Fig 2: MBD2 was the target of miR‐520c‐3p in HCC cells. (A) The potential binding sites of miR‐520c‐3p with MBD2 and GLIS3. (B, C) The luciferase activity in 293T cells after co‐transfection with the reporter plasmids and miR‐520c‐3p mimics. Groups of MBD2 WT + mock and GLIS3 WT + mock served as references for normalization. (D) Results of RNA pull‐down followed RT‐qPCR testing the enrichment level of MBD2 3′UTR in the RNA pulled down by biotin‐miR‐520c‐3p probe or control probe (Biotin‐NC, reference for calculating relative value). Binding site 1–3, different region of MBD2 3′UTR that contains one of the three putative binding sites. (E) The effect of miR‐520c‐3p mimics on the mRNA and protein levels of MBD2 in Huh7 and HepG2 cells. Groups of mock served as references for calculating relative mRNA levels. (F) The effect of circKCNN2 knockdown on the mRNA and protein levels of MBD2 in Huh7 and HepG2 cells. Groups of si‐scramble served as references for calculating relative mRNA level. (G) The expression level of MBD2 in human HCC tumor tissues and adjacent tissues. (H) The correlation between the mRNA level of MDB2 in HCC tissue and the recurrence‐free survival of HCC. (I) The correlation between the mRNA level of MDB2 in adjacent tissue and the recurrence‐free survival of HCC. All the assays were performed at least three times. N = 3 for each group. *P < 0.05, **P < 0.01, ***P < 0.001