Selection of transcription factors highly expressed in ATL. (A) Schematic diagram of the selection criteria for candidate transcription factor genes. FC, fold-change; TPM, transcripts per million. (B) Heat map representing the mRNA expression of 48 selected transcription factor genes in 3 healthy donors, 3 HTLV-1 carriers, and 9 primary ATL cases from our cohort (NCU cohort). The genes whose expression was significantly downregulated by THZ1 treatment in TL-Om1 cells were shown in red with asterisks. (C) IRF4 mRNA expression was analyzed by RNA-seq. Data are represented as box plots where the middle line indicates the median, the lower and upper hinges correspond to the first and third quartiles, the lowest datum indicates the minimum, and the highest datum indicates the maximum. ns: non-significant; *P < .05; **P < .01; ***P < .001 by an unequal variances t test. (D) Apoptosis was measured via Annexin V staining and propidium iodide followed by flow cytometry analysis in various cell lines on day 3 after transduction with a lentivirus expressing shRNA. The percentage of Annexin V-positive cells were shown as the mean ± standard deviation (SD) of duplicates. *P < .05; **P < .01 by a 2-sample, 2-tailed t-test compared with the shGFP control. (E) cDNA containing only the coding region of IRF4 mRNA was transduced via retroviral infection into TL-Om1 cells. Cells overexpressing IRF4 (IRF4 OE) or empty vector (EV) were then transduced via lentiviral infection with control shRNA (sh-LUC), sh-IRF4-1, or sh-IRF4-2, which targeted the 3′ untranslated region of IRF4 mRNA. Percentage of apoptotic cells were measured by Annexin V staining and propidium iodide followed by flow cytometry analysis, and shown as mean of SD of duplicates (left). *P < .05 by a 2-sample, 2-tailed t-test compared with the shLUC control. Whole-cell extracts were harvested and subjected to immunoblot analysis with antibodies specific for IRF4, cleaved PARP, or β-actin (internal control; right).