Erastin, 3-methyladenine (3-MA) and bafilomycin A1 (Baf-A1) were purchased from Selleck Chemicals. Anti-autophagy related (ATG)12 (cat. no. 4180) was purchased from Cell Signaling Technology, Inc. Deferoxamine (DFO), anti-p62 (cat. no. ab56416) and anti-ATG5 (cat. no. ab221604) were purchased from Abcam. Anti-beclin 1 (cat. no. B6061) and anti-microtubule-associated proteins 1A/1B light chain 3B (LC3B) (cat. no. l7543) were obtained from Sigma-Aldrich; Merck KGaA. Anti-β-actin (cat. no. AF5001), and horseradish peroxidase-conjugated anti-mouse IgG (cat. no. A0216) and anti-rabbit IgG (cat. no. A0208) were purchased from Beyotime Institute of Biotechnology.

Human MCF-7 and MDA-MB-231 cell lines were purchased from the Shanghai Institute of Cell Biology at the Chinese Academy of Sciences. All cells were cultured in DMEM (SH 30243.01; Hyclone; Cyvita) supplemented with 10% heat-inactivated fetal bovine serum (FB15015; http://zn.clarkbio.com/Clark), 2 mmol/l glutamine, penicillin (100 U/ml) and streptomycin (100 µg/ml). Cells treated with 5 mmol/l 3-MA, 1.5 µmol/l Baf-A1 or 500 µmol/l DFO for 1 h prior to erastin treatment which lasted for 24 h with different concentrations (MCF-7 cells were treated with 40 and 80 µmol/l; MDA-MB-231 cells were treated with 20 and 40 µmol/l) at 37°C and 5% CO₂ in a humid environment and used for experimentation when entering the mid-log phase. A light microscope was used at magnification, ×400 to observe cellular changes.

MCF-7 (6×10³ cells/well) and MDA-MB-231 (8×10³ cells/well) breast cancer cells were seeded into 96-well microplates, cultured at 37°C for 24 h and treated with 5 mmol/l 3-MA, 1.5 µmol/l Baf-A1 or 500 µmol/l DFO for 1 h prior to erastin treatment which lasted for 24 h with two values separately (MCF-7 cells were treated with 40 and 80 µmol/l; MDA-MB-231 cells were treated with 20 and 40 µmol/l) at 37°C and 5% CO₂ in a humid environment. Cellular viability was assessed by performing an MTT assay (DMSO was used to dissolve the purple formazan), the results of which were expressed as a ratio to the absorbance of control cells at 490 nm. Absorbances were measured at 490 nm using an automatic multi-well spectrophotometer (Bio-Rad Laboratories, Inc.). The IC₅₀ values were calculated using GraphPad Prism 6 software (GraphPad Software, Inc.), and the mean IC₅₀ value of four experiments was presented.

The LDH cytotoxicity assay kit (cat. no. C0017; Beyotime Institute of Biotechnology) was used to determine the cell death rate. According to the manufacturer's protocol, the absorbance of each sample was measured at 490 nm using an automatic multi-well spectrophotometer. The cell death ratio was calculated using the following formula: Cell death ratio (%)= (A_sample-A_control/A_max-A_control) ×100%. Where A_sample is the sample absorbance value, A_control is the absorbance value of the control group and A_max is the absorbance value of the positive group.

Intracellular Fe²⁺ levels were determined using an iron colorimetric assay kit (cat. no. K390) purchased from BioVision, Inc. According to the manufacturer's instructions, MCF-7 and MDA-MB-231 breast cancer cells were collected and separated into untreated, erastin-treated and DFO-pretreated groups before the cells were added to iron assay buffer, homogenized on ice and centrifuged at 16,000 × g for 10 min at 4°C to obtain the supernatant. Samples (50 µl/well) were then incubated with 50 µl assay buffer in a 96-well microplate for 30 min at 25°C. Samples were subsequently incubated with 100 µl iron probe in the dark for 60 min at 25°C and assessed with a microplate reader at a wavelength of 593 nm. Absorbance values were calibrated to a standard concentration curve to calculate the concentration of iron. The results are expressed as a ratio to the concentration of control cells.

