KMT2C is one of several tumor suppressor genes deleted in myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) as part of larger chromosome 7q deletions. These deletions are particularly common in therapy-related MDS/AML, raising the question of whether loss of one or more 7q genes conveys a selective advantage to hematopoietic stem cells (HSCs) in the setting of chemotherapy-induced stress. We recently showed that haploid Kmt2c deletions do indeed enhance HSC self-renewal capacity. However, the deletions do not drive HSCs into cycle; instead, a proliferative stress such as chemotherapy is required to create a context in which Kmt2c deletion convey a selective advantage. We have also identified a mechanism, to explain this phenotype. Kmt2c encodes MLL3, a SET domain protein that binds enhancers and facilitates transcription. We have shown that Kmt2c/MLL3 deletions impairs enhancer recruitment during HSC differentiation, therefore blunting HSC commitment.

Our findings suggest that acquired or pre-existing 7q (KMT2C) deletions may select for HSCs that could give rise to MDS/AML in the setting of autologous-transplantation. Granulocyte-colony stimulating factor (G-CSF) is a cytokine that is often used to expedite neutrophil recovery after chemotherapy and to mobilize HSCs for collection and transplantation. We considered the possibility that 7q deletions, and KMT2C deletions in particular, may promote disproportionate mobilization of the mutant HSCs in response to G-CSF. To test this, we treated Kmt2c f/f; Vav1-Cre and Kmt2c f/+; Vav1-Cre mice with G-CSF, and we assessed HSC mobilization to the spleen and bone marrow. A far greater proportion of HSCs with heterozygous and homozygous Kmt2c deletions mobilized in response to G-CSF, relative to wild type HSCs. Kmt2c deletion also enhanced colony forming unit frequency in the peripheral blood after G-CSF treatment. Total body HSC numbers did not change in the body after G-CSF treatment on any genetic background, indicating that Kmt2c deletions enhanced HSC mobilization in response to G-CSF rather than self-renewal.

To more faithfully recapitulate clinical conditions, we used Fgd5-CreER to delete a single Kmt2c allele in only a minority of HSCs. We then tested whether the mutant HSCs mobilized more efficiently than wild type HSCs. Surprisingly, Kmt2c deletions did not enhance HSC mobilization in this context. This raised the question of whether Kmt2c deletion in a non-HSC population could promote HSC mobilization in the Kmt2c f/+; Vav1-Cre mice. Indeed, analysis of mice chimeric for wild type and Kmt2c f/+; Vav1-Cre bone marrow suggested that Kmt2c non-cell autonomously regulates HSC mobilization. Finally, Kmt2c deletions did not enhance mobilization following exposure to plerixafor, a CXCR4 antagonist that acts directly on HSCs to promote mobilization rather than indirectly via monocyte populations, as occurs with G-CSF. Additional studies are needed to elucidate the mechanism by which Kmt2c non-cell autonomously regulates HSC mobilization. Our findings provide reassurance that, in a clinical setting, rare KMT2C-mutant HSCs will not disproportionately mobilize prior to apheresis. Furthermore, the data suggest that transient inhibition of MLL3, or its targets, may enhance HSC mobilization and negate selective advantages that 7q-deleted HSCs may acquire after chemotherapy treatment.


No relevant conflicts of interest to declare.