1. Patients and specimens
The study was approved by the Ethic Committee of Third Hospital, Peking University. Written consent was obtained from all patients prior to surgery. A total of 71 HCC tissues and matched adjacent nontumor tissues were obtained from patients who underwent surgery in the Third Hospital, Peking University. None of the patients had received chemotherapy, radiotherapy, or any other anticancer therapy before surgery. The diagnoses were histologically confirmed by experienced pathologists. Paired tissue specimens (tumor and adjacent normal tissues) were collected from the patients, immediately frozen in liquid nitrogen, and stored at -80°C until use. Clinical information and follow-up data of the patients are listed in Table 1.
2. CircRNA microarray hybridization and data analysis
Total RNAs were digested with RNase R (20 U/μL, Epicentre, Madison, WI, USA) to remove linear RNAs and enrich circular RNAs. The enriched circular RNAs were amplified and transcribed into fluorescent cRNA utilizing a random priming method (Super RNA Labelling Kit; Arraystar, Rockville, MD, USA). The labeled cRNAs were hybridized onto the Human circRNA Array (8 × 15K, Arraystar). The slides were incubated for 17 h at 65°C in a hybridization oven (Agilent Technologies, Santa Clara, CA, USA). CircRNAs, with statistically significant fold-change differential expression between HCC and paired normal tissues (FC ≥ 2 and p ≤ 0.05) were identified by Volcano plot filtering. Hierarchical clustering was performed to determine the distinguishable expression pattern of circRNAs among samples.
3. Cell culture
The human HCC cell lines (BEL-7402, BEL-7404, Huh7, Hep3B, and HepG2) and normal hepatocyte cells HL-7702 were purchased from the Shanghai Institute of Cell Biology (Shanghai, China) with a certificate of authenticity for each cell line. The samples were maintained in DMEM/RPMI1640 (Gibco BRL, Grand Island, NY, USA) supplemented with 10% heat-inactivated fetal calf serum, 2 mM L-glutamine, and 100 U/mL penicillin–streptomycin mixture (Gibco BRL, Grand Island, NY, USA) at 37°C in 5% CO2. After culturing the cells for several months, all cell lines were evaluated for genetic identification prior to the experiments. Based on the appraisal reports provided by CoBioer Biosciences Co., Ltd. (Nanjing, China), no mutations or contaminations were found in our cell lines.
4. Quantitative reverse transcription polymerase chain reaction (qRT-PCR)
All primers used for qRT-PCR were designed using primer 5.0 software (Premier, Canada) and synthesized by Tsing Ke Biotech (Chengdu, China). The cDNAs for circRNA, mRNA, or miRNA measurement were synthesized using random, oligo(dT)18, or stem-loop primers, respectively, using a RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA). qRT-PCR was performed in triplicate using Maxima SYBR Green qPCR Master Mix (Thermo Fisher Scientific) on the CFX connect real-time system (Bio-Rad, Hercules, CA, USA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the intrinsic control for circRNA and mRNA, while U6 was used as the endogenous control for miRNA. The relative expression level of each gene was calculated using the 2−ΔΔCt method.
5. circEPS15-overexpressing vector construction and transduction
The circEPS15-overexpressing plasmid was synthesized and cloned into the adenoviral vector GV138 (Genechem, Shanghai, China) with a Flag-tag. Full-length circEPS15 was serially diluted with X-tremeGENE siRNA transfection reagent (Roche, Mannheim, Germany). For western blot analysis, cell proteins were prepared 72 h post-transfection.
6. Wound healing Assay
The cells were seeded into 6-well plates and transfected following the manufacturer’s instructions. When the cell confluence reached approximately 80% at 48 h post-transfection, the cells were serum-starved for 12 h and scratched using the fine end of a 100-µL pipette tip. Wound healing was observed at different time points within the scratch line. Every 24 h, scratch lines were imaged using an inverted Leica phase-contrast microscope (DFC 300 FX, Wetzlar, Germany) under a 20× objective lens. Image processing and analysis in Java was performed to obtain the wound-healing assay results. Duplicate wells were examined for each condition, and each experiment was repeated three times.
