Madison L Hall, Aliya Quintal, Sonia N Acharya, Samantha Worme, Thai T Nguyen, Zachary Schonrock, Mitsuhiro Tsuchiya, Hanqian L Carlson, Itallia V Pacentine, Shawn B Shrestha, Jommel Macaraeg, Randall Armstrong, Nicole Szczepanski, James Adler, James McGann, Sara Evans-Dutson, Theresa A Lusardi, Brendan L O'Connell, Andrew C Adey, Galip Gürkan Yardımcı, Julia E Maxson, Theodore P Braun
Mutations in the epigenetic regulator ASXL1 are common in myeloid malignancies and portend a near-universally poor prognosis. While multiple mechanisms for mutant ASXL1-dependent oncogenesis have been proposed, none have been functionally validated in the context of the human hematopoietic stem cell, where these mutations almost certainly arise. Here, we extensively characterized a CRISPR-engineered human hematopoietic stem and progenitor cell model of ASXL1 mutations. In this context, mutant ASXL1 expression decreases differentiation, increases clonogenicity in serial replating experiments, and improves engraftment in immunodeficient mice. We also show that endogenous truncating ASXL1 mutations stabilize the protein and confirm that mutant ASXL1 resists proteasomal degradation. At the transcriptional level, these phenotypes are driven by significant repression of immediate early genes and subtle global transcriptional upregulation, especially of genes repressed during normal differentiation. Using protein-interaction screens, genomic and functional approaches, we link the phenotypic changes in ASXL1-mutant cells to increased chromatin binding of RNAPII, BRD4, and the transcription factor MECOM. The association between ASXL1 and MECOM may reflect a direct or indirect interaction. We also observe aberrant RNA polymerase II pausing dynamics, especially at immediate early genes, in ASXL1-mutated cells. Finally, we demonstrate that ASXL1-mutant AML exhibits increased MECOM activity, consistent with our gene-editing models. Collectively, these studies highlight a highly reproducible model of mutant ASXL1 in the appropriate cell context. Further, they are the first to functionally describe the mutant ASXL1 interactome in the context of human HSCs, identifying MECOM and BRD4 as actionable dependencies with therapeutic potential for ASXL1-mutant myeloid malignancies.