ADAR2 repression of the clonogenic growth of t(8,21) AML cells is associated with RNA editing–mediated protein recoding of COPA and COG3. (A) Box plots illustrating editing frequencies of 21 ADAR2-regulated editing sites in patients with t(8;21) (n = 7), inv16 (n = 8), and control non-CBF (n = 164) AML, from the TCGA RNA-seq data sets. The box extends from the 25th to 75th percentiles. The line in the middle of the box is plotted at the median. The whiskers indicate 5% to 95% percentile. (∗∗∗P < .001; ∗∗P < .01 using two-tailed Student t test.). (B) Scatter plots showing the changes in editing frequency of each of 172 protein-recoding editing targets between Kasumi-1 cells stably overexpressing the WT ADAR2 and the EV control. Red lines indicate the cutoff values that discriminate sites showing ≥5% editing change from those with <5% change in editing frequency. (C) Box plots showing the editing frequency of COG3 and COPA in patients with CBF AML (n = 15) and those with non-CBF (n = 164) AML, from the TCGA AML RNA-seq data sets. Western blot analysis of COG3 (D) or COPA (E) protein levels in Kasumi-1 cells stably overexpressing the WT or edited COG3 (COG3-WT or COG3-I635V); or WT or edited COPA (COPA-WT or COPA-I164V); or the pMSCV-flag-puro EV control via retroviral transduction. GAPDH was used as a loading control. (F-G) Sequence chromatograms illustrating the editing levels of COG3 and COPA transcripts in the same samples as described in panels D and E, respectively. The arrowhead indicates the position of the edited nucleotide. (H-I) Bar chart illustrating the number of colonies formed by the indicated group of cells as described in panels D and E. Data are presented as the mean ± SD of technical replicates from 2 representative experiment out of 3 independent experiments. ∗∗∗P < .001; ∗∗P < .01; ∗P < .05 using two-tailed Student t test.