijing, China), and the pCMV6-myc empty vector was purchased from Origene Technology (Rockville, MD, USA). Livin-siRNA and NC-siRNA were purchased from Santa Cruz Biotechnology. Transfection was carried out using the Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions.

Cells were fixed with 4% paraformaldehyde and blocked with 5% BSA. The primary antibodies against E-cadherin, vimentin and N-cadherin (1:100, respectively) were incubated overnight at 4°C. The following day, the cells were incubated with FITC-conjugated secondary antibodies in the dark at room temperature for 2 h. The coverslips were incubated with DAPI for nuclear counterstaining. The results were observed using a fluorescence microscope.

When cell confluency reached ~90% after transfection, wounds were created in the confluent cells using a 200-µl pipette tip. The cells were rinsed with PBS to remove any free-floating cells and debris. Medium was subsequently added and culture plates were incubated at 37°C. Wound healing within the scrape lines was observed at different time-points and representative scrape lines for each cell line were photographed.

Cells (3×10⁴) were plated in the upper Transwell chamber (Corning, Lowell, MA, USA) in 100 µl medium with 2% fetal bovine serum (FBS) after they were cultured to the exponential phase. The lower chamber was filled with 600 µl medium containing 20% FBS as a chemoattractant. After a 24-h incubation, the migrated cells were fixed with cold methanol and stained with hematoxylin. The stained migrated cells were scored and photographed under microscopic observation.

An invasion assay was performed using the Transwell chamber precoated with diluted Matrigel (BD Biosciences, San Jose, CA, USA). Cells (1×10⁵) in 100 µl medium containing 2% FBS were seeded into the upper chamber, whereas the lower chamber was filled with 600 µl medium containing 20% FBS. The remaining experimental procedures were in accordance with the cell migration assay.

All statistical analyses were performed using the SPSS 16.0 statistical software (SPSS, Chicago, IL, USA). The immunohistochemistry results were analyzed using the Chi-square test and the Pearson's correlation test. Differences between groups were assessed by the Student's t-test. Data were processed using GraphPad Prism 5.0 software (GraphPad Software, Inc., La Jolla, CA, USA). P<0.05 was considered to indicate a statistically significant difference.

The expression of Livin at transcriptional and translational levels was assessed in fresh breast tissues by quantitative real-time RT-PCR (qRT-PCR) and western blot analysis. As shown in Fig. 1A and B, Livin mRNA and protein expression were markedly upregulated in breast cancer tissues compared with that in paired non-tumor tissues.

To ascertain this result, an immunohistochemistry assay was employed to detect the expression of Livin in paraffin-embedded specimens. The results revealed that Livin was mainly located in the cytoplasm of cancer cells in breast cancer tissues (Fig. 2). The rate of positive expression of Livin in breast cancer tissues (59.3%) was significantly higher than that in adjacent non-tumor tissues (23.3%, P<0.001).

We detected the expression of Livin in the breast cancer cell lines MCF-7 and MDA-MB-231 and in the normal breast epithelial cell line MCF-10A using western blot analysis and qRT-PCR (Fig. 3). The levels of Livin expression were significantly elevated in the MDA-MB-231 and MCF-7 cells compared to those in the MCF-10A cells.

We analyzed the correlation between Livin expression and the clinicopathological features of the breast cancer patients (Table I). Livin expression was significantly positively associated with TNM stage and lymph node metastasis in the breast cancer cases, especially in TNBC (P<0.01). In contrast, there was no marked correlation between Livin expression and patient age, tumor size, or histological differentiation (P>0.05). Furthermore, the rate of Livin expression in TNBC tissues was slightly higher than that in non-TNBC tissues. However, this difference was not statistically significant. In addition, Livin expression was higher in TNBC MDA-MB-231 cells than that in non-TNBC MCF-7 cells. These results suggested that Livin expression is positively associated with the progression and metastasis of breast cancer and that Livin plays a more important role in TNBC.

To explore whether Livin affects the migration and invasion of breast cancer cells, we upregulated Livin expression by transfection of Livin in the non-TNBC cell line MCF-7, which has low Livin expression, and knocked down Livin expression by siRNA treatment in the TNBC cell line MDA-MB-231, which exhibits high Livin expression. Transfection efficiencies were assessed by western blot analysis and qRT-PCR, respectively (Fig. 4).

Wound healing, cell migration and Matrigel invasion assays were used to investigate the effect of Livin on cell migration and invasion abilities. The upregulation of Livin led to a significant increase in the migration and invasion of the MCF-7 cells. Conversely, the migratory and invasive abilities of the MDA-MB-231 cells were suppressed by the Livin-knockdown (Fig. 5).

EMT is highly correlated with tumor invasion and metastasis (7,8). We therefore observed the influence of Livin on EMT in breast cancer cells. We examined the levels of several EMT-related factors using western blot analysis, after altering the expression of Livin. Livin overexpression caused a decrease in the expression of the epithelial marker E-cadherin and increases in the mesenchymal markers vimentin and N-cadherin, as well as the EMT transcription factor Snail and the metastasis-associated factors MMP-2 and MMP-7. Depletion of Livin resulted in an increase in E-cadherin and decreases in vimentin and N-cadherin as well as Snail, MMP-2 and MMP-7 (Fig. 6A). We further detected the expression of the EMT marker proteins by immunofluorescence assay and the results were consistent with those of the western blot analysis (Fig. 6B).

In addition, to determine whether Livin is associated with the EMT markers, we further performed immunohistochemical analysis of the samples from 33 cases of TNBC and 117 cases of non-TNBC (Table II). Correlation analysis revealed that Livin expression was positively correlated with the expression of vimentin and N-cadherin and negatively correlated with that of E-cadherin in TNBC. In non-TNBC, enhanced expression of Livin was significantly associated with enhanced expression of vimentin and reduced expression of E-cadherin, while Livin expression was not significantly associated with the expression of N-cadherin. These findings suggested that Livin induced EMT to promote cell migration and invasion in breast cancer cells.

Recent studies have revealed that the p38 and GSK3β pathways regulate EMT to promote progression and metastasis in several types of cancer (16,17), activating p38 and inhibiting GSK3β to promote EMT. To explore the possible mechanism of Livin-induced EMT in breast cancer cells, we examined the expression of p38/GSK3β-associated factors. The overexpression of Livin in MCF-7 cells markedly enhanced the phosphorylation of p38, ATF2 and GSK3β. In contrast, the knockdown of Livin in MDA-MB-231 cells clearly inhibited the phosphorylation of p38, ATF2 and GSK3β (Fig. 7A).

Moreover, we ascertained the above results using a specific inhibitor of p38, SB203580 (Cell Signaling Technology) and an effective inhibitor of GSK3β, SB216763 (Cell Signaling Technology). SB203580 is a specific inhibitor of p38 that inhibits p38 catalytic activity by binding to the ATP binding pocket, but it does not inhibit phosphorylation of p38. SB216763 is an effective inhibitor of the GSK3β activity. Inactivation of GSK3β results in an increase in phosphorylated GSK3β. GSK3β is a downstream target of p38 and inhibiting p38 downregulates phosphorylated GSK3β expression and enhances GSK3β activity. The increased expression of p-ATF2, p-GSK3β, vimentin and N-cadherin was eliminated and the expression of E-cadherin was partially restored by SB203580 (Fig. 7B). When SB216763 was added to the medium after Livin was upregulated, the expression of p-GSK3β, vimentin and N-cadherin was further increased and E-cadherin expression was markedly reduced (Fig. 7C). These results indicated that Livin may activate the p38/GSK3β pathway to promote EMT in breast cancer cells.