TRANSCRIPTIONAL AND EPIGENETIC REGULATION OF PANCREATIC B-CELL FUNCTION AND DYSFUNCTION

Abstract

Type 2 Diabetes arises from the inability of pancreatic ß-cells to compensate for insulin resistance induced by environmental factors in genetically predisposed individuals. Goals in diabetic research are to restore the functionality of damaged ß-cells or to replace them, and accomplishing this requires better understanding of the essential determinants of the ß-cell phenotype. Pdx1, NeuroD1 and MafA are transcription factors critical in maintaining mature ß-cell function. Epigenetic changes to the chromosome architecture in ß-cells are associated with the loss of glucose sensitivity and insulin production. My hypotheses are (1) The transcriptional regulation mediated by Pdx1, NeuroD1 and MafA are key determinants of ß-cell function; (2) As ß-cell dysfunction develops, chromatin remodeling adversely affects the function of important ß-cell transcriptional regulators. To test these hypotheses, I have defined the Pdx1, NeuroD1 and MafA transcriptional regulatory network operative in ß-cells, and established the epigenetic profile of normal ß-cells and dysfunctional ß-cells with impaired glucose-stimulated insulin secretion (GSIS). Pdx1 binds more independently, and NeuroD1 and MafA have very similar binding profiles. Genes regulated by these transcription factors are over-represented in metabolic, developmental and cellular processes. Combinatorial depletion coupled with combinatorial co-occupancy of these transcription factors shows that genes with Pdx1 binding sites have the most effect on genes that are regulated. When two or three transcription factors are bound, they have the most effect on their target genes. Pdx1, NeuroD1 and MafA regulates genes involved in GSIS: They repress Itgb1bp2 and Cplx2 to enhance GSIS, and activate Tspyl1, F13a1 and Septin7 to dampen the GSIS response, emphasizing the exquisite level of regulation required for the appropriate response of ß-cells to a glucose stimulus. In order to investigate differential epigenetic states in normal and dysfunctional ß-cells, I profiled marks of active and repressive chromatin in both cellular states. There is a global increase of open chromatin in dysfunctional ß-cells, indicated by the increase in numbers of FAIRE and H3K4me3 peaks and a decrease in H3K27me3 peaks in dysfunctional ß-cells. Motif analysis of FAIRE peaks in dysfunctional ß-cells show a bias for general transcription factors, while motifs in normal ß-cells show enrichments for tissue-specific transcription factors. Genes bound by Pdx1, NeuroD1 and MafA are associated with H3K4me3 in normal ß-cells. In dysfunctional ß-cells however, there is an increase in the association of these transcription factors and their target genes with the repressive H3K27me3 mark. Bivalent domains marked by both H3K4me3 and H3K27me3 in normal ß-cells are enriched for developmental processes and cell regulatory processes. Bivalent chromatin states are modified to different extents in dysfunctional ß-cells and lead to expression changes in genes involved in pancreatic development/ function (e.g. MafB, Oxr1, Camk2b, miR-9, Sox17), genes involved in neuronal development/ function (e.g. Sema3d, Celsr1), Fox transcription factors (e.g. Foxf2, Foxq1), and the potassium channels Kcnc1 and Kcnq1. Little is known about the transcriptional and epigenetic network governing ß-cell function and dysfunction. My thesis therefore provides original insight into the Pdx1-, NeuroD1- and MafA- regulated pathways, and presents novel evidence for chromatin remodeling in ß-cell dysfunction.