Targeted Protein Degradation Reveals a Repressive Role of Mecom at the CEBPA Locus to Prevent Differentiation in High-Risk Acute Myeloid Leukemia

Increasing evidence suggests that acquiring hematopoietic stem cell (HSC) gene expression programs in acute myeloid leukemia (AML) confers a particularly poor prognosis and increases the risk of relapse. However, the precise basis by which high-risk features emerge and promote HSC gene expression programs remains undefined. Acquisition of HSC gene expression programs in a subset of AML is driven by increased expression of MECOM, a transcription factor that plays a key role in HSC self-renewal. To decipher how MECOM alters gene expression in AML, we engineered multiple human cell models of high-risk MECOM-driven AML by tagging the endogenous MECOM locus with an FKBP12^F36V degron to enable targeted and rapid degradation of MECOM by addition of the small molecule dTAG. Importantly, these models, for the first time, enable molecular profiling studies in the absence of cell state alterations that can confound the interpretation of MECOM’s direct function. Treatment of these models with dTAG^V-1 induced acute MECOM degradation resulting in a significant increase in chromatin accessibility at sites of MECOM chromatin occupancy and increased the expression of linked genes that promote differentiation. These findings suggest a direct mechanism for MECOM in repressing cis-regulatory elements and target genes in AML that prevents differentiation.

Remarkably these MECOM-repressed cis-regulatory elements and linked gene networks are conserved in primary patient AML samples. Utilizing single-cell RNA-seq and single-cell ATAC-seq data from pediatric AML cohorts, we stratified patient samples by MECOM expression status and compared chromatin accessibility and cognate gene expression levels. Strikingly, AMLs with elevated MECOM expression exhibit significantly decreased regulatory element accessibility and gene network activity relative to AMLs without increased MECOM expression. Further, when stratifying the AMLs by cell states, we found that these MECOM-repressed gene regulatory networks are depleted in primitive HSC-like blasts and are enriched in more differentiated cell states, corroborating a role for MECOM in repressing differentiation programs in primary AMLs.

Nonetheless, the functional relevance of the putative repressive activity of MECOM in AMLs remains undefined. We therefore employed pooled CRISPR inhibition (CRISPRi) to repress MECOM-bound cis-regulatory elements following dTAG-mediated MECOM degradation to attempt to rescue the loss of CD34⁺ HSC-like AML progenitors seen with MECOM loss. Functional dissection of more than 550 MECOM-occupied sites revealed that the individual repression of a single regulatory element 42 kb downstream of the hematopoietic transcription factor CEBPA is sufficient to maintain stem-like cell states in the absence of MECOM. Furthermore, we show that CRISPR activation (CRISPRa) directed towards this CEBPA cis-regulatory element in the absence of MECOM ablation is sufficient to differentiate AML progenitor cells.

To validate the importance of this CEBPA cis-regulatory element and confirm its role as the key functional node amongst MECOM-regulated loci we performed multiplexed genetic perturbation studies in primary inv(3) AML patient samples. CRISPR/Cas9-mediated perturbation of MECOM promoted differentiation of CD34⁺ stem-like blasts across multiple patient samples, which could be almost completely rescued through concomitant disruption of the CEBPA regulatory element. Additionally, we used dCas9-VPR mRNA to transiently activate this site without MECOM perturbation to induce significant differentiation of these MECOM-driven leukemia stem cells and validate a key role for MECOM in preventing differentiation by repressing a single key pro-differentiation cis-regulatory element.

Through targeted degradation of MECOM coupled to functional genomic readouts, we uncover a previously unappreciated and simple regulatory logic underlying MECOM’s role in promoting stem cell-like states in high-risk AML. Remarkably, repression of CEBPA expression by MECOM is necessary and sufficient to confer high-risk stem cell-like states in AML. These observations suggest a therapeutic opportunity to promote differentiation in high-risk AMLs, which could enable cures to be achieved in these previously intractable malignancies, akin to the remarkable success of retinoids to achieve cures in acute promyelocytic leukemia.