Abstract

Core-binding factors (CBFs) are heterodimeric transcriptional factors consisting of a DNA-binding Runx1 (CBFα) subunit and a CBFβ subunit. Cbfβ allosterically increases the affinity of Runx1 for DNA ~2.5 fold. CBF subunits are encoded by four genes in mammals. RUNX1 (AML1), RUNX2, and RUNX3 encode for CBFα subunits, and CBFB encodes the CBFβ subunit. Homozygous disruption of either the Runx1 or the Cbfb genes in mice results in essentially identical phenotypes: midgestation embryonic lethality accompanied by extensive hemorrhaging and a profound block at the fetal liver stage of hematopoiesis. In humans, chromosomal rearrangements that disrupt the Runx1 and CBFB genes are associated with a significant percentage of leukemias. CBFβ is disrupted in acute myeloid leukemia by inv(16)(p13;q22), t(16;16), and del(16)(q22). These translocations result in the production of novel fusion proteins containing most of the CBFβ protein fused to the C-terminal coiled-coil domain from smooth muscle myosin heavy chain (SMMHC) encoded by the MYH11 gene. A knock-in of the CBFB-MYH11 allele in mice resulted in embryonic lethality with a profound block in hematopoietic development, the same phenotype observed for the Runx1 and Cbfb knockouts. We recently demonstrated that the CBFβ-SMMHC fusion protein binds to the DNA binding Runt domain from Runx1 with both higher affinity and altered stoichiometry relative to native CBFβ. We also provided NMR-based evidence for multiple sites of contact between Runx1 and CBFβ-SMMHC, proving the role of the SMMHC sequence in creating this altered affinity. Here we demonstrate that CBFβ-SMMHC inhibits DNA binding of the Runx1 Runt domain by ~6-fold for the CD4 dual-site silencer element. Cross-saturation NMR mapping on the Runt domain in complex with CBFβ-SMMHC reveals that the SMMHC portion of the oncoprotein makes contacts with β-strands 1 and 2 in the Runt domain. We propose that the inhibition of DNA-binding and increased affinity combine to mediate the dysregulation of Runx-regulated genes caused by CBFβ-SMMHC. These results also clearly suggest that targeting of the CBFβ-SMMHC protein for drug development may well be a viable approach for the treatment of the associated leukemia.

Author notes

Corresponding author