Fig. 4

miR-486-3p and eIF5A are downstream targets of circ-EPB41. A Venn diagrams from the Interactome, miRanda and circBank websites for predicting miRNAs sponged by circ-EPB41. B Bioinformatics analysis (http://circnet.mbc.nctu.edu.tw/) found that miR-486-3p, miR-587 and miR-595 were targets of circ-EPB41. C Dual-luciferase reporter assays showed that co-transfection of WT and the mimic miR-486-3p markedly decreased luciferase activity in HEK293T cells. D Results of dual-luciferase reporter assays indicated that luciferase activity did not change in HEK293T cells when miR-486-3p binding sites in circ-EPB41 were mutated. Data are presented as means ± SD; ^**P < 0.01. E Fluorescence in situ hybridization found that miR-486-3p co-localized with circ-EPB41 in the cytoplasm. F RT-qPCR detection showed the expression of miR-486–3 in 90 paired NSCLC tissues. Data are presented as means ± SD; ^***P < 0.001 vs normal group. G Clustered heat map of differentially expressed mRNAs in A549 cells after downregulation of circ-EPB41. H Bioinformatics analysis (http://www.targetscan.org/vert_71/) found that eIF5A was the direct target of miR-486-3p. The MUT version of the 3ʹUTR-eIF5A is also shown. Relative luciferase activity was determined 48 h after transfection with miR-486-3p mimic/NC or with the 3'UTR-eIF5A WT/MUT in HEK293T cells. Data are presented as means ± SD; ^***P < 0.001