Dissecting the mechanisms of blocked differentiation in erythroleukemia

Acute erythroleukemia (AEL) is characterized by uncontrolled accumulation of poorly differentiated progenitor cells of the erythroid lineage. The overall aim of this thesis was to characterize molecular mechanisms that are responsible for impaired differentiation in AEL.
Our laboratory studies molecular lesions initiating and/or maintaining acute myeloid leukemia (AML). While studying a chromosomal translocation t(5;11)(q35;p15) leading to the expression of a fusion between the nuclear pore protein 98 (NUP98) and the nuclear receptor interacting SET domain protein 1 H3K36 histone lysine methyltransferase (NSD1) associated with pediatric AML, we wondered about the function of NSD1 in normal hematopoiesis. To address this question, we first assessed the effects of knocking down NSD1 in human CD34+ cells and found increased self-renewing capacity and the formation of dense reddish colonies with a proerythroblast-like morphology. Unfortunately, deeper functional studies were hampered by the fact that cells could not be expanded in culture. Therefore, a mouse model was created in which we conditionally ablated the Nsd1 gene in the hematopoietic system of the mouse. We found that inactivation of Nsd1 during late fetal development unexpectedly induced a fully penetrant and transplantable disease closely mimicking human AEL. To identify the underlying molecular mechanisms, we aimed to rescue the phenotype by viral expression of either Nsd1-WT or a catalytically SET-domain mutant Nsd1. Only the Nsd1-WT ORF transduced Nsd1-/- erythroblasts restored terminal differentiation associated with increased chromatin binding and activation of target genes of the erythroid master regulator GATA1. While the loss of the Nsd1 catalytical domain severely impaired GATA1 transactivation activity and erythroid differentiation, overexpression of exogenous GATA1 was sufficient to partially overcome the differentiation block.
To better understand the molecular mechanisms that control human AEL, we, in collaboration with the group of Thomas Mercher, have characterized the genetic and transcriptional landscape of 33 AEL patients. Strikingly, >25% of AEL patients aberrantly expressed transcriptional co-regulators such as SKI, ERG, and CBFA2T3, which when ectopically expressed in WT murine erythroid progenitors blocked erythroid differentiation and functionally interfered with GATA1 activity resulting in decreased chromatin accessibility of GATA1-binding sites, suggesting that human AEL could be driven by GATA1 dysfunction. Since multiple studies have shown that terminal erythroid maturation is controlled by GATA1 acting in transcriptionally active and repressive complexes, we primarily focused on investigating differences between normal and malignant erythroblasts, by looking into their proteome and GATA1 interactome. To identify novel regulators that control erythroid differentiation, we performed a targeted CRISPR/Cas9 screen of genes that encode for proteins overlapping between the proteome and GATA1-IP/MS analysis.