TET-dependent DNA methylation patterns in mammaliandevelopment and disease

Abstract

TET enzymes are relatively novel players in the epigenetic regulation of mammalian DNA methylation. They participate in DNA demethylation, but their precise roles in different developmental and disease scenarios are not fully understood. The aim of this work was to investigate the biological roles of TET enzymes in lineage-committed normal and cancer cells. To this end, murine primary cells with genetic deletion of TET enzymes and human cancer cells with recurrent mutations in the cofactor providing isocitrate dehydrogenases (IDH), provoking competitive inhibition of TET enzymes, were analyzed. By characterizing mouse embryonic fibroblasts adipogenic differentiation defects, inefficient activation of genes relevant to adipogenesis and widespread gene deregulation upon TET1/2-deficiency were discovered. Examination of the genome-wide DNA methylation landscape demonstrated the hypermethylation of DNA methylation canyons as a main characteristic of the TET1/2-deficient methylome. Canyons were associated with developmentally important genes and canyon collapse due to hypermethylation coincided with developmental gene deregulation, defective induction of adipogenic markers and the hypermethylation of their promoters. Together, these findings uncovered a novel epigenetic regulatory role in the maintenance of DNA methylation canyons for TET1 and TET2 that is essential for epigenetic programming during differentiation. In the second part of this thesis, published array-based DNA methylation profiles of a large acute myeloid leukemia (AML) patient cohort were used to examine mutant IDH- (mIDH) and TET-dependent DNA methylation changes. This confirmed the known association between mIDH and genome-wide hypermethylation. However, similar global methylation changes were not present in TET2 mutant patients and mIDH carrying patients lacked specific canyon hypermethylation. Intriguingly, neither overexpression of mIDH, nor treatment of a leukemia cell line with D-2-hydroxyglutarate, which is a putative TET inhibitor produced by mIDH, recapitulated the mIDH-associated hypermethylation. Instead, comparison with hematopoietic reference methylomes revealed high similarity between mIDH-associated and myeloid progenitor methylation profiles, suggesting the involvement of differentiation state rather than TET inhibition in the hypermethylation phenotype. These findings implicate a previously unnoted factor in the epigenomic changes of AML cells with mIDH, which may be critical to understand and therapeutically target mIDH-dependent pathogenesis.