α-solanine and its solvent, dimethyl sulfoxide (DMSO), were purchased from Sigma-Aldrich (St. Louis, MO, USA). α-solanine was stored at −20°C after dissolution. Human esophageal cancer cell lines (EC9706 and KYSE30) were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). The cells were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS; Gibco BRL, Gaithersburg, MD, USA) and incubated in a 5% CO₂ humidified incubator at 37°C.

EC9706 and KYSE30 cells in logarithmic phase growth were seeded into each well of a 96-well plate at a density of 1×10⁴ cells/well. After 24 h, cells of each cell line were treated with increasing concentrations (0, 2, 4 and 6 µmol/l) of α-solanine combined with increasing concentrations (0, 10, 20, 30, 40 and 50 µg/ml) of 5-FU or increasing concentrations (0, 1, 2, 4, 6 and 8 µg/ml) of Cis, respectively, for 24 h. The blank group was treated with PBS, and the control group was treated with 0 µmol/l α-solanine containing 1‰ DMSO. The viability of cells was determined according to the protocol included in the Cell Counting Kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan). Cells were cultured at 37°C in 5% CO₂ for another 3 h with medium inclusion with 10% CCK-8, and then the optical density was measured at 450 nm. The experiments were performed in triplicate.

Apoptosis was analyzed by flow cytometric assay. Flow cytometric assays were conducted using FITC Annexin V Apoptosis Detection Kit I (BestBio, Shanghai, China), according to the manufacturer's instructions. Briefly, EC9706 and KYSE30 cells were cultured in a 96-well plate, in which the medium was added with different concentrations of α-solanine (0, 2, 4 and 6 µmol/l) with or without 40 µg/ml 5-FU or 6 µg/ml Cis for 48 h. The cells were divided into a Blank group (treated with PBS containing 1‰ DMSO), additional treatment groups (treated with 40 µg/ml 5-FU or 6 µg/ml Cis and different concentrations of α-solanine), and the control group treated with 40 µg/ml 5-FU or 6 µg/ml Cis with PBS containing 1‰ DMSO. Cells from each group were harvested by trypsinization and re-suspended at a density of 1×10⁶ cells/ml in 1X binding buffer. After double staining with FITC Annexin V and propidium iodide (PI), cells were analyzed using a FACScan flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) equipped with CellQuest software (BD Biosciences).

The enzymatic activity of caspase-3/7 was measured using the Caspase-Glo 3/7 assay kit (Promega, Shanghai, China) according to the manufacturer's instructions. In brief, EC9706 and KYSE30 cells were seeded in 96-well plates and treated with or without 40 µg/ml 5-FU or 6 µg/ml Cis and different concentrations (0, 2, 4 and 6 µmol/l) of α-solanine for 48 h; the Blank group was treated with PBS containing 1‰ DMSO. Then, the cells were lysed and incubated with 100 µl of Apo-ONE Caspase-3/7 reagent (substrate and buffer in the ratio of 1:100). After a one hour incubation in the dark at room temperature (RT), the fluorescence of each well was measured at 485–520 nm by reading in an Epoch microplate reader (Biotek Instruments, Winooski, VT, USA).

EC9706 cells (5×10⁶), which were labeled with firefly luciferase, were subcutaneously inoculated into the armpit of the right forelimb of 6-week-old female BALB/c nude mice purchased from Vital River Laboratory Animal Technology Corp. (Beijing, China). After one week, the mice were randomly divided into six groups (5 mice per group): Blank group, α-solanine group, 5-FU group, Cis group, α-solanine and 5-FU group, α-solanine and Cis group. Blank group was given normal saline weekly and 1‰ DMSO every two days for 4 weeks. α-solanine was administered to the mice at a dosage of 3.5 mg/kg every other day, which lasted 4 weeks. 5-FU was administered at a dosage of 20 mg/kg weekly for 4 weeks. Cis was given to the mice at a dosage of 3 mg/kg weekly for 4 weeks. Before the mice were anesthetized with isoflurane, an aqueous solution of luciferin (150 mg/kg) was intraperitoneally injected 10 min prior to imaging. The mice were placed into the light-tight chamber of a CCD camera system (Xenogen). The luminescent area of the xenograft tumor was defined as the region of interest (ROI) and the total signal in the ROI (photon/sec/m²) was quantified using Living Image software 3D (Xenogen). All procedures involving mice were performed in compliance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health).

