miR-195-5p Targets CDK1 To Regulate New DNA Synthesis and Inhibit the Proliferation of Hepatocellular Carcinoma Cells

In cell biological functions and viability, cyclin-dependent kinase 1 (CDK1) takes an essential part. miR-195-5p is pivotal in pathogenesis and development of hepatocellular carcinoma (HCC). But in HCC, whether there is a connection between CDK1 and miR-195-5p remains an unanswered question. In view of this, this study focuses on exploring the mechanism of miR-195-5p/CDK1 in the progression of HCC. The bioinformatics method was applied to predict target mRNA and upstream miRNAs, and further analyzes the signal enrichment pathway of target mRNA. We utilized qRT-PCR and Western blot for detecting expression of genes, as well as their corresponding protein levels. Cell cycle was assayed through flow cytometry. As for the examination of DNA replication, the EDU staining was employed. Cell proliferation was determined via plate colony formation assay. The combined application of bioinformatics analysis and dual-luciferase gene assay assisted in figuring out the binding relationship between miR-195-5p and CDK1. DNA damage was marked by immunofluorescence staining. CDK1 was overexpressed in HCC cells, and enriched in cell cycle and DNA replication pathway. Silencing CDK1 modulated cell cycle of HCC cells and inhibited DNA replication and proliferation. In HCC cells, miR-195-5p targeted and reduced CDK1 expression, inhibited the G1 phase-to-S phase transition, induced DNA damage response, and inhibited DNA replication and proliferation. miR-195-5p targeted CDK1 and repressed synthesis of new DNA in HCC cells, thus restraining HCC cell proliferation.


Introduction
Hepatocellular carcinoma (HCC) is a major cause of cancer-related deaths in the world [1]. Among the current treatment methods, patients with advanced HCC have few treatment options and their prognoses are poor [2]. Therefore, further exploration of potential therapeutic targets of HCC may help improve the poor prognoses of patients. Cyclin-dependent kinase (CDK) is an important cell cycle regulating protein. In the CDK family, only CDK1 can foster cell cycle independently [3]. A recent study showed that CDK1 competitively restrained CDK2-cyclin A complex formation, thereby inhibiting re-replication of the replicated DNAs, thus successfully completing the replication in late S phase [4]. In addition to these functions, CDK1 is involved in repair of the double-strand break of homologous recombination (HR)-dependent DNA [5] and coupling DNA damage repair pathways to cell cycle processes [6]. Overall, CDK1 is a major modulator of several core biological events.
CDK1 expression is increased in human colorectal cancer [7] and prostate cancer [8]. Thus, CDK1 is involved in tumor progression. Recent studies showed that combination of CDK1 inhibitor JNJ-7706621 and paclitaxel is effective in treating transplantable liver cancer and repressing tumor growth [9]. However, whether the promotive effect of CDK1 on liver cancer progression is related to modulation of cell cycle progression, DNA replication, and DNA damage repair has not been clarified.
Recently, more and more studies have confirmed the significance of microRNA (miRNA) as regulator in various cancers, like hepatocellular carcinoma (HCC) [10]. Studies displayed the decrease of miR-195 in cancers, containing liver cancer, stomach cancer, bladder cancer, breast cancer, and adrenocortical cancer [11][12][13][14]. miR-195 has been discovered to exert a suppressive role in HCC cell colony formation in vitro and HCC tumor development in nude mice [14]. Another report also mentioned that miR-195 inhibited tumorigenesis and regulated G1/S transition of HCC cells [11]. Recent studies suggested that miR-195-5p/CDK1 may act as a possible prognostic biomarker and therapeutic target for HCC [15]. But specific biological function and regulatory mechanism of miR-195-5p/ CDK1 in HCC are rarely studied.
We unveiled that CDK1 expression was elevated in HCC cells, and miR-195-5p affected cell cycle, DNA replication, and DNA damage response via targeting CDK1, thus restraining HCC cell proliferation. CDK1 played a pivotal carcinogenic role in HCC development and progression and is modulated by miR-195-5p. These findings bolstered a better understanding of function of miRNAs in HCC and provided targets for HCC diagnosis and treatment.
The "STRING" database was applied to conduct protein-protein interaction (PPI) network analysis and calculate the core degree of genes in the network to determine the target mRNA (interaction score: 0.9). Meanwhile, the mature miRNA (normal: 50, tumor: 375), the mRNA expression data (normal: 50, tumor: 374), and the clinical data of HCC were obtained from TCGA-LIHC.
Starbase and TargetScan bioinformatics databases were utilized for predicting miRNAs upstream of target mRNA, and the results were overlapped with the differential miRNAs to obtain the miRNA with binding site with the target mRNA. "Survival" package was utilized for survival analysis and clinical stage analysis of the target mRNA in TCGA-LIHC dataset. Gene Set Enrichment Analysis (GSEA) software was applied for pathway enrichment analysis of the target mRNA. Pearson correlation analysis was measured to determine the correlation between the expression of miR-195-5p and CDK1.

