Fisetin and MTT were purchased from Sigma (St. Louis, MO, USA). RPMI-1640 and fetal bovine serum (FBS) were purchased from Gibco Life Technology (Carlsbad, CA, USA). E-cadherin, vimentin, Twist and the β-actin antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Horseradish peroxidase (HRP)-conjugated, rhodamine isothiocyanate (RITC)-conjugated or fluorescein isothiocyanate (FITC)-conjugated secondary antibodies were purchased from Bioworld Technology, Inc. (St. Louis Park, MN, USA). Matrigel and 8 μm pore-sized polycarbonate membranes were purchased from BD Biosciences (Franklin Lakes, NJ, USA). Other reagents were produced in China and purchased from local suppliers. Stock solution of fisetin was made to a concentration of 100 mM in DMSO. The final concentration of DMSO for all treatments was 0.3%.

The NPC cell line, CNE1 (well-differentiated), was obtained from the Institute of Virology, Chinese Academy of Preventive Medicine (Beijing, China). CNE1-LMP1 cells were derived from CNE1 cells transfected with EBV. LMP-1 eukaryotic expression plasmid PAT-LMP1, was established in the Department of Pathology, Guangdong Medical College (Zhanjiang, China). CNE1 cells were maintained in RPMI-1640 supplemented with 10% FBS and penicillin (100 U/ml) and streptomycin (10 μg/ml). CNE1-LMP1 cells were cultured in a similar medium containing 0.5 μg/ml puromycin. Cells were grown at 37°C in a humidified atmosphere of 5% CO₂ and 95% air.

The effect of fisetin on cell viability was assessed using the MTT (3-(4,5-dimethylthiazol2-yl)-2,5-diphenyltetrazolium bromide cell viability assay. Exponentially growing CNE1 and CNE1-LMP1 cells were seeded at 8×10³ cells per well in 96-well culture plates, cultured for 12 h and then treated with a series of concentrations (6.25, 12.5, 25, 50 and 100 μM) of fisetin dissolved in DMSO. Blank control cells were treated with DMSO alone. Incubation was performed at 37°C for 48 h. MTT solution was added to each well (1.0 mg/ml) and incubated for 4 h. The MTT-formazan product was dissolved in DMSO and the absorbance was measured at 490 nm using an ELISA plate reader. Inhibitory effects were assessed using the growth inhibition rate and 50% growth inhibition concentration (IC₅₀) values.

To study the invasive ability of tumor cells, we performed an in vitro invasion assay using a Matrigel invasion chamber. Polycarbonate membranes with a pore size of 8 μm were coated with Matrigel (10 μg/cm²) and incubated overnight. CNE1 and CNE1-LMP1 cells were treated with fisetin (0, 12.5, 25 and 50 μM) for 48 h, then 5×10⁵ cells in RPMI-1640, supplemented with 1% BSA were placed in the upper chamber. The bottom chamber contained RPMI-1640 medium supplemented with 10% FBS which was used as a chemoattractant. The cells were incubated for a further 24 h, after which the inserted membranes were fixed for 20 min in 4% paraformaldehyde and stained with eosin. Subsequently. the membranes were washed with PBS and removed from the inserts. Cells on the upper surface of the membrane were scraped with a cotton swab and the membranes were mounted using glycerol. Invaded cells on the lower surfaces of the membranes were counted within five representative fields in triplicate inserts. A transwell motility assay was performed as described above for the Matrigel invasion chamber, with the exception that the transwell insert was not coated with Matrigel.

At the time of harvest, CEN1 or CEN1-LMP1 cells were seeded onto coverslips, subsequently the cells were treated with fisetin (0, 12.5, 25 and 50 μM) for 48 h. CEN1 and CEN1-LAMP1 cells were washed three times with PBS, fixed in 1:1 acetone/methanol for 5 min, and the cells were incubated overnight with the primary antibodies (E-cadherin, vimentin, Twist, 1:100) at 4°C. The cells were then incubated with FITC-conjugated secondary antibodies or with RITC-conjugated secondary antibodies (1:100) for 45 min. Images of the antigenic sites were captured using a laser scanning confocal microscope.