Paraffin was washed away with xylene from tumor tissue sections of 4 µm. The sections were then rehydrated in graded ethanol, and pretreated in an electric waterbath at 100°C with 0.01 M citric acid buffer (pH 6.0) for 20 min for antigen retrieval. The sections were pre-incubated with 3% hydrogen peroxide in deionized water for 10 min for inactivation of endogenous peroxidase, after being washed 3 times with phosphate-buffered saline (PBS) solution and then blocked with 5% normal goat serum for 20 min. The specimens were then incubated with rabbit anti-Ki67 nuclear antigen monoclonal antibody (Zhong Shan Golden Bridge Biotechnology Co., Ltd., Beijing, China) at a dilution of 1:100. Incubation was then carried out at 4°C in a refrigerator overnight. After being washed with PBS three times, the sections were incubated with biotin-conjugated secondary antibody anti-rabbit IgG (Zhong Shan Golden Bridge Biotechnology Co., Ltd.) for 30 min. The sections were washed thrice with PBS followed by the application of biotin-streptavidin HRP for 30 min. The bound complexes were visualized by the application of a 0.05% diaminobenzidine (DAB) solution and counterstained with Harris hematoxylin. The immune-histochemical expression of Ki67 in the tumor tissues was semi-quantitatively scored on a five-point scale (22). The Ki67 staining index was scored by counting positive cells. Scoring was performed and reviewed by two experienced investigators and average scores were calculated.

Relative levels of miR-138 and survivin mRNA in the EC9706 and KYSE30 cells treated with different reagents or reagent combination were determined by quantitative real-time PCR (qRT-PCR). Total RNA was extracted using the Total RNA Kit I (R6834-01; Omega Bio-Tek, Inc., Norcross, GA, USA) from the aforementioned cells. Determination of relative levels of miRNA-138 was performed by using a High-Specificity miR-138 qRT-PCR Detection Kit (Stratagene Corp. La Jolla, CA, USA) in conjunction with an ABI 7500 thermal cycler (Applied Bio-systems, USA). All protocols were performed according to the manufacturer's instructions. We used U6 small nuclear RNA (U6 snRNA) for normalization. To determine the relative levels of survivin mRNA, the qRT-PCR assays were performed using ABI TaqMan PCR Master Mix (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA). GAPDH was used as an internal control. The corresponding CT values were recorded with ABI 7500 software, and then the relative expression levels were calculated according to the formula 2^−∆∆Ct.

Total proteins from the transfected cells were extracted using RIPA buffer containing phenylmethanesulfonyl fluoride (PMSF). A BCA protein assay kit (Beyotime, Haimen, China) was used to determine the protein concentrations. Proteins (40 µg) were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinyl difluoride (PVDF) membranes. After blocking, the membranes were incubated overnight at 4°C with diluted (1:1,000) primary antibody (polyclonal rabbit anti-survivin; Santa Cruz Biotechnology, Inc., Dallas, TX, USA). Following extensive washes, the membranes were incubated with diluted (1:5,000) horseradish peroxidase conjugated goat anti-rabbit lgG (Santa Cruz Biotechnology, Inc.). Signals were determined using DAB detection kit (Amersham Pharmacia Biotech, Piscataway, NJ, USA). GAPDH (Santa Cruz Biotechnology, Inc.) served as the endogenous reference.