Flow Cytometry
After being digested, suspended, and rinsed twice with PBS, cells were made into singlecell suspension (1 × 10 6 cells/mL). The cells were fixed with 70% cold ethanol and then stored at − 4 °C overnight. Before staining, the cells were rinsed with PBS to remove supernatant. Cells were added with PI/RNase A staining solution (KGA511-KGA512; Keygentec, China) and incubated in darkness for 30 min. Flow cytometer (Partec, Germany) was adopted to analyze cells and Flowjo software was employed to process data. The proportion of cells in different cell cycle phases in each experimental group was compared and analyzed.

Real-time Quantitative Polymerase Chain Reaction (qRT-PCR)
The Trizol Reagent (15,596,018; Thermo Fisher Scientific, USA) was employed for the extraction of RNA from cells. The reverse transfection of the extracted RNA into cDNA was accomplished by utilizing the High-Capacity cDNA Reverse Transcription Kit (4,374,967; Thermo Fisher Scientific, USA). According to instructions of the kit, qPCR analysis was done with SYBRA Green PCR Master Mix (RR820A; Takara, Japan) in QuantStudio 3 (Thermo Fisher Scientific, USA) qPCR instrument (primers: see Table 1). GAPDH was utilized as an internal reference gene for CDK1 expression, and U6 for miR-195-5p.

Western Blot
RIPA lysis buffer was used to treat cells and isolate intracellular proteins. SDS-polyacrylamide gel electrophoresis was performed on the same amount of protein samples. Then, the proteins were blotted onto the PVDF membrane which was then blocked with 5% skim milk at room temperature for 1 h. Finally, the membrane added with primary antibodies was incubated at 4 ℃ overnight. Antibodies were obtained from Abcam (UK): anti-CDK1 antibody (ab32094), anti-RPA32 antibody (ab109084), anti-p-RPA32 antibody (ab109394), and anti-GAPDH antibody (ab181602). After the primary antibody was washed, membrane was cultured with the secondary antibody IgG H&L (HRP, ab6721) for 1 h at room temperature. In the end, we observed the protein bands by utilizing the enhanced chemiluminescence kit (ECL, Thermo Fisher Scientific, USA).

EDU Staining
Cells successfully transfected with different plasmids were seeded into 24-well plates (5 × 10 3 cells/well) with 3 replicates per group. After 12 h of cell culture, the medium of the cells was replaced with 50 μmol/L EDU medium, and the cells were incubated for 2 h in CO 2 incubator. After EDU medium being discarded, cells were fixed with addition of 4% paraformaldehyde for 30 min. Next, 200 μL 2 mg/mL glycine was added, and cells were incubated with a shaker at room temperature for 5 min, followed by PBS washing for 3 times, 5 min each. Subsequently, 300 μL 0.5% Triton X-100 was supplemented, and cells were maintained for 30 min at room temperature. Next, 200 μL Apollo staining solution was supplemented, and cells were maintained at room temperature without light for another 30 min, followed by 0.5% Triton X-100 washing for 3 times, 10 min each. At last, DAPI solution was added for nuclear staining. The staining results were observed

Immunofluorescence Staining
The transfected cells were inoculated on a slide and treated with 4% formaldehyde solution with 0.1% TritonX-100 (T8787-100ML; Sigma, USA) at room temperature for 15 min. After fixation, the slide was blocked with 3% BSA solution (configured with PBS) containing 0.5% TritonX-100 for at least 1 h. After blocking, the slide was maintained with the primary antibodies (anti-γ H2AX antibody ab81299, anti-RPA32 antibody ab109084) at 4 ℃ overnight. Slide was incubated with AlexaFluor488-labeled secondary antibody (Life Technologies, USA) for 1 h at room temperature and stained with DAPI. The automatic Nikon Eclipse Ni microscope with Nikon Elements software (Nikon Instruments, Japan) was utilized for imaging and photographs were taken with the 100 × oil lens.

