Cell differentiation, reprogramming and malignant transformation are major events characterized by remarkable changes in the epigenome and involve remodeling of DNA methylation patterns. In cancer tissues, DNA methylation patterns are drastically different from those in normal tissues. Two major events are observed, (i) global DNA hypomethylation in cancer affecting predominantly repetitive DNA sequences, and (ii) gene-specific hypermethylation of CpG islands affecting hundreds of genes. The mechanisms how these cancer-associated DNA methylation patterns arise are largely unknown. In 2009, it was reported that a sixth DNA base, 5-hydroxymethylcytosine, is present in substantial amounts in certain mammalian cell types. 5-hydroxymethylcytosine (5hmC) is created from 5-methylcytosine (5mC) by enzymatic oxidation carried out by the TET family of proteins. One model proposes that 5hmC is an intermediate in DNA demethylation. Our hypothesis is that defects in the 5mC oxidation pathway are responsible for altered DNA methylation patterns in human tumors. We have established methodology for precise quantification and genome-wide mapping of 5mC and 5hmC. Our goal is to determine the level and the genomic distribution of 5hmC in normal human tissues and in malignant tumors. These data will be compared directly with the distribution of 5mC in the same tissues. We will focus primarily on two tumor types: (1) human grade II/III astrocytomas, because these tumors frequently contain mutations in isocitrate dehydrogenases (IDH1 or IDH2), an enzymatic activity potentially impacting on the 5mC oxidation pathway; and (2) myelodysplastic syndrome (MDS), because this malignancy often is characterized by mutations in one of the TET genes, TET2. The third Aim will focus on functional studies of TET and TET-associated proteins and their aberrations in cancer.