Mangiferin was purchased from Shanghai PureOne Technology (Shanghai, China). The purity of mangiferin was >95%, and the chemical structure is shown on Fig. 1A. Human Embryonic Kidney 293T cells, human ovarian adenocarcinoma OVCAR3 cells, CaOV3 cells, SKOV3 cells, endometrioid carcinoma TOV-112D cells, and clear cell carcinoma TOV-21G were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). Human ovarian adenocarcinoma A2780 cells were purchased from Procell Co. (Wuhan, China). RPMI-1640 medium and fetal bovine serum (FBS) was purchased from Gibco BRL (Grand Island, NY, USA). 3-(4,5-Dimetrylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT, cat. no. M5655), dimethyl sulfoxide (DMSO, cat. no. D2650), paraformaldehyde (PFA, cat. no. 16005), cisplatin (USP grade, cat. no. 1134357), Annexin V-FITC Apoptosis detection kit (cat. no. APOAF), HEPES (cat. no. H3375), Triton X-100 (cat. no. H9284), 2 mmol/l sodium orthovanadate (cat. no. S6508), sodium fluoride (cat. no. S7920), 1 mmol/l edetic acid (cat. no. E9884), PMSF (cat. no. 78830), aprotinin (cat. no. A11530), leupeptin (cat. no. L2884) and rhodamine-123 were purchased from Sigma-Aldrich Chemical (St. Louis, MO, USA). Hoechst 33342 (cat. no. C1025) was purchased from Beyotime Chemicals (Suzhou, China). Rabbit polyclonal pro-caspase-3 (cat. no. sc-7148; 1:1,000 dilution), mouse monoclonal pro-caspase-9 (cat. no. sc-56073; 1:1,000 dilution), rabbit polyclonal Bax (cat. no. sc-493; 1:500 dilution), mouse monoclonal Bid (cat. no. sc- 135847; 1:1,000 dilution), rabbit polyclonal Bcl-2 (cat. no. sc-492; 1:1,000 dilution), mouse monoclonal Bcl-X_L (cat. no. sc-8392; 1:1,000 dilution), rabbit polyclonal PARP (cat. no. sc-7150; 1:1,000 dilution), rabbit polyclonal cytochrome c (cat. no. sc-7159; 1:1,000 dilution), mouse monoclonal cyclin D1 (cat. no. sc-450; 1:1,000 dilution), mouse monoclonal β-catenin (cat. no. sc-59737; 1:1,000 dilution), rabbit polyclonal Notch3 (NICD domain, cat. no. sc-5593; 1:1,000 dilution), mouse monoclonal Notch1 (cat. no. sc-376403; 1:1,000 dilution) and mouse monoclonal β-actin (cat. no. sc-47778; 1:5,000 dilution) purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Mouse monoclonal PI3K p85 (cat. no. ab189403; 1:500 dilution), rabbit polyclonal phosphor-PI3K p85 (cat. no. ab182651; 1:500 dilution), rabbit polyclonal mTOR (cat. no. ab2732; 1:500 dilution) and rabbit polyclonal phospho-mTOR (S2448, cat. no. ab109268; 1:500 dilution) antibodies were from Abcam (Cambridge, MA, USA). Mouse monoclonal Akt (pan, cat. no. 2920; 1:500 dilution), rabbit polyclonal phospho-Akt (S473, cat. no. 4060; 1:500 dilution), rabbit polyclonal cleaved caspase-3 (cat. no. 9661; 1:500 dilution) and rabbit polyclonal cleaved caspase-9 (cat. no. 9509; 1:1,000 dilution), were purchased from Cell Signaling Technologies (Danvers, MA, USA). Horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG polyclonal antibody (cat. no. 115-035-003) and HRP-conjugated goat anti-rabbit polyclonal IgG (cat. no. 111-035-003) were purchased from Jackson ImmunoResearch Laboratories.

