Cell lines
The human normal osteoblast cell line hFOB1.19 and human OSA cell lines 143B, U2OS, MNNG, MG63 and Saos-2 were obtained from the American Type Culture Collection (ATCC, Manassas, VA). The cells were cultured in Dulbecco’s Modified Eagle’s medium (DMEM) (Gibco, USA) supplemented with 10% fetal bovine serum (BI, Israel), 1% penicillin G and streptomycin (Gibco, USA). All OSA cell lines were cultured in an incubator with 5% CO2 at 37 °C, whereas hFOB1.19 cells were cultured at 34 °C and 5% CO2. The MTX-resistant OSA cell line U2OS was provided by Dr. M. Serra (Istituti Ortopedici Rizzoli, Bologna, Italy). U2OS cells were continuously cultured in a medium containing 300-ng/mL MTX.
RNA sequencing
RNA sequencing was performed and libraries were constructed by Gene Denovo Biotechnology (Guangzhou, China). The samples were processed and sequenced on an Illumina HiSeqTM 2500 platform. The offline data obtained were filtered and compared with the reference genome to obtain the gene expression data, which were compared in the RPM format. Ensembl_release98 served as the reference genome [19]. RNA sequencing of circRNAs was performed after the removal of conventional ribosomal RNA and degradation of linear RNA with the RNase R enzyme. The obtained circRNAs were fragmented in a fragmentation buffer, and a library was subsequently constructed.
Real-time reverse transcription polymerase chain reaction (qRT-PCR)
Total RNA was extracted from whole cell lysates of paracancerous tissues, cancerous tissues and cells from patients with OSA using the RNAiso Plus kit (TaKaRa, Japan) according to the manufacturer’s instructions. The quality of the extracted RNA was determined on the NanoDrop 2000 spectrophotometer. Two transcription reagents were used to convert mRNA, circRNA and miRNA to cDNA. Except for miRNA, for which we used Evo M-MLV RT Kit (Accurate Biotechnology, China), we use PrimeScript RT Reagent Kit (TaKaRa, Japan). We used two reagents to perform real-time amplification for circRNA and miRNA, namely SYBR Green Premix Pro Taq HS qPCR Kit (Accurate Biotechnology, China) and SYBR Premix Ex Taq II (TaKaRa, Japan). All qRT-PCR reactions were performed utilizing the Bio-Rad CFX connect system (Bio-Rad, CA, USA). The specific program was designed as previously described [20]. The relative gene expression was determined using the 2 –ΔΔCT method. The specific primers are listed in the Additional file 1: Table S1.
Fluorescence in situ hybridization (FISH)
FISH assay was conducted for detection of the location of circEMB and miR-3184-5p in OSA cells. Cy3-labeled circEMB and FAM-labeled miR-3184-5p probes were used to detect the localization of circEMB and miR-3184-5p in OSA cells, respectively. According to the manufacturer’s instructions, the cells were inoculated in confocal dishes and fixed using 4% paraformaldehyde. After that, the cell was incubated with oligo probe overnight at 37 °C. Finally, samples were acquired by confocal microscopy (AX confocal, Nikon, Tokyo, Japan) to obtain images.
RNA interference (RNAi)
For circEMB knockdown, two shRNAs targeting the back-splice junction of circEMB were generated by IGE Co., Ltd. (Guangzhou, China). Then, shRNA against circEMB and shRNA-NC as negative control were packaged into lentiviruses by Genechem Co., Ltd. (Shanghai, China). The miR-3184-5p mimic/inhibitor, and their corresponding negative controls (NC) were purchased from IGE Co., Ltd. (Guangzhou, China). The lentivirus-miR-3184-5p inhibitor and lentivirus-sh-EGFR were also procured from Genechem Co., Ltd. (Shanghai, China). All vectors were validated via sequencing. Lentivirus infection was performed using the HiTransG reagent (Genechem, China), as per the manufacturer’s specifications.
