Human ESCC cell lines (EC109 and TE1) were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). ESCC cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin mix (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany). Cells were cultured at 37°C in a humidified atmosphere containing 5% CO₂. Cisplatin (MedChemExpress, Monmouth Junction, NJ, USA) was dissolved in dimethyl sulfoxide (DMSO; Beyotime Institute of Biotechnology, Shanghai, China) at 50 mM and stored at −80°C until use.

Using a 96-well plate, 1×10⁴ stably transfected cells in 100 µl of complete growth medium were seeded 1 day prior to treatment. Cells were treated with cisplatin (1, 2, 4, 8, 16 and 32 µM) dissolved in complete growth medium containing <1% DMSO and control cells were treated with complete growth medium containing the same concentration of DMSO.

BMI1 short hairpin RNA (shRNA) was designed and cloned into the pcDNA3.1-EGFP vector with the neomycin resistant gene. The BMI1 shRNA target sequence was as follows: 5′-GGTCATCAGCAACTTCTTCT-3′. Mel18 shRNA was designed and cloned into the psi-LVRU6GP vector with the puromycin resistant gene. The Mel18 shRNA target sequence was as follows: 5′-GGCTCTGAGTGATGATGAGAT-3′. Human full-length Mel18 (reference sequence no. NM_007144.2) was isolated from the human complementary DNA library and connected to the pEZ-M13 vector with the neomycin resistant gene. All plasmids, including Negative control (NC) shRNA, were purchased from GenePharma (Shanghai, China). Transfections of all vectors were performed using Lipofectamine^® 2000 (Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. The mass/concentration of all plasmid transfected was 2 mg/ml. Stably transfected TE1 and EC109 cells were selected and maintained in DMEM containing 700 µg/ml G418 (Beyotime Institute of Biotechnology) or 1 µg/ml puromycin (Beyotime Institute of Biotechnology). After 2 weeks, stable transfected cells were used for subsequent experiments. Transfection efficiency was evaluated by western blot analysis.

The cytotoxic effects of cisplatin in ESCC cells were measured using Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Kumamoto, Japan). Stably transfected cells were plated in 96-well plates at a density of 1×10⁴ cells/well. Following incubation at 37°C for 24 h, cells were treated with various concentrations of cisplatin (1, 2, 4, 8, 16 and 32 µM) at 37°C for 24 h. Following 24 h, the complete growth medium was replaced with serum-free medium, 10 µl CCK-8 was added to each well and the cell mixtures were incubated at 37°C for 2 h. Using a fine needle the bubbles were punctured and the absorbance was then measured at 450 nm using a microplate reader. The half maximal inhibitory concentration (IC₅₀) was defined as the concentration resulting in a 50% reduction in growth compared with the growth of the control cells. Cell viability was calculated according to the following equation: Cell viability = [A450(drug)-A450(blank)]/[A450(control)-A450(blank)]. Six replicate wells were set up in each group and three independent experiments were performed. The IC₅₀ dose-response curves were plotted with GraphPad Prism Version 7.0 (GraphPad Software, Inc., La Jolla, CA, USA).

Following stable transfection of TE1 and EC109 cells, exponentially growing cells were harvested and placed into 60 mm plates (1×10³ cells/well) and cultured with 5 µM cisplatin at 37°C. The medium was changed every 3 days. After 2–3 weeks, cells were washed in PBS three times, fixed in 4% formalin (Beyotime Institute of Biotechnology) at room temperature for 15 min and stained with 0.1% crystal violet (Beyotime Institute of Biotechnology) at room temperature for 10 min. The number of colonies was counted and analyzed using ImageJ software version 1.0 (National Institutes of Health, Bethesda, MD, USA).

Apoptosis was determined with the Annexin V-APC/7-AAD apoptosis kit (MultiSciences, Hangzhou, Zhejiang, China). ESCC cells were seeded in 6-well plates and incubated with 5 µM cisplatin and serum-free medium at 37°C for 24 h. The cells were digested with 0.25% trypsin without EDTA (Thermo Fisher Scientific, Inc.) and resuspended in binding buffer to a density of 1×10⁶ cells/ml. Annexin V-APC (5 µl) and 7-AAD (10 µl) were mixed prior to incubation in the dark at room temperature for 10 min. Apoptotic cells were detected using a flow cytometer (FACSAriaII; Becton, Dickinson and Company, Franklin Lakes, NJ, USA). FlowJo software (version 10; Becton, Dickinson and Company) was used for data analysis.

