miR-29b, miR-205 and miR-221 enhance chemosensitivity to gemcitabine in HuH28 human cholangiocarcinoma cells

Abstract

Background and aims: Cholangiocarcinoma (CCA) is highly resistant to chemotherapy, including gemcitabine (Gem) treatment. MicroRNAs (miRNAs) are endogenous, non-coding, short RNAs that can regulate multiple genes expression. Some miRNAs play important roles in the chemosensitivity of tumors. Here, we examined the relationship between miRNA expression and the sensitivity of CCA cells to Gem.

Methods: Microarray analysis was used to determine the miRNA expression profiles of two CCA cell lines, HuH28 and HuCCT1. To determine the effect of candidate miRNAs on Gem sensitivity, expression of each candidate miRNA was modified via either transfection of a miRNA mimic or transfection of an anti-oligonucleotide. Ontology-based programs were used to identify potential target genes of candidate miRNAs that were confirmed to affect the Gem sensitivity of CCA cells.

Results: HuCCT1 cells were more sensitive to Gem than were HuH28 cells, and 18 miRNAs were differentially expressed whose ratios over ± 2log2 between HuH28 and HuCCT1. Among these 18 miRNAs, ectopic overexpression of each of three downregulated miRNAs in HuH28 (miR-29b, miR-205, miR-221) restored Gem sensitivity to HuH28. Suppression of one upregulated miRNA in HuH28, miR-125a-5p, inhibited HuH28 cell proliferation independently to Gem treatment. Selective siRNA-mediated downregulation of either of two software-predicted targets, PIK3R1 (target of miR-29b and miR-221) or MMP-2 (target of miR-29b), also conferred Gem sensitivity to HuH28.

Conclusions: miRNA expression profiling was used to identify key miRNAs that regulate Gem sensitivity in CCA cells, and software that predicts miRNA targets was used to identify promising target genes for anti-tumor therapies.

Figure 1. Sensitivity of two CCA cell lines, HuCCT1 and HuH28, to gemcitabine (Gem).

HuCCT cells were significantly more sensitive than HuH28 cells to gemcitabine. *: p<0.05

Figure 2. Gem treatment slightly affected miRNA expression profiles of CCA cell lines.

(A) Scatter plot of miRNA expression log2 ratios between untreated and gemcitabine (Gem)-treated HuH28 cells. No miRNAs were differentially expressed because of the Gem treatment. (B) Scatter plot of miRNA expression log2 ratio between untreated and Gem-treated HuCCT1 cells. miR-1260 and miR-1280 were downregulated in Gem-treated HuCCT1 cells. (C) Ectopic overexpression of miR-1260 or miR-1280 by transfection of miRNA mimics did not affect the sensitivity of HuCCT1 cells to Gem. Relative cell viabilities were assessed 72 hr after Gem treatment. Final concentration of each miRNA mimic was 10 nM. Mock: receiving only transfection reagent. siCON: control treated with a non-silencing miRNA mimic.

Figure 3. Eighteen miRNAs were differentially expressed between untreated HuH28 and untreated HuCCT1 cells.

(A) Scatter plot of miRNA expression log2 ratios between untreated HuH28 and HuCCT1 cells. The threshold defining differential expression was a ratio smaller than -2log₂2 or larger than 2log₂2. (B) Normalized expression intensities and ratio values of the 18 miRNAs were reported in the table.

Figure 4. Modification of four of the candidate miRNA expression restored Gem sensitivity to HuH28 cells.

Ectopic overexpression of miR-29b (A), miR-205 (B), or miR-221 (C) via transfection of a corresponding miRNA mimic and downregulation of miR-125a-5p (D) via transfection of an anti miRNA oligonucleotide made HuH28 cells more sensitive to Gem. Relative cell viabilities were assessed 72 hr after Gem treatment. The final concentration of each miRNA mimics was 10 nM; that of the anti miRNA oligonucleotide was 40 nM. Mock: receiving only transfection reagent. siCON: control treated with a non-silencing miRNA mimic. AntiOligoCON: control treated with a non-silencing control oligonucleotide. The asterisk denotes p<0.05 as compared to non-treated control, mock and siCON or control AntiOligoCON.

Figure 5. Web-driven programs predicted putative chemo-sensitivity-related target genes of candidate miRNAs.

Genes that were among the ontology-filtered results from TargetScanHuman (left circle) and those from microT (right circle) were designated putative target genes of candidate miRNAs. Ontology filtering was performed with web-driven gene ontology software DIANA mirPath.

Figure 6. Two of software predicted miRNA target genes actually suppressed by corresponding miRNA.

(A) MMP-2 and (B) PIK3R1 were significantly suppressed by transfection of the respective miRNA mimics. Selective downregulation of MMP-2 (C) or PIK3R1 (D) by transfection of each corresponding siRNA conferred Gem sensitivity to HuH28 cells. The analysis of Western blot and relative cell viability were performed 72 hours after Gem treatment. Final concentration of siRNAs and miRNA mimics were 10 nM. Mock: receiving only transfection reagent. siCON: control treated with a non-silencing miRNA mimic. The asterisk denotes p<0.05 as compared to non-treated control, mock and siCON.

Figure 7. Caspase-3/7 activity assays were performed to assess apoptosis.

The final concentrations of miRNA mimics and anti miRNA oligonucleotide were 10; the final concentration of Gem was 1×10⁻⁴ M. Mock: receiving only transfection reagent. siCON: control treated with a non-silencing miRNA mimic. *: p<0.05 versus non-treated control. †: p<0.05 versus Gem treated.