MG-63 and U2OS human osteosarcoma cell lines were bought from the American Type Culture Collection (ATCC; Manassas, VA, USA). Cells were cultivated with ATCC-formulated Eagle's Minimum Essential Medium (cat. no. 30–2003; ATCC) supplemented with 10% fetal bovine serum (FBS; Sangon Biotech Co, Ltd., Shangai, China) in an incubator (37°C; 5% CO₂). When required, cells were treated with HIF1α (cat. no. 776–826; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) at 10, 20 and 40 ng/ml for 24 h or with the HIF inhibitor LW6 (cat. no. S8441; Selleck Chemicals, Shanghai, China) at 10 ng/ml for 24 h prior to further experiments.

Plasma samples were obtained from 62 patients with osteosarcoma and 48 healthy volunteers who were admitted to Zhongnan Hospital Affiliated to Medical College of Wuhan University (Wuhan, China) between May 2015 and January 2018. Inclusion criteria were: i) Patients who were diagnosed with osteosarcoma through pathological examination; ii) patients who were diagnosed and treated for the first time; iii) patients who were willing to join the study. Exclusion criteria were: i) Patients with multiple diseases; ii) patients who received treatment within 90 days before admission. Among the 62 patients with osteosarcoma, distant tumor metastasis was observed in 28 cases, and these patients were classified into the metastatic osteosarcoma (MO) group based on imaging findings. The remaining 34 patients were classified into the non-metastatic osteosarcoma (NMO) group. The patient group included 34 males and 28 females, aged between 12 and 44 years (mean, 27.4±4.5 years). The healthy control group included 28 males and 20 females, aged between 15 and 45 years (mean, 28.8±4.3 years). This study was approved by the Ethics Committee of Zhongnan Hospital Affiliated to Medical College of Wuhan University; all participants signed informed consent. Clinicopathological characteristics of patients included in the study are available in Table I

Total RNA was extracted from plasma (1 ml) or MG-63 and U2OS cells (3×10⁴ cells/ml) using RNAzol^® RT RNA Isolation Reagent (GeneCopoeia, Guangzhou, China). RNA concentrations were measured using a NanoDrop™ 2000 spectrophotometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA). RNA extraction was repeated until all RNA samples had an A260/A280 ratio between 1.8 and 2.0. cDNA was synthesized using RevertAid RT Reverse Transcription kit (Thermo Fisher Scientific, Inc.) under the following conditions: 25°C for 5 min, 55°C for 30 min and 75°C for 10 min. One-Step PrimeScript RT-PCR kit for Real Time RT-PCR (Clontech Laboratories, Inc., Mountainview, CA, USA) was used to prepare all PCR reaction conditions. Primer sequences were as follows: LINK-A, forward 5′-TTCCCCCATTTTTCCTTTTC-3′, reverse 5′-CTCTGGTTGGGTGACTGGTT-3′; GAPDH, forward 5′-GAAGGTGAAGGTCGGAGT-3′, reverse 5′-GAAGATGGTGATGGGATTTC-3′. qPCR reaction conditions were: Initial denaturation at 95°C for 50 sec, followed by 40 cycles of 95°C for 15 sec and 60.5°C for 34 sec. Relative expression levels were quantified using the 2^−ΔΔCq method (13) and normalized to GADPH loading control.

LINK-A lncRNA overexpression vectors (pcDNA3.1) and empty vectors were designed and synthesized by Shanghai GenePharma Co., Ltd., (Shanghai, China). MG-63 and U2OS cells were cultivated overnight to reach 80–90% confluence. Lipofectamine^® 2000 reagent (cat. no. 11668-019; Invitrogen; Thermo Fisher Scientific, Inc.) was used to perform transfection with vectors at a dose of 15 mM. Vectors were incubated with 5×10⁵ cells at 37°C for 6 h. Subsequent experiments were carried out 24 h post-transfection. Non-transfected cells were used as control cells; empty vector transfection was used as the negative transfection control (NC).

