Abstract 1

TET2 mutations are frequently found across broad spectrum of myeloid malignancies but how these mutations contribute to diseases is still unknown. Preliminary results from our laboratory have suggested that TET2 converts 5-methylcytosine (5-mC) to 5-hydroxymethylcytosine (5-hmC) and consequently, the levels of 5-hmC may be lower in genomes of mutant bone marrow cells. To facilitate study of TET2 function we developed a blot assay to detect 5-hmC in genomic DNA with a specific antiserum to 5-hmC. In a second improved assay with increased sensitivity and precision, we treated genomic DNA with bisulfite in order to convert 5-hmC to cytosine 5-methylenesulfonate (CMS) and measured 5-hmC levels indirectly using a specific anti-CMS serum.

Based on the results of this technique we demonstrate here for the first time that indeed TET2 mutations in predicted catalytic residues and other positions compromised TET2 function. We studied 102 patients with various myeloid malignancies (4/28 MDS, 14%, 26/48 MDS/MPN, 54% and 1/4 MPN, 2% and primary 2/11 AML 18% and 3/11 sAML, 27% TET2 mutants, respectively) and compared to wt cases or controls (N=17). Mutations were found throughout the entire coding region and were mostly inactivating (33/45 TET2 mutations). The levels of 5-hmC in genomic DNA from TET2 mutants were significantly decreased in comparison to wt cases and controls (p=4.5e-08 and p=1.8e-09, respectively). Particularly low levels of 5-hmC were found in patients with homozygous (UPD)/hemizygous (deletion) TET2 mutations and those with biallelic mutations. Surprisingly, 18% of all TET2 WT patients also showed low levels of genomic 5-hmC (despite normal TET2 mRNA expression), suggesting that these patients may carry not yet identified variants/lesions in TET2 or other partner proteins involved in TET2-mediated catalysis.

To further investigate the impact of TET2 mutations associated with myeloid malignancies we also introduced 9 different missense mutations corresponding to those found in patients into murine Tet2 cells; severe loss of enzymatic activity was observed in 7/9 cases as measured by greatly diminished 5-hmC levels. To study the role of Tet2 in normal hematopoiesis we depleted Tet2 in C57BL/6 mice by retrovirus-mediated transduction of shRNA against Tet2. Tet2 depletion is associated with skewing of hematopoietic differentiation towards the monocyte/macrophage lineage. To further investigate the function of TET2 we transduced the myeloid THP-1 cell line with lentiviral vector containing TET2 cDNA (TET2+) or an empty vector. This manipulation allowed us to select clones showing 19-fold increase in TET2 mRNA expression without significantly alterations of proliferation kinetics. Using this model we studied the impact of TET2 overexpression on resultant methylation pattern of CpG sites. We have applied Illumina Infinium HumanMethylation27 arrays (27,5K CpG sites/14.4K genes). Overexpression of TET2 resulted in a distinct promoter methylation patterns with 169 altered CpG sites with difference of averaged β>0.5 (considered significant as compared to control). Among these differentially methylated loci, 27 promoters were significantly hypomethylated while 42 were hypermethylated as compared to control cells. Change in methylation pattern observed through overexpression of TET2 in vitro prompted us to analyze methylation patterns in patients with and without TET2 mutations or those with decreased 5-hmC levels. Using methylation arrays a total of 62 cases were analyzed. When patients were grouped based on the levels of 5hmC, an associated methylation signature can be clearly discerned with 2512 differentially methylated loci and distinct skewing towards hypomethylation (2510 sites; e.g., TMEM102, ABCC11) vs. hypermethylation (2 sites, AIM2 and SP140), consistent with the observation made in the TET2+ cells line.

In sum, our results provide strong evidence for TET2 as the first mutated gene in myeloid malignancies that is involved in conversion of 5-mC to 5-hmC in DNA, indicating the novel role of TET2 in a substantial component of epigenetic deregulation in myeloid malignancies.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution