AUTHORS
Alba Sommerschield, MD, Samuel Ojeda2, Meng Wang, PhD3, Wallace A. Bourgeois, MD4, Francisca deLuna Vitorino, PhD5, Emily Zahn, PhD6, Benjamin A Garcia, PhD7, Eric S Fischer, PhD8, Scott A Armstrong, MD, PhD9
AFFILIATIONS
1 Boston Children's Hospital, Boston, Massachusetts, United States, Dana-Farber Cancer Institute, Boston, Massachusetts, United States, Harvard Medical School, Boston, Massachusetts, United States
2 Dana-Farber Cancer Institute, Boston, Massachusetts, United States, Harvard Medical School, Boston, Massachusetts, United States
3 Dana-Farber Cancer Institute, Boston, Massachusetts, United States, Harvard Medical School, Boston, Massachusetts, United States
4 Boston Children's Hospital, Boston, Massachusetts, United States, Dana-Farber Cancer Institute, Boston, Massachusetts, United States, Harvard Medical School, Boston, Massachusetts, United States
5 Washington University School of Medicine, St. Louis, Missouri, United States
6 Washington University School of Medicine, St. Louis, Missouri, United States
7 Washington University School of Medicine, St. Louis, Missouri, United States
8 Dana-Farber Cancer Institute, Boston, Massachusetts, United States, Harvard Medical School, Boston, Massachusetts, United States
9 Boston Children's Hospital, Boston, Massachusetts, United States, Dana-Farber Cancer Institute, Boston, Massachusetts, United States, Harvard Medical School, Boston, Massachusetts, United States
ABSTRACT
Background: The histone acetyltransferase KAT2A is an established dependency in acute myeloid leukemia (AML). Targeted protein degraders of KAT2A and its paralog KAT2B are currently in clinical trials for adult relapsed/refractory AML. However, how KAT2A degradation drives antileukemic effects, which disease subtypes respond, and whether paralog-selective degradation of KAT2A is sufficient for efficacy remain unknown. We previously reported a first-in-class molecular glue degrader that selectively degrades KAT2A with nanomolar potency (compound 4, DC50 = 22 nM; Ojeda et al., Science 2026), without appreciable degradation of KAT2B, and that confers a significant survival advantage in a PDX model of KMT2A-rearranged AML. We have now leveraged this tool compound for time-resolved transcriptional, chromatin, and proteomic profiling to demonstrate a direct role for KAT2A in sustaining the leukemic transcription program in KMT2A-driven AML.
Results: KAT2A degradation with compound 4 produced a spectrum of antileukemic activity across a genetically diverse panel of AML cell lines - not restricted to a single genetic subtype - indicating a context-dependent vulnerability. For mechanistic analysis we focused on two poor-prognosis, KMT2A-driven subtypes: KMT2A-rearranged (KMT2A-r) and KMT2A partial tandem duplication (KMT2A-PTD).
In MV4-11 (KMT2A-r) and EOL1 (KMT2A-PTD) cell lines, KAT2A degradation led to rapid depletion of H3K9ac, the predominant KAT2A-associated histone mark, and drove G1 arrest, myeloid differentiation, and apoptosis. Degradation altered the expression of a restricted gene set shared between KMT2A-r and KMT2A-PTD, including downregulation of MYC and E2F targets and G2M checkpoint, and upregulation of myeloid-differentiation and interferon programs. Chromatin profiling by ChIP-seq showed KAT2A occupied both promoters and intronic/intergenic regions. Integrating promoter binding with early expression changes defined direct KAT2A targets which included cell-cycle and chromatin regulators (CCNE1/2, HOXA9, SETD1B), as well as hematopoietic transcription factors (CEBPA/D, IRF8, MEF2D, BCL11A), whose promoters were enriched for CEBP, RUNX, and PU.1/IRF8 motifs.
Having defined the KAT2A target gene program, we next established how KAT2A sustains its transcription in MV4-11. Degradation caused a decline in nascent transcription at target genes, detectable within 15 minutes. SLAM-seq (metabolic labeling of nascent RNA) showed this decline was limited to target genes, with the broader transcriptome largely unaffected, indicating a selective rather than global effect. At target promoters, RNA Pol II occupancy declined early (1 hour) and progressively (24 hours) post-degradation. As a readout of KAT2A's enzymatic function, time-course histone post-translational modification mass spectrometry identified H3K9ac and combinatorial H3K9acK14ac as selectively depleted, and ChIP-seq confirmed H3K9ac was lost genome-wide within 1 hour of degradation. Compared to non-target genes, target gene promoters were significantly enriched for KAT2A, H3K9ac, and initiating RNA Pol II, highlighting these as highly active transcriptional loci. Together these data directly link KAT2A to the maintenance of transcription at core myeloid leukemogenic gene hubs.
KAT2A was also distributed across intronic and intergenic regions genome-wide; these elements corresponded to enhancers predicted to regulate over 8,000 genes (Activity-by-Contact modelling), a subset of which were identified as super-enhancers (SEs) by ROSE2, including at MYC and IRF8 loci. At the MYC and IRF8 SEs, KAT2A degradation led to depletion of BRD4 and MED12, components of the enhancer-promoter communication machinery, by 24 hours, identifying a previously unappreciated role for KAT2A in enhancer regulation in AML.
Conclusions: Using the paralog-selective degrader compound 4, we have established the chromatin-level basis of KAT2A dependency in KMT2A-driven AML, where it directly sustains a defined set of oncogenic transcriptional hubs driving lineage and proliferation programs shared across both KMT2A-r and KMT2A-PTD models. Critically, in these subtypes selective degradation of KAT2A without KAT2B is sufficient to disrupt these programs, impairing leukemic growth and inducing differentiation. These findings provide a mechanistic rationale for paralog-selective KAT2A degradation as a therapeutic strategy for KMT2A-driven leukemias, high-risk subtypes with few targeted options.
DISCLOSURES
Alba Sommerschield, MD: No relevant conflicts of interest to declare.