The protocols of the present study were approved by the Ethics Committee of Jiangxi Cancer Hospital (Nanchang, China). All volunteers had provided written informed consent prior to sample collection. Serum samples from healthy volunteers (sex, 24 males and 16 females; age, 54.45±7.27; age range, 42-73 years) and patients with NSCLC (sex, 34 males and 13 females; age, 60.7±10.9; age range, 42-83 years) was obtained at Jiangxi Cancer Hospital (Nanchang, China) from Jan 2017 to May 2019. The clinicopathological data of the patients with NSCLC are provided in Table I. Bone metastasis of NSCLC were diagnosed using computed tomography. The inclusion criteria for patients with NSCLC were as follows: i) Aged <85 years; and ii) imaging and pathological diagnosis of NSCLC stages I-IV according to TNM staging system (19). No patients received chemo- or radiotherapy when they were recruited. Exclusion criteria: i) Patients with type I or II diabetes or cardiovascular disease; ii) body mass index >30; and iii) patients with respiratory failure (PaO₂ <60 mmHg (20), breathing room air at rest). The inclusion criteria for healthy volunteers were: i) Aged <85 years; and ii) non-tumor population. Exclusion criteria were: i) Patients with type I or II diabetes or cardiovascular disease; ii) body mass index >30; and iii) with respiratory failure (PaO₂ <60 mmHg, breathing room air at rest). All samples were collected, followed by ELISA and reverse transcription-quantitative PCR (RT-qPCR) analysis.

The human NSCLC cell lines A549 and H3122 were obtained from the American Type Culture Collection, and grown in DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.), 1% streptomycin and penicillin (Gibco; Thermo Fisher Scientific, Inc.) in a cell incubator at 37˚C with 5% CO₂.

OPG was overexpressed using pcDNA3.1 plasmid vector (Thermo Fisher Scientific, Inc.). In brief, total RNA were extracted from A549 cells using TRIzol^® kit (Thermo Fisher Scientific, Inc.) and cDNA was synthesized by reverse transcription PCR using iScripe™ cDNA Synthesis kit (Bio-Rad Laboratories, Inc.) according to the manufacturer's protocols. The full-length coding sequence of OPG (GenBank no. U94332.1) was amplified from the cDNA using PCR and then inserted into the vector at the restriction sites BamHI and XhaI to form the recombinant plasmid pcDNA3-OPG. The specific primer sequences used for amplifying OPG cDNA were: Forward, 5'-cgggatcccgATGAACAAGTTGCTGTGCTGC-3', reverse, 5'-gctctagagcTTATAAGCAGCTTATTTTTACTGATTGG-3'. pcDNA3-OPG plasmids (5 µg) were transfected into A594 cells using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) for 48 h to overexpress OPG. pcDNA3 was used as the negative control.

OPG protein levels in serum samples from patients with NSCLC with or without metastasis and healthy individuals were assessed using an ELISA kit (cat. no. EHTNFRSF11B; Invitrogen; Thermo Fisher Scientific, Inc.) as per the manufacturer's protocol. The absorbance of samples was measured at 450 nm with a microplate reader (Thermo Fisher Scientific, Inc.).

A total of 13 NSCLC-associated miRNAs were assessed using RT-qPCR. TRIzol^® kit (Thermo Fisher Scientific, Inc.) was applied to isolate total RNA from A549 cells. Moloney's murine leukemiavirus reverse transcriptase (cat. no. 2641A; Takara Biotechnology Co., Ltd.) was used to perform reverse-transcription reaction using Oligo(dT) primers (Vazyme Biotech Co., Ltd.) and dNTPs (cat. no. 4035; Takara Bio, Inc.). The temperature protocol for reverse transcription was 70^˚C for 5 min, followed by ice bath for 2 min and then 42˚C for 60 min. The specific reverse transcription primers for miR-20a were synthesized as previously described (21). qPCR was performed using TB Green^® Premix Ex Taq™ (cat. no. RR420L; Takara Bio, Inc.) on a real-time PCR system (CFX96 Touch system; Bio-Rad Laboratories, Inc.). The thermocycling protocol for qPCR was 95˚C for 10 min, followed by 35 cycles of 95˚C for 15 sec and 58˚C for 20 sec. The expression of OPG mRNA and 13 miRNAs was normalized to those of β-actin and U6, respectively. The primer sequences involved were listed in Table II. The qPCR results were analyzed and calculated using the 2^-∆∆Cq method (22).

