Fig 2: The miR-34a/CA9 axis regulates cell proliferation and mobility of hepatocellular carcinoma (HCC). (A) Migration, invasion, and colony formation assays were conducted after infection with a lentivirus carrying CA9 shRNAs or scrambled shRNA in PLC5 and Mahlavu cells. Upper panels: Western blot analysis of CA9. Middle panels: Migration/invasion abilities of CA9 knockdown cells. Lower panel: Colony formation abilities of CA9 knockdown cells. Representative micrographs of colony formation assays shown above the plot. (B,C) Migration, invasion, and colony formation assays were conducted after transfection of miR-34a or a control plasmid in Mahlavu and PLC5 cells. miR-34a and CA9 expression levels were examined by Western blotting and a real-time PCR, respectively, as shown in the upper two panels. Middle panels: Migration/invasion abilities of miR-34a-overexpressing cells. Lower panel: Colony formation abilities of miR-34a-overexpressing cells. Representative micrographs of colony formation assays shown above the plot. (D) Control and miR-34a-overexpressing Mahlavu cells were infected with lentiviruses carrying either control or CA9-V5 for 48 h and subjected to transwell migration and invasion assays. Migrated/invaded cells are presented as multiples of the control. Three independent replicates were performed for each experiment. Error bars, ±S.D. *p < 0.05 by two-sided Student’s t-test.

Fig 3: Association of rs1048638 with survival of hepatocellular carcinoma (HCC) patients. (A) Schematic illustration of CA9 single-nucleotide polymorphisms (SNPs) evaluate in HCC. (B) Kaplan-Meier analysis of the correlation between rs1048638 genotypes and overall survival (OS) and disease-free survival (DFS) of 86 HCC patients. (C) Kaplan-Meier analysis of the correlation between CA9 expression levels and OS or DFS of 86 HCC patients.

Fig 4: Association of rs1048638 with CA9 expression in hepatocellular carcinoma (HCC) patients. (A) RNA expression scatter diagrams of CA9 mRNA versus miR-34a. Black dots represent expression levels of both genes from specimens in TCGA HCC dataset; the regression line is shown on the plot. (B) Correlations between CA9 mRNA levels and rs1048638 genotypes in 86 HCC specimens. *p < 0.05 by Chi-squared test. (C) CA9 mRNA versus miR-34a expression scatter diagrams of specimens from rs1048638-CC or rs1048638-AA/CA cohorts. Statistical significance of (A) and (C) was assayed by Spearman’s rank correlation test. p values are labeled in the plot.

Fig 5: The CA9 polymorphism determines the response to miR-34a targeting. (A) Schematic illustration of miRNA target sites and the rs1048638 position on the CA9 3'untranslated region (UTR). (B) CA9 3'UTR reporters were transfected with the indicated miRNA or scrambled control (NC) mimics into 293T cells. Luciferase activities were measured and are shown as multiples of the control. (C) A minimum free energy (MFE) analysis of the alignment between miR-34a and the CA9 3'UTR with an “A’ or “C” allele at rs1048638. (D) Upper panel: Western blot analysis of CA9 in HCC cell lines. The rs1048638 genotype of each cell line is labeled below the blot. Lower panel: Quantitative analysis of miR-34a levels in HCC cell lines. Data are presented normalized to the miR-34a level of Mahlavu cells. (E) Upper panel: Western blot analysis of CA9 in PLC5 and Huh7 cells transfected with miR-34a or the control mimic, respectively. miR-34a expression levels are shown in the middle panel. CA9 RNA expression levels are shown in the lower panel. Three independent replicates were performed in each experiment. Error bars, ±S.D. *p < 0.05 by a two-sided Student’s t-test. GAPDH was used as a loading control.