Choriocarcinoma cell lines JAR and JEG-3 obtained from American Type Culture Collection (Manassas, VA, USA) were cultured in DMEM medium with 10% fetal bovine serum at 37°C with 5% CO₂. ERK was activated by 10 µM concentration of ceramide C6 (C-C6; Santa Cruz Biotechnology, Santa Cruz, CA, USA), added 24 h after daidzein treatment for another 24 h.

Growth rates of cells were measured by 3- (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma, St. Louis, MO, USA) assay. The absorbance was measured at 490 nm using a universal microplate reader (Model ELx800; BioTek Instruments, Inc., Winooski, VT, USA).

JAR and JEG-3 cells were seeded onto 6-well plates at a density of 1,000 cells per well. After 14 days of culture, cells were washed with PBS, fixed with 4% paraformaldehyde, and subsequently stained with 0.1% crystal violet solutions.

Cells with 60–80% confluence were trypsinized, washed with cold PBS, resuspended with cold 70% ethanol then stored at −20°C overnight. Before subjected to flow cytometry analysis, ethanol was removed and pellet cells were washed with cold PBS twice, cells were resuspended in PBS with 0.5 µg/ml RNase and 50 µg/ml propidium iodide and incubated at room temperature in the dark for 30 min. Then cell samples were analyzed in a FACSCalibur flow cytometer (Becton-Dickinson, San Jose, CA, USA) and CellQuest software.

Cells were washed once with cold PBS and lysed in RIPA buffer (50 mM Tris pH 8.0, 150 mM NaCl, 0.1% SDS, 1% NP-40, and 0.5% sodium deoxycholate) containing protease inhibitors. Approximately 20 µg of protein was separated with 8–12% SDS-PAGE gel and blotted onto nitrocellulose membranes. Then membranes were blocked with 5% skim milk at room temperature for 1 h and then incubated with primary antibodies against β-actin (Abcam, Cambridge, UK), c-myc (Abcam), PCNA (Santa Cruz), cyclin D1 (Santa Cruz), p21 (Cell Signaling Technology, MA, USA), ERK (Abcam) and p-ERK (Abcam) at 4°C overnight, followed by TBST wash and 1-h incubation with HRP-conjugated secondary antibodies at room temperature. Protein bands were visualized by a Molecular Imager ChemiDoc XRS System (Bio-Rad Laboratories, Hercules, CA, USA).

Cells were fixed in 4% paraformaldehyde in PBS for 15 min, and then permeabilized with 0.1% Triton X-100 for 15 min and blocked in 1% BSA for 1 h. Cells were sequentially incubated with primary antibody (p-ERK; 1:100 in 1% BSA) at 4°C overnight, followed by Alexa Fluor 488 second antibody (1:150 in 1% BSA) for 1 h and DAPI for 5 min. Slides were analyzed and photographed by Olympus BX51 Microscope (Olympus, Tokyo, Japan).

Fifteen mice were randomly divided into three groups depending on different gavage components: negative control group (edible oil only), low dose group (10 mg/kg body weight/day of daidzein dissolved in oil) and high dose group (20 mg/kg body weight/day of daidzein dissolved in oil). For tumorigenesis assay in nude mice, 5×10⁶ JEG-3 (considering JEG-3 more easily formed subcutaneous xenografts than JAR when performing pre-experiment) were injected subcutaneously into left sides of the flank region after 1 week daily gavage feeding. Three weeks after subcutaneous injection with JEG-3 (4 weeks after daily gavage with/without different concentration of daidzein), the mice were sacrificed, then xenograft tumors were harvested, weighed and fixed with 4% paraformaldehyde. Animal care and protocols were in accordance with the guidelines of the Institutional Animal Care and Use Committee of Xi'an Jiaotong University.

Tumor sections of nude mouse xenografts were studied by immunohistochemistry (IHC) assay using EnVision™ System (Dako, Carpinteria, CA, USA). Primary antibodies used in IHC were c-myc (Abcam), PCNA (Santa Cruz), p-ERK (Abcam). Staining signals were photographed using an Olympus BX51 microscope (Olympus). Average intensity score of the positive cells (0, none; 1, weak; 2, intermediate; 3, strong) and percentage score of stained cells (1, 0–25%; 2, 25–50%; 3, 50–75%; 4, 75–100%) were multiplied to get the total staining score, ranging from 0 to 12.

All experiments were repeated at least three times. Statistical analysis was carried out using the SPSS 19.0 statistical software package (SPSS, Inc., Chicago, IL, USA) was used to analyze differences between two groups (Student's t-test) and p-values <0.05 were regarded as statistically significant.