Human Embryonic Kidney 293T (HEK-293T) cells, ovarian cancer cell lines CaOV3 and SKOV3 were cultured in DMEM containing 10% FBS, 100 µg/ml streptomycin (Life Technologies, Rockville, MD, USA), 100 U/ml penicillin (Life Technologies), and 0.03% L-glutamine (Sigma-Aldrich, St. Louis, MO, USA). TOV-112D and TOV-21G cells were cultured in RPMI-1640 containing 15% FBS, 100 µg/ml streptomycin, 100 U/ml penicillin and 0.03% L-glutamine. OVCAR3 cells were cultured in RPMI-1640 containing 20% FBS, 10 µg/ml bovine insulin, 100 µg/ml streptomycin, 100 U/ml penicillin and 0.03% L-glutamine. CHO cells were cultured in F-12K containing 10% FBS, 100 µg/ml streptomycin, 100 U/ml penicillin and 0.03% L-glutamine.

OVCAR3 cells at logarithmic growth phase were seeded in a 96-well plate and incubated at 37°C for 24 h, and different dosage of mangiferin (12.5, 25, 50 and 100 µg/ml) were added and incubated for 12, 24, 36 and 48 h, respectively. MTT [0.05 mg (10 µl of 5 mg/ml)] was added to each well and incubated at 37°C for 4 h, and then medium was removed and shaked thoroughly for 1 h. Finally, termination buffer was added into each well. The absorbance at 570 nm was measured with a spectrophoto-meter (Model 3550 Microplate Reader, Bio-Rad Laboratories, Hercules, CA, USA). Cell viability (%) = [OD 570 nm (drug)/OD 570 nm (control)] ×100%.

OVCAR3 cells were seeded into a 6-well culture plate at a density of 4×10⁵ cells/well and cultured for 24 h. Control groups were treated with 0.05% DMSO, and mangiferin groups were treated with 25 µg/ml mangiferin. Hoechst 33342 staining was applied to further detect viable cells. In brief, cells were fixed with 4% PFA for 30 min at room temperature after 48 h incubation with either mangiferin or 0.05% DMSO, and then cells were washed twice with PBS. Hoechst 33342 (5 µg/ml) was added and stained for 15 min, and then the cells were washed and analyzed immediately with a fluorescence microscope (Olympus, Tokyo, Japan). All experiments were carried out in triplicate. Experiments were repeated three times, and data are representative of replicate experiments.

To study the effects of mangiferin on cell apoptosis, OVCAR3 cells were treated with 25 µg/ml mangiferin for 24 h. Detached and adherent cells were then collected and labeled for 15 min at room temperature with Annexin V-FITC (1 µg/ml) and prodidium iodide (PI, 40 µg/ml) after washing away unbound fluorescent dye, cells were immediately analyzed on a FACS Aria II cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). Cells (1×10⁵) were measured for each sample. All experiments were carried out in triplicate. Experiments were repeated three times, and data are representative of replicate experiments.

Mangiferin-induced apoptosis of OVCAR3 cells was also assessed by measuring the activity of caspase-3. OVCAR3 cells were seeded into a 6-well culture plate at a density of 4×10⁵ cells/well for 16 h incubation. Subsequently, cells were treated with 25 µg/ml mangiferin for 0, 12, 24, 36, 48 h. Caspase-3 activity was measured by using a caspase-3 activity kit (Beyotime Institute of Biotechnology, Haimen, China). In brief, cells were lysed, the supernatant was collected, quantified, and incubated with the caspase-3-specific color substrate Ac-DEVD-pNA. Caspase-3 activity was determined by measuring optical density at OD400 nm.

Mitochondrial membrane potential was measured by fluorescent dye rhodamine-123. After treatment with 25 µg/ml mangiferin for 12, 24, 36 and 48 h, cells were collected and suspended in 1 ml of PBS containing 1 µg/ml rhodamine-123 and incubated at 37°C for 15 min. The fluorescence intensity of the cells was analyzed on a FACS Aria II cytometer (Becton Dickinson). All experiments were carried out in triplicate. Experiments were repeated three times, and data are representative of replicate experiments.

