Abstract
AML containing the t(8;21) is exceptionally sensitive to the FDA approved HDAC inhibitors. While these compounds cause extensive, S phase-dependent DNA damage by stalling replication forks, they also activate transcription by increasing acetylation of the N-terminal histone tails that are "read" by chromatin modifying enzymes. However, gene expression analysis showed that these compounds turn off many genes even though histone acetylation is commonly associated with gene activation. To address this conundrum, we performed precision nuclear run-on transcription analysis (PRO-seq) of t(8;21)-expressing cells treated with the class I HDAC inhibitors depsipeptide, vorinostat, and panobinostat and the HDAC3-selective compound RPG966 for 1 or 4 hr and found that the broad-spectrum compounds activated a relatively small number of genes while causing RNA polymerase pausing at key oncogenic loci including MYC, KIT, BCL-XL, and BCL6 . By contrast, RGFP966 mostly activated gene expression, which appeared to be associated with increased initiation. In our previous study (Zhao et al, Cell Reports, 2016) we reported genes with RNA polymerase pausing at promoters after bromodomain and extra terminal (BET) inhibitor treatment. The majority of those sites also displayed RNA polymerase pausing after treatment with broad-spectrum HDAC inhibitors. Analysis of histone modifications by western blot and ChIP-seq found large increases in marks bound by BRD4 and other BET family members. ChIP-exo and ChIP-seq also found a loss of proper localization by BRD4 after HDAC inhibitor treatment. Thus, HDAC inhibitors massive increases in global histone acetylation, which triggers "epigenetic confusion", whereby BRD4 cannot locate its normal chromatin address in high levels of nucleosome acetylation. Therefore, for genes regulated by BRD4, HDAC inhibitors act similarly to BET inhibitors, but can also cause DNA damage to yield a potent combination mechanism of action against t(8;21) AML. Interestingly, almost half of AML1-ETO fusion protein target genes also had change in promoter proximal pausing of RNA polymerase after Depsi, SAHA or RGFP966 treatment, suggesting that HDAC inhibitors may more directly perturb the effects of AML1-ETO.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.