EPIGENETIC REPRESSORS MAINTAIN TREG STABILITY IN A FOXP3-INDEPENDENT MANNER

Abstract INTRODUCTION FOXP3+ CD4+ T cells (regulatory T cells, TREGs) are expanded within the inflamed intestine of human Crohn’s disease (CD), yet FOXP3-mediated gene repression within these cells is lost. Through single cell RNA-sequencing (scRNA-seq), we have identified Treg populations in CD patients with unique inflammatory transcriptomic profiles. Furthermore, we have previously shown that BMI1, a key member of the epigenetic repressor complex PRC1, is required to maintain TREG identity by repressing proinflammatory gene networks. In the absence of BMI1, Tregs secrete proinflammatory cytokines and mice develop spontaneous colitis. Thus, we investigated the role for Polycomb family members in the regulation of TREG gene programs. METHODS We performed RNA-seq on murine BMI1-deficient TREGs isolated from our conditional BMI1-knockout mouse model. We deleted BMI1 in human TREGS using CRISPR-CAS9, followed by qPCR and RNA-seq to infer associations between PRC1 and activated inflammatory genes. To identify the mechanism of PRC1 repression of FOXP3 target genes, we performed genome-wide localization of RING1B, the catalytic subunit of the PRC1 complex, and the histone mark it recognizes, Histone 3 trimethylation at Lysine 27 (H3K27me3) in human TREGs using Chromatin Immunoprecipitation (ChIP-seq). Finally, to test the therapeutic potential of our engineered cells, we have established a stability assay aimed to challenge TREGs by treating them with TH17-polarizing conditions followed by qPCR of key transcription factors, cytokines, and FOXP3-repressed genes. We have successfully established BMI1 overexpression in TREGs using dCAS9-VPR. RESULTS Gene set enrichment analysis from the murine BMI1-deficient TREGs demonstrated a significant overlap between BMI1 (PRC1) and FOXP3-repressed genes (NES 1.88, p-adj 0.02). Furthermore, we found that these PRC1/FOXP3 repressed genes are upregulated in TREGs isolated from the ileum of Crohn’s disease patients when compared to healthy controls (NES 1.6, p-adj 0.03). BMI1-deficient human TREGS maintained baseline FOXP3 gene expression. By ChIP-seq, we observed significant enrichment for RING1B and H3K27me3 at the enhancer region for IRF4, a proinflammatory gene known to be repressed by FOXP3. Through our stability assay, we showed increased IRF4 expression accompanied by increased expression of RORC and IL17 in TREGs cultured in TH17 media compared to TREG media. CONCLUSIONS Our data suggests that PRC1 regulates inflammatory networks in FOXP3+ cells through direct binding of enhancer regions associated to TREG identity, such as IRF4. Overexpression of PRC1 complex members could represent a FOXP3-independent mechanism that maintains TREG cell stability under inflammatory conditions.