The human TNBC MDA-MB-231 cell line, human epithelial breast cancer BT-474 cell line and human embryonic kidney 293T and 293A cell lines were purchased from the American Type Culture Collection (Manassas, VA, USA), and were cultured in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% fetal bovine serum (FBS; both Thermo Fisher Scientific, Inc., Waltham, MA, USA) at 37°C in a humidified 5% CO₂ atmosphere.

Three G(N19)NGG sequences downstream from the BECN1 TSS were selected as potential targeting sites. N indicates any nucleotide (A, T, G or C). The three targeting sites were as follows: B432, GTCCAACAACAGCACCATGCAGG; B497, GGACACGAGTTTCAAGATCCTGG; and B525, GTCACCATCCAGGAACTCACAGG. To construct B432, B497 and B525 sgRNA lentiviral plasmids, respective primer pairs were used as a template for CRIF and CRIR PCR amplification in separate reactions (Table I). The PCR product was inserted into the pLKO.1 plasmid (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) using Gibson master mix (New England Biolabs, Inc., Ipswich, MA, USA). The flag-tagged Cas9 gene was cloned into the pShuttle-CMV plasmid (a gift from Dr Pu-Min Zhang) at the BglII and XhoI restriction sites. Following linearizing with PmeI, pShuttle-CMV-Cas9 was co-transformed with the pAdEasy-1 plasmid (a gift from Dr Pu-Min Zhang) into BJ5183 cells (Stratagene; Agilent Technologies, Inc., Santa Clara, CA, USA), according to the manufacturer's protocol. Following incubation overnight at 37°C, positive clones selected on LB plates containing 50 µg/ml kanamycin (Sigma-Aldrich; Merck KGaA) were identified by digesting with PacI.

The UCA™ CRISPR efficiency evaluation kit was purchased from Beijing Biocytogen Co., Ltd. (Beijing, China). Briefly, the targeting sequence of BECN1 amplified with the primer set B371_F and B793_R (Table I) was cloned into a precut pUCA plasmid (UCA-BECN1; Beijing Biocytogen Co., Ltd.). The UCA-BECN1, sgRNA and Flag-Cas9 plasmids were co-transfected into 293T cells using Lipofectamine™ 3000 (Thermo Fisher Scientific, Inc.). Approximately 48 h after transfection, cells were lysed and luciferase activity was measured using the dual-luciferase reporter assay (Promega Corporation, Madison, WI, USA) with a Lumat LB 9501 Luminometer (Berthold Technologies GmbH & Co., Bad Wildbad, Germany). CRISPR efficiency was calculated as the ratio of firefly luciferase to Renilla luciferase activity.

To package the sgRNA lentivirus, the sgRNA plasmid and packaging plasmids, pMD and SPA were co-transfected into 293T cells as described earlier. The supernatant was harvested 48 h after transfection and used to infect MDA-MB-231 cells. The recombinant Ad plasmid was linearized with PacI and transfected into 293A cells using Lipofectamine™ 3000 (Thermo Fisher Scientific, Inc.) to produce Cas9 adenovirus. Infected MDA-MB-231 cells were screened for sgRNA-positive cells by adding 2 µg/ml puromycin (Sigma-Aldrich; Merck KGaA) for one week. Puromycin-resistant cells were further infected with Cas9 adenovirus for 12 h. After 4 days of culture, the cells were well prepared for the subsequent experiments.

The genomic DNA of MDA-MB-231 cells was extracted using a Qiagen Genomic kit (Qiagen GmbH, Hilden, Germany). Next, the genomic region surrounding the CRISPR/Cas9 target site was amplified by PCR using the primer set, B371_F and B793_R (Table I). PCR was performed using GoTaq^® Green Master mix (Promega Corporation) in an ABI 2720 Thermal cycler (Applied Biosystems; Thermo Fisher Scientific, Inc.) as follows: 95°C for 2 min; 33 cycles at 95°C for 30 sec, 60°C for 30 sec and 72°C for 30 sec; and 72°C for 5 min. The 423-bp PCR product was verified using 1.5% agarose gel electrophoresis and then purified using the QIAquick^® gel extraction kit (Qiagen GmbH).

T7EI was purchased from NEB. Prior to T7EI digestion, the PCR product amplified using the primer set, B371_F and B793_R (Table I), was denatured and annealed. In brief, the PCR product was incubated in 94°C water and cooled to room temperature naturally. The PCR product was then digested with T7EI at 37°C for 2 h and resolved with 1.5% agarose gel electrophoresis.

