miR-145 suppresses Fli-1. (A) HEL cells were infected with control lentiviruses, lentiviruses overexpressing miR-143 or miR-145, or lentiviruses expressing miRNA sponges targeting miR-143 and miR-145. Western blotting was performed for Fli-1, actin, p38alpha, and GSK3beta. Fold expression was determined by the ratio of Fli-1 expression to p38alpha. These experiments were replicated twice. (B) Luciferase assays were performed in 293T cells cotransfected with control lentiviruses or lentiviruses overexpressing miR-143 or miR-145 and pRL-TK containing binding sites or murine Fli-1 and c-Myc 3′UTRs. pGL3 was used as a transfection control. Expression was normalized to pGL3 levels and subsequently normalized relative to pRL-CXCR4 expression. Values are mean ± SEM (n = 6) with propagated error. These experiments were replicated 3 times. (C) Luciferase assays were performed in 293T cells cotransfected with control lentiviruses (empty FLI1 wt) or lentiviruses overexpressing miR-145 and pRL-TK containing either the wild-type murine Fli-1 3′-UTR (FLI1 wt) or the Fli-1 3′UTR with individual or combined point mutations of miR-145 binding sites (FLI1 mut1-mut4). pGL3 was used as a transfection control. Expression was first normalized to pGL3 levels and subsequently normalized relative to FLI1 wt expression. Values are mean ± SEM (n = 6) with propagated error. These experiments were replicated 3 times. (D) Quantitative RT-PCR was performed for Fli-1 in CD34⁺ progenitors as described in Figure 1C. Expression levels were normalized to GAPDH. Values are mean (n = 4). These experiments were performed in quadruplicate and replicated twice. (E) Relative to a control cDNA (LacZ), overexpression of Fli-1 promotes megakaryocytic relative to erythroid differentiation in CD34⁺ progenitors as assessed in Figure 1A. Values are mean ± SEM with propagated error. These experiments were performed in triplicate and replicated 2 times. *P < .05. ***P < .0005.