Both ESCC cells were seeded in 6-well plates at a density of 2×10⁵ cells/well and treated with cisplatin (5 µM) at 37°C for 24 h. Cells were lysed in radioimmunoprecipitation assay lysis buffer (Beyotime Institute of Biotechnology) supplemented with phenylmethylsulfonyl fluoride (Beyotime Institute of Biotechnology) for 30 min at 4°C and centrifuged at 10,000 × g for 15 min in 4°C. The supernatant was then collected. Tumor tissues were cut into 100 mg pieces, incubated with 500 µl lysis buffer, homogenized completely with a tissue homogenizer for 5 min and lysed for 30 min on ice. Samples were then centrifuged at 10,000 × g for 15 min in 4°C, and supernatant was collected. Protein concentration was determined using bicinchoninic acid assay (Beyotime Institute of Biotechnology). Proteins (30 µg) were separated by 10% SDS-PAGE and transferred onto 0.22 µm polyvinylidene fluoride membranes (EMD Millipore, Billerica, MA, USA). Membranes were blocked with 5% skimmed milk for 2 h at 37°C and incubated with the appropriate primary antibodies at 4°C overnight. The following antibodies were used: Anti-Mel18 (cat. no. ab5267, Abcam, Cambridge, UK), anti-BMI1 (cat. no. 6964; Cell Signaling Technology, Danvers, MA, USA), anti-B-cell lymphoma-2 (Bcl-2) (cat. no. 4223; Cell Signaling Technology), anti-Bcl-2-associated X protein (BAX) (cat. no. 5023; Cell Signaling Technology), anti-caspase3 (cat. no. 9662; Cell Signaling Technology), anti-nuclear factor-κB (NF-κB) (cat. no. 8242; Cell Signaling Technology), anti-c-Myc (cat. no. 5605; Cell Signaling Technology), anti-Akt (cat. no. 4685; Cell Signaling Technology), anti-phosphorylated-Akt (cat. no. 4060; Cell Signaling Technology) and anti-GAPDH (cat. no. 5174; Cell Signaling Technology). Subsequently, membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:5,000; cat. no. ab97200; Abcam) at room temperature for 1 h. Bands were detected using enhanced chemiluminescence substrate (EMD Millipore) and Bio-Rad ChemiDoc MP High-end imaging system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Relative expression level of all proteins was normalized to endogenous control GAPDH using ImageJ version 1.0 (National Institutes of Health, Bethesda, MD, USA).

All animal procedures were approved by the Ethics Committee of Qianfoshan Hospital (Jinan, China). Mice were housed (five mice per cage) under specific pathogen-free conditions, at a constant room temperature of 22–24°C, with a 12-h light/dark cycle and had unlimited access to food and water. The significance of BMI1 and Mel18 inhibition in the sensitization of cisplatin in ESCC in vivo was studied by subcutaneous injection of cancer cells into nude mice. A total of 20 male BALB/c nude mice (3-week old) were purchased from Vital River Laboratories Co., Ltd. (Beijing, China). All 20 nude mice were randomly divided into four groups. Exponentially growing transfected EC109 cells, including Mel18 shRNA and BMI1 shRNA-transfected cells, Mel18 and BMI1 shRNA-transfected cells, BMI1 shRNA-transfected cells, and NC shRNA-transfected cells, were harvested and resuspended in sterile PBS. Equal cell numbers (5×10⁶) of each group were injected subcutaneously in the right flank of BALB/c nude mice. Tumor volume was calculated using the following formula: Tumor volume (mm³) = length × width² ×0.5. When the tumor volume reached approximately 200 mm³, cisplatin was intraperitoneally injected every 3 days with a dose of 5 mg/kg, according to the manufacturer's protocol and a previously published study (35). After 3 weeks of treatment, mice were sacrificed in a chamber with increasing concentrations of carbon dioxide. The tumor volume was measured by a caliper and calculated using the aforementioned formula.

All statistical analysis was performed using SPSS software (version 20; IBM Corp., Armonk, NY, USA). Unless otherwise indicated, data were presented as the means ± standard deviation. An unpaired Student's t-test of two independent samples was used for statistical comparison between two groups. Analysis of variance was performed to compare the mean among multiple groups and Student-Newman-Keuls method was used for pairwise comparison between different treatment groups. P<0.05 was considered to indicate a statistically significant difference.

To study the biological role of BMI1 on cell survival in ESCC with the treatment of cisplatin, two BMI1 shRNA stably transfected cell lines were established. As demonstrated in Fig. 1A, BMI1 expression was markedly lower in the transfected EC109 and TE1 cells compared with the negative control cells. To examine the effects of cisplatin on the survival of BMI1 knockdown cells, a CCK-8 assay was performed following treatment of the cells with cisplatin for 24 h. Following treatment with various concentrations of cisplatin, it was identified that the BMI1 shRNA-transfected cells demonstrated lower cell viabilities compared with the NC shRNA-transfected cells (Fig. 1B). The IC₅₀ values of cisplatin in the BMI shRNA-transfected and the NC-transfected EC109 cells were 4.695±0.287 and 6.953±0.369 µM, respectively (P<0.05). The IC₅₀ values of cisplatin in the BMI shRNA-transfected and the NC-transfected TE1 cells were 4.117±0.192 and 6.376±0.294 µM, respectively (P<0.05; Fig. 1C). These results indicate that BMI1 knockdown increases sensitivity to cisplatin.