Cell proliferation assay
The proliferation of OSA cells was detected using the cell counting kit-8 (CCK-8, GLPBIO, USA). Cells in different groups were quantified using a cell counting method. Subsequently, 2000 OSA cells (100 uL) in the logarithmic phase were seeded in a 96-well plate, and phosphate-buffer saline (PBS) was introduced into the marginal well. The growth and proliferation of cells were detected after 6, 24, 48 and 72 h by CCK-8 assay. A total of 10 uL of CCK-8 detection solution was introduced into every well and subjected to incubation at 37 °C for 2 h.
Colony formation assay
A total of 800 OSA cells were seeded in a 6-well plate, and the medium was changed every 2 days. After 14 days of culture at 37 °C, cell colonies were stained with 4% polymethanol and 0.1% crystal violet and counted.
Migration and invasion assays
transwell chamber was used to assess the migration and invasion abilities of OSA cells. Transfected OSA cells were re-suspended in 200 uL of DMEM and added to the upper compartment, whereas a medium containing 10% FBS was added to the lower compartment. The transwell chamber was incubated at 37 °C for 24 h. Thereafter, cells in the upper compartment were wiped using cotton swabs, whereas those in the lower compartment were fixed with paraformaldehyde, stained with crystal violet and counted. For invasion assay, the upper compartment was precoated with 50 uL of Matrigel (BD Bioscience, USA) to detect intrusion.
Wound healing assays
OSA cells were seeded in 6-well plates, and a scratch was created in the cell monolayer with a 200 μL pipette tip after 90% confluency was achieved. After the scratch was created, the cells were cultured in a serum-free medium and photographed at 0 and 24 h.
Flow cytometry
Flow cytometry was used to assess apoptosis and cell cycle progression. Cells in the logarithmic growth phase were digested in trypsin without ethylenediaminetetraacetic acid (EDTA) and washed thrice with PBS. Apoptosis was determined using the Annexin V-APC Apoptosis Assay kit (BestBio, China) according to the manufacturer’s instructions. For cell cycle analysis, cells were lysed in trypsin containing EDTA, washed thrice with PBS and fixed with ethanol overnight. Cell cycle progression was determined according to the instructions on the Cell Cycle Analysis kit (BestBio, China). Each experiment was repeated at least thrice.
Dual-luciferase reporter assay
The 3′-untranslated region (UTR) sequences of circEMB and EGFR and the corresponding mutants were cloned into pmiR-RB-REPORT™ vectors, which were named circEMB-WT, circEMB-MUT, EGFR-WT and EGFR-WUT. The luciferase plasmid was co-transfected with miR-3184-5p mimics or miR-NC. The relative luciferase activity was determined according to the instructions on the Dual-Luciferase Reporter Assay System (Promega, USA).
RNA immunoprecipitation
RNA immunoprecipitation (RIP) was performed using the Magna RIP Kit (Millipore, Billerica, MA, USA) according to the manufacturer’s instructions. Briefly, OSA cells were lysed in RNA immunoprecipitation (RIP) buffer supplemented with protease inhibitors and RNase. Magnetic beads were pre-incubated with anti-IgG and anti-Ago2 antibodies and were subsequently incubated with the aforementioned cell lysates overnight at 4 °C. Finally, qRT-PCR and western blotting were performed to determine the presence of binding targets.
RNA pull-down assay using a biotin-labelled probe
C1 magnetic beads were incubated with a biotinylated circEMB probe or oligonucleotide probe (IGE Biotech Co., Ltd., Guangzhou, China) at 25 °C for 2 h. The solution was co-incubated with a cell lysis solution at 4 °C overnight. The precipitates were extracted and purified using an RNeasy Mini Kit (Qiagen, USA) and assessed via qRT-PCR according to the manufacturer’s instructions.
Subcellular fractionation
The PARIS kit (Invitrogen, USA) was used to isolate nuclear and cytosolic fractions from cells according to the manufacturer’s instructions.
