Wnt/β-catenin signalling is required for pole-specific chromatin remodeling during planarian regeneration

Abstract

For successful regeneration, the identity of the missing tissue must be specified according to the pre-existing tissue. Planarians are ideal for the study of the mechanisms underlying this process; the same field of cells can regrow a head or a tail according to the missing body part. After amputation, the differential activation of the Wnt/β-catenin signal specifies anterior versus posterior identity. Initially, both wnt1 and notum (Wnt inhibitor) are expressed in all wounds, but 48 hours later they are restricted to posterior or anterior facing wounds, respectively, by an unknown mechanism. Here we show that 12 hours after amputation, the chromatin accessibility of cells in the wound region changes according to the polarity of the pre-existing tissue in a Wnt/β-catenin-dependent manner. Genomic analyses suggest that homeobox transcription factors and chromatin-remodeling proteins are direct Wnt/β-catenin targets, which trigger the expression of posterior effectors. Finally, we identify FoxG as a wnt1 up-stream regulator, probably via binding to its first intron enhancer region.

Fig. 1. The chromatin of cells in the wound remodels according to the polarity of the pre-existing tissue.

a Workflow to identify putative anterior and posterior specific enhancers at 12 hR. Next to each step the program used is indicated. b Accessibility changes of the putative anterior and posterior specific active enhancers after notum and wnt1 inhibition at 12 hR are represented in percentages in pie charts. Schematic illustration shows the representative phenotypes observed after each gene inhibition. hR, hour of regeneration. All the data is provided in Supplementary Data 1.

Fig. 2. Homeobox TFs motifs are found enriched in cis-regulatory elements of wnt1 (RNAi) downregulated genes.

a Workflow to identify differentially expressed genes, cis-regulatory elements (CRE) and transcription factors (TFs) associated with wnt1 (RNAi). Next to each workflow the program used is indicated. Motif discovery for TCF binding sites was specifically performed in downregulated wnt1 (RNAi) genes. b Volcano plot of 72 hR shows the (694) down- and (1409) upregulated genes after wnt1 inhibition, which present fold change (fc) ± 0.5 and p-value adjusted (padj) <0.05 (FDR on lima-voom empirical Bayes moderated t-test). Significant genes are colored light blue, and not significant are colored light red. Some significant genes are labeled. Data provided in Supplementary Data 3. c TF motifs presence on the putative enhancer regions of wnt1 (RNAi) downregulated genes, showing a higher representation of Homeobox. Data provided in Supplementary Data 4. d Gene Ontology (molecular function) analysis of the wnt1 (RNAi) downregulated genes containing TCF binding sites (http://pantherdb.org/). Data provided in Supplementary Data 5. e Schematic illustration of the proposed genetic program activated by wnt1 in posterior wounds.

Fig. 3. FoxG could bind to a cis-regulatory element found in wnt1 first intron.

a Schematic illustration of Smed-wnt1 gene locus, indicating exons (violet boxes) linked by introns (lines). Genome Browser screenshot showing the ATAC-seq Accessible Chromatin Regions (ACRs), ATAC-seq and H3K27ac ChIP-seq profiles of the putative enhancer regions. Putative Enhancer 1 (E1) and Enhancer 2 (E2) present a FoxG motif (SLP1) (green line). The ATAC-seq peaks corresponding to E1 and the E2 are indicated. b Alignment of WNT1 amino acid sequences show the conservation of the intron 1 position. Yellow line shows the separation between the first and the second exon. c Schematic summary of accessible chromatin regions (ACRs), enhancer activity and FoxG motif evidence in the first intron of wnt1 genes in different eumetazoan species. Green box indicates evidence and white box indicates no available data. Data is provided in Supplementary Data 7. Species used: Homo sapiens (Hsap), Mus musculus (Mmus), Danio rerio (Drer), Xenopus tropicalis (Xtro), Drosophila melanogaster (Dmel), Schmidtea mediterranea (Smed), Macrostumum ligano (Mlig), Hydra vulgaris (Hvul), and Nematostella vectensis (Nvec). Silhouettes are from https://beta.phylopic.org. Xenopus silhouette is from Sarah Werning and the licence is https://creativecommons.org/licenses/by/3.0/. Schmidthea silhouette is from Noah Schlottman and the license is https://creativecommons.org/licenses/by/3.0/.

Fig. 4. foxG is expressed in regenerating blastemas and regulates wnt1 expression.

a Schematic illustration indicating the foxG (RNAi) procedure. b WISH of wnt1 in foxG (RNAi) animals demonstrate its absence in both 12 and 72 hR wounds. Schematic illustration of wnt1 in intact animals with the analyzed zones (squares) added. Data representative of two independent experiments. c WISH of foxG in intact animals shows its expression in the posterior midline, similar to wnt1 expression. Data representative of one experiment. Orange dashed lines show the magnified area. Single cell analysis performed by showed six genes (top 16%) over represented in posterior organizing wnt1+ cells. Among these, foxG was found. d WISH of wnt1 and foxG in regenerating anterior and posterior wounds at 24 hR showing a salt-and-pepper pattern (left panel). A magnification was taken from each blastema (right panel). Purple boxes show the magnified area. Schematic illustration of regenerating animals was added. Data representative of one experiment. Scale bar: 50 µm and 10 µm for left and right panel in b, respectively; 100 µm in c; 200 µm in d.

Fig. 5. foxG (RNAi) phenocopies wnt1 inhibition.

a In vivo phenotypes after foxG (RNAi). Data representative of two independent experiments. b Immunostaining using α-SYNAPSIN (3C11) (neural system) and α-βCAT2 (β-catenin-2, digestive system) reveals rounded ventral nerve cords in foxG (RNAi) (tailless phenotype). Nuclei are stained in DAPI. WISH of posterior markers in regenerating foxG (RNAi) animals demonstrated a reduced expression. Data representative of two independent experiments. Schematic illustrations of posterior markers were added. c Immunostaining using α-SYNAPSIN (3C11) (neural system) reveals a posterior brain in the foxG (RNAi) two-headed animals. Data representative of two independent experiments. Nuclei are stained in DAPI. Posterior eyes are indicated with an orange arrow in a and c. Scale bar: 100 µm in a, immunostaining in b and c; 200 µm in WISH in b.

Fig. 6. Three-step model of posterior specification.

Our data supports a working model in which after an amputation, foxG is expressed at early stages of regeneration in the posterior-facing wounds. The FOXG transcription factor activates wnt1 expression in the organizer cells, possibly by interacting with an enhancer located in its first intron. Because of this interaction, these cells express and secret wnt1. In the 3-step model, the wnt1 responding cells, through the β-catenin-1/TCF pathway, will induce the expression of chromatin remodelers (1) and TFs (i.e. sp5, tsh, Homeobox) (2). Consequently, the chromatin will be accessible for the TFs regulating the expression of crucial genes for posterior specification (i.e. Wnt11s, fzd4-1) (3).