7. Transwell assay
A cell invasion assay was performed in a transwell chamber (24-well type, 8 mm pore size, Corning, Inc., Corning, NY, USA). BD Matrigel Basement Membrane Matrix (BD Biosciences, Franklin Lakes, NJ, USA) was used according to the manufacturer’s instructions. Next, 0.5 mL of serum-free medium was added in the upper chambers; DMEM containing 10% FBS was added to the bottom chambers. Equal numbers (1 × 105) of cells were plated in the upper chambers of the quadruplicate wells and incubated at 37°C for 72 h. The cells were then fixed with paraformaldehyde and stained with crystal violet to visualize the nuclei. Average values were calculated from the results of three independent experiments. Image processing and analysis in Java was performed to obtain the transwell assay results.
8. Cell cycle and apoptosis assay
To assess the cell cycle and apoptosis, 3 × 105 treated cells were seeded into 6-well plates and cultured for 48 h at 37°C. For cell cycle analysis, the cells were digested using trypsin (Hyclone, Logan, UT, USA), washed twice with phosphate-buffered saline (PBS), and fixed in 70% ethanol overnight at 4°C. The cells were centrifuged at 500 ×g for 5 min, washed twice with cold PBS, and centrifuged. After treating the cells with RNase A (0.1 mg/mL) and propidium iodide (PI, 0.05 mg/mL, 4A Biotech, Beijing, China) ) for 30 min at 37°C, cell cycle analysis was performed using fluorescence-activated cell sorting flow cytometry (Beckman Coulter, Palo Alto, CA, USA). To analyze apoptosis, the cells were trypsinized followed by two PBS washes. The cells were stained using the Annexin V/PI detection kit (4A Biotech) for 5 min at 25°C. Apoptotic cells were measured using flow cytometry (Beckman Coulter). All experiments were repeated at least three times.
9. Western blot analysis
To detect protein, the cytoplasmic and nuclear protein extracts were prepared from the cells. The protein concentration of each sample was determined using a NanodropTM spectrophotometer (Thermo Fisher Scientific). Protein (100 µg) from each sample were subjected to SDS-PAGE (4% stacking and 10% separating gels) and then transferred overnight onto polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). The membranes were incubated overnight with the following antibodies: poly clonal anti-human TJP1 antibody (1:200, Abgent, San Diego, CA, USA); anti-CDH2 antibody (1:1,000, Abcam, Cambridge, UK); anti-aVIM antibody (1:1,000, Cell Signaling Technology, Danvers, MA, USA); anti-EPS15 antibody (1:500, Abgent); anti-GAPDH antibody (1:1,000, Santa Cruz Biotechnology, Dallas, TX, USA); and anti-Flag antibody (1:1,000, Cell Signaling Technology). Next, the blots were incubated with peroxidase-conjugated goat anti-rabbit antibody or goat anti-mouse antibody (1:4,000, Millipore) for 1 h. Finally, the membranes were subjected to immunoblotting analysis using an ECL immunoblotting kit according to the manufacturer’s recommendations (Beyotime Institute of Biotechnology, Shanghai, China).
10. IRES prediction
IRESfinder [29] was used to identify the IRES sequence of circRNAs. Each sequence of circRNA was evaluated by the sliding window approach, with window size of 174 nt and step size of 50 nt. The region with the highest score was considered as the IRES sequence of circRNA.
11. Statistical analyses
All statistical analyses were performed using SPSS version 21.0 (SPSS, Inc., Chicago, IL, USA) and Prism version 5.0 (GraphPad Software, La Jolla, CA, USA) software. Categorical variables were expressed as a count or percentage and tested using χ2 or Fisher’s exact test, as deemed appropriate. Continuous data are shown as the mean ± standard deviation (SD) and were compared using Student’s t-test, one-way analysis of variance, or Mann-Whitney test as deemed appropriate. Correlations were calculated using Pearson’s correlation analysis. The median expression of target genes was used as the cut-off value to stratify patients into high and low expression groups. Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. All tests were 2-sided, and results with p < 0.05 were considered statistically significant.