MicroRNA-144 inhibits cell proliferation, migration and invasion in human hepatocellular carcinoma by targeting CCNB1

Abstract

Background: Hepatocellular carcinoma (HCC) is one of the most common malignancies with a high morbidity and mortality worldwide. MicroRNAs are key regulators of HCC genesis. However, the regulatory role and underlying mechanisms of microRNA in HCC is still limited.

Methods: Cyclin B1 (CCNB1) mRNA levels were examined in non-tumor and liver cancer of The Cancer Genome Atlas (TCGA) cohort. CCNB1 was knockdown to evaluate the HCC cell proliferation, migration and invasion. MicroRNA-144 targeting CCNB1 was identified with TargetScan analysis and confirmed with reporter assay. Overexpression of MicroRNA-144 was achieved using microRNA mimics and function of microRNA-144 was tested in vitro HCC cell line proliferation and in vivo tumor formation experiments.

Results: Here, we found that the high level expression of CCNB1 was closely associated with poor prognosis in HCC patients. Knockdown of CCNB1 by RNA interference significantly inhibited cell proliferation, migration and invasion in HCC. Furthermore, we found that miR-144 directly targeted CCNB1 and inhibited CCNB1 expression. Moreover, in vivo experiments of subcutaneous tumor formation further demonstrated that miR-144 delayed tumor formation by negative regulation of CCNB1.

Conclusion: Therefore, we conclude that microRNA-144/CCNB1 axis plays an important role in human HCC. Therapies targeting microRNA-144 could potentially improve HCC treatment.

Keywords: CCNB1; Hepatocellular carcinoma; Invasion; MicroRNA-144; Migration; Proliferation.

Fig. 1

Bioinformatics analysis of CCNB1 in HCC tissues and normal liver tissues using TCGA cohort public database. a Comparison of CCNB1 mRNA levels in HCC tissue and non-tumor tissues in the TCGA database. The expression of CCNB1 was normalized using a logarithm. p < 0.0001, n = 424. b Comparison of CCNB1 mRNA levels in 50 pairs of HCC tissues and adjacent non-tumor tissues. c Pearson correlation between CCNB1 mRNA level and proliferation marker Ki67 mRNA level. p < 0.0001, n = 374. d Kaplan–Meier analysis of the overall survival and disease-free survival time of patients with high or low CCNB1 expression, n = 313. e The expression of CCNB1 mRNA level in HCC patients with different TNM stages (p = 0.015, n = 292)

Fig. 2

CCNB1 knockdown promotes apoptosis and suppresses proliferation in HCC cell line HepG2 and SMMC-7721. a Real-time PCR analysis of CCNB1 mRNA levels and b western blot analysis of CCNB1 protein levels in HCC cell line Hep-G2 and SMMC-7721 with no transfection (control), transfected with control siRNA (Scramble-siRNA) or CCNB1 siRNA (CCNB1-siRNA). Bargraphs in a represent normalized data from three independent experiments. **p < 0.01 vs. control. b Actin expression levels were detected as an endogenous control. Experiments were repeated twice and representative data was shown. c, d HepG2 and SMMC-7721 cells were transfected with either Scramble siRNA or CCNB1 siRNA at 50 nM. 48 h after transfection, cells were subjected to apoptosis detection using flow cytometry by Annexin-V/Propidium Iodide (PI) staining. c Representative dot plots show the percentages of Annexin V +/PI- apoptotic cells. d Bar graphs represent the mean ± SD from 3 independent experiments. *p < 0.05, **p < 0.01 vs. Control or Scramble siRNA. e HepG2 and SMMC-7721 cells were transfected with either Scramble siRNA or CCNB1 siRNA at 50 nM. Cell proliferation at different time points was measured by MTT assay. Bar graphs represent the mean ± SD from triplicate wells and experiments were repeated three times. **p < 0.01 vs. Cell only group

Fig. 3

CCNB1 knockdown inhibits HCC cell migration and invasion. a HepG2 and SMMC-7721 cells were transfected with either Scramble-siRNA or CCNB1 siRNA at 50 nM. Micrographs of HepG2 and SMMC-7721 cells at 0 h and 48 h after wounding. b Migration distance of HepG2 and SMMC-7721 cells for 48 h was calculated and bar graphs represent the mean ± SD from 3 independent experiments. ^**p < 0.01 vs. NS-siRNA. c, d Transwell Cell invasion activity was evaluated by Matri-gel invasion assay. The number of invading cells was counted in three images per membrane by microscopy using a ×20 objective, scale bar = 100 μm. Experiments were performed in triplicate and repeated at least three times. **p < 0.01 vs. NS control

Fig. 4

miR-144 inhibits CCNB1 expression by binding to CCNB1 3′-UTR CCNB1. a Diagram of the putative binding sites of miR-144 on the 3′-UTRs of CCNB1 and the mutated sequence of 3′ UTRs of CCNB1. b The WT 3′-UTR or mutated 3′-UTR of CCNB1 was fused to the luciferase coding region and co-transfected into HEK293T cells with miR-144 mimics. Relative luciferase activity was determined 48 h after transfection. Data are expressed as mean ± SD from 3 independent experiments. **p < 0.01 vs. Vector Control. c RT-PCR analysis of the effect of CCNB1 expression in HepG2 and SMMC-7721 cells after transfection with miR-144 mimics. GAPDH expression levels were detected as an endogenous control. Data are expressed as mean ± SD. Experiments were performed in triplicate and repeated at least three times. **p < 0.01 vs. control. d Western blot analysis of the CCNB1 expression in HepG2 and SMMC-7721 cells after transfection with miR-144 mimics or miR-144 inhibitor. Actin expression levels were detected as an endogenous control. Experiments were repeated twice and representative data was shown

Fig. 5

MiR-144 mimics elicit similar proliferation repressive effect in HCC cell. a miR-144 expression levels were analyzed in liver cancer or non-tumor tissues of TCGA cohort. b CCNB1-siRNA HepG2 and SMMC-7721 cells were transiently transfected with miR-144 mimics or inhibitor, and the cell proliferation was measured by MTT at different time points. Bar graphs represent the mean ± SD from triplicate wells and experiments were repeated three times. *p < 0.05, **p < 0.01, ***p < 0.001 vs. Scramble control

Fig. 6

MiR-144 overexpression suppresses the tumor formation of SMMC-7721 cells in nude mice. a, b Photographs of nude mice and tumor tissues from different groups at day 30. c Growth curves of tumor volumes in xenografts of nude mice were determined based on tumor volume measured every 5 days for 30 days (n = 4). d Tumor weights of different groups at day 30 (n = 4). Data in a–d are representative of two independent experiments. *p < 0.05, **p < 0.01 vs. Lenti-Scramble, **p < 0.01 vs. Lenti-Scramble. e Representative immunohistochemistry images of Ki67 staining of tumor sections obtained from the Lenti-Scramble and Lenti-miR-144 mouse groups. Magnification ×200, scale bar = 10 μm