A total of 45 pairs of OS tissues and matched adjacent healthy tissues were obtained from patients (30 male patients and 15 female patients; mean age, 22.7±6.18 years) who underwent surgery at First Hospital of Lanzhou University (Lanzhou, China) between June 2018 and March 2021. The inclusion criteria were as follows: i) No other distant metastatic lesions; ii) none of the patients received chemotherapy or radiotherapy prior to surgery; iii) no history of other tumors; iv) patients had complete clinical data. The exclusion criteria were as follows: i) Patients were pregnant or lactating; ii) multiple lesions were present at the first diagnosis; iii) patients had other diseases that would affect the progression of treatment. Written informed consent was obtained from all patients, and the research method was approved by the Ethics Committee of the First Hospital of Lanzhou University (Lanzhou, China). All tissue samples were immediately treated with liquid nitrogen and then stored at -80˚C. A total of 100 mg of each tissue sample was ground in liquid nitrogen for reverse transcription-quantitative (RT-q)PCR.

Human fetal osteoblastic cells (hFOB1.19) and OS cell lines (HOS, U2OS, 143B and MG-63) were obtained from American Type Culture Collection. These cells were cultured in DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.), 100 U/ml penicillin and 100 mg/ml streptomycin (Gibco; Thermo Fisher Scientific, Inc.) at 37˚C in a humidified atmosphere with 5% CO₂.

Small interfering RNA (siRNA) against MALAT1 (si-MALAT1: sense strand: 5'-AAGAAAAAUAAAAGCUUUCCU-3', antisense strand: 5'-GAAAGCUUUUUAUUUUUCUUCC-3') and corresponding negative control siRNA (Scramble: sense strand: 5'-UUCUCCGAACGUGUCACGUTT -3', antisense strand: 5'-ACGUGACACGUUCGGAGAATT-3'), miR-590-3p mimic (miR-590-3p: 5'-TAATTTTATGTATAAGCTAGT-3') and the negative control (NC: 5'-UUCUCCGAACGUGUCACGUTT-3') mimic, miR-590-3p inhibitor (anti-miR-590-3p: 5'-ACTAGCTTATACATAAAATTA-3') and the NC inhibitor (anti-NC: 5'-CAGUACUUUUGUGUAGUACAA-3'), MALAT1 overexpression vector (MALAT1) and the empty pcDNA3.1 vector (Vector; Invitrogen; Thermo Fisher Scientific, Inc.) were purchased from Guangzhou RiboBio Co., Ltd. Cell transfection was performed using Lipofectamine^® 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.). When cell confluence reached ~90%, a mixture of 0.8 µg plasmid or 20 nM oligonucleotides and 2 µl Lipofectamine was added to each well of the 96-well plate and incubated in a CO₂ incubator at 37˚C, followed by incubation for 48 h. Meanwhile, untreated cells acted as the Control group.

Total RNA was extracted from tissues and cell lines (hFOB1.19, and OS cell lines HOS, U2OS, 143B and MG-63) using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). RNA was reversely transcribed into cDNA using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Inc.) or MicroRNA Reverse Transcription Kit (Thermo Fisher Scientific, Inc.), according to the manufacturer's protocols. The expression levels were detected using SYBR^® Premix Ex Taq™ (Takara Biotechnology Co., Ltd.) and quantified using 2^-ΔΔCq method. GAPDH and U6 were used as internal references. Primer sequences were as follows: MALAT1-forward (F), 5'-AAAGCAAGGTCTCCCCACAAG-3' and MALAT1-reverse (R), 5'-GGTCTGTGCTAGATCAAAAGGCA-3'; miR-590-3p-F, 5'-GCGTAATTTTATGTATAAGC-3' and miR-590-3p-R, 5'-GTATCCAGTGCGTGTCGTGGAGT-3'; GAPDH-F, 5'-GACTCCACTCACGGCAAATTCA-3' and GAPDH-R, 5'-TCGCTCCTGGAAGATGGTGAT-3'); U6-F, 5'-CTCGCTTCGGCAGCACA-3' and U6-R, 5'-AACGCTTCACGAATTTGCGT-3'.

A total of 2x10³ 143B and MG-63 cells were seeded into 96-well plates (Corning, Inc.) after transfection. A final 10% concentration of the CCK-8 solution (Dojindo Laboratories, Inc.) was added to each well after incubation for 24, 48 and 72 h. After 2 h incubation at 37˚C, the absorbance was measured at 450 nm using a microplate reader (Bio-Rad Laboratories, Inc.).

Briefly, 143B and MG-63 cells (1x10⁶) were digested with trypsin and washed with PBS. After being resuspended with binding buffer, the cells were stained with an Annexin V-FITC/PI kit (BD Pharmingen; BD Biosciences) for 15 min at room temperature in the dark. Then, the apoptotic rate was determined by the FACScan flow cytometry (BD Biosciences) and analyzed using FlowJo v.10.1 (FlowJo, LLC).

For cell migration assay, the transfected 143B and MG-63 cells (5x10⁴ cell/well) were seeded in serum-free DMEM and plated into the upper chamber of a 24-well Transwell with 8-µm polycarbonate membrane filters (Corning, Inc.). DMEM containing 10% FBS was added to the lower chamber. Cells were incubated for 24 h in a humidified atmosphere of 5% CO₂ at 37˚C. The cells adhering to the lower surface were fixed using 4% paraformaldehyde at 37˚C for 20 min, stained with 1% crystal violet at 37˚C for 10 min and counted under a light microscope (200x magnification; Olympus Corporation) in three random fields. For the invasion assay, the upper layer of the polycarbonate membrane was coated with 100 µl of diluted (1:1) Matrigel (BD Biosciences) and incubated at 37˚C for 4 h to dry into a gel. The invasive cells were analyzed using the same methods as aforementioned.