To determine the effects of Mel18 and BMI1 on the chemosensitivity of ESCC cells, the present study established four stably transfected ESCC cell lines for each TE1 and EC109 cells, including Mel18 shRNA and BMI1 shRNA-transfected cells, Mel18 and BMI1 shRNA-transfected cells, BMI1 shRNA-transfected cells, and NC shRNA-transfected cells. As presented in Fig. 2A-C, the transfection efficiency was verified by western blot analysis.

Subsequently, a survival assay was performed with the transfected cells treated with different concentrations of cisplatin. As expected, the cell viability was significantly lower for the BMI1 shRNA-transfected cells compared with the NC cells following treatment with cisplatin in EC109 cells (P<0.05; Fig. 2D and E; right panel). A combination of Mel18 and BMI inhibition significantly enhanced the sensitivity to cisplatin compared with BMI inhibition alone in EC109 cells (P<0.05, Fig. 2D and E; right panel). Overexpression of Mel18 combined with BMI1 inhibition did not enhance the sensitivity to cisplatin compared with BMI inhibition alone in EC109 cells (Fig. 2D and E; right panel). Similar results were observed in TE1 cells (Fig. 2D and E; left panel). In comparison with the NC group, the three remaining groups demonstrated significantly lower IC₅₀ values (all P<0.05; Fig. 2E). The EC109 cells with the most significant reduction in IC₅₀ were those in which both Mel18 and BMI1 had been inhibited (P<0.01; Fig. 2E). Similar results were observed in TE1 cells (Fig. 2E).

To further determine the effects of Mel18 and BMI1 inhibition on cisplatin sensitization, the current study investigated the long-term effects of BMI1 and Mel18 inhibition by colony formation assay. It was revealed that depletion of BMI1 increased sensitivity to cisplatin (Fig. 3A and B). Furthermore, depletion of Mel18 and BMI1 in the two ESCC cell lines further increased sensitivity to cisplatin compared with individual depletion of BMI1 (Fig. 3A and B). In summary, the current results indicate that combined inhibition of Mel18 and BMI1 sensitizes ESCC cells to cisplatin. By contrast, overexpression of Mel18 combined with BMI1 inhibition did not markedly reduce cell survival compared with individual knockdown of BMI1.

Cisplatin-induced cytotoxic effects predominantly occur via apoptosis (36). To evaluate the effect of BMI1 and Mel18 knockdown on cisplatin-induced apoptosis, the apoptotic rates of stably transfected ESCC cells were examined following treatment with 5 µM of cisplatin for 24 h. Consistent with the cytotoxic effects observed with the cell viability assay, combined inhibition of BMI1 and Mel18 increased the rate of apoptosis compared with the BMI1 inhibition group (P<0.05) or the NC group (P<0.01), as presented in Fig. 4A and B. Similar results were revealed in the EC109 and TE1 cells. In summary, Mel18 and BMI1 inhibition can prompt an increase in the rate of apoptosis with BMI1 inhibition alone in ESCC cells.

Cisplatin-induced apoptosis predominantly occurs via phosphoinositide 3-kinase/Akt, NF-κB and c-Myc dysregulation (36). Mel18 and BMI1 can regulate proliferation, apoptosis, angiogenesis, tumorigenesis and development via these pathways (28,37,38). To further investigate the mechanism underlying the enhancement of cisplatin-induced apoptosis by the silencing of BMI1 and Mel18, the present study examined the expression levels of proteins associated with these pathways, including total NF-κB, total-Akt, phosphorylated-Akt, c-Myc, caspase-3, BAX and Bcl-2, in the stably transfected cells. The protein expression levels were evaluated by western blot analysis. As presented in Fig. 5, the expression levels of caspase-3, BAX markedly increased and Bcl-2 markedly decreased in cells transfected with Mel18 shRNA and BMI1 shRNA compared with cells transfected with NC or BMI shRNA (all P<0.05), which indicated that BMII1 and Mel18-induced apoptosis may be closely associated with the mitochondrial apoptotic pathway. Inhibition of BMI1 was identified to reduce the expression levels of proteins associated with these signaling pathways to different extents, which is consistent with previous studies (6,7,9). Compared with BMI1 inhibition or NC group, the effect of inhibiting BMI1 and Mel18 on c-Myc was more notable (P<0.05), while the effect on NF-κB and Akt was limited.

The effect of cisplatin treatment in combination with BMI1 and Mel18 inhibition on the growth of esophageal tumors was further determined in vivo. An ESCC xenograft model was established by subcutaneous injection of ESCC cells into the right flank of nude mice. The treatments were initiated once tumor volumes reached ~200 mm³ (Fig. 6A). With the same does of cisplatin, the size of EC109 ×enograft tumors in the BMI1 shRNA group were significantly smaller compared with the NC group (P<0.05; Fig. 6B). In addition, the tumor volume was markedly lower in the BMI1 and Mel18 shRNA group compared with the BMI1 inhibition alone group (P<0.05; Fig. 6B). The expression levels of Mel18 and BMI1 in xenograft tumors were verified by western blot analysis (Fig. 6C). These results suggest that Mel18 inhibition can cooperate with BMI1 inhibition to enhance chemosensitivity of EC109 cells in vivo.