CNE1 and CNE-LMP1 cells were treated with fisetin (0, 12.5, 25 and 50 μM) for 48 h. The cells were washed twice with ice-cold PBS and lysed in cold lysis buffer. Lysates were incubated for 20 min on ice and centrifuged at 12,000 × g for 15 min, after which time the supernatant was collected. The protein was electrophoresed using sodium dodecyl sulfate polyacrylamide gel electrophoresis and subsequently transferred onto polyvinylidene fluoride membranes. The membranes were blocked with 50 g/l non-fat, dried milk in PBST (PBS supplemented with 0.5 ml/l Tween-20) for 2 h at room temperature and then they were respectively incubated overnight at 4°C with the primary antibodies against human E-cadherin (1:400), vimentin (1:400) or Twist (1:400), followed by incubation with HRP-conjugated secondary antibodies at room temperature for 1 h. Enhanced chemiluminescence (ECL) was used to detect the results.

The statistical analysis was performed with SPSS 13.0 (SPPS Inc., Chicago, IL, USA). Data were expressed as the means ± SD. For the analysis of differences between two groups, the Student’s t-test was performed. For multiple groups, ANOVA was performed followed by the Student-Newman-Keuls test. P<0.05 was considered to indicate a statistically significant difference.

The MTT cell viability assay was performed after the CNE-LMP1 and CEN1 cells were treated for 48 h with fisetin at various concentrations. The results demonstrated that with an increase in the concentration of fisetin, the growth of CNE1-LMP1 and CNE1 cells was significantly decreased (Table I). The IC₅₀ of fisetin to CNE1-LMP1 cells was 86.1 μM, but the IC₅₀ in CNE1 cells was 100.1 μM. Compared with CNE1 cells, fisetin treatment caused a more marked decrease in cell viability in CNE1-LMP1 cells.

Matrigel-coated chambers were used to measure the effect of fisetin on cell invasion. The results demonstrated that invaded CNE1-LMP1 cells decreased dose-dependently following pretreatment with fisetin (Fig. 1A). Compared with the DMSO control, the number of invaded CNE1-LMP1 cells was decreased 5.6-fold following pretreatment with 50 μM fisetin (Fig. 1B). In order to detect the effect of fisetin on cell motility, we performed a transwell motility assay. A similar effect was observed for the motility assay. The number of CNE1-LMP1 cells that migrated through the bottom of the filter was significantly decreased following pretreatment with fisetin (Fig. 1C). Compared with the DMSO control, the number of migrated CNE1-LMP1 cells was decreased 3.1-fold following pretreatment with 50 μM fisetin (Fig. 1D). These results demonstrated that fisetin significantly inhibited the invasion and migration of CNE1-LMP1 cells. Compared with CNE1-LMP1 cells, CNE1 cells exhibited weaker migration and invasion ability. Fisetin did not affect the invasion ability of CNE1 cells at low concentrations, and only exhibited minor inhibition to the motility of CNE1 cells with a concentration of <50 μM (Fig. 1). These results suggest that fisetin is more potent in suppressing the invasion and migration of CNE1-LMP1 cells compared with CNE1 cells.

We examined the levels of epithelial and mesenchymal markers in CNE1-LMP1 and CNE1 cells. Western blot analysis and immunofluorescent staining revealed that compared with CNE1 cells, CNE1-LMP1 cells exhibited a weaker expression of the E-cadherin protein and a more marked expression of the vimentin protein (Figs. 2 and 3). It was demonstrated that LMP1 induced the EMT emergence of CNE1-LMP1 cells, but no EMT occurred in CNE1 cells. Western blot analysis showed that the expression of E-cadherin protein in CNE1-LMP1 cells increased following treatment with fisetin, whereas that of vimentin protein markedly decreased (Fig. 2). Immunofluorescent staining results also confirmed the marked expression of E-cadherin from cell membranes in addition to the weaker vimentin expression in the cytoplasm of CNE1-LMP1 cells treated with fisetin (Fig. 3). These results suggest that fisetin may reverse the molecular changes associated with EMT in CNE1-LMP1 cells.

Western blot analysis and immunofluorescent staining demonstrated that CNE1-LMP1 cells constitutively express Twist, but CNE1 cells exhibit little expression of Twist (Figs. 4 and 5). As shown by western blot analysis, fisetin dose-dependently inhibited the expression of Twist protein expression in CNE1-LMP1 cells, and the inhibitory effect was distinguished at a concentration of <50 μM (Fig. 4). Immunofluorescent staining verified the downregulation of Twist expression in CNE1-LMP1 cells following treatment with fisetin. This was most distinctive following treatment with 50 μM fisetin (Fig. 5).