HDAC1 Cooperates with CBFβ-SMMHC in Regulating Gene Expression in Inversion 16 Acute Myeloid Leukemia

The fusion of the genes for core binding factor beta and smooth muscle myosin heavy chain (CBFB-MYH11) is the recurrent mutation found in inversion 16 (inv(16)) acute myeloid leukemia (AML). The expressed fusion protein, CBFβ-SMMHC, binds to the transcriptional regulator RUNX1, and this interaction is required for leukemogenesis. Recent data shows CBFβ-SMMHC and RUNX1 are associated with promoters of both transcribed and repressed genes, implying that the CBFβ-SMMHC:RUNX1 complex directly regulates target gene expression. However, it is not known whether other transcriptional co-factors are also required for this activity.

Histone deacetylase 1 (HDAC1) removes acetyl groups from histone tails to regulate the accessibility of chromatin to transcriptional machinery. It is recruited to chromatin by transcription factors, including RUNX1. HDAC1 also colocalizes with RUNX1 and CBFβ-SMMHC to promoter regions in ME-1 cells, a human inv(16) cell line. Based on this, we hypothesized that HDAC1 could bind to the RUNX1: CBFβ-SMMHC complex and plays a role in transcriptional regulation in inv(16).

To test if CBFβ-SMMHC and HDAC1 form a complex, we transfected COS-7 cells with expression plasmids for HDAC1-FLAG and CBFβ-SMMHC and performed immunoprecipitations (IP) with nuclear extracts. IP with anti-MYH11 showed an interaction between HDAC1 and CBFβ-SMMHC, as did IP with anti-FLAG. Importantly, we found that HDAC1 and CBFβ-SMMHC co-immunoprecipitate in mouse leukemia cells from our knockin model which expresses CBFβ-SMMHC from the endogenous CBFβ promoter (CBFβ-SMMHC+). Confirming the specificity of this interaction, we found that IP with anti-MYH11 in ME-1 cells shows HDAC1 interaction, but the same IP in the t(8;21) AML cell line Kasumi-1 did not show HDAC1 interaction even though similar levels of HDAC1 are expressed.

We next tested whether RUNX1 mediates the interaction between CBFβ-SMMHC and HDAC1. We performed IP experiments using a CBFβ-SMMHC mutant lacking RUNX1 binding (CBFβ-SMMHC_N63K,N104K,_Δ179-221). This mutant was co-immunoprecipitated with HDAC1, but not RUNX1, indicating that CBFβ-SMMHC's interaction with HDAC1 does not require RUNX1. We tested a construct lacking the c-terminal 95 amino acids, CBFβ-SMMHC_ΔC95, the domain known to interact with the related protein, HDAC8. We found that HDAC1 immunoprecipitated with CBFβ-SMMHC_ΔC95, suggesting that HDAC1 binds to a unique region of CBFβ-SMMHC.

To test if HDAC1 plays a role in CBFβ-SMMHC-mediated gene expression, we performed chromatin immunoprecipitations on mouse CBFβ-SMMHC+ primary leukemia cells with antibodies against HDAC1, RUNX1, and MYH11, followed by real-time PCR for the promoter regions of three CBFβ-SMMHC target genes: MPO, CSF1R, and CEBPD. We observed all three proteins enriched on the target gene promoters as compared to immunoglobulin controls. This indicates that HDAC1 localizes with CBFβ-SMMHC and RUNX1 on target gene promoters in mouse primary leukemia cells. To test if HDAC1 is required for expression of these target genes, we used shRNA to knockdown Hdac1 expression. Mouse CBFβ-SMMHC+ leukemia cells were transduced with one of 2 different shRNAs against Hdac1 or with a control construct. We found that expression of all three genes was decreased with Hdac1knockdown, implying that HDAC1 is required for CBFβ-SMMHC induced changes in gene expression. These results also suggest that HDAC1 may have a role in transcriptional activation for certain genes, which is in contrast to its traditional role as a transcriptional repressor.

These findings imply that HDAC1 activity is required for the maintenance of CBFβ-SMMHC expressing leukemia cells, and that HDAC1 inhibitors may be effective against inv(16) AML. To test this possibility we performed colony-forming assays using mouse leukemia cells grown in the presence of two different HDAC inhibitors, entinostat which is specific for HDAC1, and vorinostat, a nonspecific HDAC inhibitor. Our preliminary results indicate that both entinostat and vorinostat reduce the ability of primary CBFβ-SMMHC+ mouse leukemia cells to form colonies as compared to the vehicle control, while having minimal effects on growth of normal hematopoietic cells. In summary, we demonstrated that HDAC1 forms a complex with CBFβ-SMMHC and is required for its regulation of target gene expression, and that HDAC inhibitors may be effective for the treatment of inv(16) AML patients.