Although molecularly and clinically heterogeneous, AML is characterised by aberrant transcription and abnormal epigenetic regulation. The transcriptional co-activators CBP and p300 have well characterised roles in hematopoiesis and hematopoietic stem cell (HSC) function. These are mutated in AML and also bind a number of AML-associated oncogenes. However, despite this, their roles in the induction and maintenance of AML are poorly understood. To address this question, we have combined genetic and pharmacological inhibition of CBP and p300 in AML and normal hematopoiesis.Using a murine model where Cbp was conditionally deleted from murine hematopoietic stem and progenitor cells (HSPC), either prior to, or following expression of a number of AML-associated oncogenes, we assessed the role of Cbp in the induction and maintenance of AML. We demonstrated that although not an absolute requirement, Cbp confers a selective advantage for robust immortalisation in vitro, and that Cbp is also an important requirement for the generation and maintenance of AML in vivo. Furthermore, redundancy between Cbp and p300 in myeloid transformation was demonstrated, as p300 knockdown further decreased proliferation in Cbp-/- AML cells.We next validated CBP/p300 as potential therapeutic targets, using a pharmacological strategy. Using a selective small molecule inhibitor (C646) of the lysine acteyltranferase (KAT) activity of CBP/p300, we demonstrated a significant decrease in growth and clonogenic potential across multiple AML subtypes in vitro. This was mediated through induction of apoptosis and cell cycle arrest. Importantly, no alteration in the growth of normal murine and human hematopoietic progenitors was detected at similar doses. We further demonstrated that inhibition of CBP/p300 KAT activity in human leukemia cells alters a transcriptional programme associated with genomic integrity, linking transcriptional changes to the cellular phenotype. Finally, we demonstrated the efficacy of the HAT inhibitors to decrease clonogenic growth across a panel of primary AML patient samples, representing multiple genetic subtypes. Taken together, these data suggest that CBP/p300 are involved in leukemogenesis across multiple subtypes in AML and that targeting these proteins may be possible with an acceptable side-effect profile. Disclosures:No relevant conflicts of interest to declare.
Read full abstract