Functional implications of specific RNA editing. (A) Scatter plot showing the frequency of editing sites within 3' UTRs between control and Adar1 knockout (KO) in c-Kit⁺ cells. (B) A volcano plot showing that expression of selected genes with 3' UTR editing frequencies in control group were greater than that in Adar1 KO group by at least 0.1. Genes with significantly increased or decreased expression in the control group compared with the Adar1 KO group are indicated with red and blue dots, respectively. P < .05 indicates a statistically significant difference. (C) Spearman correlation coefficients between the frequencies of stage-specific editing events within 3' UTRs and mRNA expression levels. (D) Representative examples of the Spearman correlation analysis between the frequencies of editing sites within 3' UTRs and mRNA expression levels. (E) The proportion of the nonsynonymous shift in coding region affected by editing sites (left) and the proportion of types of RNA editing events that cause nonsynonymous shifts in the CDS region. (F) A heat map showing group-specific/stage-specific editing events that cause nonsynonymous shift in the CDS region. (G) The editing frequency of 8 candidate genes among 3 hematopoietic groups. (H) The CFU experimental procedure. Lentivirus containing individual candidate genes expressing the WT transcript or the corresponding edited transcript was transduced into c-Kit⁺ cells. The transduced c-Kit⁺ BM cells were then sorted for CFU assays. (I) Colony formation of c-Kit⁺ BM cells after transduction with the indicated lentivirus expressing the WT or edited gene (referring to Figurepanel G). The number of colonies in the different groups were counted after 7 days of incubation, including E (BFU-E), GM (CFU-GM, CFU-G, CFU-M), and mix (CFU-GEMM) (n = 6 per group, 2 independent experiments, Error bars represent SD). **P < .01; ***P < .001.