Targeted Therapy for AML Expressing RUNX1 Mutation

RUNX1 is the DNA-binding subunit of the core binding factor (CBF) complex and a transcription factor (TF) involved in normal and malignant hematopoiesis. Transcriptionally active full-length RUNX1b isoform is transcribed from 6 exons, encoding a DNA binding 'Runt' homology domain and a more C-terminal transcription activation domain. Within its CTCF-anchored loop or TAD (topologically-associating domain), RUNX1 gene has a highly conserved +24-kb enhancer (eR1 in mouse), which is part of a large super-enhancer (>170 kb) covering the entire intron 1 between proximal P2 and distal P1 promoter of RUNX1 and controlling the transcription of RUNX1. TFs bind to super-enhancers/enhancers that loop over to interact with promoters to transcriptionally regulate oncogenes, such as RUNX1 and its targets, as well as multiple other AML-relevant oncogenes involved in growth and survival of AML blast progenitor cells (BPCs). Heatmap of Hi-C interaction scores within the RUNX1 TAD showed triangle shaped regions of high interaction scores highlighting interactions between +24-kb enhancer and the P2 promoter of RUNX1. To functionally validate these findings, we cloned and transfected into HEK293 cells expressing RUNX1, a luciferase reporter construct containing the DNA of the P2 promoter with or without the +24 kb-enhancer of RUNX1. Our findings confirm that, as compared to the promoter or the enhancer alone, the +24 kb-enhancer-driven P2 promoter exhibited superior luciferase induction representing increased transcriptional activity of RUNX1. Somatic mutations in RUNX1 commonly occur in MDS, as well as in the secondary (s) or de novo AML (~10%). Majority of mutant (mt) RUNX1 are missense, large deletions or truncation-mutations, behaving mostly as loss of function mutations. Presence of mtRUNX1 confers relative therapy-resistance and poorer survival in patients with AML. Lack of specific targeted therapy coupled with resistance to standard chemotherapy may account for the poorer outcome in AML expressing somatic or germline mtRUNX1. Therefore, there is a strong rationale and an unmet need to develop novel targeted therapies for AML expressing somatic or germline mtRUNX1. In the present studies, we demonstrate for the first time that shRNA mediated knockdown of RUNX1 expression inhibits growth and induces differentiation and apoptosis of cultured cell line (OCI-AML5) and patient-derived (PD) AML BPCs expressing mtRUNX1. After engraftment, doxycycline-inducible shRNA-mediated in vivo knockdown of RUNX1 in OCI-AML5 cells, compared to controls, markedly reduced AML burden and improved survival of the immune-depleted (NSG) mice engrafted with luciferase transduced OCI-AML5 cells. Interaction of TFs with DNA and transcriptional co-factors such as the histone acetyltransferase p300 at enhancers and promoters leads to acetylation of histone proteins and TFs. BET (Bromodomain Extra-terminal) protein (BETP) BRD4 binds to acetylated histones and TFs at super-enhancers and recruits transcriptional co-factors to promote RNA pol II (RNAP2)-mediated transcription of oncogenes and their targets. ChIP-Seq analysis with anti-BRD4 and anti-H3K27Ac antibody showed that Runx1 super-enhancer and P2 promoter are occupied by BRD4. shRNA mediated knockdown of BRD4 attenuated expression of RUNX1 and its targets, MYC, PU.1 and myeloperoxidase (MPO), inhibited growth and induced apoptosis of OCI-AML5 cells. We also found that BETP-PROTACs (proteolysis targeted chimera) ARV-825 and its pharmacologically superior version ARV-771 (20 to 250 nM) caused degradation and depletion of BRD4, reduced BRD4 occupancy at the+24-kb-enhancer and P2 promoter of RUNX1, repressing RUNX1 and its targets, and potently induced apoptosis of OCI-AML5, Mono-Mac-1 and PD AML BPCs expressing mtRUNX1. Treatment with ARV-771 (30 mg/kg SQ daily for 3 weeks also inhibited AML growth and improved survival of NSG mice engrafted with luciferase-transduced OCI-AML5 cells. By utilizing the RNA-Seq mRNA signature from RUNX1-depleted AML cells we queried for expression mimickers (EMs) through LINCS1000-CMap (Connectivity Mapping) analyses. This uncovered several novel agents that inhibit RUNX1 expression and induce in vitro lethality and reduce in vivo leukemia burden in AML BPCs expressing mtRUNX1. These findings highlight RUNX1-targeted novel agents to be developed and tested as potential therapy for AML expressing mtRUNX1.