T24 and 5637 cell lines (Shanghai Cell Bank, Chinese Academy of Sciences, Shanghai, China) were cultured in RPMI-1640 medium (Beijing Solarbio Science & Technology, Beijing, China), supplemented with 10% fetal bovine serum (FBS; Beijing TransGen Biotech Co., Ltd., Beijing, China) and the J82 cell line (Shanghai Cell Bank, Chinese Academy of Sciences) was cultured in minimum essential medium (Beijing Solarbio Science & Technology), supplemented with 10% FBS. All cells were maintained at 37°C in 5% CO₂. When cells were grown to 90% confluence, the 5 μM γ-secretase inhibitor (GSI; cat. no. #565750; Calbiochem, Merck, Darmstadt, Germany) was added for 24 or 48 h (only for western blot).

Total RNA was extracted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA), according to the manufacturer’s instructions. The RNA quality was measured by distinct 18S, 28S and total RNA separated by electrophoresis in agarose gels and the proper ratio of absorbance at 230/280 nm (UV-8000 Double Beam UV/VIS model; Metash Instruments, Shanghai, China).

Total proteins were extracted by radioimmunoprecipitation assay cell lysis reagent supplemented with proteinase and phosphatase inhibitors (Beijing Solarbio Science & Technology) at 4°C for 30 min. The cell extracts were centrifuged at 12,000 ×g for 20 min at 4°C, and the supernatants containing total proteins were mixed with an equal volume of 5X SDS loading buffer (Beijing Solarbio Science & Technology). The samples were heated to 95°C for 5 min.

Total RNA (1 μg) was used for RT using a Reverse Transcription system (PrimeScript™ RT Reagent kit; Takara Bio, Inc., Shiga, Japan), according to manufacturer’s instructions. RT-qPCR was performed using SYBR® Premix Ex TaqTM II (TliRNase H Plus; Takara Bio, Inc.) on the ABI PRISM® 7500 real-time PCR system (Applied Biosystems, Life Technologies, Thermo Fisher Scientific, Waltham, MA, USA), with β-actin as an internal control. Thermocycling was performed in a final volume of 20 μl consisting of 40 cycles at 95°C for 5 sec then 55°C for 30 sec, following an initial denaturation step at 95°C for 10 sec. The sequences of the PCR primers of E-cadherin, N-cadherin, vimentin and smooth muscle actin are listed in Table I. The results were analyzed using the 2^−ΔΔCt method. ΔCt = average gene Ct-Δ average of β-actin Ct. ΔΔCt was calculated as follows: ΔΔCt = ΔCt of sample group-ΔΔCt of control group.

The protein concentrations were quantified using a Micro Bicinchoninic Acid Protein Assay kit (Pierce Biotechnology, Inc., Rockford, IL, USA). Identical quantities of total protein from each sample (30 μg) were loaded and separated by 12% SDS-PAGE and transferred onto a polyvinylidene fluoride membrane. The membranes were incubated with the following primary antibodies at 4°C overnight: Anti-N-cadherin (1:800; cat. no. ab18203; Abcam, Cambridge, MA, USA), anti-vimentin (1:500; cat. no. 3877; Cell Signaling Technology, Inc., Danvers, MA, USA), anti-α-smooth muscle actin (1:1,000; cat. no. P62736; Abgent, Inc., San Diego, CA, USA), anti-E-cadherin (1:800; cat no. 5296; Cell Signaling Technology, Inc.), and anti-β-actin (1:1,500; cat. no. ab6276; Abcam). The membranes were then incubated with horseradish peroxidase-conjugated AffiniPure goat anti-rabbit lgG(H+L) secondary antibody (1:10,000; cat. no. ZF-0316-1; ZSGB-BIO, Beijing, China) or horseradish peroxidase-conjugated AffiniPure goat anti-mouse lgG(H+L) secondary antibody (1:10,000; cat. no. ZF-0315-1; ZSGB-BIO) at room temperature for 2 h. The signals were detected using an enhanced chemiluminescence system (cat. no. 32109; Thermo Fisher Scientific) and the films (cat. no. 6535876; Eastman Kodak Company, Rochester, NY, USA) were scanned and analyzed by AlphaEaseFC version 1.1 software (Alpha Innotech, San Leandro, CA, USA).