Using MTT assay, we investigated the viability inhibition effect of daidzein in CC cell lines JAR and JEG-3. Both cell lines were treated with daidzein at concentration of 12.5, 25, 50, 100, 200 and 400 µM. The results showed that daidzein had a concentration- and time-dependent effect on JAR cell viability. With the increasing of concentration or time of treatment with daidzein, the viability decreased. Also, the IC₅₀ at 24, 48 and 72 h were 101.60, 92.56 and 87.73 µM, respectively (Fig. 1A). Similar effect were seen in JEG-3, while the IC₅₀ was 103.10, 94.14 and 68.82 µM, respectively (Fig. 1B).

In order to determine whether the cell viability reduction is due to proliferation inhibition, we performed colony formation assay on both cell types with daidzein treatment. The colony-forming capacity of JAR and JEG-3 decreased as the concentration of daidzein increased. While 25 µM daidzein seemed not to change much, treatment with 50 and 100 µM daidzein markedly suppressed proliferation. However, when the concentration reached 200 µM, colonies hardly formed (Fig. 2A and B).

Cell proliferation inhibition is usually associated with cell cycle arresting at a specific stage. To examine if daidzein causes cell cycle arrest, we performed fluorescence-activated cell sorting (FACS) analysis. In the asynchronized steady state, the percentage of proportion of G1 phase cells was greater after treated with daidzein. In JAR cells, G1 phase percentage of proportion was 61.66, 65.38 and 72.64% treated with 25, 50 and 100 µM, respectively, while it was only 48.02 without (Fig. 2C). As to JEG-3 cells, daidzein increased G1 phase percentage from 50.85% (0 µM) to 56.34, 60.00 and 61.07% (25, 50 and 100 µM, respectively, Fig. 2D). Then we detected the expression of G1 phase related proteins (c-myc, PCNA and cyclin D1) and p21, one of CDKIs family, by western blot analysis. The expression levels of c-myc, PCNA and cyclin D1 were decreased significantly, and the decrease was greater with the increase of concentration (p<0.05, respectively). On the contrary, the expression level of p21 increased, and the differences were also significant (p<0.05, respectively, Fig. 2E and F). These results suggest that choriocarcinoma cells are arrested at the G1 phase by daidzein.

When detecting expression of cell cycle associated protein, we also found that daidzein suppressed ERK1/2 by reducing phosphorylation, which suggested ERK1/2 might participate in daidzein proliferation inhibition capacity (p<0.05, respectively, Fig. 3A and B). As known, ERK1/2 is involved in phosphorylation and subsequent nuclear translocation, so we wondered whether daidzein might also effect nuclear translocation of ERK1/2. Therefore, we undertook to examine whether daidzein influenced this process by using immunofluorescence staining (IF), and the results showed that compared with DMSO, phospho-ERK1/2 translocated into the nucleus less after daidzein treatment (Fig. 3C and D) in both cell-lines. Taken together, our results suggested daidzein could inhibit ERK1/2 phosphorylation and p-ERK1/2 nuclear translocation.

The changes of phosphorylation and translocation of ERK1/2 were obtained during daidzein's proliferation inhibition as described above, thus we hypothesized ERK1/2 involved in daidzein-caused choriocarcinoma cell proliferation suppression. Ceramide C6 is ERK pathway agonist, and could significantly increase phosphorylation of ERK which was inhibited by daidzein in both cell lines (p<0.05, respectively, Fig. 4A and B). At the same time, colony formation assay showed that C-C6 obviously increased the amount of cell colonies which were decreased by daidzein (Fig. 4C and D). Compared to daidzein only, C-C6 decreased G1 phase percentage from 74.56 to 63.02 and 60.75 to 54.02 in JAR and JEG-3, respectively on the basis of daidzein (Fig. 4E and F). Besides, C-C6 also abolished daidzein's increase on protein expressions of c-myc, PCNA and cyclin D1 and decreased p21 (Fig. 4G and H). In summary, daidzein's proliferation inhibition on choriocarcinoma cells can be abolished by ERK pathway agonist C-C6, therefore, we consider that ERK participated in this process.

To determine whether daidzein's anti-proliferation activity in vitro could be extended in vivo, subcutaneous xenograft model was introduced in this study. Average weights of tumors in three groups showed negative correlation with daidzein concentration (Fig. 5A). The low dose group was lighter than that of control group, and weight in high dose group was even lighter than low dose group (p<0.05, Fig. 5B). We then stained c-myc, PCNA and p-ERK expression in these JEG-3 xenograft tissues using IHC assays (Fig. 5C). The reduced c-myc and PCNA expression was accompanied with less p-ERK staining in low dose group and even less in high dose group than control. The IHC scores of c-myc, PCNA and p-ERK in control group were 10.60±1.95, 8.20±2.49 and 7.40±3.13, those in low dose group were 5.60±1.67, 4.20±1.10 and 3.60±1.82, while those in high dose group were 2.60±1.95, 2.20±1.30 and 0.80±0.84, respectively, with significant differences, and the scores of c-myc and PCNA showed positive correlation with p-ERK (Fig. 5D). These data suggested that daidzein could also inhibit choriocarcinoma growth in vivo by downregulating c-myc and PCNA, and this process was probably mediated by p-ERK.