KAT6A chimeras form a self-reinforcing epigenetic module with NURF and MLL/COMPASS to sustain AML

KAT6A-CBP (K/C) and KAT6A-P300 (K/P) fusions are recurrent genetic alterations in acute myeloid leukemia (AML) associated with poor prognosis. Despite their strong oncogenic potential, the mechanisms underlying their genomic targeting and leukemogenic function remain unclear. A major challenge has been their large size, which has impeded preclinical model development and mechanistic studies.

We employ a domain-focused truncation strategy to generate de novo murine models of K/C and K/P fusions, which faithfully recapitulate the morphological, immunophenotypic, and transcriptomic features of KAT6A-rearranged AML. Genomic profiling reveals that KAT6A fusions preferentially localize to H3K4me2/3-marked regions, while biochemical analyses uncover that KAT6A interacts with the Nucleosome Remodeling Factor (NURF), a key H3K4me2/3 reader. Disrupting NURF-chromatin interactions via depletion or small-molecule inhibition of its subunit, Bromodomain PHD Finger Transcription Factor (BPTF), impairs K/C recruitment and disrupts MLL/COMPASS-mediated H3K4me2 deposition, defining a functional epigenetic module involving KAT6A chimeras, NURF, and MLL/COMPASS. Notably, CBP/P300 inhibition reduces histone acetylation and chromatin accessibility, further impairing the recruitment of this epigenetic module. Targeting this module via NURF or CBP/P300 inhibition demonstrates efficacy in K/C leukemia models, with enhanced therapeutic effects observed when combined.

Our study identifies a self-reinforcing epigenetic module of histone modifiers and readers in KAT6A-rearranged AML, providing mechanistic insights into the genomic targeting of KAT6A chimeras and highlighting promising combinatorial therapeutic strategies.