The genomic DNA of a single cell colony was extracted using a Qiagen Genomic kit (Qiagen GmbH) and amplified using the primer set, B371_F and B793_R (Table I). Following purification, the PCR product was ligated into the pGEM-T Easy vector (Promega Corporation). The ligation product was used to transform competent E. coli DH5α (Thermo Fisher Scientific, Inc.), and positive clones selected on LB plates containing 100 µg/ml ampicillin (Sigma-Aldrich; Merck KGaA) were sequenced using the universal T7 primer.

Cells were lysed using radioimmunoprecipitation assay buffer (Applygen Technologies Inc., Beijing, China) supplemented with protease and phosphatase inhibitors (Roche Diagnostics GmbH, Mannheim, Germany). The concentration of the protein was determined by Pierce™ bicinchoninic acid protein assay kit (Thermo Fisher Scientific, Inc.). Protein lysate (20 µg per lane) was separated by 10% SDS-PAGE gel and transferred onto nitrocellulose membranes (EMD Millipore, Billerica, MA, USA). The membranes were blocked with 5% skimmed milk (BD Biosciences, San Jose, CA, USA) in TBST for 2 h at room temperature and incubated overnight at 4°C with the following primary antibodies: BECN1 (dilution, 1:2,000; cat. no. sc-11427; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), GAPDH (dilution, 1:2,000; cat. no. sc-365062; Santa Cruz Biotechnology, Inc.), N-cadherin (dilution, 1:1,000; cat. no. 13116; Cell Signaling Technology, Inc., Danvers, MA, USA), Snail (dilution, 1:1,000; cat. no. 3879; Cell Signaling Technology, Inc.), Claudin-1 (dilution, 1:1,000; cat. no. 13255; Cell Signaling Technology, Inc.), E-cadherin (dilution, 1:1,000; cat. no. 3195; Cell Signaling Technology, Inc.) and Vimentin (dilution, 1:8,000; cat. no. ab92547; Abcam, Cambridge, UK). Next, the membranes were incubated with horseradish peroxidase-conjugated anti-rabbit or anti-mouse secondary antibodies (dilution, 1:5,000; cat. no. 7074 and 7076; Cell Signaling Technology, Inc.) for 1 h at room temperature. Bands were visualized using an enhanced chemiluminescence reagent (EMD Millipore).

Total RNA was isolated using TRIzol reagent (Thermo Fisher Scientific, Inc.). Next, reverse transcription of 2 µg total RNA was performed using Superscript™ III transcriptase (Thermo Fisher Scientific, Inc.). Regarding qPCR, cDNA was amplified using THUNDERBIRD SYBR qPCR mix (Toyobo Life Science, Osaka, Japan) in a Chromo4™ system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). PCR started with a 10-min denaturation step at 95°C, followed by 40 step cycles at 95°C for 15 sec and 60°C for 60 sec. The expression of N-cadherin, Vimentin, Claudin-1, Snail and E-cadherin was analyzed using the primers shown in Table I. GAPDH was used as an internal control for normalization. The 2^−∆∆Cq method was used to calculate the relative expression levels (19).

For the CCK-8 assay, MDA-MB-231 cells (1,000 cells/well) were seeded into 96-well plates and incubated for 7 consecutive days. CCK-8 assays were performed in quintuplicate and proliferation of cells was measured by adding 10 µl CCK-8 reagent (Dojindo Molecular Technologies, Inc., Kumamoto, Japan) to each well at different time-points. Plates were incubated at 37°C for 2.5 h and absorbance was evaluated using a microplate reader (Bio-Rad Laboratories, Inc.) at 450 nm.

For the colony formation assay, MDA-MB-231 cells (1,000 cells/well) were plated into 6-well plates. Cells were cultured for up to 14 days and fresh DMEM supplemented with 10% FBS (both Thermo Fisher Scientific, Inc.) was changed every 3 days. Cells were fixed in paraformaldehyde for 10 min at room temperature and stained with crystal violet for 15 min at room temperature. Colonies with >50 cells were counted.

For the apoptosis assay, cells were stained using the PE Annexin V Apoptosis Detection kit I (BD Biosciences), according to the manufacturer's protocol. In brief, cells were digested with trypsin without EDTA, washed two times with PBS and resuspended in 100 µl binding buffer, followed by simultaneous staining with 5 µl PE Annexin V and 5 µl 7-AAD for 15 min at room temperature in the dark, and then analyzed using a flow cytometer (BD Biosciences).

