Polycomb Repressive Complex 1 Controls Maintenance of Fungiform Papillae by Repressing Sonic Hedgehog Expression

Abstract

How tissue patterns are formed and maintained are fundamental questions. The murine tongue epithelium, a paradigm for tissue patterning, consists of an array of specialized fungiform papillae structures that harbor taste cells. The formation of fungiform papillae is preceded by pronounced spatial changes in gene expression, in which taste cell genes such as Shh, initially diffused in lingual epithelial progenitors, become restricted to taste cells when their specification progresses. However, the requirement of spatial restriction of taste cell gene expression for patterning and formation of fungiform papillae is unknown. Here, we show that a chromatin regulator, Polycomb repressive complex (PRC) 1, is required for proper maintenance of fungiform papillae by repressing Shh and preventing ectopic SHH signaling in non-taste cells. Ablation of SHH signaling in PRC1-null non-taste cells rescues the maintenance of taste cells. Altogether, our studies exemplify how epigenetic regulation establishes spatial gene expression patterns necessary for specialized niche structures.

Keywords: PRC1; Polycomb; Sonic Hedgehog; filiform papillae; fungiform papillae.

Figure 1.. Ablation of Ring1a/b in the Non-gustatory Lingual Epithelium Results in a Progressive Loss of Fungiform Papillae and Ablation of Filiform Papillae

(A) Developmental timeline and gene expression pattern in the murine lingual epithelium (see text for details). R, repressor. (B) Expression of the basal epithelial K14-Cre driver in control neonatal (P0) lingual epithelium, visualized by the Rosa26-mT/mG reporter. (C) Immunofluorescence (IF) analysis of the H2AK119ub mark in the lingual epithelium of control and Ring1a/b 2KO E16 embryos. (D–I) H&E analysis of control and Ring1a/b 2KO lingual epithelium (D, F, and H). (E, G, and I) IF analysis of taste cell markers SOX2 and K8 in control and Ring1a/b 2KO lingual epithelium at E16 (D and E), E17 (F and G), and P0 (H and I). Arrowheads indicate taste cell clusters. Arrows indicate the non-gustatory epithelium. Dashed lines label the basement membrane. All IF and bright-field scale bars are 50 μm.

Figure 2.. Cell-Cycle Repressors and SHH Signaling Components Are Upregulated in PRC1 Null Non-gustatory Epithelium

(A) Differential expression analysis in E16 Ring1a/b 2KO versus control in FACS-purified lingual epithelial cells. Genes with absolute fold change ≥ 1.8 and a false discovery rate (FDR) < 0.05 were considered upregulated (red) or downregulated (green). n = 2 embryos per genotype. (B) Gene Ontology (GO) analysis of E16 Ring1a/b 2KO upregulated genes. Selected terms are labeled in red. (C) qPCR analysis of selected genes in GO-enriched categories in Ring1a/b 2KO upregulated genes. (D) Percentage of H2AK119ub-bound genes among Ring1a/b 2KO upregulated genes. (E) GO analysis of E16 Ring1a/b 2KO downregulated genes. Selected terms are labeled in green. (F) qPCR analysis of selected genes in GO-enriched categories in Ring1a/b 2KO downregulated genes. (G) IF analysis of downregulated lingual keratinocyte markers K6, PAX9, and K13 in E16 control and Ring1a/b 2KO embryos. Dashed lines label the basement membrane. IF scale bars are 50 μm. Data in graphs (C) and (F) are mean ± SEM. Data were analyzed by two-tailed Student’s t test, *p < 0.05, **p < 0.01, ***p < 0.001. n = 3 embryos per genotype.

Figure 3.. Loss of Lingual Papillae in Ring1a/b 2KO Mice Is Not Caused by Cdkn2a Upregulation Alone

