Fig. 5: RUNX1-IT1 is abundant in cytoplasm and sponges miR-632 in HCC cells.

RNA FISH assays a and subcellular fractionation assays b indicated that RUNX1-IT1 was predominately located in the cytoplasm. Magnification is ×400, and scale bars = 20 μm. c By applying bioinformatics tools (MIRDB, and DINAN tools-LncBase Predicted v.2), we found that there were two putative binding sites between 3′-UTR of RUNX1-IT1-wild type (wt) and miR-632. RUNX1-IT1-mutant (mut) means mutation of binding sites in the 3′-UTR of RUNX1-IT1. d The expression of miR-632 in tumour tissues (n = 87) was dramatically higher than that in adjacent non-tumour tissues (n = 87). ***P < 0.001 by Student’s t test. e Pearson correlation analysis verified that there existed a negative association between miR-632 and RUNX1-IT1 in HCC tissues. f MiR-632 was elevated in HCC cell lines. n = three independent experiments, **P < 0.01, ***P < 0.001 by Student’s t test versus L02. g Real-time PCR showed that miR-632 was negatively regulated by RUNX1-IT1. n = three independent experiments, ***P < 0.001 by Student’s t test versus vector or ANOVA versus shVector. h MiR-632 expression was greatly increased by miR-632 mimics, whereas significantly decreased by the inhibitors, and the expression of RUNX1-IT1 was negatively regulated by miR-632. n = three independent experiments, **P < 0.01, ***P < 0.001 by ANOVA. i Luciferase reporter gene assays showed that miR-632 negatively regulated the luciferase activity of RUNX1-IT1-wt-3′-UTR, rather than of RUNX1-IT1-mut-3′-UTR. n = three independent experiments, **P < 0.01, ***P < 0.001 by ANOVA. j The anti-Ago2 RIP assay with miR-632 mimics showed that both miR-632 and RUNX1-IT1 were enriched in Ago2 precipitate compared with IgG. n = three independent experiments. ***P < 0.001 by ANOVA. k RUNX1-IT1 was highly enriched in the sample pulled down by biotinylated wt miR-632 rather than mut miR-632. n = three independent experiments. ***P < 0.001 by ANOVA versus Bio-NC.