Superior efficacy of co-targeting GFI1/KDM1A and BRD4 against AML and post-MPN secondary AML cells

Warren Fiskus,Naval Daver,Michael R Green,Guillermo Garcia Manero,Tapan M Kadia,Prithviraj Bose,Christopher P Mill ,Naveen Pemmaraju,Sunil Sharma,Gautam Borthakur,Dimuthu Perera,Christopher R Vakoc,Lucia Masárová ,Joseph D Khoury,Bernardo H Lara ,Cristian Coarfa,Christine Birdwell,Steven M Kornblau ,Taghi Manshouri,Koichi Takahashi,Behnam Nabet,Courtney D Dinardo ,Srđan Verstovšek ,Matthew C Stubbs ,Xiaoping Su,Guillermo Montalban‐Bravo ,Kapil N Bhalla

doi:10.1038/s41408-021-00487-3

Abstract

There is an unmet need to overcome nongenetic therapy-resistance to improve outcomes in AML, especially post-myeloproliferative neoplasm (MPN) secondary (s) AML. Studies presented describe effects of genetic knockout, degradation or small molecule targeted-inhibition of GFI1/LSD1 on active enhancers, altering gene-expressions and inducing differentiation and lethality in AML and (MPN) sAML cells. A protein domain-focused CRISPR screen in LSD1 (KDM1A) inhibitor (i) treated AML cells, identified BRD4, MOZ, HDAC3 and DOT1L among the codependencies. Our findings demonstrate that co-targeting LSD1 and one of these co-dependencies exerted synergistic in vitro lethality in AML and post-MPN sAML cells. Co-treatment with LSD1i and the JAKi ruxolitinib was also synergistically lethal against post-MPN sAML cells. LSD1i pre-treatment induced GFI1, PU.1 and CEBPα but depleted c-Myc, overcoming nongenetic resistance to ruxolitinib, or to BETi in post-MPN sAML cells. Co-treatment with LSD1i and BETi or ruxolitinib exerted superior in vivo efficacy against post-MPN sAML cells. These findings highlight LSD1i-based combinations that merit testing for clinical efficacy, especially to overcome nongenetic therapy-resistance in AML and post-MPN sAML.

Highlights

Introduction During complete remission inAML, AML stem/progenitor cells (LSCs) in the measurable residual disease (MRD) resist differentiation, retain leukemia-initiating potential and mediate relapse of AML1–3
We employed the dTAG system[34] to degrade LSD1FKBP12(F36V) in OCI-AML5 cells following knockout of the endogenous LSD1 to assess within hours loss of LSD1
Utilizing the dTAG system to degrade LSD1 and CRISPR-Cas[9] to edit and KO LSD1, findings presented here highlight gene-expression perturbations that lead to differentiation of AML and post-MPN sAML blasts

Summary

Introduction

AML, AML stem/progenitor cells (LSCs) in the measurable residual disease (MRD) resist differentiation, retain leukemia-initiating potential and mediate relapse of AML1–3. This phenotype of LSCs is mainly orchestrated by dysregulated superenhancers (SEs)/enhancers (Es) and the ensuing dysregulated transcriptome/proteome, which directly results from genetic alterations and/or perturbed levels and activity of epigenetic and transcriptional regulators in AML4–6. LSD1 (lysine specific demethylase 1; KDM1A) is a FAD-dependent amine-oxidase family member and a component of the co-repressor complexes involving HDAC1/2 and CoREST (RCOR1) or NuRD (MTA2)[14]. Two antiparallel α-helices divide and project away from the AOL domain as the Tower domain, which provides the binding interface with CoREST (RCOR1 and HDAC1/2) and allosterically regulates the catalytic activity and stability of LSD114–16.

Methods

Results

Conclusion