Epac1 links prostaglandin E2 to β-catenin-dependent transcription during epithelial-to-mesenchymal transition

Abstract

In epithelial cells, β-catenin is localized at cell-cell junctions where it stabilizes adherens junctions. When these junctions are disrupted, β-catenin can translocate to the nucleus where it functions as a transcriptional cofactor. Recent research has indicated that PGE2 enhances the nuclear function of β-catenin through cyclic AMP. Here, we aim to study the role of the cyclic AMP effector Epac in β-catenin activation by PGE2 in non-small cell lung carcinoma cells. We show that PGE2 induces a down-regulation of E-cadherin, promotes cell migration and enhances β-catenin translocation to the nucleus. This results in β-catenin-dependent gene transcription. We also observed increased expression of Epac1. Inhibition of Epac1 activity using the CE3F4 compound or Epac1 siRNA abolished the effects of PGE2 on β-catenin. Further, we observed that Epac1 and β-catenin associate together. Expression of an Epac1 mutant with a deletion in the nuclear pore localization sequence prevents this association. Furthermore, the scaffold protein Ezrin was shown to be required to link Epac1 to β-catenin. This study indicates a novel role for Epac1 in PGE2-induced EMT and subsequent activation of β-catenin.

Keywords: EMT; Epac; Ezrin; PGE2; β-catenin.

Figure 1. Effect of PGE₂ on EMT in A549 cells

A. Gene expression of E-cadherin and ZEB1 following 18 hours stimulation with PGE₂ (10 μg/ml). B. Representative western blot image of E-cadherin expression in a subconfluent culture of A549 cells stimulated for 18 hours with PGE₂. C. Immunofluorescence images of E-cadherin after18 hours stimulation with PGE₂. The white line indicates the migrating border in a scratch wound assay. White arrows in indicate areas of cell-cell contact, which are decreased in cells on the migrating border in the right image. Scale bar represents 20 μm. D. Quantification of E-cadherin expression in migrating border cells and cells incorporated in an epithelial sheet. Each points represents the average integrated density value (IDV) of 20 cells. E. Immunofluorescence images of N-cadherin/E-cadherin and Vimentin/E-cadherin after18 hour stimulation with PGE₂. Scale bar represents 20 μm. Data represent mean ± SEM of 5 separate experiments. # p < 0.05, ## p < 0.01 compared to DMSO treated cells. *** p < 0.001 between the indicated groups. F. Quantification of N-cadherin and Vimentin expression in cells treated for 18h with DMSO or PGE₂.

Figure 2. Effect of PGE₂ on β-catenin nuclear translocation and transcriptional activity in A549 cells

A. Immunofluorescence images of β-catenin after 18 hours stimulation with PGE₂. White arrows indicate areas of cell-cell contact. Presence of β-catenin at these areas is decreased in PGE₂-treated cells, whereas nuclear localization is increased. Scale bar represents 40 μm. B. Quantification of β-catenin nuclear localization. Each points represents the average integrated density value (IDV) of 20 cells. C. TCF luciferase gene reporter assay of β-catenin transcriptional activity (TOPFlash) after 18 hours stimulation with PGE₂. D. TOPFlash assay of cells co-incubated with antagonists of the EP₄ receptor (L161,982; 3 μM), the EP₁ and EP₂ receptor (AH6809; 10 μM) or PI3 kinase (LY29004, 50 μM). E. Measurement of cyclic AMP production. Cells were co-treated with IBMX to inhibit phosphodiesterase activity. Forskolin (10 μM) was used as a positive control for cyclic AMP production by adenylyl cyclase. Data represent mean ± SEM of 5-9 separate experiments. # p < 0.05, ## p < 0.01, ### p < 0.001 compared to DMSO treated cells. * p < 0.05, ** p < 0.01 *** p < 0.001 between the indicated groups.