The human survivin 3′ untranslated region (3′UTR) fragment containing putative binding sites for miR-138 were amplified by PCR from human genomic DNA. The mutant survivin 3′UTRs were obtained by overlap extension PCR. The fragments were cloned into a pmirGLO-M-survivin. For the luciferase reporter assay, the EC9706 cell line was transiently co-transfected with miRNA (miR-138 agomir or scrambled-miR-138 negative control) and reporter vectors (wild-type reporter vectors or mutant-type reporter vectors), using Lipofectamine^® 2000. Luciferase activities were measured using a Dual-Luciferase assay kit (Promega) according to the manufacturer's instructions at 48 h post-transfection.

Statistical testing was conducted with the assistance of SPSS 17.0 software (SPSS, Inc., Chicago IL, USA). All data are expressed as means ± standard deviation (SD). One-way ANOVA and LSD tests were used to analyze data. Results were considered significantly significant at P-values <0.05.

We found that the treatment of different concentrations of α-solanine (2, 4 and 6 µmol/l) for 24 h did not alter the viability of both cell lines. The results revealed that the concentrations of α-solanine used in the experiments in vitro did not cause obvious cytotoxicity to the EC9706 and KYSE30 cells. The results of CCK-8 assay showed that α-solanine enhanced the cytotoxic effect of 5-FU and Cis on EC9706 and KYSE30 cells. As depicted in Fig. 1, in both cell lines, the semi-lethal concentration (LC₅₀) of 5-FU and Cis were nearly 30 and 4 µg/ml, respectively. However, when 4 µmol/l α-solanine was added, the LC₅₀ of 5-FU and Cis reduced to nearly 20 and 2 µg/ml, respectively, in the EC9706 and KYSE30 cells (P<0.05).

We studied the effects of α-solanine on the apoptosis of EC9706 and KYSE30 cells treated with 5-FU or Cis. From the flow cytometric assay results, we found that when the cells were treated with 40 µg/ml 5-FU, the apoptosis rate of the EC9706 and KYSE30 cells ranged from 20 to 25%, and following treatment with 4 µg/ml Cis, it was near 25%. However, when different concentrations of α-solanine (2, 4 and 6 µmol/l) were added, the apoptosis rates of both cell lines were significantly increased to a range of 35 to 40% (Fig. 2A and B, P<0.05). We also found that α-solanine and 5-FU or Cis can activate caspase-3/7 more effectively than 5-FU or Cis separately in EC9706 and KYSE30 cells (Fig. 2C and D, P<0.05). These results indicated that α-solanine may exert its function as chemosensitivity enhancer through apoptosis pathway.

To further detect the chemosensitivity enhancing effect of α-solanine on EC9706 cells, we performed a nude mouse experiment and immunohistochemical test. The bioluminescence signal was relatively weaker in the group treated with 5-FU or Cis and α-solanine than groups that were treated with 5-FU or Cis separately (Fig. 3A and B, P<0.05). However, there were no significant differences between the group that was treated with normal saline and 1‰ DMSO and that with α-solanine in regards to the bioluminescence signal. The immunohistochemistry and the semi-quantitative scoring results showed that α-solanine enhanced the proliferation inhibition effect of 5-FU or Cis in the EC9706 cells (Fig. 3C-E).

We performed qRT-PCR and western blot analysis tests to detect the expression of miR-138 and survivin in EC9706 and KYSE30 cells after treatment with different concentrations of α-solanine (0, 2, 4 and 6 µmol/l). Expression of miR-138 had an obvious upregulation in EC9706 and KYSE30 cells after treatment with α-solanine (2, 4 and 6 µmol/l) compared with that of 0 µmol/l α-solanine (Fig. 4A, P<0.05). The results of qRT-PCR revealed that survivin mRNA expression was significantly downregulated in the EC9706 and KYSE30 cells after treatment with α-solanine (2, 4 and 6 µmol/l) compared with that of 0 µmol/l α-solanine (Fig. 4B, P<0.05). Results of western blotting showed a significant downregulation of survivin protein expression in the EC9706 and KYSE30 cells treated with α-solanine (2, 4 and 6 µmol/l) when compared with that of 0 µmol/l α-solanine (Fig. 4C and D, P<0.05). These results demonstrated that α-solanine can induce the upregulation of miR-138 and downregulation of survivin protein expression in EC9706 and KYSE30 cells.