Transfected MG-63 and U2OS cells were harvested and serum-free single cell suspensions with a cell density of 5×10⁴ cells/ml were prepared. Cell migration and invasion were examined by the following steps: The upper Transwell chamber (3 µm pore size) was filled with 100 µl serum-free cell suspension and the lower chamber was filled with ATCC-formulated Eagle's Minimum Essential Medium containing 20% FBS. Cells were cultivated in an incubator (37°C; 5% CO₂) for 24 h and 0.5% crystal violet (Sigma-Aldrich; Merck KGaA) staining was performed for 20 min at room temperature. Free cells were removed using a cotton swab. For the invasion assay, the upper chamber was pre-coated with Matrigel (cat. no. 356234; EMD Millipore, Billerica, MA, USA). Membranes were collected and invading or migrating cells were observed under a light microscope (magnification, ×40; Olympus Corporation, Tokyo, Japan).

Proteins were extracted from cells (3×10⁴ cells/ml) using ReadyPrep™ Protein Extraction kit (Total Protein; Bio-Rad Laboratories, Inc., Hercules, CA, USA). Protein concentration was measured with the bicinchoninic acid assay. Following denaturing, proteins (35 µg) were separated by 10% SDS-PAGE. Proteins were transferred onto polyvinylidene difluoride membrane by semi dry method. Membranes were blocked in 5% milk for 2 h at room temperature. Membranes were then incubated with the rabbit anti-human primary antibodies against HIF1α (1:1,500; cat. no. ab2185, Abcam, Shanghai, China) and GAPDH (1:2,000; cat. no. ab9485; Abcam) at 4°C overnignt, followed by secondary antibody horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G (1:1,000; cat. no. MBS435036; MyBioSource, San Diego, CA, USA) at room temperature for 2 h. Signals were developed using ECL™ (Sigma-Aldrich; Merck KGaA) and scanned by MYECL™ Imager (Thermo Fisher Scientific, Inc.). Densitometric analysis was performed using ImageJ v1.46 software (National Institutes of Health, Bethesda, MD, USA). GAPDH was used as endogenous control for data normalization.

All experiments were performed in triplicates and data were recorded as the mean ± standard deviation. SPSS 19.0 (IBM Corp., Armonk, NY, USA) was used to perform all statistical analyses. Comparisons among multiple groups were performed by one-way analysis of variance followed by Tukey test. Receiver operating characteristic (ROC) curve analysis was performed using patients with osteosarcoma as true positive cases and healthy controls as true negative cases. P<0.05 was considered to indicate a statistically significant difference.

RT-qPCR results revealed that, compared with the control and NMO groups, plasma LINK-A lncRNA expression was significantly upregulated in the MO group (Fig. 1). In addition, plasma levels of LINK-A lncRNA were also slightly higher in the NMO group than in control group (Fig. 1).

Based on the diameter of the primary tumor, patients with MO and NMO were divided into 0–2 cm group (n=22), 2–4 cm group (n=20) and >4 cm group (n=20) (Fig. 2). RT-qPCR results indicated no significant differences in plasma levels of LINK-A lncRNA among these groups. No significant differences in plasma levels of LINK-A lncRNA were found in MO patients with different tumor size (data not shown).

To investigate the diagnostic value of plasma LINK-A lncRNA for osteosarcoma, ROC curve analysis was performed patients with NMO or MO as true positive cases and healthy controls as true negative cases. For metastatic osteosarcoma, the area under the curve (AUC) was 0.9141, with standard error of 0.03214 and 95% confidence interval of 0.8511–0.9771 (P<0.0001; Fig. 3A). For non-metastatic osteosarcoma, the AUC was 0.6351, with standard error of 0.06082 and 95% confidence interval of 0.5159–0.7543 (P=0.038; Fig. 3B).

Compared with the Control and NC groups, LINK-A and HIF1α expression levels were significantly upregulated following cell transfection (Fig. 4A). LINK-A lncRNA overexpression led to significantly increased expression of HIF1α protein in MG-63 and U2OS cells (Fig. 4C). By contrast, treatment with HIF1α at doses of 10, 20 and 40 ng/ml failed to significantly affect endogenous LINK-A lncRNA expression in untransfected cells (Fig. 4D) as well as in cells overexpressing LINK-A lncRNA (Fig. 4D).

Compared with the Control group and the NC group, LINK-A lncRNA overexpression led to significantly promoted migration (Fig. 5) and invasion (Fig. 6) of MG-63 and U2OS cell lines. In addition, treatment with HIF inhibitor LW6 at a dose of 10 ng/ml significantly attenuated the enhancing effects of LINK-A overexpression on cancer cell migration (Fig. 5) and invasion (Fig. 6).