Bioinformatic analysis
Multiple databases, including ENCORI (http://starbase.sysu.edu.cn/), TargetScan (http://www.targetscan.org/vert_72/), miRanda (http://www.microrna.org/), RNAhybrid (http://bibiserv.techfak.uni-bielefeld.de/rnahybrid/) and RNA22 (https://cm.jefferson.edu/rna22/), were used to predict the interaction between circEMB and miRNA [21,22,23,24]. The interaction between miR-3184-5p and EGFR was predicted using TargetScan, miRanda, RNAhybrid and miRDB (http://mirdb.org/) [25]. Pathway enrichment analysis was performed using Reactome in the KOBAS software (KOBAS, Surrey, UK).
Western blot (WB) analysis
Total protein was isolated from tissues and cells lysed in radioimmunoprecipitation assay (RIPA) buffer (Fudebio, China) and quantified via bicinchoninic acid (BCA) assay. Subsequently, SDS-PAGE (Fudebio, China) was performed to separate proteins of varying molecular weights. The separated proteins on the gel were transferred to a polyvinylidene difluoride (PVDF) membrane (Invitrogen). The membrane was blocked with 5% skim milk at room temperature for 1.5 h and incubated with primary antibodies overnight at 4 °C. The following day, the membrane was incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:10000) (Proteintech, China) at room temperature for 1 h, and protein bands were visualised via enhanced chemiluminescence (Millipore, USA).
Haematoxylin and eosin (HE) and immunohistochemical (IHC) staining
The extracted fresh animal tissues were washed with normal saline to remove excess blood and immediately fixed in 4% paraformaldehyde overnight. The tissues were embedded in paraffin using an automatic dehydrator and a paraffin-embedding machine and stained with haematoxylin and eosin (H&E) to visualise the histopathological features. Immunohistochemical (IHC) analysis was performed as described previously [20].
Construction of a xenograft-bearing nude mouse model
All animal experiments were approved by the Ethics Committee of Zhujiang Hospital, Southern Medical University. Nude mice were purchased from the Experimental Animal Center of Southern Medical University. The effects of circEMB on the proliferation of OSA cells were examined using tumour xenograft-bearing mouse models. Stable lentivirus-infected cell lines were isolated, resuspended in 100 μL of PBS (143B, 1 × 106) and injected into the right hindlimb of female nude mice (age, 4 weeks). Malignant growth was examined using an in vivo imaging system. After 4 weeks, mice were euthanised via carbon dioxide inhalation, and tumour tissues were harvested.
To determine the effects of circEMB and its signalling axis on the metastatic capability of OSA cells, 2 × 106 cells (200 uL; 143B) was injected into the tail vein of male nude mice (age, 4 weeks). After 4 weeks, tumour growth in the lung was detected using an in vivo imaging system, and mice were sacrificed via cervical dislocation. Thereafter, lung tissues were harvested for HE staining.
Molecular docking
AutodockTools (version 1.5.6) was used for molecular docking. The 2D structure of MTX (the ligand) was downloaded from PubMed and converted into a 3D structure using Chem3D. The crystal structure of EGFR was downloaded from Protein Data Bank (PDB) (PDB code: 5UG9) and was used as the protein target for docking simulation. Docking results with the highest score were visualised and analysed using PyMOL.
Statistical analysis
The SPSS Statistics (version 22.0) software (Chicago, Illinois, USA) was used for statistical analysis. The Mann–Whitney U test or two-tailed Student’s t-test was used to determine the significance of between-group differences, whereas ANOVA followed by Tukey’s test was used to determine the significance of differences among multiple groups. Survival curves were approximated using the Kaplan–Meier approach, and survival data were compared using the log-rank test. Pearson correlation analysis was performed to determine the link between gene expression and clinicopathological features. A P-value of < 0.05 was considered significant. Data are expressed as the mean ± standard deviation (SD).