Clinical specimens

A total of 41 pairs of OS specimens and adjacent normal mesenchymal tissue samples were obtained from the metaphyseal regions of long bones of OS patients at Changzhou No.2 People’s Hospital, the Affiliated Hospital of Nanjing Medical University. The adjacent normal tissue samples were ≥ 5 cm away from the edge of OS tissues. The inclusion criteria for OS patients: (1) The clinicopathological diagnosis was confirmed by two pathologists. (2) Patients who had not received radiotherapy, chemotherapy or other treatment before the surgery. The exclusion criteria for OS patients: (1) OS patients who had received radiotherapy, chemotherapy or other treatment before the surgery. (2) Patients who were unsuitable for surgery. (3) Patients who had serious infection. Tissues resected from patients were immediately frozen in liquid nitrogen. Clinical experiment was conducted by the permission of the Ethics Committee of Changzhou No.2 People’s Hospital, the Affiliated Hospital of Nanjing Medical University, and written informed consent was signed by all the patients. The correlation between circ_0136666 expression and clinicopathological characteristics of osteosarcoma patients is shown in Table 1.

Table 1 The correlation between circ_0136666 expression and clinicopathological characteristics of osteosarcoma patientsCell lines

hFOB 1.19, U2OS and SaOS2 obtained from Shanghai Academy of Sciences (Shanghai, China) were cultured with Dulbecco’s modified Eagle’s medium (DMEM) medium (Gibco, Carlsbad, CA, USA) plus 10% fetal bovine serum (FBS, Gibco) and 1% antibiotic mixture (Sangon Biotech, Shanghai, China). hFOB 1.19 cells were maintained at 34℃ with 5% CO2, and two OS cell lines were cultivated at 37 ℃ with 5% CO2.

Real-time quantitative polymerase chain reaction (RT-qPCR)

RNA extraction from tissues and cells was performed using Trizol reagent (Invitrogen, Carlsbad, CA, USA). MiR-1244 was reversely transcribed into DNA using stem-loop primer, and miR-1244 level was assessed using the stem-loop primer SYBR Green RT-qPCR Kit (Synbio, Suzhou, China). The reverse transcription of circ_0136666 and CEP55 was carried out using TaqMan Reverse Transcription Reagents (Invitrogen), and the qPCR reaction was conducted using SYBR Green detection reagent (Cowin Biotech, Beijing, China). Primers are listed in Table 2. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or U6 was utilized as control for circRNA/mRNA or miRNA. Fold change was analyzed by the 2−∆∆Ct method.

Table 2 Specific primers for RT-qPCRRNase R treatment

Total RNA samples were digested with 100 μg/mL RNase R (Applied Biological Materials, Vancouver, Canada) for 20 min at 37℃. The levels of circ_0136666 and its matched linear form protein kinase, DNA-activated, catalytic subunit (PRKDC) were examined by RT-qPCR.

Cell transfection

The specific small interfering RNA (siRNA) of circ_0136666 (si-circ_0136666), negative control of siRNA (si-NC), the specific short hairpin RNA (shRNA) of circ_0136666 (sh-circ_0136666), sh-NC, mimics of miR-1244 (miR-1244), miRNA NC (miR-NC), inhibitor of miR-1244 (anti-miR-1244), anti-NC, CEP55 re-constructed overexpression plasmid (CEP55) and pcDNA vector (Mock group) were acquired from Genepharma (Shanghai, China) and Sangon Biotech. OS cells were seeded into 6-well plates at the density of 3 × 105 cells/well. Next day, Lipofectamine 3000 (Invitrogen) was utilized to introduce RNA or plasmid into OS cells when cell confluence reached about 70%. After transfection for 6 h, the culture supernatant was replaced by fresh complete medium. After transfection for 24 h, transfection efficiencies were assessed by RT-qPCR and Western blot assay.

5-Ethynyl-2′-deoxyuridine (EDU) assay

DNA synthesis was monitored via EDU incorporation using commercial KeyFluor488 Edu Kit (keyGEN Biotech, Jiangsu, China). 4,6-diamino-2-phenyl indole (DAPI) was used to mark cell nucleus. The fluorescence images were captured using the fluorescence microscope (Olympus, Tokyo, Japan). The relative rate of EDU incorporation was analyzed.

Colony formation assay

OS cells were seeded onto 12-well plates at the density of 200 cells per well. Culture media was replaced every 5 d. After incubation for 2 weeks, cell colonies included more than 50 cells were fixed using 4% paraformaldehyde (Sangon Biotech) and stained by 0.1% crystal violet (Sangon Biotech). The number of colonies was manually counted.

Transwell assays

In transwell invasion assay, the upper chambers were added with diluted 40 µL Matrigel (1:8; BD Biosciences, San Jose, CA, USA) at 37℃ for 30 min for solidification to pre-coat the upper chambers. In transwell migration assay, un-coated upper chambers were directly utilized for further analysis. A total of 200 μL cell suspension (without serum; in transwell migration assay: 1 × 104 cells; in transwell invasion assay: 8 × 104 cells) was added to the upper chambers, and 10% FBS-added culture medium was pipetted into the lower chambers. FBS acted as chemokine in this experiment. Un-migrated or un-invaded OS cells were wiped out using cotton swab. Migrated or invaded OS cells were dyed using 0.1% crystal violet (Sangon Biotech). Cell number was manually counted using an optical microscope (Olympus, Osaka, Japan) at the magnification of 100×.