Human miR-20a mimics (5'-UAAAGUGCUUAUAGUGCAGGUAG-3') and miR-20a inhibitor (5'-2'-O-methyl-CUACCUGCACUAUAAGCACUUUA-3') were synthesized by Invitrogen (Thermo Fisher Scientific, Inc.), and applied to overexpress or inhibit miR-20a expression in A549 cells using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's protocol. The overexpression/inhibition was identified using RT-qPCR analysis after transfection (80 ng miR-20a mimics or inhibitor) for 48 h (23). miRNA mimic Control (5'-UUCUCCGAACGUGUCACGUUU-3'; Invitrogen; Thermo Fisher Scientific, Inc.) and miRNA inhibitor control (5'-2'-O-methyl-AAACGUGACACGUUCGGAGAA-3'; Invitrogen; Thermo Fisher Scientific, Inc.) were used as control after transfection (80 ng) for 48 h.

Total protein was extracted from A549 cells using RIPA Buffer (Beijing Solarbio Science & Technology Co., Ltd.) and protein concentration was quantified using a bicinchoninic acid protein assay kit (Beijing Solarbio Science & Technology Co., Ltd.). In total, ~80 µg total protein was separated by 12% SDS-PAGE and transferred onto PVDF membranes (Thermo Fisher Scientific, Inc.) as previously described (24). To measure the protein expression levels of OPG, the membranes were incubated with anti-OPG (cat. no. ab11994; 1:1,000) and anti-β-actin (cat. no. ab179467, 1:5,000 dilution) primary antibodies (both from Abcam) for 2 h at room temperature. HRP-conjugated goat anti-mouse IgG Fc (cat. no. ab205719; 1:20,000 dilution; Abcam) or goat anti-rabbit IgG Fc (cat. no. ab97200; 1:20,000 dilution; Abcam) was used to as the secondary antibody to incubate the membranes for 2 h at room temperature. An ECL chemiluminescence kit (EMD Millipore) was used to visualize the specific blots and autoradiograms were quantified by densitometry with Quantity One^® software (version 4.6.9; Bio-Rad Laboratories. Inc.) after normalizing to β-actin.

Transwell chambers (8-µm pore size) coated with Matrigel were purchased from BD Biosciences (cat. no. 354481) and applied to evaluate the cell invasion capacity (25). Briefly, after transfection with pcDNA-OPG in the presence or absence of miR-20a inhibitor for 48 h, 3.0x10⁴ A549 cells suspended in serum-free DMEM medium were seeded onto the upper chambers. Complete DMEM medium with 10% FBS was added into the lower chambers. Cells that had invaded into the lower chamber 48 h later were fixed with 4% formaldehyde for 15 min and stained with crystal violet (0.1%) for 10 min at room temperature. The number of invaded cells was counted from three visual fields using an Olympus light microscope (Olympus Corporation) under x200 magnification.

In total, 24 BALB/c nude female mice (weight, 20.6±2.04 g; age, 4-6 weeks; randomly divided into three groups of eight mice) were obtained from Shanghai Model Organisms Center, Inc. Mice were maintained in sterile conditions (temperature, 22-25˚C, humidity, 45-65%) under a 12-h light/dark cycle with free access to water and food. All experimental protocols were approved by the Animal Committee of Jiangxi Cancer Hospital (Nanchang, China). A549 cells transfected with pcDNA-OPG with or without miR-20a inhibitor for 48h (5x10⁶/200 µl PBS) were injected through the tail vein of the nude mice (26,27). Lung tissues were collected 10 weeks later for analysis of lung metastasis by gross anatomy after euthanasia. Euthanasia was performed by CO₂ inhalation at 20% V/min for 5 min followed by cervical dislocation. Humane endpoints included dehydration, weight loss >15%, hunched appearance and severe lameness. None of mice exhibited these humane endpoint criteria.