OVCAR3 cells were treated with 0, 12.5, 25, 50 and 100 µg/ml mangiferin for 24 h, respectively, and then both adherent and floating cells were collected. The cell pellets were resuspended with lysis buffer and lysed at 4°C for 15 min, and the lysis buffer consisted of 50 mmol/l HEPES (pH 7.4), 1% Triton X-100, 2 mmol/l sodium orthovanada, 100 mmol/l sodium fluoride, 1 mmol/l edetic acid, 1 mmol/l PMSF, 10 mg/l aprotinin and 10 mg/l leupeptin. After 12,000 × g centrifugation for 15 min, the protein content of supernatant was determined by Bradford protein assay (cat. no. P0006, Beyotime, Suzhou, China). Equal amounts of the total protein were separated by 4–12% NuPAGE Bis-Tris gels (cat. no. NP0327BOX, Life Technologies, Carlsbad, CA, USA) and transferred to PVDF membranes (cat. no. ISEQ00010, Millipore, Bedford, MA, USA), the membranes were soaked in blocking buffer (5% BSA, cat. no. V900933, Sigma-Aldrich Chemicals). Then proteins were detected using primary and secondary antibodies and visualized with ECL (cat. no. 345818, Millipore). Experiments were repeated three times, and data are representative of replicate experiments.

The PCR product of the Notch3 intracellular domain (NICD) was cloned into pTY linkers. Third-generation vectors were used in this experiment. NICD lentiviral vector were transient transfected into HEK-293T cells. Briefly, HEK-293T cells were co-transfected with appropriate amounts of vector plasmids, including helper construct, envelope plasmid, tat plasmid and pTY linker containing NICD. The viruses were collected, concentrated, and 1×10⁶ lentiviral particles were transduced to OVCAR3 cells in the presence of 5 µg/ml Polybrene (cat. no. sc-134220, Santa Cruz Biotechnology).

Acute toxicity testing was performed to determine median lethal dosing (LD₅₀) of mangiferin. After 16 h fasting, BALB/c mice were randomly divided into 6 groups (n=10). Graded doses of mangiferin (10, 20, 50, 100, 200, and 500 mg/kg) were separately administered intraperitoneally to the mice. All mice were allowed free access to food and water, and then mortality in each group was assessed after 24, 48 and 72 h administration of mangiferin. Percentage mortality in each group was calculated and plotted against log₁₀ of mangiferin dose. A regression line was fitted by the method of least squares, and LD₅₀ values were calculated. Animal handling was in accordance with the ethics committee of the Sichuan Academy of Medical Science and Sichuan Provincial People's Hospital, and all animals were kept in 12 h light/dark cycle with free access to water and food, which is in consistent with IVC requirement in Sichuan Academy of Medical Science and Sichuan Provincial People's Hospital.

The in vivo efficacy of mangiferin was studied in xenograft models. In brief, to initiate tumor xenografts, 5×10⁶ OVCAR3 cells were injected into BALB/c nude female mice. After implantation, tumors were allowed to grow to a size of 100–550 mm³ before starting daily treatment. Mice were randomly divided into five groups: 1) Blank control group: mice intraperitoneally administered PBS after tumor formation by OVCAR3 cells; 2) High dose mangiferin group: mice intraperitoneally administered 100 mg/kg mangiferin after tumor formation by OVCAR3 cells; 3) Medium dose mangiferin group: mice intraperitoneally administered 50 mg/kg mangiferin after tumor formation by OVCAR3 cells; 4) Low dose mangiferin group: mice intraperitoneally administered 10 mg/kg mangiferin after tumor formation by OVCAR3 cells; 5) Combined therapy group: mice intraperitoneally administered both cisplatin (10 mg/kg) and mangiferin (50 mg/kg) after tumor formation by OVCAR3 cells.

The administration of mangiferin lasted for another two weeks. Before and after the mangiferin treatment, the body weight of the mice was measured daily then, 5 mice in each group were sacrificed, and 5 mice in each group were raised for survival assay. For the sacrificed mice, the subcutaneous tumors were peeled off and weighed. The volume of tumors was determined in three dimensions with vernier calipers according to the formula: Tvol = length × width × depth ×0.5, and then relative tumor volume was calculated as relative increase or decrease in mean tumor volume from initiation of treatment (V0) up to value at a given time (Vt) and RTV = Vt/V0. Inhibitory rate of tumor volume = (Vcontrol × Vt)/Vcontrol ×100%. After 14 days treatment, mice were sacrificed by cervical dislocation, and subcutaneous tumor masses were determined. Inhibitory level of tumor weight = Wcontrol - Wt)/Wcontrol ×100%.

All data were expressed as mean values ± SEM from at least three independent experiments. Statistical significance was determined by two-way ANOVA and Student's t-test. Experiments were repeated three times, and data are representative of replicate experiments.