Plate Colony Formation Assay
0.25% trypsin was used for the digestion of cells (logarithmic growth stage) which were then blown into single cells. Cells were then suspended in cell growth medium with 10% FBS for reserve. Each dish in each group was inoculated with 1 × 10 3 cells and added with 10 mL 37 ℃ pre-heated medium, and the cells were dispersed evenly by gently rotating. The cells were cultured in an incubator (37 ℃, 5% CO 2 ) with saturated humidity. As long as visible colonies were found in the petri dish, culturing was put to an end. After removing the supernatant, we utilized PBS to wash the cells twice. Then cells were added with 5 mL of pure methanol and fixed for 15 min. Next, the fixing solution was removed and we added appropriate amount of crystal violet for staining cells for 20 min. Then the staining solution was washed off slowly with running water and drying in air. A camera was used to photograph, and the number of cell colonies was counted.

Dual-Luciferase Reporter Assay
The 3′UTR terminal sequences of wild-type and mutant CDK1 were imported into the pmiR-Reporter vectors (Addgene, USA) to construct reporter plasmids. miR-195-5pmimic or NC-mimic and the reporter gene plasmid were co-transfected into HepG2, and the fluorescence intensity of transfected groups was assessed through luciferase activity assay kit (E1501; Promega, USA) after 24 h of culture.

Experimental Design
Through bioinformatics and cell function assays, CDK1 was found to be notably highly expressed in HCC. Hence, we constructed different cell lines (si-NC, si-CDK1 were transfected into HCC cells, respectively), which were cultured in DMEM medium with 10% FBS (37 ℃, 5% CO 2 ). Then we utilized flow cytometry, EDU staining, and colony formation assay to detect the effects of CDK1 on HCC cell cycle, DNA replication level and proliferation. Additionally, we also found that miR-195-5p could target and negatively regulate CDK1, so we designed a rescue assay to detect the effect of miR-195-5p and CDK1 on HCC cell function. Experimental groups were set as follows: oe-NC + NC-mimic, miR-195-5p-mimic + oe-NC, miR-195-5p-mimic + oe-CDK1 groups, which were transfected into HCC cells and cultured. Then the effect of miR-195-5p/CDK1 axis on HCC cell functions was investigated by flow cytometry, EDU staining, cell colony formation, immunofluorescence and Western blot assays.

Data Analysis
GraphPad Prism 6 software (GraphPad Software, USA) was utilized for statistical analyses. All experiments were conducted at least 3 times, with results expressed as mean ± SD. We made a comparison between groups using the t-test or one-way analysis (Python) of variance. p < 0.05 meant statistically significant difference, and * in the figure meant p < 0.05.

CDK1 May Be a Target of HCC
A total of 194 differential mRNAs were obtained by the differential analysis of combined gene microarrays. The top 20 upregulated as well as downregulated differential genes were selected to plot a heat map (Fig. 1a). Subsequently, the degree value of interaction between each gene and other genes was calculated through the PPI network analysis, and it was found that CDK1 was at the core of the whole interaction network, with the highest degree value (Fig. 1b). Therefore, analysis of expression levels between the tumor group and the normal group displayed that CDK1 was conspicuously increased in tumor tissues (Fig. 1c). Also, it has been reported that CDK1 is up-regulated in tumor tissues [16]. The survival analysis on the TCGA-LIHC data set illustrated that patients with overexpressed CDK1 had a longer survival time than patients with decreased CDK1 level (Fig. 1d). It has been documented that CDK1 overexpression has an adverse effect on the prognosis [17]. Based on clinical information of patients, we found that CDK1 had notable differences in different stages of HCC (Fig. 1e). CDK1 level increased with progression of tumor stage in the early stage, but decreased in stage IV. The results demonstrated the potential of CDK1 as an oncogene in HCC in vivo.

