CBFA2T3::GLIS2 directly represses differentiation and is required for AMKL disease maintenance Free

Acute megakaryoblastic leukemia (AMKL) is a rare form of AML that largely affects infants and toddlers. Among the most aggressive subtypes is that driven by the oncogenic fusion transcription factor CBFA2T3::GLIS2. Despite intensive chemotherapy and bone marrow transplantation, the 5-year overall survival for this leukemia remains approximately 20%.

Previous studies have demonstrated that the CBFA2T3::GLIS2 fusion oncoprotein is sufficient to induce leukemogenesis. The necessity of the fusion in leukemia maintenance, however, remains under explored. To address this question, we first generated a panel of Cas9 expressing fusion-positive AML cell lines and introduced gRNAs to genetically ablate the fusion transcription factor. Our results demonstrated a significant effect on cell growth in various assays, both in vitro and in vivo, confirming the requirement of the fusion for disease maintenance. Phenotypically, loss of the fusion led to differentiation either towards megakaryocytic/erythroid, monocytic, or a mixture of both lineages, depending upon the cell line analyzed.

At a molecular level, the oncogenic fusion leads to a single open reading frame wherein the N-terminus of CBFA2T3, a chromatin-associated factor which can recruit various other transcription factors and epigenetic regulators, is fused to the DNA-binding domain of the GLIS2 transcription factor. How this promotes leukemogenesis and the core transcriptional program directly regulated by the fusion are not well understood. One model proposes that the fusion, albeit directly or indirectly, suppresses GATA1 and upregulates ERG, thereby blocking differentiation and promoting self-renewal, respectively.

Our initial investigation involved performing RNA-Seq after Cas9-mediated genetic ablation of CBFA2T3::GLIS2 in our various Cas9 AML models. Knockout of the fusion resulted in approximately equal numbers of genes upregulated and downregulated. Importantly, we identified a core transcriptional network that is altered upon loss of the fusion, which included transcription factors involved in myeloid, erythroid, and megakaryocytic lineage progression, which were largely repressed by the fusion.

While powerful, an inherent limitation in this system is that genetic ablation happens on the timescale of days, and thus whether these effects are directly mediated by the fusion or are indirect, secondary consequences could not be elucidated. We therefore generated CBFA2T3::GLIS2 dTAG AMKL models which allow for rapid degradation of the fusion within an hour to assess the most immediate effects on the transcriptional and chromatin state that occur with fusion loss. To do so, we deployed CRISPR-directed homologous recombination to append an FKBP12^F36V to the C-terminus of the endogenous CBFA2T3::GLIS2 locus. Subsequently, treatment with the bivalent dTAG compound induced rapid proteasomal degradation of the fusion protein. Characterization of these engineered human cell lines reproducibly confirmed a strongly deleterious impact on growth following degradation of the fusion via a differentiation to death phenotype, as well as upregulation of GATA1 and downregulation of ERG as observed with CRISPR-mediated fusion knockout.

With these systems in hand, we then characterized the most immediate transcriptional changes upon loss of the fusion (i.e., the direct target genes) using PRO-Seq to capture nascent RNA transcription. Furthermore, we mapped areas of high fusion occupancy on chromatin and changes in the epigenetic state after fusion degradation using ChIP-Seq and ATAC-Seq. Strikingly, our PRO-Seq analysis demonstrated that the fusion plays primarily a repressive role with 10-fold more genes being repressed rather than activated. Among these are the key megakaryocytic, erythroid, and monocytic lineage-promoting transcription factors identified above. Importantly, this largely repressive role is masked in bulk RNA-sequencing likely due to indirect effects that are captured. Current efforts are focused on defining how exactly the fusion mediates gene repression at a chromatin level as well as identifying key cooperating factors. This may open new avenues to block this process and thereby promote leukemic cell differentiation for therapeutic gain.