A Ctnnb1 enhancer regulates neocortical neurogenesis by controlling the abundance of intermediate progenitors

Abstract

β-catenin-dependent canonical Wnt signaling plays a plethora of roles in neocortex (Ncx) development, but its function in regulating the abundance of intermediate progenitors (IPs) is elusive. Here we identified neCtnnb1, an evolutionarily conserved cis-regulatory element with typical enhancer features in developing Ncx. neCtnnb1 locates 55 kilobase upstream of and spatially close to the promoter of Ctnnb1, the gene encoding β-catenin. CRISPR/Cas9-mediated activation or interference of the neCtnnb1 locus enhanced or inhibited transcription of Ctnnb1. neCtnnb1 drove transcription predominantly in the subventricular zone of developing Ncx. Knock-out of neCtnnb1 in mice resulted in compromised expression of Ctnnb1 and the Wnt reporter in developing Ncx. Importantly, knock-out of neCtnnb1 lead to reduced production and transit-amplification of IPs, which subsequently generated fewer upper-layer Ncx projection neurons (PNs). In contrast, enhancing the canonical Wnt signaling by stabilizing β-catenin in neCtnnb1-active cells promoted the production of IPs and upper-layer Ncx PNs. ASH2L was identified as the key trans-acting factor that associates with neCtnnb1 and Ctnnb1's promoter to maintain Ctnnb1's transcription in both mouse and human Ncx progenitors. These findings advance understanding of transcriptional regulation of Ctnnb1, and provide insights into mechanisms underlying Ncx expansion during development.

Fig. 1. neCtnnb1 is a neocortex-specific enhancer of Ctnnb1.

a Schematic representation of the upstream region of mouse Ctnnb1 gene and the location of putative enhancer neCtnnb1 (2.28 kb, orange shading), which is evolutionarily conserved among amniotes and enriched with H3K27ac, DNase I hypersensitivity and P300 in developing forebrains. Data were obtained from ENCODE. Locations of ChIP sites in (h) were indicated. b Hi-C data of in-vitro differentiated NPCs were obtained from the 3D genome browser. Boundaries of the TADs and locations of Ctnnb1 and neCtnnb1 (red bar) were marked below. c Top: a schematic illustration showing the transgenic construct carries three tandem neCtnnb1 sequences, the hsp68 mini-promoter, the LacZ reporter gene and inducible CreERT2 (iCre). Bottom: X-Gal staining (blue) in E11.5 and E15.5 neCtnnb1-LacZ-iCre embryos and brains. Signals at E15.5 dorsal forebrains (fb) were indicated with red arrows and enlarged. d Immunofluorescence of PAX6 (gray), TBR2 (green), and β-Gal (red) on coronal sections of E15.5 neCtnnb1-LacZ-iCre Ncx. e Bar plots showing TBR2 + β-gal+ cells relative of TBR2 + cells (left) and PAX6 + TBR2 + β-gal+ cells relative of PAX6 + TBR2 + cells (right) in E15.5 neCtnnb1-LacZ-iCre Ncx. n = 2 for neCtnnb1-LacZ-iCre brains. Each point represents an individual brain. f The 3C assay performed using Neuro-2a neuroblastoma cells. The Ctnnb1 promoter (pCtnnb1) is the anchor point from which long-range DNA interactions across the neCtnnb1 interval were measured. Numbers over the data points represent fragment locations. g Schematics show the ChIP-qPCR evaluating the physical association between the pCtnnb1 and neCtnnb1. h ChIP-qPCR measuring enrichment of pCtnnb1 and neCtnnb1 after Flag-tagged dCas9 were targeted to neCtnnb1 and pCtnnb1 respectively. Each point represents an independent experiment. I, k Schematics showing the CRISPR activation (CRISPRa, i) and CRISPR interference (CRISPRi, k) experiments. j, l RNA levels of Ctnnb1 in Neuro-2a cells transfected with indicated CRISPRa (j) or CRISPRi (l) vectors for two days. In h quantification data are shown as means ± SEM, statistical significance was determined using two-way ANOVA followed by Tukey’s multiple comparisons test. In j and l, quantification data are shown as means ± SD, statistical significance was determined using one-way ANOVA analysis, *P < 0.05, **P < 0.01, and ***P < 0.001. ns, not significant. Scale bars, 1 mm (c), 50 μm (d). fb, forebrain; CP, cortical plate; VZ, ventricular zone; SVZ, subventricular zone.

