MicroRNA‐191 promotes hepatocellular carcinoma cell proliferation by has_circ_0000204/miR‐191/KLF6 axis

Abstract Objectives MicroRNAs are powerful regulators in hepatocellular carcinoma (HCC) tumorigenesis. MicoRNA‐191 (miR‐191) has been reported to play an important role in HCC, However, the regulatory mechanism is still unclear. In this study, we investigated the role of miR‐191 in HCC and studied its underlying mechanisms of action. Materials and methods The expression of miR‐191 in HCC tissues was determined by quantitative real‐time PCR (qRT‐PCR). The role of miR‐191 in HCC cells was examined by using both in vitro and in vivo assays. Downstream targets of miR‐191 were determined by qRT‐PCR and Western blot analysis. Dual‐luciferase assays were performed to validate the interaction between miR‐191 and its targets. Results The expression of miR‐191 was significantly higher in HCC patients and a higher miR‐191 expression predicted poorer prognosis. Analysis of The Cancer Genome Atlas data sets suggested that miR‐191 positively correlated with cell cycle progression. Gain and loss of function assays showed that miR‐191 promoted cell cycle progression and proliferation. Luciferase reporter assay showed that miR‐191 directly targeted the 3'‐untranslated region of KLF6 mRNA. Furthermore, circular RNA has_circ_0000204 could sponge with miR‐191, resulting in inactivation of miR‐191. Conclusions Our study sheds light on the novel underlying mechanism of miR‐191 in HCC, which may accelerate the development of cancer therapy.

regulating numerous biological processes. 8 Accumulating evidence has suggested that altered miRNA levels are found in various types of human cancers and might play critical roles in tumorigenesis. 9 miR-191, a highly conserved miRNA, was found to be abnormally expressed in more than twenty different cancer types and has shown to be a major player in the regulation of some of these. 10 In patients with HCC, miR-191 was reported as a serum exosomal miRNA and a potential oncogenic target for HCC therapy. 11,12 In HCC cancer cells, miR-191 was shown to be regulated mainly by DNA methylation and involved in the regulation of epithelial mesenchymal transition (EMT). 13 These studies suggested an important role of miR-191 in HCC development. However, the molecular mechanisms by which miR-191 exerted its effects in HCC tumorigenesis as well as its significance remain largely unknown.
Circular RNA is another type of non-coding RNAs. In recent studies, circular RNAs have been reported to participate into tumorigenesis of various types of tumours, including HCC. 14,15 For example, circular RNA SMARCA5 inhibited the proliferation and migration of HCC cells by sponging with miR-17-3p and miR-181b-5p. 16 Moreover, circular RNA circMTO1 was found to function as the sponge of miR-9, thereby suppressing HCC progression. 17 Circular RNA HIPK3 regulated cell proliferation and migration by sponging miR-124 to increase Aquaporin-3 (AQP3) expression. 18 These studies suggested that circular RNA functioned as a miRNA sponge in HCC cells.
Krüppel-like factor (KLF) belongs to a family of zinc finger transcription factors that control essential cellular processes. 19,20 KLF6, a member of the KLF family, functions as an essential player in cell growth and cell cycle progression. 21,22 In cancer studies, KLF6 has been reported to mediate growth suppression by upregulation of p21, whereas downregulation of KLF6 activated c-MYC transcription in prostate cancer. [23][24][25] In HCC, KLF6 was reported to be downregulated in tumour tissues and resulted in cell cycle progression arrest and cell death. 26,27 Besides, Recent studies revealed that KLF6 suppressed the expression of cyclin D, resulting in G1 cell cycle arrest in hepatocellular carcinoma-derived cells. 28

| Cell culture and cell transfection
The Hep3B and HepG2 cells used in this study were obtained from Chinese Academy of Science. Cells were cultured in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% foetal bovine serum (FBS), 100 U/mL ampicillin and 100 g/mL streptomycin (Life Technologies), at 37%, and 5% CO 2 . The HEK 293T cell line was used to produce lentivirus, and cells were transfected with pCDH-pri-miR-191 or purogreen-miZip-191 (shRNA) or pZW-hsa_ circ-0000204 or control plasmid and two package plasmids and then incubated for at least 10 hours at 37%. For the selection of stable cell lines, overexpression or knock-down lentivirus was transduced into liver cancer cells in the presence of polybrene (5μg/mL; Sigma), and cells were selected using medium containing 2 μg/mL puromycin for a total of two weeks. pcDNA-KLF6 was transient transfection into liver cancer cells for expressing KLF6.

