Uncovering the potential of the PA2G4-MYC axis as a target in 3q26 acute myeloid leukemia.

Abstract Type: Oral PresentationSession Title: Acute myeloid leukemia - Biology & translational research 1Background:Despite advancements in next-generation technologies, treating subtypes of acute myeloid leukemia (AML)characterized by aberrantly activated transcription factors (TFs) remains challenging. However, the integrationof orthogonal omics approaches with functional high-throughput screening of small-molecule libraries hasaccelerated the discovery of chemical leads to probe gene dependencies, including TFs traditionally considered‘undruggable’.Aims:Our aim is to identify regulators of the EVI1/MECOM gene, which is the most lethal oncogenic TF and is highlyexpressed in AML with chromosome 3q26 abnormalities.Methods:We intersected phenotypic, gene expression-based, and proteomic approaches to pinpoint EVI1 regulators. Wescreened more than 10000 compounds to test their efficacy in suppressing proliferation of 3q26 AML modelswith high EVI1 expression (EVI1High). Using different EVI1Null transcriptional signatures, we defined “on” and“off” states of EVI1 AML. We then gathered in silico gene-expression based prediction approaches to identifycompounds able to induce the transition to an EVI1 “off” status. We dissected the EVI1 co-transcriptionalcomplex using rapid immunoprecipitation mass spectrometry of endogenous proteins (RIME). Subsequently,target hits were validated through in vitro experiments, in vivo studies on 3q26 AML patient-derived xenograft(PDLX), and N-of-1 clinical trial.Results:Histone deacetylase inhibitors (HDACis) have emerged as the compound class able to impede leukemiaproliferation by suppressing EVI1. In AML models, HDACis, such as AR-42, belinostat, and entinostat, inducedapoptosis and reduced EVI1 expression, with greater sensitivity in EVI1High cells. Based on these findings, westarted an N-of-1 clinical trial proposing entinostat in combination with azacitidine for 3q26 AML patients.Notably, EVI1 expression decreased in blasts within 6 hours of entinostat administration and clinicalimprovements correlated with EVI1 depletion. To exclude azacitidine contribution, we ran a co-clinical trial inPDLX 3q26 model, which demonstrated an EVI1-mediated reduction in AML blasts only with entinostat, notwith azacitidine alone. Moreover, longitudinal single-cell RNA sequencing of 3q26 AML blasts from patientsand 3q26 PDLX twins treated with azacitidine-entinostat showed suppression of Myc target genes, like in celllines treated with HDACis. This suggests that multiple HDACis regulate the Evi1-Myc axis in 3q26 AML. HDACi-mediated or genetic silencing of EVI1 reduced MYC expression. We then defined the EVI1 co-transcriptionalcomplex and its impact on Myc signaling modulation by RIME and identified eleven EVI1 interactors targetedby Myc signaling. We decided to validate PA2G4, an RNA-binding protein modulated by HDACis. Modifyingthe chemical structure of a validated PA2G4 inhibitor, WS6, we generated a corresponding biotinylated formand demonstrated that biotin pulldown recovers the EVI1-PA2G4 complex, indicating that direct targeting ofPA2G4 may interfere with EVI1 function. Silencing or inhibiting PA2G4 decreased EVI1 and consequently MYCprotein levels, similar to the effects of HDACis. Finally, WS6 suppressed leukemic growth in vivo by targeting anEVI1-MYC signature, suggesting that PA2G4 is a druggable mediator within the EVI1 complex.Summary/Conclusion: We advocate for a clinical treatment approach that harnesses the ability of HDACis tosuppress EVI1 in 3q26 AML patients. Additionally, our findings highlight PA2G4 as a potential therapeutic target for 3q26 AML patients and urge the development of PA2G4-MYC-EVI1 complex disruptors in this AMLsubtype.Keywords: Acute myeloid leukemia, Genomics, Chemoresistance, EVI1