Figure 3. Role of Epac1 in PGE₂-induced β-catenin nuclear translocation in A549 cells

A. Gene expression of Epac1, but not Epac2, is increased by PGE_2. B. Quantification of β-catenin nuclear localization. Each points represents the average integrated density value (IDV) of 20 cells. C. Immunofluorescence images of β-catenin in cells treated with PGE₂. Co-incubation with a specific Epac1 antagonist (CE3F4, 20 μM) abolished PGE₂-induced β-catenin nuclear translocation. Scale bar represents 40 μm. D. TOPFlash assay of cells co-incubated with specific antagonists for Epac1 (CE3F4) and Epac2 (ESI-05; 10 μM). E. PGE₂-induced gene expression of ZEB1 is attenuated by co-incubation with CE3F4. Data represent mean ± SEM of 5-9 separate experiments. # p < 0.05, ### p < 0.001 compared to DMSO treated cells. *** p < 0.001 between the indicated groups.

Figure 4. Epac1 knockdown prevents PGE₂-induced β-catenin transcriptional activity in A549 cells

A. TOPFlash assay of cells transfected with non-targeting siRNA or Epac1 siRNA for 48h and subsequent PGE₂ treatment. B. Gene expression of Epac1 and ZEB1 in cells transfected with Epac1 siRNA in combination with PGE₂ treatment. C. Knockdown of Epac1 protein in Epac1 siRNA transfected cells. Data represent mean ± SEM of 5-9 separate experiments. ## p < 0.01, ### p < 0.001 compared to non-targeting siRNA transfected cells. ** p < 0.01, *** p < 0.001 between the indicated groups.

Figure 5. Role of Epac1 in PGE₂-induced cell migration of A549 cells

A. Representative images of a wound healing assay 24 hours post scratch. The black lines indicate borders of scratch on 0 hours. The white lines indicate borders of scratch on 24 hours of DMSO treated cells. B. Quantification of wound closure of PGE₂ treated cells in co-incubation with the Epac1 antagonist CE3F4 or in Epac1 silenced cells. C. xCELLigence assay of real-time cell migration of cells stimulated with PGE₂ (left panel) or in co-incubation with the Epac1 antagonist CE3F4. x-Axis indicates time in hours. Data represent mean ± SEM of 6 separate experiments. ## p < 0.01 compared to DMSO treated cells. * p < 0.05 between the indicated groups.

Figure 6. Expression of a mutant Epac1 with a deletion of the 764-838 domain has aberrant localization and prevents PGE2-induced β-catenin transcriptional activity in A549 cells

A. Immunofluorescence images of FLAG-tagged Epac1 wildtype (WT) and Epac1 Δ764-838. White arrows indicate absence of nuclear localization for the Epac1 deletion mutant. Scale bar represents 40 μm. B. TOPFlash assay of cells transfected with Epac1 Δ764-838 and subsequent PGE₂ treatment. C. Western blot image confirming successful expression of FLAG-tagged Epac1 Δ764-838. D. Western blot images of a FLAG immunoprecipitation showing co-immunoprecipitation of Epac1 and β-catenin in FLAG-Epac1 WT transfected cells, which is attenuated in and FLAG-Epac Δ764-838 transfected cells. For input we used 5% of the amount of protein used for co-immunoprecipitation. E. Quantification of β-catenin co-immunoprecipitation with FLAG immunoprecipitation. β-Catenin was normalized for the amount of FLAG immunoprecipitated. Data represent mean ± SEM of 3-9 separate experiments. # p < 0.05, ### p < 0.001 compared to empty vector transfected cells. ^§ p < 0.05 compared to FLAG-Epac1 WT transfected cells. *** p < 0.001 between the indicated groups.

Figure 7. Ezrin knockdown prevents PGE₂-induced β-catenin transcriptional activity in A549 cells

A. Knockdown of Ezrin protein in Ezrin siRNA transfected cells. B. TOPFlash assay of cells transfected with non-targeting siRNA or Ezrin siRNA for 48 hours and subsequent PGE₂ treatment. C. FLAG co-immunoprecipitation of Epac1, β-catenin and Ezrin in FLAG-Epac1 WT transfected and Ezrin silenced cells. For input we used 5% of the amount of protein used for co-immunoprecipitation. D. Quantification of β-catenin and Ezrin co-immunoprecipitation with FLAG immunoprecipitation. β-Catenin and Ezrin were normalized for the amount of FLAG immunoprecipitated. Data represent mean ± SEM of 3-9 separate experiments. ## p < 0.01 compared to non-targeting siRNA transfected cells. ^§ p< 0.05, ^§§ p < 0.01 compared to scrambled siRNA transfected cells. *** p < 0.001 between the indicated groups.