Degrading SMARCA2/4 Relieves the Differentiation Block in AML Via Chromatin Alterations Affecting S100A8/9

Introduction: Acute Myeloid Leukemia (AML) is characterized by the unfettered proliferation and poor differentiation of hematopoietic stem and progenitor cells causing cytopenias and reduced survival. The mammalian BAF complexes (also known as SWI/SNF) are important regulators of cell proliferation and differentiation. SMARCA2 and 4 are mutually exclusive ATPase subunits within the BAF complexes that use helicase activity to remodel chromatin via nucleosome sliding and removal. The association of these BAF subunits with regulatory elements of oncogenic transcription factors makes them intriguing targets for AML. Degradation of SMARCA2/4 is a potential targeted therapeutic approach that needs further evaluation in AML/MDS.

Results: Gene expression analysis from highly purified primary AML stem cells (Lin-ve, CD34+, CD38-) and control samples showed significant overexpression of SMARCA2/4 in the AML samples when compared to healthy control stem cells. A novel and highly selective SMARCA2/4 and PBRM1 PROTAC degrading agent AU-15330 (Xiao et al., Nature 2022) was used to investigate the therapeutic effectiveness of targeting these BAF components in leukemic pathogenesis. We demonstrate that AU-15330 was able to effectively degrade SMARCA2/4 and inhibit human leukemia cell lines; MV4-11 being most responsive with inhibition at low nanomolar concentrations.

When evaluating patient derived AML cells in colony forming unit assays, AU-15330 treatment increased erythroid and myeloid differentiation in patient AML samples (N=12) with mutations including DNMT3A, TP53, RUNX1, WTI, NPM1, FLT3-ITD etc. Minimal effects were seen on healthy PBMCs and CD34+ HSPCs (N=5). A xenograft of patient derived AML cells in irradiated mice was performed to determine the requirement of SMARCA2/4 in leukemia stem cell maintenance. Mice treated with AU-15330 had greater myeloid differentiation, reduced leukemic cells in their bone marrows and reduced leukemia stem cell activity, as seen in secondary transplants.

A multi-omic analysis of transcription and chromatin dynamics in MV4-11 cells was used to uncover the underlying mechanism of SMARCA2/4 and PBRM1 mediated leukemic pathogenesis. Globally, a loss of chromatin accessibility as well as epigenetic enhancer marks (H3K27ac) were determined by ATAC-seq and CHiP-seq analysis. To determine the effect of SMARCA2/4 degradation on pBAF complex association with chromatin, single cell imaging was performed of U2OS cells with HALO-tagged PBRM1 treated with the degrader or DMSO. SMARCA2/4 degradation does not collapse the pBAF complex but causes destabilization of pBAF's association with chromatin, determined by decreased residence time on binding to the nucleosome when compared to DMSO treated cells. This suggests that degradation of SMARCA2/4 impairs functionality of the pBAF complex on DNA.

Several leukemia related transcription factors as well as inflammatory genes were downregulated after SMARCA2/4 degradation in AML cells. S100A9 is an inflammatory protein that heterodimerizes with S100A8 and binds to the TLR4, RAGE and CD33 receptors, promoting the inflammasome and NF-kB expression. These pathways have previously been linked to MDS/AML pathogenesis. Our data show a decrease in S100A9 transcription as well as a significant decrease in S100A8/9 protein secretion by AU-15330 treated MV4-11 cells. HiCHiP analysis showed decrease in enhancer promoter loops near the S100A9 locus with SMARC2/4 degradation. Differentiation and growth inhibition seen after AU-15330 treatment were partially reversed after treatment with recombinant S100A8/9; demonstrating this is a critical downstream target.

Conclusions: Taken together, these studies elucidate a novel role of SMARCA2/4, as part of the BAF complex, in promoting leukemogenesis via regulation of alarmins (S100A8/9). Our data also demonstrates the in vivo and in vitro efficacy of degrading SMARCA2/4 in relieving the differentiation block seen in MDS/AML.