GA-A (96%) was purchased from Chengdu Must Bio-technology Co., Ltd., (Chengdu, China), dissolved in dimethylsulfoxide (DMSO) and maintained at 4°C. AG490, MTT, and dichloro-dihydro-fluorescein diacetate (DCFH-DA) were obtained from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). The propidium iodide (PI) staining kit and the JC-1 fluorescent dye were purchased from Beyotime Institute of Biotechnology (Nanjing, China).

MDA-MB-231 breast cancer cells were obtained from the American Type Culture Collection (Manassas, VA, USA) and maintained with Leibovitz L-15 (L-15) medium (pH=7.3; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 100 mg/ml streptomycin, 100 IU/ml penicillin and 10% fetal bovine serum (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). Cells were maintained at 37°C and 5% CO₂, and the medium was refreshed every 3 days. The GA-A was diluted in L-15 immediately prior to experiments. The control group was treated with L-15 (0.01%) alone. Cells were first treated with GA-A (0.1, 0.2 and 0.4 mmol/l), then with 0.4 mmol/l GA-A and 20 µM AG490 at 37°C for 24 h. Cells were harvested at 24 or 48 h.

Cell viability was investigated using an MTT assay. Cells (5×10³) were treated with GA-A (0.01, 0.02, 0.05, 0.1, 0.2, 0.4, 0.6 and 0.8 mmol/l) for exactly 24 h then 20 ml MTT at 5 mg/ml for 4 h at 37°C. The supernatant was then removed and 150 ml DMSO was added. After 15 min, the 490 nm optical density value was measured. The results were presented as the ratio between the control group and research group.

A Transwell chamber (8 µm pore polycarbonate; Corning Incorporated, Corning, NY, USA) with Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) was used in this experiment. Cells were seeded (1×10⁶) into the upper chambers of 24-transwell chamber wells and treated with GA-A (0.1, 0.2 and 0.4 mmol/l) at 37°C for 24 h. For the invasion assay, the cells were plated in the upper chamber in the serum-free L-15 medium. The medium containing 20% of FBS in the lower chamber served as a chemoattractant. Following incubation for 24 h, cells in the upper chambers were removed with a cotton swab. Subsequently, cells were fixed with 100% methanol and then stained with 1% crystal violet in 2% ethanol for 10 min at room temperature. Under a Nikon light microscope (×100), migrated cells were counted and images were captured at five different fields of the chamber.

Cells were seeded into 6-well plates (5×10⁵ cell/well) and incubated for 24 h prior to treatment. GA-A (0, 0.1, 0.2 and 0.4 mmol/l, respectively) or DMSO were added to the plate and incubated at 37°C for 1 day. Subsequently, adherent cells and non-adherent cells were digested in trypsin at 37°C for 2 min. The cell suspension was incubated with 10 ml PI and 5 ml annexin V-fluorescein isothiocyanate (from the kit aforementioned) for 10 min in the dark at 37°C and then analyzed using a flow cytometer (BD Canto II) and Diva Software v7.0 (BD Biosciences). The apoptotic index in Fig. 3 was calculated as: Apoptotic cells/total cells ×100%.

Intracellular ROS production was investigated using DCFH-DA. Cells were seeded at a density of 1×10⁶ cells/well and cultured for 24 h. Cells were then harvested in trypsin at 37°C for 2 min, washed with PBS for 5 min and re-suspended in 500 µl PBS containing 20 µM DCFH-DA then incubated at 37°C for 20 min in the dark. The cells were harvested and analyzed by flow cytometry (BD Canto II) and Diva Software v7.0.

