The expression profile dataset GSE20437 was obtained from the GEO database (http://www.ncbi.nlm.nih.gov/geo/), which is a public and free database for gene expression data (32,33). GSE20437 includes 24 healthy and 18 breast cancer tissue samples (34). GEO2R (http://www.ncbi.nlm.nih.gov/geo/geo2r/) was used to identify differentially expressed genes (DEGs) between healthy and breast cancer tissue. P<0.05 and |log fold-change|>1 were considered the criteria to classify significant DEGs between healthy and breast cancer tissue samples.

A PPI network was constructed using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING; https://string-db.org/) and Cytoscape software 3.6.1 (www.cytoscape.org). Hub genes were identified from the PPI network.

To understand the biological significance of DEGs, GO enrichment (http://geneontology.org/) and KEGG analyses (https://www.genome.jp/kegg/) were conducted using Database For Annotation, Visualization And Integrated Discovery (DAVID; http://david.ncifcrf.gov) (35), which is a free analysis online tool that provides a convenient method for identifying the biological role of DEGs.

The OncoLnc (http://www.oncolnc.org) database was used to perform survival analysis based on DEGs (36). OncLnc contains clinical data of 8,647 patients from 21 studies on cancer and precomputed survival analyses for users to explore survival associations in cancer. The difference in the expression level of BTG2 between healthy and breast cancer tissues was further analyzed with the Oncomine database (www.oncomine.org) (37). BTG2 expression was assessed in breast cancer tissues relative to that in healthy tissues. UALCAN (http://ualcan.path.uab.edu), which is an open web-portal for cancer subgroup gene expression, was then used to perform survival analysis of BTG2 in different cancer subgroups based on race, menopause status and cancer type (38).

The luminal A breast cancer and paracarcinoma tissues (collected >5 mm from the tumor border) were collected from 8 patients with luminal A breast cancer at the affiliated Weihai Second Municipal Hospital of Qingdao University (Weihai, China) between July 2019 and November 2019. All patients were female and aged between 18 and 60 years with ER-positive, PR-positive (>20%), HER2-negative and Ki67 (<30%). Patients were excluded if they had received chemotherapy or radiotherapy prior to surgical resection. The patients signed an informed consent form prior to study commencement, and the study was approved by the Ethics committee of Clinical Trails of the affiliated Weihai Second Municipal Hospital of Qingdao University (Weihai, China; approval no. 2019-ER-04).

The MCF-7 breast cancer cell line representing luminal A cancer was purchased from the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences and cultured in DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) at 37˚C under 5% CO₂ (27,28). BTG2 overexpression (OE-BTG2) and empty (OE-NC) vectors were designed and synthesized by Wanleibio Co., Ltd. Prior to transfection, cells with 70-90% density were washed twice with serum-free Opti-MEM (Invitrogen; Thermo Fisher Scientific, Inc.). OE-BTG2 and OE-NC vectors (50 nM) were subsequently transfected into MCF-7 cells using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) at 37^˚C for 4 h according to the manufacturers protocol. The cells were cultured at 37˚C for 24 h and then collected for further study.

TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) was used to extract total RNA from MCF-7 cells or breast cancer/paracarcinoma tissues, according to the manufacturer's protocol. The extracted RNA was reverse transcribed into cDNA using a PrimeScript RT Reagent Kit with gDNA Eraser (cat. no. RR047A; Takara Biotechnology Co., Ltd.) according to the manufacturer's protocol. Subsequently, SYBR Green PCR Master Mix kit (Toyobo Life Science) was used for PCR-mediated amplification. Relative mRNA expression was calculated using the 2^-ΔΔCq method (39). The primer sequences used for qPCR were as follows: BTG2-forward (F), 5'-CATCATCAGCAGGGTGGC-3'; BTG2-reverse (R), 5'-CCAATGCGGTAGGACACC-3'; β-actin-F, 5'-CTTAGTTGCGTTACACCCTTTCTTG-3'; and β-actin-R, 5'-CTGTCACCTTCACCGTTCCAGTTT-3'. The reactions were performed using the following thermocycling conditions: Initial denaturation at 95˚C for 5 min, followed by 32 cycles of 95˚C for 30 sec, 56˚C for 40 sec and 72˚C for 40 sec. All quantifications were normalized to the internal reference gene β-actin.

