A total of 4 human pancreatic cancer cell lines (BxPC-3, SW1990, PANC-1, and AsPC-1) and one normal human pancreatic duct epithelial cell line (HPDE6-C7) were purchased from the ATCC (American Type Culture Collection). All cells were maintained in DMEM containing 10% FBS and placed in a humidified incubator at 37°C with 5% CO₂.

3-(4,5-Di-2-yl)-2,5-ditetrazolium bromide (MTT) assay was performed to evaluate the cytotoxic effect of ruscogenin. The cells were seeded in 96-well plates (3×10³ cells/well), cultured for 24 h and treated with ruscogenin (0.001–100 µM). At the end of the incubation period, the cells were incubated with 1 mg/ml MTT solution. Three hours later, the absorbance was measured at 450 nm and the data were assessed using an ELX-800 spectrometer reader (Bio-Tek Instruments).

Cell death was detected with trypan blue staining and by flow cytometry as described previously (14,15).

The number of iron oxide nanoparticles was evaluated using Prussian blue staining. Briefly, 4% paraformaldehyde was used to fix the cells and they were subsequently washed with PBS. The cells were incubated with Prussian blue and rewashed with PBS. Finally, a light microscope was used to evaluate intracellular iron oxide distribution.

Flow cytometry with DHE (dihydroethidium) was conducted to measure ROS levels. The cells were collected, stained with DHE and incubated in the dark at 37°C for 20 min. Finally, the Cell Quest software (BD Biosciences) was used to analyze the cell population of each sample in a flow cytometer.

The following primary antibodies were used: Anti-transferrin (cat. no. ab9538), anti-transferrin receptor (cat. no. ab1086), anti-SLC40A1 (cat. no. ab85370), anti-DMT1 (cat. no. ab55735), anti-ferritin (cat. no. ab75973) and β-actin (cat. no. ab8227) (purchased from Abcam). The proteins were separated and transferred to Hybond membranes (Amersham). Then, membranes were blocked using 5% fat-free milk. Subsequently, the membranes were blocked using 5% fat-free milk. The membranes were incubated with primary antibodies overnight at 4°C and the following day secondary antibodies were added for 1 h at room temperature. Finally, the proteins were visualized using an enhanced chemiluminescence system and following X-ray film exposure.

The shRNA transfection was performed as described in our previously studies (16). The negative control (sh-NC) and transferrin (sh-transferrin) shRNA sequences, and negative control (vector) and ferropotin plasmids were designed and synthesized by Shanghai Gene Chemical Technology Co., Ltd. (Shanghai, China). The RNAs were introduced to cells at a final concentration of 50 nM. Transfections were performed using the Lipofectamine 3000 kit (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Cells in the logarithmic growth phase were cultured with lentiviral vector solution for 6 h prior to the addition of lentiviral vectors, and incubated for a further 6 h. At 36 h following infection, cells that exhibited stable transferin knockdown or feroportin overexpression were positively selected using hygromycin B.

All the in vivo experiments were approved by the Research Ethics Committee of the First Affiliated Hospital of the Nanchang University and conducted according to the Guide for the Animal Care and Use Committee of the Nanchang University. A total of 15 six-week-old female BALB/c nude mice (25.2–25.9 g) were purchased from SLAC Animal Laboratories and housed in a specific pathogen-free environment (12-h light/dark cycle at 25°C and 60% relative humidity; the mice were provided with food and water ad libitum in the animal research center of Nanchang University). Approximately 1×10⁶ BxPC-3 cells were subcutaneously implanted into the right flank of the nude mice. The tumors were allowed to grow to approximately 120 mm³ in size and a total of 15 tumor-bearing mice were divided randomly into 3 groups (n=5 per group). Group 1 was injected with 0.1 ml PBS as control; group 2 and 3 received 5 and 10 mg/kg ruscogenin, respectively twice per week. The tumor volume and animal body weight were assessed every 4 days. Finally, the mice were sacrificed by CO₂ asphyxiation and the kidney and liver tissues were collected for toxicity analysis.

All data are presented as mean ± SD. An independent samples Student's t-test was used for direct comparisons between two groups, and an F-test and one-way analysis of variance (ANOVA) followed by the Tukey-Kramer test was used for multigroup comparisons. A P-value less than 0.05 (P<0.05) was considered for significant differences.

