Table 1.
Descriptive characteristics of NPC patients and normal controls.
Figure 1.
Expression of p-GSK3β (Ser9) and EZH2 in NPC tissues and normal controls.
(A) Representative immunohistochemical staining results of EZH2 and p-GSK3β (Ser9) in NPC and control tissues; (B) The mean number of EZH2 positive cells is greater in NPC tissues than in normal controls; (C) The mean number of p-GSK3β positive cells is greater in NPC tissues than in normal controls; (D) Coincident high or low EZH2 and p-GSK3β (Ser9) immunohistochemical staining in 2 representative NPC tissues; (E) Immunoreactivity of EZH2 is positively associated with p-GSK3β (Ser9) immunoreactivity in NPC tissues; (F) Immunoreactivity of EZH2 is positively associated with higher stage of NPC.
Figure 2.
Evidence that GSK3β is able to bind to EZH2 in NPC cell lines.
(A) Schematic diagram of the putative GSK3β phosphorylation motif sequence alignment in EZH2; (B) Evidence that GSK3β is able to bind to EZH2, as determined by immunoprecipitation and immune blotting. Lysates from CNE-1 and CNE-2 cells were used for immunoprecipitation. IB: immune blot; IP: immunoprecipitation.
Figure 3.
Inactivation of GSK3β upregulated EZH2 production in NPC cells.
(A) The transfection efficiency was evaluated by testing the protein level of GSK3β. Representative western blot analysis of p-GSK3β (Ser9) after GSK3β-CA or KD transfection or lithium treatment (20 mmol/L) in NPC cells; (B) GSK3β-CA transfection (2 μg/mL) significantly reduced p-GSK3β (Ser9) production in CNE-1 and CNE-2 cells, whereas GSK3β-KD transfection (2 μg/mL) or lithium treatment (20 mmol/L) significantly increased p-GSK3β (Ser9) production in CNE-1 and CNE-2 cells; (C) Representative western blot analysis of EZH2 after GSK3β-CA or KD transfection in NPC cells; (D) GSK3β-CA transfection (2 μg/mL) significantly reduced EZH2 production in CNE-1 and CNE-2 cells, whereas GSK3β-KD transfection (2 μg/mL) significantly increased EZH2 production in CNE-1 and CNE-2 cells; (E) Representative western blot analysis of EZH2 after lithium treatment (20 mmol/L) in NPC cells; (F) Lithium treatment (20 mmol/L) significantly increased EZH2 production in CNE-1 and CNE-2 cells. The data indicate the means (SEM) of 3 independent experiments. NC: normal control; CA: constitutively active GSK-3β plasmid; KD: kinase-dead GSK-3β plasmid.
Figure 4.
GSK3β inactivation promoted the stability of EZH2 protein in vitro.
(A, B) Representative western blot analysis of EZH2 after GSK3β-KD (2 μg/mL) or control plasmid transfection in CNE-1 cells; CNE-1 cells were treated with cycloheximide (20 μM) after transfection, and EZH2 protein level in the indicated time point was detected by western blot analysis which containing equal amounts of protein. (C) The half-life of EZH2, as suggested by the relative EZH2 intensity, was significantly longer in GSK3β-KD group than in normal control (p<0.05). The data indicate the means (SEM) of 3 independent experiments. KD: kinase-dead GSK3β plasmid; NC, normal control plasmid.
Figure 5.
Inhibition of GSK-3β activity enhanced migration of NPC cell lines.
(A) Representative images showing the cell covered area on the culture plate containing NPC cells after transfection with GSK3β plasmid. Inhibition of GSK-3β by GSK3β-KD transfection enhanced migration of CNE-1 and CNE-2 cells, whereas activation of GSK-3β by GSK3β-CA transfection suppressed migration of CNE-1 and CNE-2 cells; (B) Quantitative analyses for the cell covered areas showed that the migrated cells in the GSK3β-KD group increased significantly, whereas those in GSK3β-CA group decreased significantly when compared to the control. Migration of NPC cells was evaluated by scratch assay after GSK3β-KD or CA plasmid (2 μg/mL) transfection for 24 or 48 h. The data indicate the means (SEM) of 3 independent experiments. NC: normal control; CA: constitutively active GSK3β plasmid; KD: kinase-dead GSK3β plasmid.
Figure 6.
Inhibition of GSK-3β activity enhanced invasion of NPC cells.
(A) Representative photos showing the NPC cell density on the filter after transfection with GSK3β plasmid. Inhibition of GSK-3β by GSK3β-KD transfection enhanced invasion of CNE-1 and CNE-2 cells, whereas activation of GSK-3β by GSK3β-CA transfection suppressed invasion of CNE-1 and CNE-2 cells; (B) Quantitative analyses of the number of invaded cells showed that the invaded cells in GSK3β-KD group increased significantly, whereas those in GSK3β-CA group decreased significantly when compared to the control. Invasion of NPC cells was evaluated by transwell assay after transfection with GSK3β-KD or CA plasmid (2 μg/mL) for 72 h. The data indicate the means (SEM) of 3 independent experiments. NC, normal control; CA: constitutively active GSK-3β plasmid; KD: kinase-dead GSK-3β plasmid.
Figure 7.
GSK3β-enhanced migration and invasion of NPC cells were abrogated by EZH2 siRNA transfection.
(A–C) The siRNA knockdown efficiency was evaluated by testing the protein level of EZH2. EZH2 siRNA transfection significantly reduced EZH2 expression in CNE-1 and CNE-2 cells; (D,E) Inhibition of EZH2 by EZH2 siRNA (50 nmol/L) transfection significantly inhibited migration of CNE-1 and CNE-2; (F,G) EZH2 siRNA (50 nmol/L) transfection significantly inhibited GSK3β-KD-dependent migration of CNE-1 and CNE-2; (H,I) Inhibition of EZH2 by EZH2 siRNA (50 nmol/L) transfection significantly inhibited invasion of CNE-1 and CNE-2; (J,K) EZH2 siRNA transfection significantly inhibited GSK3β-KD-dependent invasion of CNE-1 and CNE-2. Migration and invasion of NPC cells were evaluated after EZH2 siRNA (50 nmol/L) and/or GSK3β-KD plasmid (2 μg/mL) were transfected for 48 or 72 h. The data indicate the means (SEM) of 3 independent experiments. NC: normal control; KD: kinase-dead GSK-3β plasmid.