EZH2 regulates neuroepithelium structure and neuroblast proliferation by repressing p21

Abstract

The function of EZH2 as a transcription repressor is well characterized. However, its role during vertebrate development is still poorly understood, particularly in neurogenesis. Here, we uncover the role of EZH2 in controlling the integrity of the neural tube and allowing proper progenitor proliferation. We demonstrate that knocking down the EZH2 in chick embryo neural tubes unexpectedly disrupts the neuroepithelium (NE) structure, correlating with alteration of the Rho pathway, and reduces neural progenitor proliferation. Moreover, we use transcriptional profiling and functional assays to show that EZH2-mediated repression of p21(WAF1/CIP1) contributes to both processes. Accordingly, overexpression of cytoplasmic p21(WAF1/CIP1) induces NE structural alterations and p21(WAF1/CIP1) suppression rescues proliferation defects and partially compensates for the structural alterations and the Rho activity. Overall, our findings describe a new role of EZH2 in controlling the NE integrity in the neural tube to allow proper progenitor proliferation.

Keywords: EZH2; gene silencing; histone methylation; neural development; neuroblast proliferation.

Figure 1.

EZH2 depletion leads to small and structurally disorganized neural tubes. (a) Diagrams show regions occupied by proliferating neural progenitors (ventricular zone, VZ) and post-mitotic neurons (mantle zone, MZ) in HH10 (only progenitors) and HH30 chick embryo spinal cord (left panel). HH30 wild-type embryo neural tubes were immunostained using anti-EZH2 antibody (right panel). (b) Ten HH10 or HH30 embryo neural tubes were dissected out for each replicate. mRNA was extracted and retrotranscribed for qPCR analysis. The graph shows EZH2 mRNA, normalized by GAPDH mRNA levels. Results are means of three independent experiments. Error bars indicate s.d. *p < 0.05. (c) HH11–12 embryos were co-electroporated with a mixture of shRNA-EZH2(1) and shRNA-EZH2(2) (shEZH2) or shRNA-control (shCtrl) cloned into pSHIN vector expressing GFP. Transversal sections of electroporated neural tubes as indicated above stained with DAPI 48 h PE. Graphs show the quantification of the size of the control side and shEZH2-electroporated side. To do that, we measured the dorsal, medial and ventral distances to the lumen on each side, relative to the length of the central line of the lumen. Data represent mean of n = 10–20 sections (from three to five embryos). Error bars indicate s.d. **p < 0.01; ***p < 0.001. (d–f) Transversal sections of neural tubes from HH11–12 embryos electroporated in ovo with shCtrl or shEZH2 and stained for H3S10p (d), Sox5 (e) or caspase 3 (f) 48 h PE. Graphs show the quantification of the corresponding immunostaining. Data represent mean of n = 20–30 sections (from four to six embryos). Error bars indicate s.d. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 2.

EZH2 preserves NE structure. (a–c) Transversal sections of neural tubes from HH11–12 embryos electroporated in ovo with shCtrl or shEZH2 and stained with phalloidin-rhodamine (a), aPKC (b) or N-cadherin (N-CAD, c) antibodies and DAPI 48 h PE. The results are representative of at least four independent experiments. White arrows mark progenitor cells that have lost the apicobasal polarity and have occupied the lumen. Green arrow indicates area where the apical membrane was severely disrupted along the luminal surface. (d) Graph shows the quantification of the morphological alterations observed in electroporated neural tubes. Data represent the percentage of n = 51 sections (from three to four embryos). (e) HH11–12 embryos were electroporated in ovo with shCtrl or shEZH2 together with pSRE-luc plasmid [–26] and β-gal reporter used as internal control; 48 h PE the neural tubes were dissected out, the tissue was disaggregated and the luciferase activity measured using the Life Technologies dual kit. Data represent mean of three experiments from five to seven embryos. Error bars indicate s.d. ***p < 0.001. (f) HH11–12 embryos were electroporated in ovo with shCtrl or shEZH2; 48 h later GFP+ neural cells were FACS purified. Total protein extracts were prepared and the levels of paxillin (pY31) were determined by western blot. β-Tubulin levels were used to normalize the protein content.

