Polycomb Repressive Complex 2-mediated Regulation of Intestinal Epithelial Homeostasis and Differentiation in Health and in Coeliac Disease

The surface epithelium of the small intestine harbours multiple cell types that are responsible for nutrient intake and defence from pathogens. Homeostasis of the intestinal epithelium is an interplay between intestinal stem cell self-renewal and differentiation. The canonical Wnt/β-catenin signalling pathway regulates this process but epigenetic factors in this process are not fully understood. Mouse small intestinal crypts that contain stem cells can be cultured in three dimensional matrix. This support structure for the stem cells allows them to divide and differentiate toward enterocytes when a proper mix of growth factors is present in the culture medium. The cells form crypt like buds that mimic the small intestine’s form and function. These organoids or mini-guts as they are called undergo the same differentiation from stem cells to enterocytes, goblet cells, Paneth cells and others and finally to apoptosis as in vivo intestine.

We hypothesized that Polycomb Repressive Complex 2 (PRC2) is affecting epithelial homeostasis regulated by the Wnt signalling pathway. Abnormal changes in the crypt-villus axis are implicated both in coeliac disease and in intestinal malignancies. We subjected the mini-guts to chromatin immunoprecipitation- and global run on sequencing methods to study the function of PRC2 in the small intestine. These next generation sequencing tools allowed us to assess genome wide changes during differentiation from stem cells to enterocytes.

We found that histone H3 lysine 27 trimethylation (H3K27me3) repressive mark, catalysed by PRC2, was present in crypt and villus expressed genes that are responsible e.g. proliferation and differentiation. We identified 90 PRC2 target genes that were silenced along the crypt-villus axis. This would suggest that Wnt- mediated homeostasis is also involves PRC2. In colorectal adenoma and carcinomas we also noticed an indication that PRC2 target genes had different expression patterns between healthy and diseased. In addition, 12 previously unknown PRC2-regulated transcription factors were identified that are involved in M-cell differentiation. These include a novel M-cell maturation related transcription factor Esrrg that is regulated by PRC2. Esrrg is required for M-cell development and the loss of Esrrg prevents M-cells to effectively express Sox8 and Gp2 maturation markers.

In coeliac disease, we could see out of bounds expression of PRC2 member SUZ12 in TA-zone while the patients were on gluten containing diet. Wnt-signalling was overactive in the small intestine of coeliac disease patient. In gluten challenge trial we could see differentially expressed gene sets between patients on gluten free diet and gluten containing diet. Gluten challenge also made the intestinal Wnt signalling pathway hyperactive leading to immature crypt gene expression and less mature epithelium. This pathogenic Wnt-state seems to be on in coeliac patients even when they are on a gluten free diet. We also could model a tool to predict disease onset from molecular markers instead of classical histological reading.

PRC2 specific H3K27me3 signature is present on fundamental developmental genes. Combined, our results would indicate that PRC2 maintains crypt and villus H3K27me3 signature and the targets we identified could be essential factors in intestinal stem cell maintenance and differentiation and enterocyte and M-cell maturation. Coeliac disease crypt hyperplasia and other intestinal malignancies that include epithelial hyperplasia or neoplasia could be driven by PRC2 mediated pro-stemness in the epithelium.