An acetyl-histone vulnerability in PI3K/AKT inhibition-resistant cancers is targetable by both BET and HDAC inhibitors

Acquisition of resistance to phosphatidylinositol 3-kinase (PI3K)/AKT-targeted monotherapy implies the existence of common resistance mechanisms independent of cancer type. Here, we demonstrate that PI3K/AKT inhibitors cause glycolytic crisis, acetyl-coenzyme A (CoA) shortage, and a global decrease in histone acetylation. In addition, PI3K/AKT inhibitors induce drug resistance by selectively augmenting histone H3 lysine 27 acetylation (H3K27ac) and binding of CBP/p300 and BRD4 proteins at a subset of growth factor and receptor (GF/R) gene loci. BRD4 occupation at these loci and drug-resistant cell growth are vulnerable to both bromodomain and histone deacetylase (HDAC) inhibitors. Little or no occupation of HDAC proteins at the GF/R gene loci underscores the paradox that cells respond equivalently to the two classes of inhibitors with opposite modes of action. Targeting this unique acetyl-histone-related vulnerability offers two clinically viable strategies to overcome PI3K/AKT inhibitor resistance in different cancers. [Display omitted] •PI3K/AKT inhibitors suppress glycolysis, acetyl-CoA supply, and histone acetylation•BRD4 binding at growth factor loci increases amid global decline of H3K27ac•Inadequate HDAC occupancy permits locus-specific H3K27ac gain upon AKT inhibition•PI3K/AKT inhibitor resistance is vulnerable to both bromodomain and HDAC inhibitors Acquisition of resistance to PI3K/AKT-targeted monotherapy regardless of cancer type implies the existence of common mechanisms. Wu et al. identify an intrinsic mechanism that not only drives PI3K/AKT inhibitor resistance but also can be vulnerably targeted paradoxically by both BET and HDAC inhibitors.