Epigenetic priming of AML blasts for all-trans retinoic acid-induced differentiation by the HDAC class-I selective inhibitor entinostat.

All-trans retinoic acid (ATRA) has only limited single agent activity in AML without the PML-RARα fusion (non-M3 AML). In search of a sensitizing strategy to overcome this relative ATRA resistance, we investigated the potency of the HDAC class-I selective inhibitor entinostat in AML cell lines Kasumi-1 and HL-60 and primary AML blasts. Entinostat alone induced robust differentiation of both cell lines, which was enhanced by the combination with ATRA. This "priming" effect on ATRA-induced differentiation was at least equivalent to that achieved with the DNA hypomethylating agent decitabine, and could overall be recapitulated in primary AML blasts treated ex vivo. Moreover, entinostat treatment established the activating chromatin marks acH3, acH3K9, acH4 and H3K4me3 at the promoter of the RARβ2 gene, an essential mediator of retinoic acid (RA) signaling in different solid tumor models. Similarly, RARβ2 promoter hypermethylation (which in primary blasts from 90 AML/MDS patients was surprisingly infrequent) could be partially reversed by decitabine in the two cell lines. Re-induction of the epigenetically silenced RARβ2 gene was achieved only when entinostat or decitabine were given prior to ATRA treatment. Thus in this model, reactivation of RARβ2 was not necessarily required for the differentiation effect, and pharmacological RARβ2 promoter demethylation may be a bystander phenomenon rather than an essential prerequisite for the cellular effects of decitabine when combined with ATRA. In conclusion, as a "priming" agent for non-M3 AML blasts to the differentiation-inducing effects of ATRA, entinostat is at least as active as decitabine, and both act in part independently from RARβ2. Further investigation of this treatment combination in non-M3 AML patients is therefore warranted, independently of RARβ2 gene silencing by DNA methylation.