Loss of BCLAF1 Impairs AML Pathogenesis

Hematopoietic stem cells (HSCs) have intrinsic self-renewal properties to maintain their population and support life-long hematopoiesis. These self-renewal programs can be pathologically co-opted in acute myelogenous leukemia (AML) cells to promote initiation and maintenance of the malignant population. We recently identified that BCLAF1, a transcriptional regulator, has critical functions in HSC self-renewal. Specifically, loss of BCLAF1 impairs HSC repopulation and multilineage hematopoiesis. Both fetal and adult Bclaf1-deficient HSCs are defective in reconstituting hematopoiesis in competitive transplantation experiments. BCLAF1 has been shown to be upregulated in pediatric and adult AML, and previous literature demonstrated that suppression of BCLAF1 is detrimental to AML. However, the function of BCLAF1 in AML pathogenesis has not been determined. Our objective was to define the activity and mechanism of BCLAF1 in leukemic transformation and progression. Based on our previous findings in HSCs, we hypothesized that BCLAF1 is essential for AML leukemic transformation by promoting self-renewal of malignant cells. To investigate this, we used conditional and inducible models of Bclaf1-deletion in MLL-AF9-expressing AML. Bclaf1^f/f or Vav-Cre:Bclaf1^f/f hematopoietic progenitors were transduced with a retrovirus expressing MLL-AF9, then transplanted into lethally irradiated recipients. As expected, all mice that received transduced control Bclaf1^f/f cells developed AML. In contrast, only 50% of mice that received transduced Bclaf1-deficient (Vav-Cre:Bclaf1^f/f) cells developed AML. To further assess BCLAF1 function in AML, we treated Bclaf1^f/f and Mx-Cre:Bclaf1^f/f mice with pIpC to induce Bclaf1 deletion, then transduced hematopoietic progenitor cells with MLL-AF9 and transplanted them into lethally irradiated recipients. Interestingly, mice that received Bclaf1^f/f or Mx-Cre:Bclaf1^f/f transduced cells developed AML at similar incidences. However, genotyping demonstrated all Mx-Cre:Bclaf1^f/f AMLs were haploinsufficient for Bclaf1 (Bclaf1^f/-). No Bclaf1-deficient (Bclaf1^-/-) AMLs propagated from Mx-Cre:Bclaf1^f/f, despite 85% of hematopoietic progenitors losing both floxed alleles after pIpC treatment. These primary haploinsufficient Mx-Cre:Bclaf1^f/- and control Bclaf1^f/f AMLs were adoptively transferred into new recipient mice that were then treated with pIpC two weeks later. Again, AML progressed equivalently in recipients of Bclaf1^f/f or Mx-Cre:Bclaf1^f/- AML cells, and genotyping showed all Mx-Cre:Bclaf1^f/- AMLs remained haploinsufficient. To validate these results, we expanded haploinsufficient Mx-Cre:Bclaf1^f/- cells in vitro and treated them with Tat-Cre recombinase, which successfully deleted the remaining floxed allele. The resultant Bclaf1-deficient (Mx-Cre:Bclaf1^-/-) AML cells had reduced colony formation in semisolid culture compared to haploinsufficient (Mx-Cre:Bclaf1^f/-) AML cells. These findings support that loss of BCLAF1 is deleterious for AML. Next, we used RNAi to suppress BCLAF1 in human K562 AML cells. K562 cells expressing shBclaf1 were selectively depleted compared to cells expressing a non-targeting shRNA. In complementary studies, we used CRISPR to knock-out BCLAF1 in K562 AML cells. K562 cells expressing Cas9 and a Bclaf1 gRNA formed fewer colonies than cells expressing Cas9 and a non-targeting gRNA. Recovered colonies expressed BCLAF1 despite loss of BCLAF1 in bulk CRISPR-generated BCLAF1-deficient K562 cells prior to subcloning. Given that BCLAF1 is a transcriptional regulator, we investigated the gene expression program in Bclaf1-deficient versus wild-type murine MLL-AF9-expressing AML cells. Bclaf1-deficient AML cells had a dramatically altered genetic program. Differentially expressed genes in Bclaf1-deficient AML cells were functionally associated with altered metabolic, apoptotic, and myeloid differentiation programs. Ongoing studies will further delineate BCLAF1-mediated transcriptional programs and chromatin binding patterns of BCLAF1 in AML. Collectively, these studies demonstrate that BCLAF1 has critical functions in AML and that loss of BCLAF1 impairs AML pathogenesis.

No relevant conflicts of interest to declare.