β-cells cis-regulatory networks and type 1 diabetes

Abstract

[eng] Type 1 Diabetes (T1D) is a ­cell­targeted autoimmune disease, leading to a reduction in pancreatic ­cell mass that renders patients insulin­dependent for life. In early stages of the disease, cells from the immune system infiltrate pancreatic islets in a process called insulitis. During this stage, a cross­talk is established between cells in the pancreatic islets and the infiltrating immune cells, mediated by the release of cytokines and chemokines. Studying the gene regulatory networks driving cell responses during insulitis, will allow us to pinpoint key gene pathways leading to ­cell loss­of­function and apoptosis, and also to understand the role cells have in their own demise. In the present thesis, we used two different cytokine cocktails, IFN­ and IFN­ + IL­1, to model early and late insulitis, respectively. After exposing cells and pancreatic islets to such proinflammatory cytokines, we characterized the changes in their chromatin landscape, gene networks and protein profiles. Using both models, we observed dramatic chromatin remodeling in terms of accessibility and/or H3K27ac histone modification enrichment, coupled with up­regulation of the nearby genes and increased abundance of the corresponding protein. Mining gene regulatory networks of ­cells exposed to IFN­ revealed two potential therapeutic interventions which were able to reduce interferon signature in cells: 1) Inhibition of bromodomain proteins, which resulted in a down­regulation of IFN­­induced HLA­I and CXCL10 expression; 2) Baricitnib, a JAK1/2 inhibitor, which was able to reduce both IFN­­induced HLA­I and CXCL10 expression levels and cell apoptosis. In cells exposed to IFN­ + IL­1, we were able to identify a subset of novel regulatory elements uncovered upon the exposure, which we named Induced Regulatory Elements (IREs). Such regions were enriched for T1D­associated risk variants, suggesting that cells might carry a portion of T1D genetic risk. Interestingly, we identified two T1D lead variants overlapping IREs, in which the risk allele modulated the IRE enhancer activity, exposing a potential T1D mechanism acting through cells. To facilitate the access to these genomic data, together with other datasets relevant for the pancreatic islet community, we developed the Islet Regulome Browser (http://www.isletregulome.org/), a free web application that allows exploration and integration of pancreatic islet genomic data.