Flow cytometry

After transfection for 72 h, a total of 5 × 104 OS cells were harvested, washed and suspended in binding buffer (BD Biosciences). OS cells were stained by 5 μL Annexin V-fluorescein isothiocyanate (Annexin V-FITC; BD Biosciences) and 5 μL propidium iodide (PI; BD Biosciences) for 15 min. Unstained cell samples and cell samples stained with FITC or PI alone were utilized to determine the threshold. The proportion of apoptotic OS cells (FITC positive, PI positive or negative) was considered as apoptosis rate. Cell samples (1.5 × 104 cells) were loaded onto the FACS CantoII flow cytometer (BD Biosciences), and the apoptosis rate was analyzed by BD FACSDiva software (BD Biosciences).

Determination of cellular glycolysis

The consumption of glucose and the production of lactate were evaluated using Glucose Uptake Colorimetric Assay kit (Biovision, Milpitas, CA, USA) and Lactate Assay Kit II (Biovision).

Western blot assay

OS cells were collected and then washed using phosphate-buffered saline buffer (PBS; Sangon Biotech) for three times. Cell lysates were prepared using whole cell lysis buffer (Invitrogen). Protein samples were subjected to 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and dry-transferred onto polyvinylidene difluoride (PVDF) membrane (150 V/2 h; Bio-Rad, Hercules, CA, USA). The non-specific sites in the membrane were sealed using 5% non-fat milk. The diluted primary antibodies of anti-CEP55 (ab170414; Abcam, Cambridge, MA, USA) at the dilution of 1:10,000, anti-hexokinase 2 (HK2; ab227198; Abcam) at the dilution of 1:20,000, anti-pyruvate kinase M 2 (PKM2; ab137852; Abcam) at the dilution of 1:3000 and anti-β-actin (ab8226; Abcam) at the dilution of 1:20,000 were incubated with the membrane overnight. Afterwards, the membrane was labeled with diluted horseradish peroxidase (HRP)-labeled secondary antibody (Abcam) at the dilution of 1:5000. Immunoreactive protein bands were assessed by the enhanced chemiluminescence (ECL) kit (Pierce, Waltham, MA, USA). The quantification of protein bands was performed using Image Lab analysis software (Bio-Rad).

Subcellular localization

The cytoplasmic and nuclear RNA fractions were isolated using Cytoplasmic and Nuclear RNA Purification Kit (Norgen Biotek, Thorold, Canada).

Bioinformatics prediction

Circinteractome (https://circinteractome.irp.nia.nih.gov) and TargetScan (http://www.targetscan.org) bioinformatics databases were utilized to predict the interactions between circ_0136666 and miRNAs and between miR-1244 and mRNAs.

Dual-luciferase reporter assay

To test the interaction between miR-1244 and CEP55 or circ_0136666, dual-luciferase reporter assay was applied to analyze the effect of miR-1244 on the activity of circ_0136666 or CEP55 responsive element. Partial fragment of circ_0136666 or CEP55, including the wild-type or mutant type binding sites with miR-1244, was inserted into pmirGLO vector (Promega, Madison, WI, USA) to generate circ_0136666 WT, circ_0136666 MUT, CEP55 3′ untranslated region (3′UTR) WT and CEP55 3’UTR MUT. OS cells were co-transfected with miR-1244 or miR-NC and luciferase plasmids for 24 h, and the luciferase intensities in different groups were determined by the Dual-Luciferase reporter assay system kit (Promega).

RNA-pull down assay

MiR-1244 or miR-NC was biotinylated to obtain Bio-miR-1244 or Bio-miR-NC probe, which was subsequently mixed with C-1 magnetic beads (Life Technologies, Carlsbad, CA, USA) to obtain probe-labeled beads. Cell lysates were mixed with probe-labeled beads at 4 ℃ overnight. After the elution, the level of circ_0136666 was determined by RT-qPCR.

Xenograft tumor model

A total of 10 male BALB/c nude mice (5-week-old) were acquired from Vital River Laboratory Animal Technology (Beijing, China) and then divided into two groups (n = 5 in each group). U2OS cells (5 × 106 cells/200 μL PBS) stably transfected with sh-NC or sh-circ_0136666 were subcutaneously inoculated into the right flank of mice. Tumor width and length were measured every week using digital calipers, and tumor volume was calculated as length × width2 × 0.5. After injection for five weeks, the mice were killed, and tumors were weighed. Immunohistochemistry (IHC) assay was conducted to analyze the protein level of proliferation marker Ki-67 in tumor tissues using the antibody against Ki-67 (ab15580; Abcam) at the dilution of 1:500. Tumor tissues were utilized to measure the levels of circ_0136666, miR-1244 and CEP55 protein. The protocols in animal experiments were approved by the Animal care Committee of Changzhou No.2 People’s Hospital, the Affiliated Hospital of Nanjing Medical University.

Statistical analysis

Data were processed using GraphPad Prism 7.0 software (GraphPad, La Jolla, CA, USA). D’Agostino-Pearson omnibus normality test was used to determine the normality of data distribution, and the homogeneity of variances was tested by Levene test. Differences were analyzed by paired or unpaired Student’s t test (in two groups) or one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test (in multiple groups). Data were represented as mean ± standard deviation (SD). Linear correlation was analyzed by Pearson’s correlation analysis. The comparisons were considered as statistically significant with P < 0.05.