Acute myeloid leukemia (AML) is characterized by loss of myeloid differentiation coupled with an excess proliferation and self-renewal of hematopoietic stem and progenitor cells. We and others have shown that ERG is an oncogene. AMLs with high expression of ERG have a stem cell phenotype and are associated with a worse prognosis. Various leukemogenic complexes recruit ERG to drive leukemia. As ERG is a transcription factor, it is difficult to impossible to target. Yet identifying ERG-associated targeted protein is of high priority.
We discovered that the proline at 199 (P199) in the pointed (PNT) domain of ERG protein is essential for ERG’s leukemogenic potential. By a BioID- proximity ligation assay, we discovered that ERG interacts with chromatin modifiers. Most significantly, the mutation (P199L) severely compromised ERG`s interaction with components of the NCoR-HDAC3 co-repressive complex, resulting in a 40% reduction in the number of spectral counts compared with wild-type ERG. We established that ERG interacts with the NCoR-HDAC3 co-repressive complex in a P199-dependent manner to maintain hematopoietic stem and progenitor cells; HoxB8 cells in an undifferentiated state. We further show that chemical and genetic inhibition of HDAC3 suppresses the growth in vitro and abrogates the propagation of ERG-dependent human leukemia in vivo.
In conclusion, we suggest that the interaction of ERG with NCOR-HDAC3 co-repressive complex is required for its oncogenic activity, and inhibition of HDAC3 could be advantageous in AML characterized by high ERG expression, suggesting the possibility of a novel therapy for these very high-risk hematopoietic malignancies.