Somatic mutations and chromosomal translocations of genes have emerged as major drivers in a range of hematopoietic malignancies. While ASXL1 is mutated in all forms of myeloid malignancies, ASXL2 is specifically mutated in t(8;21) AML patients. ASXL1 and ASXL2 mutations are mutually exclusive in t(8;21) AML. Despite the importance of ASXL2 mutations in clinical, it's role in leukemogenesis remain unknown. In the current study, we sought to dissect the role of ASXL2 in normal hematopoiesis and to identify the molecular mechanisms by which Asxl2 loss contributes to myeloid malignancies.
In the current study, we utilized a mouse model of Asxl2 to characterize the hematopoietic features of in vivo. Asxl2-/- mice were characterized by pancytopenia and dysplastic features, including hyposegmented (bilobed) neutrophils with fine nuclear bridging (consistent with pseudo Pelger-Huët) and increased number of polychromatophilic red blood cells (RBCs), reminiscent of myelodysplastic syndrome (MDS). Flow cytometric analyses revealed that Asxl2-/- mice had an increased proportion of granulocytic/monocytic cells (Gr-1+/Mac1+) in the PB, BM and spleens compared to WT mice. The histologic analysis of the Asxl2+/- and Asxl2-/- spleen sections showed disrupted splenic architecture with an increased proportion of myeloid cells and massive accumulation of myeloperoxidase (MPO) positive cells in WT spleens. Asxl2-/- mice had an increased long-term (LH)-HSCs and granulocyte-macrophage progenitor (GMP) cells compared to WT mice.Consistently, the paired-daughter cell assays revealed that Asxl2-/- CD34-LSK BM cells had a higher proportion of cells with symmetric self-renewal capacity (SS, 62%) than WT cells (33%). In contrast, a significant reduction in the cells with symmetric differentiation potential was observed in Asxl2-/- HSCs (18%) compared to WT HSCs (40%), indicating a critical role of ASXL2 in the balance between the symmetric and asymmetric division of HSCs. Transplantation assays revealed that recipients transplanted with Asxl2-/- and Asxl2+/- bone marrow cells had shortened lifespan due to the development of MDS or AML, suggesting a cell-autonomous effect of Asxl2-loss in HSC/HPC functions. Furthermore, Asxl2-loss further increase the colony-forming potential and colony replating capacity of AML1-ETO expressing HSCs in vitro, suggesting a cooperative effect between AML1/ETO9a and Asxl2+/-to promote HSC self-renewal. RNA-seq analysis showed a unique signature of Asxl2-/- LK cells compared to WT LK cells. Gene set enrichment analysis revealed that altered expressed genes in Asxl2-/-LK cells were enriched in myeloid cell differentiation, hematopoiesis, apoptosis, and chromatin/nucleosome assembly signature. ChIP-seq analysis showed that differentially expressed genes were associated with dysregulated histone enhancer markers, including H3K27ac, H3K4me1, and H3K4me2. Further analysis demonstrated that the alteration of H3K27ac enrichment had a greater impact on gene expression, in comparison to H3K4me1/2. KEGG pathway analysis showed that genes with differential H3K27ac signals were enriched for hematopoietic cell lineage, cancer signaling pathway and myeloid leukemia development. IPA analysis further confirmed that genes with altered enrichment levels of were enriched in myeloid cell differentiation and apoptosis pathways. Altogether, these data suggest that ASXL2 regulates gene expression mainly through enhancer markers.
Our results demonstrate that ASXL2 plays an important role in normal hematopoiesis, and Asxl2-loss in mice is sufficient to cause MDS-like disease and leukemia transformation. These results indicate that ASXL2 functions as a tumor suppressor in myelopoiesis. The Asxl2 knock-out mice present an ideal model for unveiling the mechanisms underlying the Asxl2-loss mediated multiple-step pathogenesis of myeloid malignancies and for testing novel therapeutic agents for myeloid malignant patients with ASXL2 alterations. Further studies to dissect the possible roles of ASXL2alterations in leukemogenesis and to identify therapeutic vulnerabilities they may create are ongoing.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.