Figure 1.
An integrative approach to the identification of TS-miRNAs using function- and expression-based screening in HCC cell lines.
A, Strategy for the identification of TS- miRNAs by function- and expression-based screening in 6 HCC cell lines. B, Results of function-based screening in 6 HCC cell lines. The numbers of viable cells 4–5 days after transfection with 5 nM of 470 dsRNAs mimicking mature miRNAs loaded on Pre-miR™ miRNA Precursor Library-Human V3 (Ambion) or control non-specific dsRNA were evaluated by the WST-8 assay in duplicate. Relative in vitro cell growth ratios in these cell lines were calculated by normalization of each result to the cell numbers in control cells transfected with non-specific miRNA (see Dataset S1). The lower solid arrow indicates the 470 miRNAs examined. Closed arrows in each graph indicate candidate miRNAs with marked growth inhibitory effects (growth ratio <0.8) compared with the control counterpart in our function-based screening. The arrowhead indicates the results for seven candidate miRNAs conclusively selected in the function- and expression-based screening (see Fig. 1D). C, Summary of expression profiles of 45 candidate miRNAs, expression levels of which were downregulated (>2-fold) in all 6 cell lines as compared with those of two normal liver tissues (C20 and C40, see Dataset S2). Stars indicate the results of the same seven candidate miRNAs as in Fig. 1B. D, Seven candidates for tumor-suppressive miRNAs involved in HCC pathogenesis identified through function- and expression-based screening approach. MiRNAs frequently downregulated (>50%) in tumors compared with their paired non-tumorous liver tissues in clinical HCC cases are in bold face (a). Expression level of mature form for each miRNA was evaluated inTaqMan® MicroRNA Assays (Applied Biosystems) as described in Materials and Methods (b).
Figure 2.
Growth suppressive effects of miR-195 and miR-497 on HCC cell lines lacking their expression.
A, Growth curves of HCC cell lines after tansfection of 5nM of miCENTURY OX miNatural (Cosmo Bio) mimicking miR-195 (black triangle) or miR-497 (black circle), or control Luc, (white circle) assessed by WST-8 assay. Points, the mean of triplicate determinations in these experiments; bars, SD; Asterisks (*), P<0.05 versus Luc transfected cells in a statistical analysis with the Mann-Whitney U test. B, Results for the population in each phase of the cell cycle assessed by FACS (left) and phase-contrast micrographs (right) using HCC cell lines 72 hours after transfection of miCENTURY OX miNatural mimicking miR-195 (middle) or miR-497 (lower), or control Luc (upper).
Figure 3.
Exploration of possible direct miR-195 and miR-497 targets by Ago2-IP experiment.
A, GSEA profiles of the Running Enrichment Score (ES) of genes enriched in the top 8% and 36% of candidates as target genes for miR-195 (upper) and miR-497 (lower) by Ago2-IP experiments along the rank of transcripts differentially expressed after transfection with miR-195 (left) or miR-497 (right) versus Luc. B, Outline for selecting final candidates as direct targets of miR-195 (upper) and miR-497 (lower) determined by a combination of Ago2-IP candidates and repressed genes by miR-195 or miR-497 (>2-fold decrease) in gene expression array experiments. C, Significantly enriched IPA canonical pathways for candidate direct target genes of miR-195 and miR-497. Numbers(a), ratio(b) and genes (c) of candidate direct targets among all genes involved in the canonical pathway. D, Representative results of Western blotting of known targets, CDK6, CCND1, and E2F3, as positive controls and predicted targets, CDK4, CCNE1, CCND3, CDC25A, and BTRC, for miR-195 and miR-497, 48 hours after transfection with miCENTURY OX miNatural mimicking miR-195, miR-497, or control Luc. E, 3′UTR reporter assays of a miR-195- and miR-497-nonexpressing cell line, Hep G2, 48 hours after cotransfection with pMIR-REPORT luciferase vectors containing 3′-UTR target sites (upper) of CDK6, CCND1, E2F3, CDK4, CCNE1, CCND3, CDC25A, or BTRC, miCENTURY OX miNatural mimicking miR-195, miR-497, or negative control, and pRL-hTK internal control vector (lower). Horizontal open bar, gray boxes, and black bars with arrowheads in upper panel indicate 3′-UTR, possible target sites, and regions examined in the 3′UTR reporter assay, respectively, for each gene.
Figure 4.
Significance of identified miR-195 and miR-497 targets in HCCs. A
, Schema of miR-195 and miR-497 target genes regulating promotion of G1 to S in the cell cycle. B, Results of the knock down of each miR-195 and miR-497 target gene by siRNAs. Mean of relative growth ratio of Hep G2 72 hours after transfection of 10 nM of each siRNA (siNC, siBTRC, siCDK4, siCCND3, siCCNE1 or siCDC25A) and non-treated Hep G2 (Non treatment) were assessed by WST-8 assay in a triplicate manner (upper). The population in each phase of the cell cycle assessed by FACS were assessed by FACS 48 hours after siRNA transfection (lower) in a duplicate manner. Bars, SD. Asterisks (*), P<0.05 versus siNC transfected cells in a statistical analysis with the Mann-Whitney U test. C, Expression status of each target gene in primary HCC tumors (closed boxes) and paired non-tumorous liver tissues (open boxes) as compared with normal liver tissue (C20), which were used in experiments shown in Fig 1D. Asterisk (*), frequencies of primary HCC cases, in which a remarkable up-regulation of possible miR-195 and/or miR-497-target mRNAs was observed in tumor tissues compared with paired non-tumorous liver tissues (>2-fold expression). NA, not available due to lack of available cDNAs for qRT-PCR analyses in the LC96 sample.