­Histone deacetylase enzyme activity is not the universal anti-cancer target of HDAC inhibitors [RNA-seq_MC38]

As a family of anti-cancer drugs, histone deacetylase inhibitors (HDIs) show efficacies against a wide range of cancers, but their mechanism of action is unclear. Here, we performed unbiased bioinformatics analysis and found that genetic variants or expression levels of HDACs do not correlate with the carcinogenesis of most cancer types and do not predict cancer patient survival. Whole-genome CRISPR library screens did not identify HDACs as genes affecting cellular responses to HDIs such as SAHA or MS-275. Overexpression of dominant-negative, catalytically inactive Class I HDACs (HDAC1, 2, 3, and 8) caused a similar degree of histone hyperacetylation as Class I-specific HDI FK228 in cancer cell lines without similar anti-cancer effects, transcriptomic changes, or cellular responses to FK228. The anti-cancer effects of FK228 can be further dissociated from the histone hyperacetylation in an in-situ liver cancer mouse model. Importantly, chemical manipulations of several pan-HDIs abolished HDAC-inhibiting activity but retained the majority of anti-cancer effects in a mouse xenograft cancer model. These results suggest that HDAC enzyme activity is not a universal de facto target of HDIs for their anti-cancer efficacy. These findings have profound implications for understanding the pharmacodynamics of HDIs and developing next-generation compounds for this promising family of anti-cancer drugs.