To determine the toxic effects of ruscogenin on pancreatic cancer cells, MTT assay was utilized to investigate changes in ruscogenin-induced cell viability in the control cells (HPDE6-C7) and in pancreatic cancer cells (BxPC-3, SW1990, PANC-1, and ASPC-1). The cell viability of BxPC-3, SW1990, PANC-1 and ASPC-1 cells was significantly decreased by ruscogenin in a concentration- and time-dependent manner compared with that noted in HPDE6-C7 cells (Fig. 1A-E). Analysis of the cell viability data derived from BxPC-3, SW1990, PANC-1 and ASPC-1 cells treated with ruscogenin demonstrated that the IC₅₀ values noted for ruscogenin treatment of BxPC-3, SW1990, PANC-1 and ASPC-1 cells for 48 h were 7.32, 8.14, 37.62 and 28.19 µmol/l, respectively. Therefore, 7 µmol/l was used as the optimal IC₅₀ value of ruscogenin for BxPC-3 and SW1990 cells in the following studies. In addition, trypan blue staining assay further demonstrated that ruscogenin induced pancreatic cancer cell death in a concentration-dependent manner (Fig. 1F). Therefore, these results indicated that ruscogenin impaired pancreatic cancer cell viability and triggered pancreatic cancer cell death.

The concentration of intracellular ferrous iron was estimated to determine whether ruscogenin induced ferroptosis in pancreatic cancer cells. Ruscogenin induced the release of intracellular ferrous iron in a dose- and time-dependent manner (Fig. 2A). To further elucidate the significance of iron increase, the cells were pretreated with 500 µmol/l of the iron chelator deferoxamine (DFO) for 1 h prior to incubation with ruscogenin. The results indicated that DFO apparently repressed ruscogenin-induced increase of intracellular ferrous iron (Fig. 2B). In addition, the results of the lactate dehydrogenase (LDH) release assay demonstrated that pancreatic cancer cell death induced by ruscogenin was significantly impaired in the presence of DFO (Fig. 2C). In contrast to DFO, pretreatment of the cells with ferric ammonium citrate (FAC) (500 µmol/l) caused a significant increase in the levels of pancreatic cancer cell death induced by ruscogenin (Fig. 2D). The data demonstrated that ruscogenin triggered pancreatic cancer cell death by increasing intracellular iron concentration.

Various studies have shown that the cytotoxicity of ferroptosis is dependent on ROS levels (15,17). The ROS levels produced were measured and the results indicated that they were increased following ruscogenin treatment in the BxPC-3 (Fig. 3A) and SW1990 cells (Fig. 3B). Moreover, DFO apparently repressed ruscogenin-induced ROS generation (Fig. 3C and D).

It is well known that transferrin and ferroportin actively regulate cellular iron levels by transporting iron into and out of the cells, respectively (18,19). Western blot analysis was used to explore whether ruscogenin affects the expression of these two key iron regulatory proteins. The results indicated that transferrin expression was significantly upregulated, while ferroportin expression was significantly reduced following ruscogenin treatment (Fig. 4). Moreover, the expression levels of the transferrin receptor, DMT1 and of the protein ferritin did not change significantly following ruscogenin treatment (Fig. 4), which further proved that the effects of ruscogenin treatment on transferrin and ferroportin expression were not caused by the general change in the expression levels of the iron transport regulatory proteins. Taken collectively, the results indicated that ruscogenin treatment affected iron transport in the cells and further caused pancreatic cancer cell death via ferroptosis.

The effect of transferrin expression on ruscogenin-induced pancreatic cancer cell death was further explored. As shown in Fig. 5A, expression of transferrin was knocked down, and it was found that transferrin knockdown decreased ruscogenin-induced pancreatic cancer cell death (Fig. 5B). In addition, ferroportin was overexpressed (Fig. 6A) and the data indicated that its overexpression blocked ruscogenin-induced cell death (Fig. 6B). These results demonstrated that transferrin and ferroportin are involved in ruscogenin-induced pancreatic cancer cell death.

Finally, BxPC-3 human pancreatic cancer xenografts were established to investigate the anticancer effects of ruscogenin in vivo. Treatment of the animals with 5 and 10 mg/kg ruscogenin indicated significant antitumor effects with regard to xenograft growth (Fig. 7A and B). It is important to note that ruscogenin treatment was well tolerated without significant weight loss compared with the corresponding effects noted in the control group (Fig. 7C). Moreover, the levels of alanine aminotransferase and aspartate aminotransferase enzymes (Fig. 7D) were not increased following ruscogenin treatment indicating no apparent organ injury. The results of the histopathological analysis further demonstrated no evidence of toxicity in normal tissues (Fig. 7E).