Fig. 2. Knock-out of neCtnnb1 in mice inhibits the production of upper-layer PNs and Ctnnb1 transcription.

a Generation and genotyping of neCtnnb1^KO mice. WT, wild-type; gRNA, guide RNA. b Nissl-staining on adult WT and neCtnnb1^KO sagittal brain sections. c Comparison of neocortical thickness of adult WT and neCtnnb1^KO mice. n = 3 for WT brains and n = 5 for neCtnnb1^KO brains. Each point represents an individual brain. d, e Immunofluorescence (d) and quantification (e, top) of SATB2 + , CUX1 + , FOXP2 + and CTIP2 + neurons on boxed area of sagittal sections (e) of adult WT and neCtnnb1^KO Ncx. Each point represents an individual brain. f In situ hybridization (ISH) of Ctnnb1 on E14.5 WT (top) and neCtnnb1^KO (bottom) coronal brain sections, with boxed regions magnified on the right. g, h Quantification of normalized ISH signal densities in boxed regions of (f). n = 3 for WT brains and n = 3 for neCtnnb1^KO brains. Each point represents an individual brain. i RT-qPCR showing expressions of Ctnnb1 in E15.5 WT and neCtnnb1^KO neocortex. n = 4 for WT Ncx and n = 6 for neCtnnb1^KO Ncx. Each point represents an individual brain. j β-Gal immunostaining on the E15.5 coronal sections of WT_BAT (left) and neCtnnb1^KO_BAT (right) Ncx. Boxed regions are enlarged at the bottom left corners. k Quantification of normalized signal density of β-Gal in boxed regions of (j). n = 3 for WT brains and n = 3 for neCtnnb1^KO brains. Each point represents an individual brain. Quantification data are shown as means ± SEM. Statistical significance was determined using an unpaired two-tailed Student’s t-test (c, e, h and i). *P < 0.05, **P < 0.01, and ***P < 0.001. ns, not significant. Scale bars, 500 μm (f), 200 μm (b), 100 μm (d and j), 50 μm (magnified views in f and j). CP, cortical plate; VZ, ventricular zone; SVZ, subventricular zone; Ctx, cortex; LV, lateral ventricular; LGE, lateral ganglionic eminences; Hipp, Hippocampal primordium.

Fig. 3. neCtnnb1 knockout hampered the production and transit-amplification of IPs.

a–k Immunofluorescence (a) and quantification (c–k) of EdU + (red), PAX6 + (blue) and TBR2 + (green) cells on boxed area of coronal sections (b) of E16.5 WT and neCtnnb1^KO Ncx. Pregnant mice were injected with EdU 2 h before sacrifice. Each point represents an individual brain. l–t Immunofluorescence (l) and quantification (n–t) of Ki67 + (red), TBR2 + (blue) and BrdU + (green) cells on boxed area of coronal sections (m) of E16.5 WT and neCtnnb1^KO Ncx. Pregnant mice were injected with BrdU at E15.5. m–t Quantification of data in l. Each point represents an individual brain. Quantification data are shown as means ± SEM. Statistical significance was determined using an unpaired two-tailed Student’s t-test (c–k, n–t). *P < 0.05, **P < 0.01, and ***P < 0.001. ns not significant. Scale bars, 100 μm (a and l), 50 μm (magnified views in a, l).

Fig. 4. Stabilization of β-catenin in neCtnnb1-active cells promotes the production and transit-amplification of IPs in developing neocortices.

a Schematics show mating of Ctnnb1^lox(ex3) mice with neCtnnb1-LacZ-iCre mice. Pregnant mice were injected with tamoxifen (TAM) at E13.5 and sacrificed at E16.5. b Top: immunoblotting of β-catenin, c-MYC, Cyclin D1 and ACTIN using protein lysates of Ctrl - Ctnnb1^lox(ex3) and WNT-iGOF - neCtnnb1-LacZ-iCre;Ctnnb1^lox(ex3) neocortices. Bottom: histograms show protein levels of c-MYC and Cyclin D1 normalized to ACTIN levels. Values of Ctrl neocortices were set as ‘1’. C.A. β-catenin, the constitutively active (non-degradable) β-catenin caused by deletion of Ctnnb1’s exon 3; Endo. β-catenin, endogenous β-catenin. c Nissl-staining on E16.5 Ctrl and WNT-iGOF coronal brain sections. d Comparison of neocortical thickness of E16.5 Ctrl and WNT-iGOF mice. n = 16 for Ctrl brains and n = 12 for WNT-iGOF brains. Each point represents an individual brain. e Comparison of cortical plate thickness of E16.5 Ctrl and WNT-iGOF mice. n = 10 for Ctrl brains and n = 7 for WNT-iGOF brains. Each point represents an individual brain. f Comparison of dorsal ventricle surface length of E16.5 Ctrl and WNT-iGOF mice. n = 10 for Ctrl brains and n = 7 for WNT-iGOF brains. Each point represents an individual brain. g–i Immunofluorescence (g) and quantification (h, i) of SATB2 + and CTIP2 + of coronal sections of E16.5 Ctrl and WNT-iGOF Ncx. j–n Immunofluorescence (j) and quantification (k–n) of of TBR2 + (red) and PAX6 + (green) cells on boxed area of E16.5 coronal sections (p) of Ctrl and WNT-iGOF Ncx. n = 6 for Ctrl brains and n = 6 for WNT-iGOF brains. Each point represents an individual brain. o–r Immunofluorescence (o) and quantification (p–r) of TBR2 + (red) and Ki67 + (green) cells on boxed area of coronal sections (p) of E16.5 Ctrl and WNT-iGOF Ncx. n = 5 for Ctrl brains and n = 5 for WNT-iGOF brains. Each point represents an individual brain. Quantification data are shown as means ± SEM. Statistical significance was determined using an unpaired two-tailed Student’s t-test (d, e, f, h, i, k–n and p–r). *P < 0.05, **P < 0.01, and ***P < 0.001. ns not significant. Scale bars, 500 μm (c), 100 μm (g, j, o and magnified views in c). 50 μm (magnified views in j and o).