| Clinical samples
All tissue samples used in this study were collected from the Affiliated Hospital of Nanjing University of Chinese Medicine.
Written informed consent was obtained from all study participants. This study was approved by the Ethics Committee of Nanjing University of Chinese Medicine Authority.

| Human HCC data set and bioinformatics analysis
In this study, HCC microarray data sets GSE10694 29  prognosis. 31 In order to gain insight into the biological mechanism involved in HCC progression through miR-191, Pearson correlation tests were performed in tumour samples of the TCGA data sets. 32 Moreover, pathway enrichment analysis was performed using Funrich software version 3.0 (http://funri ch.org/index.html). HCC circular RNA microarray data sets GSE94508 33 (Agilent-069978 Arraystar Human CircRNA microarray V1) were downloaded from the GEO database. Differences in expression were analysed by GEO2R(https ://www.ncbi.nlm.nih.gov/geo/geo2r ). Circular-miRNA prediction was performed by the 34 (https ://circi ntera ctome.nia.nih.gov/).

| Edu assay
Cell proliferation was evaluated by Edu (5-ethynyl-20-deoxyuridine) assay using a Cell-Light EdU DNA Cell Proliferation Kit (RiboBio). In brief, Hep3B and HepG2 cells (1×10 3 ) were seeded into wells of 8well plates. After incubation at 37% and 5% CO 2 for 48 h, cells the proliferation rate was calculated according to the manufacturer's instructions. Representative images were taken using Eclipse Ni-U fluorescence microscopy.

| CCK-8 assay
The CCK-8 assay was performed in six repeated wells by using the Cell Counting Kit-8 Assay kit (Doindo, Japan), following the manufacturer's guidelines. In brief, 1×10 3 cells were seeded in 96-well plates, and the OD 450 absorption value was measured using an automatic microplate reader (Synergy4; BioTek).

| Colony formation assay
A total of 0.5×10 3 Hep3B and HepG2 cells were plated in wells of a 6-well plate and cultured for about 2 weeks. Then, images were taken, and the number of colonies per well was counted by image J.

| Western blot analysis
Western blot analyses were performed as previously described. 35 Primary antibodies used were as follows: KLF6 (1:1000 dilution,

| RNA extraction and real-time RT-PCR
Total RNA was extracted using TRIzol reagent (Invitrogen). Targets were analysed by the SYBR Green qRT-PCR assay according to the manufacturer's instructions (Applied Biosystems). The primers used are listed in Additional file Table S1. For miRNA detection, the reverse transcribed cDNA was synthesized using the All-in-One™ miRNA First-Strand cDNA Synthesis Kit (GeneCopoeia). Next, miR-191 expression was determined with the All-in-One™ miRNA qRT-PCR Detection Kit (GeneCopoeia) and U6 snRNA was used as an endogenous control.

| Animal studies
All animal studies were performed in accordance with protocols approved by the Animal Experimentation Ethics Committee of Nanjing University of Chinese Medical Center. For tumour growth analysis in a xenograft model, 5-6 weeks old immunodeficient mice were used.
A total of 5 × 10 5 cells were subcutaneously injected into the right collar of the mice (n = 4 per group). Tumour growth was analysed by measuring tumour length (L) and width (W) and by calculating the volume (V) using the formula, V = L × W 2 /2. Immunohistochemical analysis for targets was performed as previously described. 35

| Statistical analysis
Statistical analyses were conducted using SPSS16.0 or GraphPad Prism 5.0 software. Comparisons between groups were analysed by the t test, and multiple group comparisons were analysed using oneway ANOVA P < 0.05 was considered statistically significant. Data were presented as the mean ± standard deviation (SD).

| miR-191 is upregulated and positively correlated with cell cycle progression in hepatocellular carcinoma tissues
To determine the role of miR-191 in HCC tumorigenesis, we first examined miR-191 expression levels in HCC tissues. The data showed that miR-191 was highly expressed in HCC specimens when compared to adjacent non-cancerous tissues (n = 30) ( Figure 1A,B and Table 1). In addition, based on the analysis of GEO data sets, GSE10694 29 and GSE6857, 30 we found that miR-191 was frequently

| miR-191 affects cell proliferation in vivo
To explore the effects of miR-191 in vivo, we established a mouse xenograft model to investigate whether miR-191 could promote tumour growth. HepG2 cells overexpressing or knocked down for miR-191 were subcutaneously injected into nude mice. We demonstrated that after 35 days, the tumour size was decreased in the miR-191 knock-down group when compared with the control group (Figure 4A,B). Similar results were obtained for tumour weight ( Figure 4C). Moreover, tumour sections from the miR-191 knock-down group exhibited weaker Ki67 staining when compared to those from the control group, suggesting that knock-down of miR-191 inhibited tumour growth ( Figure 4D). In addition, opposite effects were obtained from the miR-191 overexpression group.
Thus, these results showed that miR-191 promoted tumour growth in vivo.

