The HDACi mocetinostat increases the endogenous expression of CD48 and NK cell cytotoxicity. (A) Flow cytomery analysis of CD48 expression in 721.221 (left), NB4 (middle), and Kasumi-1 (right) cells treated with dimethylsulfoxide (DMSO) (gray line) or with the HDACi mocetinostat (Moc) (black line). Gray shaded histograms, background staining with an isotype-matched control antibody. (B-D) ³⁵S-labeled NB4 (B) or 721.221 cells (C) were treated with DMSO or with the HDACi Moc and incubated with primary NK cells (E:T 20:1 for 721.221 cells and E:T 10:1 for NB4). (D) ³⁵S-labeled NB4 cells, which were treated with DMSO or with the HDACi Moc, were incubated with YTS eco cells at the indicated effector to target (E:T) ratios. All killing experiments were repeated twice. (E) ChIP-seq data obtained from the NCBI Gene Expression Omnibus, accession number GSE23730, of Kasumi-1 cells immunoprecipitated with an AML1-ETO antibody (upper graph labeled AE) or with an ETO antibody (lower graph labeled ETO). Binding sites at the CD48 gene are shown through the UCSC Genome Bioinformatics (http://genome.ucsc.edu/). (F) A sequence of the CD48 intron (arrow) is predicted to be targeted by AML1-ETO based on the ChIP-seq data. Possible binding sites of AML1 are marked by a gray color. (G) Kasumi-1 cells were subjected to ChIP with anti-ETO, anti-AML1-ETO, or control HA antibodies. After precipitation, qRT-PCR assays were performed with primers that specifically amplify the putative AML1-ETO–binding site (CD48 site), a control binding site (negative site), and an identified target of AML1-ETO as a positive control (LAT2). ChiP values are presented as a fraction of the input. Error bars represent the standard deviation of the means of 3 independent experiments. *P < .05; **P < .01, Student t test. ns, not significant.