Fig. 5. ASH2L associates with neCtnnb1 and pCtnnb1 to regulate Ctnnb1’s transcription.

a ChIP-seq tracks for enrichment of indicated trans-acting factors in neCtnnb1 in developing Ncx. neCtnnb1 is highlighted. b ChIP-seq tracks for ASH2L in indicated murine cells. neCtnnb1 is highlighted. c ChIP-qPCR of E15.5 WT and neCtnnb1^KO Ncx using an anti-ASH2L antibody. n = 3 for WT Ncx and n = 3 for neCtnnb1^KO Ncx. Each point represents an individual brain. d RNA levels of Ash2l (left) and Ctnnb1 (right) in WT and neCtnnb1^KO neocortical NPCs transfected with indicated vectors for two days. NPCs were derived from E12.5 Ncx. n = 3 for WT brains and n = 3 for neCtnnb1^KO brains. Each point represents an individual brain. e RNA levels of Ash2l (left) and Ctnnb1 (right) in Neuro-2a cells transfected with indicated vectors for two days. Each point represents an independent experiment. f Schematic representation the location of neCtnnb1 (pink shading) and conserved region (orange shading) marked by enrichment of H3K27ac, DNase I hypersensitivity and ASH2L in E14.5 brains. Data were obtained from ENCODE. Locations of sgRNA target sites and ChIP primer were indicated. g, h RNA levels of Ctnnb1 in Neuro-2a cells transfected with indicated CRISPRa (g) and CRISPRi (h) vectors for two days. Each point represents an independent experiment. i Luciferase reporter assay in Neuro-2a cells transfected with indicated vectors for two days. Each point represents an independent experiment. Quantification data are shown as means ± SEM. Statistical significance was determined using two-way ANOVA followed by Sidak’s multiple comparisons test. (c, d, e and i); or using one-way ANOVA analysis (g and h). *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. ns, not significant.

Fig. 6. neCTNNB1 is the human CTNNB1 enhancer.

a Schematic representation of human CTNNB1 gene and the location of putative neCTNNB1 enhancer (blue shading) marked by H3K27ac and H3K4me1 peaks and ATAC signals. b Hi-C data of H1-derived human neural progenitor cells (hNPC) were obtained from the 3D genome browser. Boundaries of TADs and locations of neCTNNB1 and CTNNB1 (red bars) are indicated below. b' Regions magnified from (b). neCTNNB1 interacts with CTNNB1 in H1-derived hNPC. c Virtual 4C of H1-derived hNPCs revealing interactions between neCTNNB1 with pCTNNB1. d ChIP-qPCR of hNPC using an anti-ASH2L antibody. e RNA levels of ASH2L (left) and CTNNB1 (right) in hNPC transfected with indicated virus for two days. Each point represents an independent experiment (d, e). f The working model. neCtnnb1, the upstream enhancer of Ctnnb1, controls production and transit-amplification of IPs, hence proper productions of UL PNs. ASH2L associates with neCtnnb1 and positively regulates transcription of Ctnnb1 in the neCtnnb1-dependent manner. In e, quantification data are shown as means ± SD. statistical significance was determined using two-way ANOVA followed by Sidak’s multiple comparisons test (e) *P < 0.05, **P < 0.01, and ***P < 0.001. ns not significant. PFC dorsal lateral prefrontal cortex.