| KLF6 is the downstream target that is directly regulated by miR-191 in HCC
Having elucidated the functional phenotype of miR-191 in HCC, we next identified its targeting genes to gain further insights into the molecular mechanism of miR-191 on HCC cells. To our knowledge, miRNAs usually function as a gene silencer, suppressing expression of its targets. 36 Here, we primarily found negatively correlated genes of miR-191 in the liver cancer TCGA database.
Our analysis showed that a total of 2202 transcripts negatively correlated with miR-191(r < −0.1). We searched for miR-191 targeted genes using miWalk 3.0, an integrated database predicted gene-miRNA interaction and the results showed that there were 7055 predicted targets. Moreover, we also found a data set that demonstrated that 178 transcripts were upregulated after F I G U R E 1 miR-191 is upregulated and positively correlated with cell cycle progression in hepatocellular carcinoma tissues. A, Expression levels of miR-191 in 30 pairs of HCC tissues and adjacent non-tumour tissues. The data were analysed by a minus delta Ct method. B, Bars represent the relative miR-191 expression comparing expression in HCC tissues C, versus adjacent non-tumour tissues (P) using alogarithmic scale. C, D, Expression levels of miR-191 in 78 paired, 240 paired HCC tissues and 10 normal tissues, respectively, obtained from the GEO database (GSE 10694 and GSE6857). E, The prognostic significance of miR-191 for 372 HCC patients assessed by Kaplan-Meier analysis (logrank P = 0.0071). F, Expression level of genes positively correlated with miR-191 in HCC tissues obtained from TCGA LIHC data sets. G, The positive correlation between the expression of miR-191 and cell cycle related genes (such as CDC7, CDC25A and CDCA8) in liver cancer TCGA data sets. H, The most five significant biological pathways were enriched by miR-191 positive correlation genes (***P < 0.001)  knock-down of miR-191 in HepG2. 11 These three data sets were overlapped to identify the genes that met the following crite-  Figure 5F). Together, these findings suggested that KLF6 was a direct target of miR-191.  Figure 6A). Cell cycle-related protein levels, which are regulated by KLF6 were also changed as we expected, such as c-Myc and CCND1/ CCND2 ( Figure 6B). Similarly, the increase in KLF6 blocked the promotion of cell cycle progression accelerated by miR-191 ( Figure 6C). Taken together, these results indicated that KLF6 mediated miR-191 promotion of HCC cell cycle progression and proliferation.

| miR-191 is sponged by circular RNA has_ circ_0000204
Recent studies suggested that miRNA could be sponged by circular RNA, resulting in miRNA inactivation. 37 According to GEO data sets GSE94508, 33 a circular RNA profile of five paired HCC patients, there were six circular RNAs had the potential binding site with miR-191 and downregulated in HCC tissues ( Figure 7A,B). Next, we performed can be attributed to the regulation of genes controlling hepatocyte growth and differentiation. 23 In other cancers, it was demonstrated that upregulation of KLF6 inhibited cell cycle progression, which was consist of our findings.
In HCC, miR-191 was highly expressed in HCC tissues and a higher expression of miR-191 predicted poorer survival. 11 In our studies, we expanded the number of samples to identify these views by using tissues we collected as well as by the GEO database data sets we analysed. As expected, our results were consisted with the data presented previously. However, a few studies were performed revealing the regulatory mechanism of miR-191 upregulation.
Yinghuaet al. proposed that miR-191 was upregulated due to promoter hypomethylation, and hypomethylation of miR-191 promoted miR-191 expression, thus, induced HCC cell EMT. 13  could not explore all circular RNAs to influence miR-191 to alter cellular function in HCC. In addition, we also believed that the function of has_circ_0000204 was not limited to regulation of cell cycle and cell proliferation because more than one miRNA binding site was present.

| CON CLUS IONS
The inspiration of this study was that miR-191 expression mark-

ACK N OWLED G EM ENTS
This study was supported by grants from the "Six Talent

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.