Figure 3.

EZH2 regulates genes involved in proliferation and NE structure. (a) Schematic of microarray analysis design. (b) Relative mRNA levels were analysed for the indicated genes by qPCR in control (shCtrl) and EZH2-depleted cells (shEZH2) from electroporated neural tubes. Data represent mean of three independent experiments (five embryos for each replicate). Error bars indicate s.d. (c) Graph showing the percentage of upregulated and downregulated genes in the microarray experiment classified by FC. (d) Scheme showing the GO term enrichment of the genes deregulated (FC > 1.5) in EZH2-depleted cells.

Figure 4.

EZH2 targets p21^WAF1/CIP1 promoter. (a) p21^WAF1/CIP1 ISH 48 h after in ovo electroporation of shCtrl or shEZH2 in HH11–12 embryo neural tubes. (b) HH11–12 embryos were electroporated in ovo with shCtrl or shEZH2 together with p21^WAF1/CIP1 promoter fused to the luciferase and pCMV-β-gal reporter used as internal control; 48 h PE the neural tubes were dissected, the tissue was disaggregated and the luciferase activity measured using the Life Technologies dual kit. Data represent mean of three experiments from four to six embryos. Error bars indicate s.d. ***p < 0.001. (c) ChIPs analysed by qPCR from HH11–12 electroporated neural tube cells with shCtrl or shEZH2 (48 h PE) using EZH2 or H3K27me3 antibodies at the p21^WAF1/CIP1, NeuroD1 and Hes5 promoters. Results are represented as percentage of input. Map of the genes and the position and length of the amplicons are shown in the bottom part of the figure. Three biological replicates, with ten HH12 embryo neural tubes each, were used in each experiment. Error bars indicate s.d. *p < 0.05.

Figure 5.

shp21 rescues neural progenitor proliferation and apicobasal polarity. (a) p21^WAF1/CIP1 ISH 48 h after in ovo electroporation of shCtrl, shEZH2 or shEZH2 and shp21 in HH11–12 embryo neural tubes. (b–e) Transversal sections of neural tube from HH11–12 embryos electroporated in ovo with shCtrl, shEZH2 or shEZH2 and shp21 and stained with H3S10p (b), caspase 3 (CASP3, c) and N-cadherin (N-CAD, e) antibodies or with phalloidin-rhodamine (d) 48 h PE. Graphs below the panels show the quantification of the corresponding immunostaining. Data represent mean of n = 20–40 sections (from four to six embryos). Error bars indicate s.d. *p < 0.05. (f) Graph showing the percentage of immunostained sections of Ctrl, shEZH2 and shEZH2 + shp21 that presents the indicated phenotypes. (g) Transversal sections of neural tubes from HH11–12 embryos electroporated in ovo with pCIG vector (Ctrl), pCIG together with p21^WAF1/CIP1 or cytoplasmic p21^WAF1/CIP1 (p21-ΔNLS) and stained with DAPI and N-cadherin 48 h PE. Ectopic lumens are amplified. Mitotic cells at the new lumens are visualized by DAPI staining (yellow arrows). The data are representative of at least three independent experiments. (h) HH11–12 embryos were electroporated in ovo with shCtrl, shp21, shEZH2 or shEZH2 and shp21 together with pSRE-luc plasmid and β-gal reporter used as internal control; 48 h later the neural tubes were dissected, the tissue was disaggregated and the luciferase activity measured using the Life Technologies dual kit. Data represent mean of three experiments from five to seven embryos. Error bars indicate s.d. *p < 0.05; ***p < 0.001. (i) 293 T cells were transfected with RhoA-V14, p21^WAF1/CIP1 or RhoA-V14 and p21^WAF1/CIP1 expression vectors together with pSRE-luc plasmid and β-gal reporter used as internal control. 24 h later the luciferase activity measured using the Life Technologies dual kit. Data represent mean of four experiments. Error bars indicate s.d. **p < 0.01; ***p < 0.001.