Total protein was extracted from MCF-7 cells or breast cancer/paracarcinoma tissues using RIPA lysis buffer (Thermo Fisher Scientific, Inc.), and quantified using Pierce BCA Protein Assay Kit (cat. no. 23225; Thermo Fisher Scientific, Inc.). Total protein (40 µg per lane) was separated by 7.5-15% SDS-PAGE and transferred to a 0.45-µm PVDF membrane (ABclonal Biotech Co., Ltd.). The membrane was pre-blocked with 5% non-fat dry milk for 1 h at room temperature before incubation with the corresponding primary antibodies, including anti-BTG2 (cat. no. ab197362; 1:1,000; Abcam) and anti-β-actin (cat. no. ab8227; 1:2,000; Abcam) overnight at 4˚C. After washing the membranes three times with TBS-0.1%Tween-20, the membranes were incubated with HRP-conjugated goat anti-rabbit (cat. no. ab205718; 1:5,000; Abcam) at 37˚C for 1 h. Protein bands were visualized with an enhanced chemiluminescence reagent (Thermo Fisher Scientific, Inc.) using the ChemiDoc™ XRS+ imaging system (Image Lab 4.0; Bio-Rad Laboratories, Inc.).

Non-transfected, OE-NC or OE-BTG2 transfected MCF-7 cells (~8,000/well) were seeded in a 96-well plate in 100 µl DMEM supplemented with 10% FBS and incubated for 24, 48, 72 or 96 h at 37˚C in the presence of 5% CO₂. Subsequently, a total of 10 µl MTT (Sigma-Aldrich; Merck KGaA) was added into the 96-well plate, which was placed in an incubator at 37˚C in the presence of 5% CO₂ for 4 h. After removing the medium, 100 µl DMSO was added to each well to dissolve formazan crystals. Finally, the plate was placed in a microplate reader (BioTek Instruments, Inc.) for measurement of the absorbance at 570 nm.

Cell invasion was calculated based on the number of cells that passed through the polycarbonate membrane that separated the upper and lower chambers of an 8.0-µm Transwell chamber (Corning, Inc.). Briefly, Matrigel was thawed and diluted in serum-free DMEM (1:3) on ice. The Transwell chambers were placed in a 24-well plate, and 40 µl diluted Matrigel was added, followed by incubation at 37˚C for 2 h. Subsequently, 2x10⁵ MCF-7 cells overexpressing BTG2 or transfected with the pcDNA3.1 empty vector [overexpression (OE)-negative control (NC)] or non-transfected cells were suspended in 200 µl serum-free DMEM and added to the upper chamber, while 800 µl DMEM containing 10% FBS was added to the lower chamber. The 24-well plate was placed in an incubator at 37˚C in the presence of 5% CO₂ for 24 h. Subsequently, the upper chamber was removed while the lower chamber was washed with PBS three times. The cells in the lower chamber surface of the membrane were subsequently stained with 1% crystal violet at 25˚C for 10 min and the number of invading cells was counted under a light microscope (magnification, x200; Olympus Corporation).

A total of 5x10⁵ MCF-7 cells/well were seeded into a six-well plate. Subsequently, a 1-ml sterile pipette tip was used to create a linear scratch in the cell monolayer. Fresh serum-free DMEM was then added to each well, and the cells were incubated at 37˚C in a 5% CO₂ incubator for 24 h. A light microscope (magnification, x100) was used to observe the progressive change in the scratch width after 24 h. The migration distance was measured using ImageJ software 1.8.0 (National Institutes of Health).

Data are presented as the mean ± SEM. All experiments were duplicated and repeated at least three times. Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software, Inc.). Unpaired Student's t-tests were used for the comparison between two groups. One-way ANOVA with the post hoc Tukey's test was used for the comparison of the mean values between multiple groups. Multiple regression analysis was used for the survival analyses. P<0.05 was considered to indicate a statistically significant difference.

In the present study, 236 DEGs were identified in dataset GSE20437, which comprised epithelial samples from patients with breast cancer and patients that were cancer-free and receiving prophylactic mastectomy. The up- and downregulated genes are displayed in the volcano plot of Fig. 1. Results of the KEGG analysis demonstrated that the DEGs were enriched in ‘tight junction’, ‘DNA replication’, ‘base excision repair’, ‘pathways in cancer’ and ‘human T-cell leukemia virus type 1 infection’ (Fig. 2).

