Induction of Truncating ASXL1 Mutations in Human CD34+ HSPCs Mimics Human ASXL1-Mutated Clonal Hematopoiesis and Progression to Myeloid Malignancies

Introduction: Mutations in additional sex combs like 1 (ASXL1) are recurrently found in myeloid malignancies including myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML), but can also be found in pre-malignant states such as clonal hematopoiesis of indeterminate potential (CHIP). However, the mechanisms by which these mutations contribute to disease initiation and progression remain unresolved. Initial observations using murine models suggested a loss of function of the ASXL1 protein as relevant to disease. However more recently, several studies have suggested that recurrent mutations of ASXL1 result in a change-of function caused by a truncated ASXL1 protein. Most of these studies have been performed in a murine system and have used over-expression of truncated ASXL1.

Methods: To interrogate the functional relevance of recurrent mutations in ASXL1 in human HSPCs, we generated a novel CRISPR/Cas9-mediated model that allows for the introduction of truncations or complete knockout of ASXL1 into CD34+ cord blood cells. Importantly, this is the first model which efficiently introduces these mutations into the native ASXL1 locus in human cord-blood and bone-marrow derived HSPCs while simultaneously introducing a fluorescent marker corresponding to the genotype, so individual cells can be tracked in vitro and in vivo.

Results: Using this system, we demonstrate that truncated, but not deleted, ASXL1 shows a proliferative advantage, decreased ability to differentiate along the megakaryocyte and erythroid lineages, as well as increased serial replating in vitro. Interestingly, truncation of ASXL1 alone did not result in a myeloid differentiation block in vitro, in line with observations of mutant HSPCs in CHIP. In vivo, CD34+ cells harboring truncation of ASXL1 exhibited myeloid skewing and a proliferative advantage in competitive engraftment studies using a xenograft NSGS mouse model (both p<0.05). Human HSPCs expressing truncated ASXL1 showed polyclonal engraftment without development of overt myeloid disease or impaired lifespan in the majority of the host animals. However, similar to humans with ASXL1 CHIP, some of the engrafted animals (2/30) developed clonal expansion of ASXL1-truncated human HSPCs resulting in a myeloid-biased, proliferative disease-like state with infiltration of non-hematopoietic organs and severe cytopenia which ultimately led to shortened survival of affected mice. Functionally, ASXL1-truncated HSPCs displayed dysregulated gene expression with depletion of RUNX1, TAL1, and GATA1 target genes (Benjamini-Hochberg adjusted p<0.0001). In line with increased activity of the Polycomb repressive deubiquitinase complex, truncation of ASXL1, but not deletion, led to decreased H2AK119Ub in our system (MFI ratio 0.7 for truncated ASXL1 vs. control, p<0.05). Orthogonally, we were able to confirm these observations in several isogenic cell lines engineered to contain truncating ASXL1 mutations as well as correcting the mutation in ASXL1 found in K562 cells (Y591*) using CRISPR/Cas9. Finally, we aimed to determine whether the introduction of additional mutations commonly observed to co-occur with ASXL1 could alter the behavior of human HSPCs and their progression from healthy hematopoiesis to myeloid disease. Indeed, introduction of mutations in the native RUNX1 locus on the background of truncated ASXL1 showed aggressive expansion of immature myeloid progenitors in our xenograft model. This system allows us now to determine the relevant genomic target regions of truncated ASXL1 in human HSPCs, investigate the mechanism of action directly in the relevant biological system and identify therapeutic strategies targeting mutant ASXL1.

Conclusion: In summary, our work demonstrates the utilization of a novel CRISPR/Cas9 model of human ASXL1-mutant HSPCs that closely resembles the phenotype seen in individuals with ASXL1 mutant CHIP and patients with myeloid malignancies carrying mutant ASXL1. Additionally, our model allows us to directly compare the deletion of ASXL1 with the truncation, adding to the evidence that truncated ASXL1 phenocopies pre-malignant and malignant hematopoiesis with recurrent ASXL1 mutations. Finally, we provide the first model for the spontaneous progression of pre-malignant human CHIP to myeloid malignancy, which should enable further studies of leukemogenesis and potentially disease prevention.