Murine lung tissues were fixed with 4% paraformaldehyde for 30 min at room temperature, dehydrated with an ascending gradient of ethanol and embedded in paraffin. The tissues were then cut into 5-µm sections, deparaffinized with xylene for 20 min and stained with hematoxylin for 1 h and then with eosin for 1 h (Beyotime Institute of Biotechnology) at room temperature. The sections were photographed using a light microscope (BX41; Olympus Corporation) and observed under x200 magnification.

The SPSS 16.0 statistical analysis software (SPSS, Inc.) was used for data analysis. Values are expressed as the mean ± standard deviation from ≥ three separate experiments. Unpaired Student's t-test (Fig. 1, Fig. 2 and Fig. 3) and one-way analysis of variance followed by Scheffé test (Figs. 4 and 5) were applied to analyze differences between two groups and those among ≥ three groups, respectively. Data in Table I was analyzed using Fisher's exact test. P<0.05 was considered to indicate a statistically significant difference.

Previous studies have demonstrated that upregulated OPG expression in lung cancer tissues is associated with cancer metastasis and worse survival (2,28). In the present study, OPG levels were detected in the serum of patients with NSCLC. ELISA results suggested that OPG levels were significantly elevated in the serum of patients compared with those in healthy individuals (Fig. 1A). Furthermore, OPG levels were significantly increased in the serum samples of patients with bone metastasis compared with those in the serum of patients in non-metastatic group (Fig. 1B). In addition, no association was found between OPG expression and the clinicopathological features of patients with NSCLC, specifically sex, age, history of smoking, histological type and TNM stage (Table I).

The biological effects of OPG on NSCLC cell lines were then explored further. A549 and H3122 cells were first transfected with the appropriate pcDNA control or pcDNA-OPG plasmids, before RT-qPCR and western blot analyses were performed to verify overexpression in the cell lines. The mRNA levels of OPG were significantly increased in H3122 and A549 cells transfected with pcDNA-OPG compared with those transfected with pcDNA control (Fig. 2A). This observation was confirmed further by western blot analysis in A549 cells (Fig. 2B). Functionally, OPG overexpression significantly increased the invasive ability of A549 cells compared with that in the pcDNA control group (Fig. 2C and D).

Next, the potential mechanism of the function of OPG in NSCLC cell invasion were explored. Given the important role of miRNAs in NSCLC progression (29) and the previously reported association between OPG and miRNAs (30), it was investigated whether OPG promoted NSCLC cell invasion in a miRNA-dependent manner. A total of 13 NSCLC-associated miRNAs (29,31-33) were selected as candidate targets that can be regulated by OPG (Table SI). As presented in Fig. 3A, the levels of miR-20a and miR-21 were found to be significantly upregulated (fold change >2.0 and P<0.05) whereas the levels of miR-200 and miR-34 were significantly downregulated (fold change <0.5 and P<0.05) after OPG overexpression in A549 cells compared with pcDNA control. The present study focused on miR-20a and its possible biological function in A549 cells was investigated because there was no effect of miR-21 in on OPG-induced cell invasion (data not shown). Significantly increased expression of miR-20a in A549 cells after OPG overexpression was confirmed compared with those transfected with the pcDNA control (Fig. 3B). Functionally, miR-20a overexpression in A549 cells significantly promoted cell invasion compared with transfected with the mimic control (Figs. 3C, D and S1).

The results aforementioned suggested that OPG possessed the potential to activate miR-20a expression in NSCLC cells and promote cell invasion, whilst that miR-20a overexpression alone also contributed to NSCLC cell invasion. It was therefore speculated that OPG may promote NSCLC cell invasion by regulating miR-20a expression. OPG overexpression significantly promoted A549 cell invasion, whereas miR-20a inhibition by its specific inhibitor significantly reversed this effect (Figs. 4A, B and S1). Similarly, OPG overexpression also significantly promoted H3122 cells invasion, which was significantly reversed by miR-20a inhibition (Figs. 4C, D and S1).

Results from the in vivo experiments appeared to be consistent with the in vitro data. The number of lung metastases in mice in the OPG overexpression group was increased compared with those in the pcDNA control group. Conversely, co-transfection with the miR-20a inhibitor resulted in the reversal of this increase (Fig. 5A). Histological analysis verified that this OPG-induced potentiation of lung metastasis was partially repressed by inhibiting miR-20a (Fig. 5B).