GO analysis demonstrated that the DEGs were enriched in the category biological process, including ‘response to mechanical stimulus’, ‘response to radiation’, ‘response to estradiol’, ‘ventricular cardiac muscle cell differentiation’ and ‘positive regulation of transcription from RNA polymerase II promoter’ (Fig. 3A). The results of GO analysis in the category cellular component demonstrated that the DEGs were enriched in the ‘integral component of plasma membrane’, ‘d DNA polymerase complex’, ‘nucleolus’, ‘nucleoplasm’ and ‘nucleus’ (Fig. 3B). Furthermore, results of the GO analysis in the category molecular function indicated that the DEGs were enriched in ‘DNA binding’, ‘sequence-specific DNA binding’, ‘transcription factor activity’, ‘transcriptional activator activity’ and ‘protein heterodimerization activity’ (Fig. 3C).

In total, 236 DEGs were uploaded to the STRING online database, and Cytoscape was subsequently used to identify the cluster. Among the DEGs, nuclear receptor subfamily 4 group A member (NR4A)1, immediate early response 2, dual-specificity phosphatase 1, activating transcription factor 3, NR4A2, protein FOSB, BTG2, proto-oncogene c-JUN and proto-oncogene c-FOS exhibited the closest association. The PPI network of these nine genes is presented in Fig. 4.

To further evaluate the prognostic value of the aforementioned hub genes, a survival analysis was conducted using the OncoLnc database. The results revealed that low expression of BTG2 was significantly associated with low survival rate of patients with breast cancer (Fig. 5A). In contrast, the low expression of other genes, including DUSP1, FOS, FOSB, JUN, MR4A1, MR4A2 and ATF3, was not associated with a low survival rate of patients with breast cancer (Fig. 5B-I). Furthermore, the Oncomine database revealed that the BTG2 expression was lower in breast cancer tissues containing luminal breast cancer compared with that in normal counterparts (Fig. 6). Further survival analysis based on UALCAN revealed that the expression level of BTG2 and menopause status may have an impact on the survival of patients with breast cancer that also exhibit mutations in breast cancer susceptibility protein (Fig. 7A and C). In contrast, the expression level of BTG2 and cancer type or race were not significantly associated with the survival of patients with breast cancer (Fig. 7B and D).

Western blotting and RT-qPCR were carried out to identify the expression levels of BTG2 in MCF-7 cells following transfection of BTG2. As presented in Fig. 8, the protein and mRNA expression level of BTG2 was low in MCF-7 cells of the OE-NC and control groups, while high BTG2 expression was detected in BTG2-overexpressing MCF-7 cells. Thus, the transfection of plasmids overexpressing BTG2 in MCF-7 cells was successful, and the transfected cells were used for further experiments.

Subsequently, the effect of BTG2 overexpression on MCF-7 cell proliferation was investigated. As demonstrated in Fig. 9, the proliferation of MCF-7 cells in the OE-BTG2 group was significantly lower than that of the OE-NC and control groups. In addition, there was no significant difference between the OE-NC and control groups. These results demonstrated that overexpression of BTG2 suppressed the proliferation of MCF-7 cells.

Crystal violet staining demonstrated that the number of MCF-7 cells that crossed the polycarbonate membrane of the Transwell invasion chamber in the OE-BTG2 group was significantly reduced, compared with the empty vector and blank control groups (Fig. 10A and B).

Furthermore, a scratch assay was used to further identify the effect of BTG2 overexpression on the migration of MCF-7 cells. Results displayed in Fig. 10C and D revealed that the wounded scratch area of MCF-7 cells in the OE-BTG2 group was markedly larger than that of the OE-NC and control groups after 24 h. These results suggested that the expression level of BTG2 may be associated with the inhibition of invasion and migration in MCF-7 cells.

Finally, the protein and mRNA expression of BTG2 in luminal A breast cancer tissue was investigated. As demonstrated in Fig. 11, compared with the paracarcinoma tissues of patients, the expression of BTG2 in luminal A breast tumor tissue was downregulated at the mRNA and protein level, which was consistent with the results in vitro. These results indicated that BTG2 may be a promising target and biomarker for luminal A breast cancer therapy in the future.