(A) BrdU incorporation assay in P0 control and Ring1a/b 2KO tongues. Integrin β4 (ITGβ4) marks the basement membrane. (B) Quantification of BrdU incorporation. Two-tailed Student’s t test, ***p < 0.0001. n = 3 animals per genotype. Data are mean ± SEM. (C) Expression levels of Arf and Ink4a relative to the housekeeping gene Ppib in E16 control and Ring1a/b 2KO FACS-purified lingual epithelium. Two-tailed Student’s t test, **p < 0.001, ***p < 0.0001. ND, signal not detected, n = 4 embryos per genotype. Data are mean ± SD. (D) Integrative Genomics Viewer (IGV) browser view of H2AK119ub and input for the Cdkn2a locus. Arrow indicates the transcription start site (TSS). (E) P0 H2AK119ub ChIP-qPCR analysis on FACS-purified lingual epithelium of control and Ring1a/b 2KO mice for Cdkn2a. Two-tailed Student’s t test, *p < 0.05. n = 2 ChIP replicates. Data are mean ± SEM. (F) IF analysis of mitotic marker phospho-histone H3 (PHH3) in P0 control, Ring1a/b 2KO, and Ring1a/b/Cdkn2a 3KO tongues. (G) Quantification of PHH3-positive cells. One-way ANOVA, **p < 0.005, ***p < 0.0005. n = 2 animals per genotype. Data are mean ± SEM. See STAR Methods for quantification details. (H) IF analysis of taste cell markers SOX2 and K8 in P0 control, Ring1a/b 2KO, and Ring1a/b/Cdkn2a 3KO tongues. (I) Counts of K8(+) cell clusters per millimeter in P0 control, Ring1a/b 2KO, and Ring1a/b/Cdkn2a 3KO tongues. The bar represents the average of all animals, and data points are the quantification average of individual animals (n). One-way ANOVA (Kruskal-Wallis), *p < 0.05. NS, non-significant. Data are mean ± SEM. See STAR Methods for quantification details. Dashed lines label the basement membrane. IF scale bars are 50 μm.

Figure 4.. PRC1 Restricts SHH Signaling in the Non-gustatory Epithelium

(A and A′) In situ hybridization of Shh in E16 control and Ring1a/b 2KO tongues, co-stained with taste cell marker K8 and basal epithelial marker K5 (A). K8 channel alone (A′). (B and B′) In situ hybridization of Gli1 in E16 control and Ring1a/b 2KO tongues, co-stained with taste cell marker K8 and basal epithelial marker K5 (B). K8 channel alone (B′). (C) Analysis of the SHH signaling reporter Gli1-LacZ in E16 control and Ring1a/b 2KO tongues. Nuclei are counterstained with fast red dye. (D) IGV browser view of H2AK119ub and input for the Shh locus. The arrow indicates the transcription start site (TSS). (E) Map of ChIP-qPCR primer locations on the Shh locus and surrounding areas. (F and G) H2AK119ub ChIP-qPCR analysis of Shh in P0 control and Ring1a/b 2KO FACS-purified lingual epithelium. n = 2 ChIP replicates (F). H2AK119ub ChIP-qPCR analysis of Shh in FACS-purified TCF/LEF:H2B-GFP(+) taste cells and TCF/LEF:H2B-GFP(—) lingual keratinocytes. n = 2 ChIP replicates (G). Two-tailed Student’s t test, *p < 0.05, **p < 0.01, ***p < 0.001. ND, signal not detected. Image arrowheads indicate taste cell clusters, and arrows indicate the non-gustatory epithelium. Data in (F) and (G) are mean ± SEM. Dashed lines label the basement membrane. Bright-field and IF scale bars are 50 μm.

Figure 5.. PRC1-Mediated Repression of SHH Signaling in the Non-gustatory Epithelium Is Critical for Fungiform Papillae Maintenance

(A) In situ hybridization analysis of Gli1 in P0 control, Ring1a/b 2KO, and Ring1a/b/Smo 3KO tongues, co-stained with taste cell marker K8 and basal epithelial marker K5. (B) IF analysis of taste cell markers SOX2 and K8 in P0 control, Ring1a/b 2KO, and Ring1a/b/Smo 3KO tongues. (C) Counts of K8(+) cell clusters per millimeter in P0 control, Ring1a/b 2KO, and Ring1a/b/Smo 3KO tongues. The bar represents the average of all animals, and data points are the quantification average of individual animals (n). One-way ANOVA (Kruskal-Wallis), *p < 0.05, ****p < 0.0001. Data are mean ± SEM. See STAR Methods for quantification details. (D) Model for PRC1 regulation in the non-gustatory epithelium (see text for details). Dashed lines label the basement membrane. Arrowheads indicate taste cell clusters. Arrows indicate the non-gustatory epithelium. IF scale bars are 50 μm.