Prostaglandin E2 alters Wnt-dependent migration and proliferation in neuroectodermal stem cells: implications for autism spectrum disorders

Abstract

Prostaglandin E2 (PGE2) is a natural lipid-derived molecule that is involved in important physiological functions. Abnormal PGE2 signalling has been associated with pathologies of the nervous system. Previous studies provide evidence for the interaction of PGE2 and canonical Wnt signalling pathways in non-neuronal cells. Since the Wnt pathway is crucial in the development and organization of the brain, the main goal of this study is to determine whether collaboration between these pathways exists in neuronal cell types. We report that PGE2 interacts with canonical Wnt signalling through PKA and PI-3K in neuroectodermal (NE-4C) stem cells. We used time-lapse microscopy to determine that PGE2 increases the final distance from origin, path length travelled, and the average speed of migration in Wnt-activated cells. Furthermore, PGE2 alters distinct cellular phenotypes that are characteristic of Wnt-induced NE-4C cells, which corresponds to the modified splitting behaviour of the cells. We also found that in Wnt-induced cells the level of β-catenin protein was increased and the expression levels of Wnt-target genes (Ctnnb1, Ptgs2, Ccnd1, Mmp9) was significantly upregulated in response to PGE2 treatment. This confirms that PGE2 activated the canonical Wnt signalling pathway. Furthermore, the upregulated genes have been previously associated with ASD. Our findings show, for the first time, evidence for cross-talk between PGE2 and Wnt signalling in neuronal cells, where PKA and PI-3K might act as mediators between the two pathways. Given the importance of PGE2 and Wnt signalling in prenatal development of the nervous system, our study provides insight into how interaction between these two pathways may influence neurodevelopment.

Figure 1

Expression of EP receptors’ mRNA and protein in NE-4C cells. (A) Real-time PCR was used to determine the RQ value for EP1, EP2, EP3α, EP3β, EP3γ and EP4 receptors, which was found to be 2, 16, 1, 2, 46 and 46 respectively. The error bars represent + SEM. (B) Western blot analysis of the EP1, EP2, EP3 and EP4 receptors expression (65, 68, 62 and 53 kDa, respectively). β-Actin was used to indicate equal loading. (C) Immunocytochemistry revealed the subcellular localization of EP1-4 receptors with specific organelles visualized through the use of anti-PDI endoplasmic reticulum marker, anti-Lamin A + C nuclear envelope marker, β-Actin cell membrane marker, and anti-58 K Golgi marker. The scale bar represents 10 μm.

Figure 2

PGE₂-dependent effect on final distance travelled from origin. (A) Final distance from origin was 65.6, 56.2, 21.3, 45.0 μm, respectively. The error bars represent + SEM and values were considered significant at *p < 0.05, **p < 0.01. (B) The Dispersion XY position plots illustrate the effect of PGE₂ on Wnt-induced behaviour, where addition of PGE₂ to Wnt-activated cells increased the final distance. Addition of H89 (PKA blocker) and Wort (PI-3K) blocker reduces the effect PGE₂. Measurements represent an average of 150 cells from three independent experiments (N = 3).

Figure 3

PGE₂-dependent effect on path length and average speed. (A) Path length travelled was 459, 409, 103, 362 μm, respectively. (B) Average speed of migration was 11.0, 10.5, 1.7, 7.2 nm/s, respectively. The error bars represent + SEM, **p < 0.01, ***p < 0.001. Results represent an average of 150 cells from three independent experiments (N = 3).

Figure 4

PGE₂-dependent effect on proliferation behaviour. (A) Over the experimental duration of 24 hours, the number of cells changed by a fold of 4.60, 4.59, 0.86, 1.03, respectively. (B) Cell viability across treatment conditions was not significantly different. (C) Percentage of successful split ratio was 100%, 98%, 0%, 15%, 0%, and 0% respectively. The error bars represent + SEM, ***p < 0.001. Measurements represent an average of 150 cells from three independent experiments (N = 3). (D) WntA treatment resulted in an arrested state indicted by the black arrows and corresponded with a significant decrease in cell proliferation (***p < 0.001). Scale bar represents 100 μm.

Figure 5

PGE₂-dependent effect on phospho-histone H3 (Ser10) expression. Western blot analysis was used to determine Phospho-Histone H3 (Ser10) protein (17 kDa). The expression of Phospho-Histone H3 (Ser10) represented in fold change was 1, 1.04, 1.35, 1.52, 1.36, and 1.58, respectively. The error bars represent + SEM and values were considered significantly different from untreated *p < 0.05, **p < 0.01. Average measurements represent protein from three independent experiments (N = 3). β-Actin was used to indicate equal loading.

Figure 6

PGE₂-dependent effect on β-catenin expression in NE-4C cells. Western blot analysis was used to determine two forms of active β-catenin: (A) non-phospho-(active) β-catenin (Ser33/37/Thr41) and (B) phospho-β-catenin (Ser552) (92 kDa). Addition of PGE₂ to NE-4C cells did not yield a significant difference in levels of either active form of β-catenin compared to control. The error bars represent + SEM and values were considered significantly different from control at *p < 0.05, **p < 0.01. Average measurements represent protein from three independent experiments (N = 3). β-Actin was used to indicate equal loading.

Figure 7

PGE₂-dependent effect on β-catenin expression in Wnt-activated NE-4C cells. Western blot analysis was used to determine two forms of active β-catenin: non-phospho-(active) β-catenin (Ser33/37/Thr41) and phospho-β-catenin (Ser552) (92 kDa). (A) The expression of active β-catenin represented in fold change was 1, 2.09, 1.61, and 1.98, respectively. The error bars represent + SEM and values were considered significantly different from control at *p < 0.05. Only PGE₂ + WntA condition was significantly different from WntA only condition. (B) There was no significant difference in phospho-β-catenin (Ser552) expression between the conditions. Average measurements represent protein from three independent experiments (N = 3). β-Actin was used to indicate equal loading.

Figure 8

PGE₂-dependent effect on Wnt-target genes. Real-time PCR was used to determine the RQ value for Ctnnb1, Ptgs2, Ccnd1, and Mmp9. The expression of Ctnnb1 represented in fold change was 1, 0.97, 1.25, 1.55, 0.84, and 0.60, respectively. The fold change expression of Ptgs2 was 1, 0.56, 2.99, 4.59, 2.16, and 4.22. The fold change expression of Ccnd1 was 1, 3.68, 1.50, 1.99, 0.74, and 1.42. Mmp9 fold change expression was 1, 1.08, 2.19, 3.00, 2.16, and 2.68, respectively. The error bars represent + SEM and values were considered significantly different from control at *p < 0.05, **p < 0.01, and ***p < 0.001. Average measurements are from three independent experiments (N = 3).

Figure 9

A proposed model for PGE₂-Wnt interactions in Wnt-induced NE-4C cells. From the compilation of our results (bolded) and other studies, a schematic model is drawn of the mechanism by which PGE₂ might interact with the canonical Wnt pathway.