For cell cycle analysis, cells were harvested and fixed in cold 70% alcohol at 4°C overnight. Cells were collected by centrifugation (800 × g for 5 min at room temperature), washed twice in PBS and incubated with FxCycle™ PI/RNase Staining Solution (Thermo Fisher Scientific, Inc.) for 30 min at room temperature in the dark. Cell cycle distribution was then evaluated using a flow cytometer and CellQuest™ software (version 6.0; BD Biosciences).

For the wound healing assay, cells were seeded into 6-well plates and an artificial wound was made using a sterile 200 µl pipette tip. Following rinsing with PBS, cells were incubated in serum-free DMEM (Thermo Fisher Scientific, Inc.). Images of the scratch gap were captured at 0 and 24 h under a light microscope (magnification, ×100; Eclipse Ti-U; Nikon Corporation, Tokyo, Japan). The percentage of wound healing was evaluated as follows: [1− (width of the wound at 24 h/width of the wound at 0 h)] ×100.

Transwell chambers (8 µm-pore size; Corning Incorporated, Corning, NY, USA) coated with or without diluted Matrigel (BD Biosciences) on the upper chamber were used for invasion or migration assays. Briefly, 5×10⁴ MDA-MB-231 cells in serum-free DMEM (Thermo Fisher Scientific, Inc.) were seeded into the upper chamber and medium supplemented with 10% FBS was added to the lower chamber. Following culture for 8 or 24 h, cells remaining on the upper side of the filter membrane were gently wiped off using a cotton swab, while cells that migrated or invaded the lower chamber were fixed with paraformaldehyde for 10 min at room temperature and stained with crystal violet for 15 min at room temperature. Cells were counted in five different randomly selected fields under a light microscope (magnification, ×100; Nikon Corporation).

Gene expression data (accession nos., GSE81838 and GSE65194) were retrieved from the NCBI Gene Expression Omnibus (GEO) database. Details of patients and analysis of microarray data were as previously described (20,21). GSE81838 consisted of 10 paired TNBC tissues and adjacent normal stromal tissues, and GSE65194 contained 41 TNBC samples and 11 normal breast tissue samples. Expression data from GSE81838 were normalized using the Robust MultiChip Averaging (RMA) algorithm implemented in the Bioconductor package Affy and log2-transformed. Data from GSE65194 were analyzed using standard AffyCDF or Brainarray HGU133Plus2_Hs_ENTREZG version 13 custom chipset definition file and normalized using GeneChip RMA (GC-RMA).

SPSS version 19.0 software (IBM Corp., Armonk, NY, USA) was used for statistical analyses. Data are presented as the mean ± standard deviation (SD) and error bars represent SD values from three independent experiments, unless otherwise indicated. Multiple-group comparisons were performed using one-way analysis of variance, followed by the Bonferroni's post hoc test. Comparisons between two groups were calculated using the two-sample t-test. P<0.05 was considered to indicate a statistically significant difference.

Becn1 bi-allelic knockout mice die early in embryogenesis (9), which hampers further investigation of the functions of BECN1 in autophagy and tumorigenesis. In order to determine if cells survive following BECN1-knockout, the CRISPR/Cas9 technique was used to establish a BECN1-knockout cell line. Using the free online bioinformatical sgRNA design tool (http://crispr.mit.edu/) developed by Dr Feng Zhang, the genomic and cDNA sequences downstream of the translation start site (TSS) were analyzed. As demonstrated in Fig. 1A, three sequences with the highest scores were selected to construct sgRNA lentivirus vectors. In order to assess vector efficiency, a luciferase reporter assay was performed (Fig. 1B). Luciferase assay data indicated that sgRNA#1 had the highest efficiency, which was also confirmed by western blot analysis (Fig. 1C).

Based on the aforementioned results, sgRNA#1 was selected to package the lentivirus that was used to infect MDA-MB-231 cells. Puromycin-resistant cells were further infected with Cas9 adenovirus. Following serial dilution and culturing for up to two weeks, >20 single cell colonies were subjected to T7 Endonuclease I (T7EI) digestion. Five clones were subjected to a second round of T7EI digestion (Fig. 2A). Clones A, B and D (renamed KO#1, KO#2 and KO#3, respectively) were selected for western blotting and sequencing verification. Western blotting verified that BECN1 was not expressed in these three clones (Fig. 2B). Sequencing indicated that the loss of BECN1 expression was due to multiple InDels occurring at the BECN1 locus, all of which led to frame shifts and subsequent gene knockout (Fig. 2C). Taken together, the results indicated that stable cell lines that could survive following BECN1-knockout were successfully constructed.

