Differentiation Induction In Acute Myeloid Leukemia Using Site-Specific DNA-Targeting

Hoxa9 and Meis1 are overexpressed in >70% of acute myeloid leukemia (AML) and associated with poor prognosis and survival. Hoxa9 and Meis1 interact with DNA and PBX to achieve transcription of differentiation-blocking genes. We tested transcriptional repression at Hoxa9-PBX-Meis1 genomic binding sites to induce differentiation in a model of human AML

We designed a DNA-recognition strategy based on the known structure of the Hoxa9-PBX-DNA complex by fusing the DNA binding helices of Hoxa9 and PBX to create concise homeodomain fusion proteins that target the Hoxa9-PBX DNA recognition sequence. To confer transcription-repressing properties to the proteins, we attached a transcriptional repressor (sin3 interacting) domain and ectopically expressed this protein in Hoxa9-Meis1 immortalized murine progenitors. Introduction of this transcription repressor protein significantly enabled cell differentiation versus control (51.2% Mac-1^high Gr-1^high cells versus 11.3% for control). Multiple gene transcripts indicative of differentiation, such as GCSFR, myeloperoxidase, neutrophil elastase, and the calcium binding protein, S100A8, were also elevated in repressor-expressing cells. Furthermore, direct transcriptional targets of Hoxa9 (e.g. SOX2, CD34, FOXP1, FLT3R, DNAJC10) were down regulated in repressor-expressing cells. Importantly, a mutant repressor lacking the DNA-interacting amino acids did not affect transcription of Hoxa9 targets, demonstrating on-target specificity. Repressor-expressing cells also exhibited lower surface expression of c-Kit and Flt3 receptors and when transplanted into mice resulted in a significant increase in disease latency with a 94 day median latency versus 62 day latency for the control group (p value = 0.002).

Our results demonstrate that site-specific DNA-targeting using homeodomain fusion proteins can enable AML cell differentiation and significantly increase disease latency.