Cells in the logarithmic growth phase were subcultured in 96-well plates at a cell density of 5,000 cells/well. In each experiment, five wells corresponded to each control group: Mitoxantrone-treated, zero-adjustment wells and control wells. The mitoxantrone wells contained 1, 4, 16, 64 and 256 ng/ml of drug (Sigma-Aldrich, St. Louis, MO, USA). Following 24 h of drug treatment, 20 μl MTT (Beijing Solarbio Science & Technology) solution was added to each well and the plate was incubated for an additional 4 h, prior to the addition of 150 μl dimethylsulfoxide (Beijing Solarbio Science & Technology) to dissolve the formazan crystals. The absorbance values were measured at 490 nm (Anthos Zenyth 340rt; Biochrom Ltd., Cambridge, UK). The ratio of inhibition was calculated using the following formula: Inhibition ratio = (1 − Ab_experiment / Ab_control) × 100%.

Transwell plates were coated with 40 μl 50 mg/l Matrigel (BD Biosciences, Franklin Lakes, NJ, USA). Cells (300 μl) in the logarithmic growth phase were added to the Transwell plates at a final cell density of 1×10⁵ cells/ml. Culture media, supplemented with 20% FBS (800 μl), was added to each well. The plate was incubated at 37°C and 5% CO₂ for 24 h. The supernatants were subsequently removed from the Transwells, followed by rinsing, fixation with 2 ml 3.7% paraformaldehyde and staining with 2 ml 0.05% crystal violet (Beijing Solarbio Science & Technology). The inner cells were removed using a cotton swab and the number of cells, which were below the microporous membrane, were counted using a microscope (XL-71; Olympus Corporation, Tokyo, Japan). The average cell number from 10 randomly selected fields (magnification, ×200) was calculated.

Values are expressed as the mean ± standard error and analyzed using SPSS 12.0 (SPSS, Inc., Chicago, IL, USA). Western blot images were semi-quantitatively analyzed using ImageJ version 2.1 software (National Institutes of Health, Bethesda, MD, USA). The means of two samples were analyzed using the independent sample t-test. P<0.05 was considered to indicate a statistically significant difference between values.

In order to determine whether the GSI caused morphological changes in bladder cancer cells, T24, 5637 and J82 cells were examined microscopically to assess cell growth and morphological changes following treatment with GSI for 24 h (Fig. 1). Treatment with GSI significantly reduced the cell density. Intercellular junction formation was also reduced, with the majority of cells exhibiting a shrunken appearance and underwent apoptosis.

Following treatment with the GSI to inhibit the Notch signaling pathway, the mRNA expression levels of mesenchymal biomarkers, including N-cadherin, vimentin and α-smooth muscle actin, were downregulated compared with those in the control group and there was a significant upregulation of the mRNA expression of E-cadherin (Fig. 2; P<0.05).

Western blot analysis demonstrated that the protein expression levels of N-cadherin, vimentin and α-smooth muscle actin were significantly downregulated following treatment with GSI compared with those in the control group, whereas the protein expression of of E-cadherin was upregulated (Fig. 3).

This observed downregulation of mesenchymal biomarkers and upregulation of epithelial biomarkers indicated that inhibition of Notch signaling inhibited the EMT.

Following treatment with the GSI (5 μM) for 48 h, the T24, 5637 and J82 bladder cancer cells were cultured with different concentrations of the bladder cancer therapeutic mitoxantrone and the cell viability was detected using an MTT assay. The results revealed that the GSI-treated cells exhibited an increased sensitivity to mitoxantrone-induced cytotoxicity compared with that of the control (untreated) cells (Fig. 4; P<0.05).

Similarly, a Transwell assay demonstrated that the GSI-treated cells were more sensitive to mitoxantrone compared with the untreated cells, as indicated by decreased invasiveness. Treatment with GSI reduced the number of cells which passed through the Transwell membrane, which indicated a deficit in the invasive capacity of those cells which lacked Notch signaling (Fig. 5). Therefore, the present study demonstrated that inhibiting the Notch signaling pathway inhibited the invasion of bladder cancer cells.