Based on our previous study on mangiferin (30), we hypothesized that mangiferin might inhibit the viability of ovarian cancer cells. Therefore, to substantiate our hypothesis and to explore the inhibitory role of mangiferin in ovarian cancer cells, we performed MTT assay on ovarian adenocarcinoma cell lines (CaOV3, SKOV3, OVCAR3 and A2780), endometrioid carcinoma cell line TOV-112D and clear cell carcinoma cell line TOV-21G (Fig. 1B). The IC₅₀ values were 38.14 µg/ml for CaOV3, 57.67 µg/ml for SKOV3, 31.67 µg/ml for A2780, 24.13 µg/ml for OVCAR3, 53.65 µg/ml for TOV-112D and 56.69 µg/ml for TOV-21G. These data strongly suggested that mangiferin could inhibit the viability of ovarian carcinoma cell lines, and OVCAR3 cells were most sensitive to mangiferin treatment. To ascertain whether mangiferin treatment was dose- and time-dependent, OVCAR3 cells were treated with different dosage of mangiferin (0, 12.5, 25, 50,100 µg/ml) for 12, 24, 36, and 48 h (Fig. 1C). As predicted, accompanied with increased dosage and extended treatment, the inhibitory ratio increased, suggesting that mangiferin inhibited the proliferation of OVCAR3 cells in a dose- and time-dependent manner. In addition, it has been widely acknowledged that the standard first-line treatment for advanced stage ovarian carcinoma is the combined chemotherapy of platinum based drugs as cisplatin coupled with paclitaxel (31). However, most patients will relapse due to drug resistance. Therefore to validate whether mangiferin treatment could enhance the sensitivity of OVCAR3 cells to cisplatin, we evaluated the inhibitory rates of combined therapy of both mangiferin and cisplatin. Simultaneously with 1 µg/ml cisplatin, different dosages of mangiferin (0, 12.5, 25, 50 and 100 µg/ml) were added into the culture medium of OVCAR3 cells (Fig. 1D). As expected, increased inhibitory ratios were observed, suggesting that mangiferin treatment could improve the sensitivity of ovarian carcinoma cells to cisplatin in dose- and time-dependent manner. However, 100 µg/ml mangiferin did not result in apparent apoptotic morphology in non-cancerous Chinese Hamster Ovary (CHO) cells after 24 or 48 h treatment (Fig. 1E and F). These results suggested that mangiferin might selectively induce OVCAR3 cells apoptosis but could not trigger apoptosis of normal ovary cells.

To address whether mangiferin could induce apoptosis, we observed the cell morphology under phase contrast microscopy (Fig. 2A). It was obvious that cells treated with 25 µg/ml mangiferin for 48 h showed shrinkage of cytoplasm and nucleus, membrane blebbing, and shattering, which are typical in apoptotic morphology. Furthermore, less viable cells were observed in the cisplatin plus mangiferin treated group. To substantiate our observations, we preformed Hoechst 33342 staining, and found that the remaining cells in mangiferin treated group showed condensed nuclei, indicating those cells were underging apoptosis (Fig. 2B). In order to determine the apoptotic percentages of each group, cells were treated with 25 µg/ml mangiferin for 24 h, and then sorted by flow cytometry. As shown in Fig. 2C, there were more early apoptotic (Annexin V⁺/PI⁻) cells and late apoptotic cells (Annexin V⁺/PI⁺) in mangiferin treated group compared to those of the controls. Statistics on the early apoptotic cells, late apoptotic cells and viable cells are in accordance further demonstrating that mangiferin induced apoptosis of OVCAR3 cells and increased the sensitivity of OVCAR3 cells to cisplatin treatment (Fig. 2D). Taken together, based on the qualitative and quantitative analysis of the apoptotic events, mangiferin was able to induce apoptosis in ovarian carcinoma OVCAR3 cells.

