Distinct Roles of TET Proteins in the Regulation of Normal and Disordered Human Erythropoiesis

TET family proteins (TET1, TET2 and TET3) have recently emerged as important epigenetic modifiers by catalyzing the conversion of 5-methylcytosine (5mc) to 5-hydroxymethylcytosine (5hmc). Although they have been documented to play important roles in a variety of biological processes, their function in erythroid differentiation has yet to be defined. In the present study, we show that of the three TET family members, TET2 and TET3 but not TET1 are expressed in erythroid cells and that TET3 is more abundantly expressed than TET2. Using shRNA-mediated knockdown approach we explored the role of TET proteins in erythroid differentiation of CD34+ human cells. We first showed that consistent with their role in the production of 5hmc, knockdown of either TET2 or TET3 led to a decrease in global 5hmc levels as assessed by mass spectrometric analysis. However, knockdown of TET2 or TET3 resulted in distinctly different phenotypic changes during erythropoiesis. Knockdown of TET3 in human CD34⁺ cells resulted in impaired cell growth which is accompanied by increased apoptosis of late stage erythroblasts. Knockdown of TET3 also led to generation of bi/multinucleated polychromatic/orthochromatic erythroblasts which is accompanied by impaired enucleation. To explore the molecular mechanisms for the observed phenotypic changes, we performed RNA-seq analysis on control and TET3 knockdown erythroblasts at same stages of development. Bioinformatics analysis revealed that the expression levels of several apoptosis-promoting genes such as FOXO1, TNFRSF10B, TGFB1 and BTG1 are increased and that of a mitosis/cytokinesis associated gene KLHDC8B is decreased in polychromatic and orthochromatic erythroblasts following TET3 knockdown. Measurement of 5hmc and 5mc at promoter region of KLHDC8B locus revealed decreased 5hmc level concurrent with increased 5mc level. Importantly, knockdown of KLHDC8B in CD34+ cells as with knockdown on TET3 led to generation of increased numbers of bi/multinucleated polychromatic/orthochromatic erythroblasts and impaired enucleation implying a role for this protein in cytokinesis of late stage but not early stage erythroblasts. These findings demonstrate that TET3 regulates erythropoiesis in a stage-specific manner by targeting different set of genes. Importantly, knockdown of TET2 led to phenotypic changes that were very different from that seen following knockdown of TET3 but the observed changes are similar to the erythroid development defects noted in myelodysplastic syndromes (MDS). These include hyper-proliferation of early stage erythroid cells; delayed terminal erythroid differentiation and increased apoptosis of late stage erythroblasts. Together with the fact that TET2 gene mutation is one of the most common mutations in MDS and dyserythropoiesis is a hallmark of this disorder， our findings suggest that TET2 gene mutations can directly account for dyserythropoiesis of MDS. Our findings demonstrate distinct and important roles for TET2 and TET3 in regulating erythropoiesis and provide significant new and novel insights into epigenetic regulation of erythropoiesis at distinct development stages. The findings are likely to be very useful for furthering our understanding of epigenetic regulation of normal and disordered human erythropoiesis.