CDK1 Regulated Cell Cycle, DNA Replication, and Proliferation of HCC cells
To investigate the signaling pathways where CDK1 may be involved in, we performed GSEA on CDK1. The result of GSEA showed that CDK1 was noticeably enriched in cell cycle and DNA replication pathways (Fig. 2a), which were significantly correlated with cell proliferation. Therefore, to figure out the impact of CDK1 on cell cycle and proliferation capacity of HCC cells, we conducted the following experiments. Firstly, the mRNA expression level of CDK1 in normal human hepatocytes L-02 and HCC cell lines (HepG2, SNU-398, Hep3b) was detected via qRT-PCR. With a contrast to the normal liver cells, HCC cells had a higher level of CDK1 expression (Fig. 2b). As revealed by Western blot, CDK1 protein expression was noticeably enhanced in HCC cell lines compared to normal human liver cell line. Among them, CDK1 presented the highest expression level in HepG2 cell line (Fig. 2c). Therefore, HepG2 cell line was selected for subsequent experiments. We constructed two CDK1-silenced plasmids and their negative controls. qRT-PCR result showed that si-CDK1-1 had lower expression efficiency (Fig. 2d). Therefore, we selected si-CDK1-1 for transfection and named it si-CDK1. Flow cytometry results displayed that proportion of G0/G1 phase cells was dramatically increased and that of S phase cells was decreased in si-CDK1 group but not in control group (Fig. 2e). EDU staining results displayed that DNA replication level was decreased in si-CDK1 group in comparison to the control group (Fig. 2f). Cell colony formation unveiled that proliferation level of the si-CDK1 group was remarkably reduced compared to control group (Fig. 2g). Thus, CDK1 silencing could block process from G0/G1 phase to S phase and inhibit DNA replication, thus inhibiting HCC cell proliferation.

miR-195-5p Targeted And Repressed CDK1 Expression
TarBase and TargetScan bioinformatics databases were utilized to predict upstream miRNAs of CDK1. The predicted results were intersected with differential miRNAs, Fig. 1 CDK1 may be a target of HCC. a Heat map of the top 20 differentially expressed genes in multimicroarrays combined analysis (red: upregulated differential genes; green: downregulated differential genes). b Degree statistics in gene PPI network graph (abscissa: degree value; ordinate: gene name). c Box plot of CDK1 level in the normal (green) and the tumor (red) groups. d The impact of CDK1 level on overall survival curve of patients (blue: CDK1 low expression; red: CDK1 high expression). e Box plot of CDK1 level in different tumor stages (red: stage I; green: stage II; blue: stage III; turquoise: stage IV). * p < 0.05 and 3 differential miRNAs were obtained (Fig. 3a). Pearson correlation analysis was conducted between CDK1 and these 3 miRNAs, and it was found that miR-195-5p had the strongest negative correlation with it (Fig. 3b, c). Expression analysis results of the tumor group and the normal group disclosed that miR-195-5p was noticeably lessexpressed in tumor tissues (Fig. 3d). Subsequently, the binding sites of miR-195-5p and CDK1 3′-UTR were predicted by bioinformatics methods (Fig. 3e). The dual-luciferase  and CDK1. The results showed that fluorescence intensity in WT-CDK1 + miR-195-5pmimic group was dramatically downregulated compared with control group. However, there was no notable difference in fluorescence intensity between MUT-CDK1 + miR-195-5p-mimic and the control group, indicating that CDK1 could be targeted by miR-195-5p (Fig. 3f). qRT-PCR and Western blot disclosed that enforced miR-195-5p level repressed mRNA and protein expression level of CDK1 (Fig. 3g-i). These experimental results illustrated that CDK1 was a downstream target gene of miR-195-5p, and miR-195-5p constrained CDK1 expression.

miR-195-5p Affected Cell Cycle, DNA Replication, and Proliferation of HCC Cells by Targeting CDK1
To clarify the effect of miR-195-5p/CDK1 axis on HCC cell functions, we designed a rescue assay with the following experimental groups: oe-NC + NC-mimic, miR-195-5pmimic + oe-NC, and miR-195-5p-mimic + oe-CDK1 groups. Flow cytometry revealed that in comparison to control group, the proportion of G0/G1 phase cells in the miR-195-5p-mimic + oe-NC group was notably elevated, while the proportion of S phase cells being reduced. However, proportion of cells at G0/G1 phase in the miR-195-5pmimic + oe-CDK1 group was noticeably inhibited, while proportion of cells at S phase recovered prominently (Fig. 4a). EDU staining unveiled that in comparison to control group, DNA replication level decreased in the miR-195-5p-mimic + oe-NC group, while proportion of DNA replication positive cells recovered in miR-195-5p-mimic + oe-CDK1 group (Fig. 4b). Cell colony formation assay displayed that enforced miR-195-5p level restrained proliferation of HepG2 cell line compared to control group, while introduction of oe-CDK1 co-transfection promoted the proliferation of HepG2 cell line (Fig. 4c). Experimental results demonstrated that miR-195-5p affected HCC cell cycle by negatively regulating CDK1, thereby inhibiting DNA replication and reducing the proliferation of HCC cells.