Bioinformatics analyses using TargetScan and miRanda predicted that the 3′UTR of survivin contains binding sites for miR-138 (Fig. 5A). The Dual-Luciferase reporter assay was performed with systems containing wild-type (pmirGLO-Wt-survivin) or mutant-type (pmirGLO-Mt-survivin) 3′UTR of survivin, respectively. Co-transfection with the miR-138 mimic significantly suppressed the luciferase activity of the reporter containing the wild-type 3′UTR of survivin (Fig. 5B, P<0.05). These results indicated that miR-138 negatively regulates survivin expression by directly binding on the sequence in the 3′UTR of its mRNA.

To validate the relationship between miR-138 and survivin, we transfected miR-138 mimic, si-survivin and miR-138 inhibitor into EC9706 and KYSE30 cells separately and treated them with 40 µg/ml 5-FU or 6 µg/ml Cis and/or 4 µmol/l α-solanine. Western blotting results showed that survivin expression levels in the α-solanine groups (cells treated with 4 µmol/l α-solanine), miR-138 groups (cells transfected with miR-138 mimics) and si-survivin groups (cells transfected with siRNA-survivin) were significantly suppressed than the Blank group (cells treated with 1‰ DMSO). However, enhanced survivin expression was found in the miR-138 inhibitor group (cells transfected with miR-138 inhibitor and treated with 4 µmol/l α-solanine) when compared with the other groups except the Blank group (Fig. 6A and B, P<0.05). Then we conducted CCK-8 assay to study the effects of α-solanine, miR-138, survivin and miR-138 inhibitor on the chemosensitivity of EC9706 and KYSE30 cells to 5-FU and Cis. CCK-8 assay revealed that the viability of EC9706 and KYSE30 cells were significantly inhibited after treatment with 5-FU (40 µg/ml) or Cis (6 µg/ml) for 24 h in the α-solanine, miR-138 and si-survivin groups compared with the blank groups; in the miR-138 inhibitor group, the change was slight (Fig. 6C and D, P<0.05). These results indicated that both the upregulation of miR-138 and the knockdown of survivin enhanced the chemosensitivity of 5-FU and Cis in the EC9706 and KYSE30 cells. These results also showed that inhibition of miR-138 counteracted the function of α-solanine to enhance the chemosensitivity of 5-FU and Cis in the EC9706 and KYSE30 cells. In short, α-solanine enhanced the chemosensitivity of 5-FU and Cis in EC9706 and KYSE30 cells by upregulation of miR-138 and downregulation of survivin.

To further validate the relationship between α-solanine, miR-138 and survivin on the effect of 5-Fu and Cis in EC9706 cells, we performed western blotting, CCK-8 and flow cytometric assays. The western blot assay was performed to detect survivin expression in cells after being treated with 5-FU or Cis and being simultaneously or separately exposed to α-solanine (4 µmol/l), miR-138 mimics and the vector containing survivin but lacking the 3′UTR sequence (pcDNA3.1-survivin). The results showed that expression of survivin was significantly suppressed in the cells transfected with miR-138 mimics or treated with α-solanine. However, compared with the baseline level, the expression of survivin, regardless of miR-138 and α-solanine, was significantly enhanced in cells transfected with pcDNA3.1-survivin in the 5-FU and Cis groups (Fig. 7A and B, P<0.05). Results of the CCK-8 assay and flow cytometric assay showed that the presence of the miR-138 mimic or α-solanine significantly promoted the apoptosis of EC9706 cells when treated with 5-FU (40 µg/ml) or Cis (6 µg/ml). Similarly, overexpression of survivin offset the effects of miR-138 or α-solanine on the cells (Fig. 7C and D, P<0.05). These results further suggest that expression of survivin counteracted the effect of α-solanine and miR-138 regarding the chemosensitivity of EC9706 cells.