Transfection with ATG5 siRNA (5 µg/µl) was performed using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. ATG5 siRNA (5′-GACGUUGGUAACUGACAAATT-3′) and scrambled siRNA (5′-UUCUCCGAACGUGUCACGUTT-3′) were purchased from Shanghai GenePharma Co., Ltd. MCF-7 and MDA-MB-231 cells were separately seeded into a 6-well plate at 6×10⁴ and 8×10⁴ cells per well overnight. The following day, cells were transfected with ATG5-specific siRNAs and scrambled siRNAs for 48 h. The efficiency of transfection was determined by western blotting.

MCF-7 and MDA-MB-231 breast cancer cells treated with eastin alone, 5 mmol/l 3-MA, 1.5 µmol/l Baf-A1, ATG5 siRNA or 500 µmol/l DFO for 1 h prior to erastin were collected and homogenized as described previously (22). Homogenates were centrifuged at 1,000 × g for 10 min at 4°C to obtain the supernatant, and the protein content of the supernatant was determined using BCA Protein assay kit (Beyotime Institute of Biotechnology). Equal quantities of protein including anti-ATG5 (55 kd), anti-ATG12 (55 kd), anti-P62 (62 kd), anti-Beclin1 (69 kd) and anti-LC3B (14 and 16 kd) were electrophoresed on 8-12% sodium dodecyl sulfate-polyacrylamide gels based on the molecular weight of the target protein and transferred to PVDF membranes (EMD Millipore). The membranes were then blocked with 5% skimmed milk in PBS for 2 h at room temperature and incubated overnight at 4°C with primary antibodies, including anti-ATG5 (1:1,000), anti-ATG12 (1:1,000), anti-P62 (1:1,000), anti-Beclin-1 (1:1,000) and anti-LC3B (1:1,000). The membranes were washed with PBS-0.1% Tween-20 (PBS-T) buffer for 30 min at 25°C prior to incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG (1:1,000) or anti-mouse IgG (1:1,000) at 25°C for 2 h. Membranes were subsequently washed with PBS-T buffer and immunoreactive proteins were visualized on a chemi-luminescence developer (Chemiscope 5300; Clinx Science Instruments, Co., Ltd.) with an enhanced chemiluminescence reagent (P10300; NCM Biotech). Densitometry was performed using ImageJ software v1.46 (National Institutes of Health).

All data was obtained from at least four independent experiments and are expressed as the mean ± standard deviation. Statistical analyses were performed with Microsoft Excel 2010 (Microsoft Corporation) and GraphPad Prism 6 software (GraphPad Software, Inc.). Statistical comparisons were made using one-way ANOVA with Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

To investigate the toxic effect of erastin on breast cancer cells, an MTT assay was performed to examine the viability of erastin-treated MCF-7 and MDA-MB-231 cells. Following treatment with 10, 20, 40 and 80 µmol/l erastin for 24 h, cell viability was significantly reduced in MCF-7 and MDA-MB-231 cells compared with untreated cells (Fig. 1A). The IC₅₀ values of erastin in MDA-MB-231 and MCF-7 cells at 24 h were 40 and 80 µmol/l, respectively. Therefore, erastin at these concentrations was used for subsequent experiments.

A LDH release assay was performed to examine erastin-induced death in MDA-MB-231 and MCF-7 cells. MDA-MB-231 and MCF-7 cells were treated with erastin at their respective IC₅₀ concentrations for 3, 6, 12 and 24 h. The results at 24 h demonstrated significant increases in cellular death in each cell line following treatment with 10, 20, 40 and 80 µmol/l erastin compared with untreated cells (Fig. 1B). These data indicated that erastin inhibited the viability of breast cancer cells and triggered breast cancer cell death

To further investigate whether erastin activates autophagy in breast cancer cells, western blotting was performed to analyze the expression of autophagy-associated proteins. Following erastin treatment, the expression of autophagy-associated proteins in breast cancer cells, including beclin-1, ATG5, ATG12 and LC3B, were all significantly increased. Additionally, p62 levels, as a substrate of autophagy, were significantly downregulated in erastin-treated cells (Fig. 1C and D).

