Dynamic Change in the Stochiometry of ETO2 and SCL Governs the Transition to Terminal Differentiation in Erythroid Progenitors.

As highly proliferative erythroid progenitors commit to terminal differentiation, they also progressively undergo growth arrest. To determine the mechanisms underlying the appropriate timing of erythroid gene expression and those associated with growth cessation, we analyzed the dynamical composition of the multiprotein complex nucleated by the bHLH transcription factor SCL, a crucial regulator of erythropoiesis that absolutely requires interaction with other factors to activate transcription. In progenitor cells, the SCL complex marks a subset of erythroid specific genes (alpha-globin, P4.2, glycophorin A) that are transcribed later in differentiating cells, conducting cells toward terminal maturation. To unravel the regulation of transcription by SCL, we used tagging/proteomics approaches in two differentiation-inducible erythroid cell lines, coupled with binding assays to immobilized DNA templates and chromatin immunoprecipitation. Our analyses reveal that the core complex comprised of known proteins (SCL, GATA-1, LMO2, Ldb1 and E2A) and two additional E protein family members, HEB and E2-2, did not change with differentiation. Strikingly, this complex recruits HDAC1-2 in undifferentiated cells which were exchanged with TRRAP, a chromatin remodelling factor, upon differentiation, suggesting an epigenetic regulation of erythroid differentiation mediated by the core SCL complex. Finally, we identified the corepressor ETO2 targeted via this complex through direct interaction with E2A/HEB. In vivo, ETO2 represses the transcription of SCL target genes both in transient assays and in chromatin. During erythroid differentiation, ETO2 remains associated with the SCL complex bound to erythroid promoters. However, the stoichiometry of ETO2 and SCL/HEB changes as SCL and HEB levels increase with erythroid differentiation, both in nuclear extracts and on DNA. To determine the functional consequence of this imbalance in activator to co-repressor ratio, we delivered ETO2 siRNA in primary hematopoietic cells and found an accelerated onset of SCL target genes on induction of erythroid differentiation, and conversely, these genes were decreased following ectopic ETO2 expression. Strikingly, inhibition of ETO2 expression in erythroid progenitors arrests cell proliferation, indicating that ETO2 is required for their expansion. We therefore analyzed gene expression in purified erythroid progenitors and differentiating erythroid cells (E1-E5) and found an inverse correlation between the mRNA levels of ETO2 and cyclin-dependent kinase inhibitors. Moreover, ETO2 siRNA treatment of primary erythroid progenitors results in increased p21 CDKI and Gfi1b expression, as assessed by real-time PCR. Finally, we show by chromatin immunoprecipitation that Gfi-1b, p21 and p27, are direct targets of the SCL- ETO2 complex. We therefore conclude that ETO2 regulates the erythroid lineage fate by repressing SCL marked erythroid genes in undifferentiated cells, and by controlling the expansion of erythroid progenitors. Our study elucidates the dual function of ETO2 in the erythroid lineage and sheds light on epigenetic mechanisms coordinating red blood cell proliferation and differentiation.