Since MDA-MB-231 cells with BECN1-knockout were able to survive, the effect of BECN1-knockout on cell growth was analyzed by comparing the viability of BECN1-knockout and wild-type (WT) MDA-MB-231 cells between 24 and 168 h after inoculation. As demonstrated in Fig. 3A, no significant difference in the viability of BECN1-knockout MDA-MB-231 cells was observed when compared with WT cells at 24 h. However, when compared with WT MDA-MB-231 cells, BECN1 KO#1 cells exhibited markedly decreased proliferation after 72 h, and the proliferation of BECN1 KO#2 and KO#3 cells was significantly inhibited after 48 h in culture (P<0.001). The colony formation assay is an in vitro approach to evaluate the ability of a single cell to grow into a colony and essentially tests every cell within a population for its ability to undergo 'unlimited ' division (22). It was also determined that BECN1-knockout inhibited the colony formation capacity of MDA-MB-231 cells (P<0.001; Fig. 3B). Therefore, BECN1-knockout MDA-MB-231 cells retained growth capability but exhibited an attenuated proliferation rate.

Next, flow cytometry was performed to investigate whether the anti-proliferative effect of BECN1-knockout was due to apoptosis or cell cycle arrest. There was no significant difference in apoptosis in the BECN1-knockout MDA-MB-231 cells compared with WT (Fig. 4A and C). However, all three BECN1-knockout cell lines accumulated at the G₀/G₁ phase with decreased cell proportions in the S and G₂/M phases compared with WT (Fig. 4B and D). The percentage of G₀/G₁ cells significantly increased from 31.7±1.6% (WT) to 42.5±1.3% (KO#1; P<0.001), 44.1±1.2% (KO#2; P<0.001) and 39.9±1.3% (KO#3; P<0.01). These results suggested that BECN1-mediated autophagy or other pathways serve an important role in cancer cell proliferation via regulation of cell cycle progression.

Subsequently, whether BECN1 affects the tumor metastasis characteristics of migratory and invasive ability in MDA-MB-231 cells was investigated. The wound healing assay revealed that MDA-MB-231 cells with BECN1-knockout exhibited a notably slower scratch closure rate than WT cells (Fig. 5A), which suggested mobility inhibition. The Transwell migration assay further confirmed that BECN1-knockout markedly reduced the migratory ability of MDA-MB-231 cells (Fig. 5B). Furthermore, the invasive potential of BECN1-knockout MDA-MB-231 cells was significantly decreased (Fig. 5C), which was confirmed by the Transwell invasion assay. In brief, the results of the present study demonstrated that BECN1-knockout markedly decreased the migratory and invasive potential of MDA-MB-231 cells in vitro.

To test whether the effect of BECN1-knockout on migration and invasion is associated with EMT, the present study assessed the cell morphology under a phase contrast microscope. As demonstrated in Fig. 6A, BECN1-knockout changed the mesenchymal morphology of MDA-MB-231 cells into a partial epithelial-like morphology with decreased cell scattering and increased cell-cell adhesion. Consistent with the morphology changes, expression of mesenchymal markers, including N-cadherin, Vimentin and Snail decreased, while expression of the epithelial marker Claudin-1, but not E-cadherin, increased following BECN1-knockout (Fig. 6B–D), suggesting a partial reversal of EMT. Taken together, the results of the present study demonstrated that BECN1 regulates breast cancer cell migration and invasion via partial modulation of the EMT signaling pathway.

To determine whether the expression of BECN1 is altered in TNBC tissues, a GEO dataset (GSE81838) containing ten patients with TNBC was analyzed (20). The results demonstrated that BECN1 was significantly upregulated in half of the TNBC tissues compared with adjacent normal stromal tissues (Fig. 7A; P<0.05). Meanwhile, another GEO dataset (GSE65194) containing 41 TNBC samples and 11 normal breast tissue samples was analyzed (21). BECN1 expression was significantly increased in tumor tissues compared with healthy breast tissues (Fig. 7B; P<0.001). The results suggested that BECN1 may serve as an oncogene in TNBC.