A direct fate exclusion mechanism by Sonic hedgehog-regulated transcriptional repressors

Abstract

Sonic hedgehog (Shh) signaling patterns the vertebrate spinal cord by activating a group of transcriptional repressors in distinct neural progenitors of somatic motor neuron and interneuron subtypes. To identify the action of this network, we performed a genome-wide analysis of the regulatory actions of three key ventral determinants in mammalian neural tube patterning: Nkx2.2, Nkx6.1 and Olig2. Previous studies have demonstrated that each factor acts predominantly as a transcriptional repressor, at least in part, to inhibit alternative progenitor fate choices. Here, we reveal broad and direct repression of multiple alternative fates as a general mechanism of repressor action. Additionally, the repressor network targets multiple Shh signaling components providing negative feedback to ongoing Shh signaling. Analysis of chromatin organization around Nkx2.2-, Nkx6.1- and Olig2-bound regions, together with co-analysis of engagement of the transcriptional activator Sox2, indicate that repressors bind to, and probably modulate the action of, neural enhancers. Together, the data suggest a model for neural progenitor specification downstream of Shh signaling, in which Nkx2.2 and Olig2 direct repression of alternative neural progenitor fate determinants, an action augmented by the overlapping activity of Nkx6.1 in each cell type. Integration of repressor and activator inputs, notably activator inputs mediated by Sox2, is probably a key mechanism in achieving cell type-specific transcriptional outcomes in mammalian neural progenitor fate specification.

Keywords: Fate specification; Mouse; Neural development; Transcriptional regulation.

Fig. 1.

Characteristics of Nkx2.2, Nkx6.1 and Olig2 target genes. (A) Immunofluorescence assay on transverse E10.5 neural tube section at forelimb level with indicated antibodies. (B) Venn diagram intersection of Nkx2.2-, Nkx6.1- and Olig2-binding regions. (C) Venn diagram intersection of Nkx2.2, Nkx6.1 and Olig2 target genes. (D) Gene Ontology analysis summary for genes targeted by different combinations of factors. (E) Target gene Venn diagram highlighting neural progenitor fate determinants and Sonic hedgehog pathway components. (F) Genome browser snapshots showing indicated ChIP-seq signal. Cons, Phastcon 30 conservation score. (G,H) Venn diagrams for binding region overlap (G) and target gene overlap (H) between Nkx2.2, Nkx6.1, Olig2 and Gli3.

Fig. 2.

Nkx2.2, Nkx6.1 and Olig2 overexpression assay in neural progenitors. (A) Schematic describing transgene structures. (B) Schematic of overexpresssion experiment design. Cell aggregates were generated from mESCs and subjected to neural differentiation and transgene activation. RA, all-trans retinoic acid; Dox, doxycycline. (C) Hierarchical clustering of genes displaying a twofold or greater change in mRNA-seq data relative to the parental cell line 12 h after Dox-mediated activation of transcriptional repressors. Fold change to the parental cell line is shown. (D-F) RT-qPCR time course repression assay. See A for color designations. x-axis, hours post-Dox induction; y-axis, fold change from Dox induction (t=0). Error bars indicate s.e.m. based on three biological replicates. Asterisks indicate significant difference from non-transgenic control based on the s.e.m.

Fig. 3.

Analysis of enriched binding motifs. (A-C) Occurrence of ChIP-recovered and in vitro-determined motifs. c, CisGenome motif recovery; d, DREME motif recovery; P, protein binding microarray. Gray, E14.5 brain DNaseI-seq control dataset; black, ChIPseq data. (Left) Motif distribution histogram relative to binding peak center. x-axis, bp from peak center; y-axis, cumulative motif occurrence. Gray, E14.5 brain DNaseI-seq control; light blue, ChIPseq data.

Fig. 4.

Analysis of Sox2 inputs into ventral repressor-bound regions. (A) Venn diagram intersection between Sox2-binding regions and the union of Nkx2.2-, Nkx6.1- and Olig2-binding regions. (B) Schematic showing neural progenitor differentiation protocol. Each condition analyzed in C-H is annotated with a solid box of the corresponding color. (C-H) Aggregate analysis of H3K27ac modification status along neural progenitor differentiation paths. Black, ESCs; green, pre-neural induction; blue, dorsal neural progenitors; red, ventral neural progenitors. Also see B for color coding. (C-E) Individual plot for Nkx2.2-, Nkx6.1- and Olig2-binding regions. (F) Sox2-binding regions that do not overlap with Nkx2.2, Nkx6.1 or Olig2. (G) Sox2-binding regions that overlap with Nkx2.2, Nkx6.1 or Olig2. (H) Nkx2.2-, Nkx6.1- or Olig2-binding regions that do not overlap with Sox2 binding.