The treated cells (1×10⁷ cells/6 ml L-15 with 10% FBS in a 90-mm dish) were collected and washed twice using cold PBS for 5 min. Cells were lysed in 200 µl lysis radioimmunoprecipitation assay buffer (Beyotime Institute of Biotechnology). The lysate was incubated on ice for 30 min, vortexed and centrifuged at 14,000 × g for 15 min at 4°C. The supernatant was collected and protein concentration was determined using a Bradford Assay. Following the addition of SDS-PAGE sample loading buffer (Beyotime Institute of Biotechnology), the protein samples (30 µg) underwent electrophoresis using a 10% SDS-PAGE and were then transferred to a polyvinylidene fluoride membrane (EMD Millipore, Billerica, MA, USA). After blocking for 4 h at room temperature in a solution of 5% non-fat dry milk in Tris-buffered saline containing 0.1% Tween-20, the membranes were incubated overnight at 4°C with the primary antibodies against phosphorylated (p)-JAK2 (catalog no. 3771), JAK2 (catalog no. 3230), p-STAT3 (catalog no. 9145), STAT3 (catalog no. 9139), Bcl-xL (catalog no. sc-8392), Bak (catalog no. sc-517390), Mcl-1 (catalog no. sc-12756), Bax (catalog no. sc-7480), Cytochrome c (catalog no. sc-13156) and β-actin (catalog no. sc-47778) at a concentration of 1:1,000 in Tris-buffered saline with 0.1% Tween-20 containing 5% non-fat dry milk. JAK2, p-JAK2, p-STAT3, STAT3 primary antibodies were obtained from Cell Signaling Technology, Inc. (Danvers, MA, USA). Bcl-xL, Bak, Mcl-1, Bax, Cytochrome c and β-actin primary antibodies were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). Following washing four times for 5 min, the membranes were incubated with a horseradish peroxidase-conjugated secondary anti-rabbit IgG antibody (1:5,000 dilution; catalog no. 7074; Cell Signaling Technology, Inc.) at room temperature for 1 h and were then washed six times for 10 min. Signals were detected with an enhanced chemiluminescence detection kit (Applygen Technologies, Inc., Beijing, China).

Statistical analysis was performed using SPSS 14.0 software (SPSS Inc., Chicago, IL, USA). Values presented as the mean ± standard deviation and were analyzed using one-way analysis of variance, followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

MDA-MB-231 cells were stimulated by GA-A (0.1–0.8 mmol/l) to determine its effect on cell viability. GA-A significantly decreased cell viability compared with the control group, and the effects were not only dose dependent but also time dependent (P<0.05; Fig. 2A). The half-maximal inhibitory concentrations of GA-A at 24 and 48 h were 0.707 and 0.163 mmol/l, respectively.

Based on the results of the Transwell invasion assay, GA-A demonstrated an inhibitory effect on the invasion of MDA-MB-231 cells. Fig. 2B reveals that following incubation for 24 h, GA-A (0.1, 0.2 and 0.4 mmol/l) markedly decreased the number of invasive cells. As GA-A concentration increased, the MDA-MB-231 cells became less invasive. These results were consistent with the Transwell invasion assay, which indicated that GA-A suppressed the invasion of MDA-MB-231 cells.

MDA-MB-231 cells were untreated or treated with GA-A (0.1, 0.2 and 0.4 mmol/l) for 24 h. Following treatment with 0.1, 0.2 and 0.4 mmol/l of GA-A for 24 h, the apoptotic index increased by 11.34±3.41, 20.89±3.32 and 38.13±3.91%, respectively (P<0.01, vs. control). The effect of GA-A on apoptosis was dose-dependent. These results indicate that GA-A induces apoptosis of MDA-MB-231 cells (Fig. 3B).

As presented in Fig. 4, ROS level markedly increased following GA-A treatment. ROS production following 0.4 mmol/l GA-A increased markedly from 27.22% in the control group to 76.35%.

Western blot analysis was applied to determine cyclin D1, p21 and p27 protein expression to further investigate the characteristics of the observed G0-G1 phase control. It was revealed that cyclin D1 expression significantly decreased compared with the control, whereas p21 and p27 expression significantly increased (P<0.05; Fig. 5), which indicates that distinctions exist between cell-cycle-associated protein, at least partially, leading to G0-G1 phase arrest by GA-A.

Western blot analysis was adopted to assess the phosphorylated forms of JAK2 and STAT3 in breast cancer cells stimulated by GA-A, from which the phosphorylation of these factors was revealed to significantly decrease following GA-A treatment (with the exception of p-JAK2 at 0.1 mmol/l GA-A; P<0.05). In addition, the expression levels of Mcl-1 and Bcl-xL were significantly downregulated by GA-A dose-dependently compared with the control, in accordance with decreased phosphorylated JAK2 and STAT3 (P<0.05; Fig. 6). Additionally, GA-A treatment significantly enhanced the levels of mitochondrial apoptotic pathway-associated proteins (Bak, Bax, and cytosolic cytochrome C) compared with the control and in a dose dependent manner (P<0.05), suggesting the mitochondrial apoptosis occurred.

The combined treatment of GA-A and AG490 further decreased JAK2/STAT3 expression levels compared with the control (P<0.01; Fig. 7). This result suggests that the anti-breast cancer effects of GA-A may involve other mechanisms, including JAK2/STAT3. Thus, the result supports the hypothesis that the antitumor effects of GA-A are associated with certain molecules, including Bcl-xL, Bak, Mcl-1 and Bax, and the JAK2/STAT3 signaling pathways.