miR-195-5p Promoted DNA Damage Response in HCC by Targeting CDK1
Studies have shown that CDK1 is involved in DNA damage repair, especially HR-mediated DNA double-strand break repair, and its main role is to promote the terminal resection [5]. Therefore, we set out to investigate whether CDK1-involved DNA damage repair is regulated by miR-195-5p. To this end, immunofluorescence staining was performed to label γH2AX and RPA32. Taking the oe-NC + NC-mimic group as control, DNA damage response was triggered dramatically in miR-195-5p-mimic + oe-NC group, and total fluorescence intensity of γH2AX and RPA32 increased significantly. In contrast, the DNA damage response was noticeably downregulated in miR-195-5p-mimic + oe-CDK1 group (Fig. 5a, b). We also revealed that in relevant to oe-NC + NC-mimic group, miR-195-5pmimic + oe-NC group cells presented excessive phosphorylation of RPA32, which was a marker of the activation of DNA damage response. However, after oe-CDK1 co-transfection, the expression of P-RPA32 protein decreased significantly (Fig. 5c). Hence, miR-195-5p may be a tumor repressor. It promoted DNA damage response in HCC by targeting CDK1, thus constraining HCC cell proliferation.

Discussion
Through bioinformatics analysis and cell experiments, we denoted that CDK1 was overexpressed in HCC tissues and cells, which was consistent with results reported previously [18]. We analyzed effect of CDK1 on biological behaviors of HCC and unveiled that CDK1, as an oncogene, affected cell cycle and promoted DNA replication and proliferation of HCC cells. CDK1, as reported previously, exerts as a oncogene in other cancers. For example, CDK1 is an oncogene whose expression increases with progressive progression of epithelial ovarian cancer [16]. CDK1 level is related to tumor size and grade of pancreatic ductal adenocarcinoma, and patients with increased CDK1 level had a shorter survival time [17]. CDK1 plays an essential role as an oncogene in HCC.
In order to further explore CDK1-related genes, we found the miR-195-5p/CDK1 signaling axis through bioinformatics analysis and that in HCC tumor tissues, miR-195-5p was dramatically less-expressed. Meanwhile, we confirmed the miR-195-5p/CDK1 targeting relationship and the repression of CDK1 level by miR-195-5p. hsa-miR-195-5p overexpression in HCC cells reduces PHF19 level, thereby inhibiting malignant behaviors of hepatocytes in vitro [19]. Consistent with it, we manifested that miR-195-5p blocked transition of cell cycle from G1 phase to S phase by targeting CDK1 and constraining CDK1 level, thus inhibiting the DNA replication of HCC cells and proliferation of HCC cells.
CDK1 is significantly pluripotent. In addition to regulating cell cycle, it also participates in DNA damage repair and activation of CHK1-dependent cell cycle checkpoints by phosphorylation of BRCA1 [20]. To investigate how miR-195-5p/CDK1 affects DNA damage repair in HCC cells, immunofluorescence experiments were conducted. It was found that miR-195-5p induced significant DNA damage response by targeting CDK1, and the expression of p-RPA32 protein, a marker protein of DNA damage response, was increased. In conclusion, inhibition of CDK1 by miR-195-5p overexpression may trigger replicationrelated DNA damage response, leading to DNA repair failure.

Conclusion
In a word, CDK1 has an oncogenic modulatory impact on HCC. MiR-195-5p could modulate HCC cell cycle through targeting CDK1, induce DNA damage response, and inhibit DNA replication and cell proliferation of HCC cells. The limitation of this study is that the effect of CDK1 on HCC by in vivo experiments and the downstream signal pathway of miR-195-5p/CDK1 axis in HCC have not been researched. These results provided a deeper understanding about the mechanism of miR-195-5p/CDK1 in HCC, which is an effective potential therapeutic target for the progression of HCC. The RPA32 protein was labeled by immunofluorescence staining. c RPA32 and p-RPA32 protein levels were detected via Western blot