3MA and Baf-A1 are often used to inhibit autophagy activation (5). 3MA inhibits autophagy initiation (23), and Baf-A1 disturbs the fusion of autophagosomes with lysosomes (24). The inhibitory effects of 3MA or Baf-A1 on autophagy activation are also decided by their concentrations and treatment times. Higher concentrations of 3MA or Baf-A1 and longer incubation times lead to obvious changes in the baseline level of autophagy-related proteins (23,24). In the present study, the cells were incubated with 5 mmol/l 3MA and 1.5 µmol/l Baf-A1 for 25 h, which did not exhibit marked inhibition on the baseline levels of autophagy-related proteins (Fig. 2A-D). To further elucidate erastin activated autophagy, cells were treated with 5 mmol/l 3-MA for 1 h prior to erastin treatment. Cells were additionally pre-treated with 1.5 µmol/l Baf-A1, an inhibitor of autophagy, for 1 h prior to erastin treatment. The results of the LDH release assay revealed that cells pre-treated with 3-MA or Baf-A1 were significantly more resistant to erastin-induced death compared with cells only treated with erastin (Fig. 2E).

As aforementioned, the results of the current study demonstrated that autophagy-associated proteins (beclin1, ATG5, ATG12, LC3B and P62) were affected by erastin. However, 3-MA treatment prevented cell death and inhibited the upregulation of autophagy-associated proteins induced by erastin. Western blotting revealed that the expression levels of beclin1, ATG5, ATG12 and LC3B were significantly downregulated, while P62 was significantly upregulated when cells were pre-treated with 3-MA compared with those only treated with erastin (Fig. 2A and B). Baf-A1 treatment demonstrated similar effects to that of 3-MA at a concentration of 1.5 µmol/l on the expression of P62. However, in contrast to 3-MA, the main effect of Baf-A1 is to inhibit the fusion of autophagosome and autolysosome (24), which resulted in a significant increase of LC3B in cells pre-treated with Baf-A1 compared with cells only treated with erastin (Fig. 2C and D). Therefore, it was unnecessary in the present study to detect the expression of ATG5, ATG12 and Beclin-1.

To further investigate erastin induced autophagic death in breast cancer cells, ATG5 was knocked-down with siRNA. The results demonstrated that ATG5-knockdown significantly decreased the cell death ratio of erastin-treated cells (Fig. 2F) and significantly inhibited the erastin-induced changes of autophagy-related proteins (Fig. 2G and H).

To investigate whether the levels of intracellular iron are intrinsically associated with erastin-induced autophagy, an iron assay was performed to assess intracellular iron accumulation after treatment with erastin. The results revealed that when compared with the control group, iron levels significantly increased in cells incubated with erastin for 24 h. Furthermore, when the concentration of erastin was increased from 40 to 80 µmol/l, iron levels significantly increased further. The data indicated that erastin-induced an increase in intracellular irons in each breast cancer cell line in a concentration- and incubation time-dependent manner (Fig. 3A).

To elucidate the importance of iron accumulation, cells were treated with iron chelator DFO at 500 µmol/l for 1 h and subsequently incubated with erastin for 24 h. Similarly with 3MA and Baf-A1, the dose of DFO used in the current study was 500 µmol/l, which did not show any inhibitory effect on cellular viability or the baseline level of autophagy-related proteins (Fig. 3B-E). The results demonstrated that DFO significantly inhibited the erastin-induced increase of intracellular Fe²⁺ iron in MDA-MB-231 and MCF-7 cells (Fig. 3D). Additionally, the LDH release assay demonstrated that erastin-induced breast cancer cell death was significantly inhibited in the presence of DFO (Fig. 3B). Consistently, observation under a light microscope demonstrated that DFO pre-treatment markedly reversed the erastin-induced cellular changes (Fig. 3F).

To determine the relationship between autophagy and erastin-induced increases of intracellular iron, western blotting was performed to assess the changes of autophagy-associated proteins in each breast cancer cell line following pre-treatment with DFO. The results demonstrated that DFO significantly inhibited the erastin-induced changes in the autophagy-associated protein expression levels (Fig. 3C and E). These data suggested that DFO treatment inhibited erastin-induced autophagy by mitigating erastin-induced iron levels.