Acute myeloid leukemia (AML) is often associated with specific, recurrent chromosomal abnormalities, such as the inversion of chromosome 16 (Inv(16)) which is associated with subtype M4 with eosinophilia. This inversion creates a fusion between CBFB and MYH11, which encode Core Binding Factor beta and Smooth Muscle Myosin Heavy Chain, respectively. The resulting fusion gene, CBFB-MYH11, is known to be the initiating factor in Inv(16) AML, but its mechanism is not clear. Previous studies indicated that repression of RUNX1 is a potential mechanism. However, we found that Cbfb-MYH11 has activities independent of Runx1 repression. During primitive hematopoiesis, we showed that expression of Cbfb-MYH11 in knockin mouse embryos (Cbfb+/MYH11) caused defects in differentiation that were not seen in embryos nullizygous for Runx1 (Runx1−/−), indicating that Cbfb-MYH11 has activities in addition to the repression of Runx1. Moreover, we found that the defects in the primitive hematopoiesis were rescued in the Cbfb+/MYH11; Runx1−/− embryos, which suggests that Runx1 is required for Cbfb-MYH11 activity during primitive hematopoiesis.
We next asked whether Cbfb-MYH11 was similarly dependent on Runx1 during definitive hematopoiesis. For this purpose we used mice expressing another allele of Runx1 in which a 3'-truncated Runx1 is fused to the b-galactosidase gene, lacZ (Runx1lzd). This Runx1 allele has been reported to have dominant negative activities. Using an in vitro promoter assay, we found that co-expression of Cbfβ with Runx1 and Runx1-lzd resulted in decreased activation of the MCSFR promoter as compared to co-expressing Cbfβ and Runx1, indicating that Runx1-lzd has dominant negative activities. In addition, we found that expression of a single Runx1-lzd allele rescued the primitive blood defect in the Cbfb+/MYH11 embryos. Runx1+/lzd; Cbfb+/MYH11 embryos showed almost normal definitive hematopoiesis providing further evidence that Runx1-lzd has dominant negative activity.
Previously we showed that induction of Cbfb-MYH11 results in a distinct population of pre-leukemic cells. By combining the Runx1-lzd allele with an inducible allele of Cbfb-MYH11, we examined the requirement for Runx1 activity in the production of pre-leukemic cells. We found that 7 days after induction of Cbfb-MYH11, Runx1+/lzd; Cbfb+/MYH11 mice showed a statistically significant decrease in the number of pre-leukemic cells as compared to Runx1+/+; Cbfb+/MYH11 mice. We also found a statistically significant decrease in BrdU incorporation in the bone marrow of Runx1+/lzd; Cbfb+/MYH11 mice as compared to Runx1+/+; Cbfb+/MYH11 mice. This indicates that Runx1 is important for Cbfb-MYH11 activity in adult hematopoietic cells. Consistent with this idea, we found that adult mice expressing Cbfb-MYH11 and the Runx1-lzd allele showed a significant delay in the development of leukemia as compared to their Cbfb+/MYH11; Runx1+/+ littermates. Collectively, this work implies that RUNX1 is important for CBFB-MYH11 activity and that inhibitors of RUNX1 have potential use for the treatment of Inv(16) leukemia.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.