To unravel the mechanism of mangiferin-induced apoptosis, and to determine whether it relied on the caspase pathway, a series of western blotting were performed. Accumulating evidence revealed that during apoptosis, cytochrome c is released from mitochondria to cytosoplasm (32). We, therefore, assessed the cytosol and mitochondria cytochrome c, and found that mangiferin treatment could promote the translocation of cytochrome c from mitochondria. Furthermore, activation of caspase plays a central role in apoptosis, and caspase-3 serves as the convergence point of different apoptotic signaling pathway. As shown in Fig. 3A, caspase-3 and −9 were activated upon mangiferin treatment with increased cleaved caspase-3 and −9. As Bid induces conformational change of Bax, and Bax induces the release of cytochrome c from mitochondria during apoptosis, we found mangiferin treated cells had higher expression level of Bax and Bid compared with that of the control, suggesting more apoptotic events mediated by mangiferin. However, as the protein Bcl-2 interferes with the activation of caspases by preventing the release of cytochrome c, cells treated with mangiferin showed a comparatively lower expression level of Bcl-2. Moreover, decreased Bcl-X_L and increased PARP cleavage were found in mangiferin treated cells. To further validate that caspase pathway was involved in mangiferin-induced apoptosis, we analyzed the caspase-3 activity and determined the integrity of mitochondria membrane. As expected, caspase activity increased in accordance with the dosage of mangiferin (Fig. 3B). Whereas, mitochondria membrane integrity was decreased with the increased dosage, suggesting that mangiferin stimulated the release of cytochrome c (Fig. 3C).

As discussed above, because Notch signaling pathway plays a crucial role in the pathogenesis and chemoresistance of ovarian cancer, we evaluated whether mangiferin could inhibit the expression of Notch1 and Notch3. As shown in Fig. 4A, mangiferin could reduce protein levels of Notch3, but has no effect on Notch1 (data not shown). Besides, as amplification and overexpression of cyclin D1 have been found in several human malignancies (33), the expression levels of cyclin D1 and β-catenin were determined and were also found to be inhibited by mangiferin. Previous report clearly showed crosstalk between Notch and PI3K/Akt (34), therefore, we examined whether mangiferin has any inhibitory effect on PI3K/Akt. As expected, decreased phosphorylation of PI3K p85 (Y607), Akt (S473) and mTOR (S2448) were found in mangiferin treated cells. To further substantiate the inhibitory role of mangiferin on Notch signaling pathway, Notch3 intracellular domain (NICD) were overexpressed in OVCAR3 cells by lentivirus infection (Fig. 4B). Then we found that mangiferin-induced inhibition of cyclin D1, β-catenin and PI3K/Akt/mTOR could be partially abrogated by Notch3 overexpression, highlighting the critical role of Notch3 in the anti-neoplastic effect of mangiferin (Fig. 4C).

Further studies on cell morphology revealed that Notch3 overexpression played a protective role from apoptosis for cells treated with mangiferin. Cell treated with mangiferin showed shrinking and bubbling compared with cells treated with both mangiferin and Notch3 (NICD) overexpressed lentivirus (Fig. 5A). Besides, fluorescent staining revealed that cells showed more condensed nuclei after mangiferin treatment. In contrast, more viable cells were found in mangiferin treated Notch3 overexpressed cells (Fig. 5B). Quantification of apoptotic cells by flow cytometry clearly showed that overexpression of Notch3 rescued the cells from apoptosis (Fig. 5C and D).

Based on our in vitro results, we decided to evaluate the anti-neoplastic properties of mangiferin in an in vivo tumor xenograft model system. Therefore, OVCAR3 cells were xenografted into nude mice, and then examined the growth rates of these tumors with or without mangiferin treatment. On the basis of caliper measurements (Fig. 6A), after 14 days of treatment, with the increase dose of mangiferin, the xenograft tumor volume decreased significantly compared to that of the blank control group. Similarly, mangiferin plus cisplatin treated mice showed reduced xenograft tumor volume, suggesting mangiferin could increase the sensitivity of ovarian cells to cisplatin. Moreover, tumor weights were also decreased after mangiferin therapy (Fig. 6B). In addition, calculation of the relative tumor volume demonstrated that mangiferin treatment could greatly inhibit tumor growth (Fig. 6C). Furthermore, judged from the survival curve, mangiferin could remarkably prolong the life span of OVCAR3 cell xenograft mice compared with blank control group (Fig. 6D). Therefore, accompanied with decreased tumor size, body weights of mice receiving mangiferin treatment, showed gradual but remarkable recovery after mangiferin treatment (Fig. 6E). In summary, our in vivo data clearly indicated that mangiferin could inhibit tumor growth, supporting the potential use of mangiferin as a chemotherapeutic agent in ovarian cancer.