Total protein was extracted from 143B and MG-63 cells using RIPA buffer (Thermo Fisher Scientific, Inc.). Proteins were quantified using BCA Protein Assay Kit (Pierce; Thermo Fisher Scientific, Inc.). A total of 50 µg protein was separated by SDS-PAGE on 10% gels and transferred to PVDF membranes (MilliporeSigma). The membranes were blocked with 5% skim milk for 2 h at room temperature and then incubated overnight at 4˚C with primary antibodies against E-cadherin (cat. no. ab40772), Vimentin (cat. no. ab92547), N-cadherin (cat. no. ab76011), Snail (cat. no. ab216347) and GAPDH (cat. no. ab9485) (all 1:2,000; all from Abcam). The membranes were washed twice in TBS-1% Tween 20 and incubated with horseradish peroxidase-conjugated goat-anti-rabbit secondary antibody (cat. no. ab97051; 1:100,000; Abcam) for 2 h at room temperature. Finally, the protein bands were detected using an ECL system (Pierce; Thermo Fisher Scientific, Inc.), and the intensity of the target protein was determined by ImageJ software v1.8.2 (National Institutes of Health). GAPDH was used as an internal reference.

The binding sequences of MALAT1 and miR-590-3p were predicted using online software starBase v2.0 (https://starbase.sysu.edu.cn/). The sequences of wild-type (wt) MALAT1 (MALAT1-wt) and mutant (mut) MALAT1 (MALAT1-mut1 and MALAT1-mut2) containing miR-590-3p binding sites were cloned into pGL3 luciferase reporter vector (Promega Corporation). Then, pGL3 luciferase reporter vectors were co-transfected with miR-590-3p or NC into 143B and MG-63 cells using Lipofectamine 2000. Luciferase activity was examined using Dual-Luciferase Reporter Assay System (Promega Corporation) at 48 h after transfection according to the manufacturer's instructions. The activity of Renilla luciferase was used for normalization.

RIP was performed with EZ-Magna RNA-Binding Protein Immunoprecipitation Kit (car. no. 17-701; MilliporeSigma) in accordance with the manufacturer's protocol. Briefly, 143B and MG-63 cells were transfected as aforementioned with miR-590-3p or NC, lysed at 4˚C using RIP lysis buffer for 5 min, and then incubated overnight at 4˚C with magnetic beads conjugated with argonaute 2 (AGO2; cat. no. ab186733; 1:50; Abcam) antibody. After purification from AGO2 immunoprecipitation complex, the relative MALAT1 expression was analyzed by RT-qPCR.

Biotin-labeled miR-590-3p probe (Bio-miR-590-3p) and the corresponding control probe (Bio-NC) were purchased from Guangzhou RiboBio Co., Ltd. In brief, the biotinylated probe at a final concentration of 100 nM was incubated with 100 µl M-280 Streptavidin Dynabeads (Invitrogen; Thermo Fisher Scientific, Inc.) at room temperature for 2 h to form probe-coated beads. Then, 143B and MG-63 cells were transfected with Bio-miR-590-3p or Bio-NC, according to the aforementioned protocol. Subsequently, the cells were lysed with Pierce IP lysis buffer (Thermo Fisher Scientific, Inc.) at 4˚C for 5 min and incubated with probe-coated beads at 4˚C for 3 h. After RNA isolation, the enrichment of MALAT1 was examined by RT-qPCR.

All data are presented as the mean ± standard deviation of three independent experiments. Pearson correlation analysis was used to analyze the association between the expression levels of MALAT1 and miR-590-3p in OS tissues. Statistical analysis was performed using unpaired or paired Student's t-test for two groups and one-way ANOVA followed by Turkey's post hoc test for multiple groups. Notably, the paired Student's t-test was used to compare tumor tissues and adjacent non-tumor tissues of the same patients. Statistical analysis was implemented using GraphPad Prism 7 software (GraphPad Software, Inc.). P<0.05 was considered to indicate a statistically significant difference.

To investigate the expression of MALAT1 in OS tissues and cells, RT-qPCR was performed in OS tissues and cell lines. The results demonstrated that MALAT1 was significantly increased in OS tissues compared with adjacent non-tumor tissues (Fig. 1A). Next, the expression of MALAT1 was significantly upregulated in OS cells (HOS, U2OS, 143B and MG-63) compared with human fetal osteoblastic cells (hFOB1.19) (Fig. 1B). These data suggested that dysregulation of MALAT1 may be associated with OS.

Loss-of-function experiments were performed to examine the function of MALAT1 in OS cell proliferation and apoptosis. Since the expression levels of MALAT1 were higher in the 143B and MG-63 cell lines than those in the other OS cell lines (HOS and U2OS), 143B and MG-63 cells were selected for subsequent experiments. Firstly, the expression of MALAT1 was detected in 143B and MG-63 cells transfected with Scramble or si-MALAT1, and the results demonstrated that transfection with si-MALAT1 significantly decreased MALAT1 expression compared with Scramble (Fig. 2A). CCK-8 assay showed that MALAT1 knockdown significantly inhibited the proliferation of 143B and MG-63 cells (Fig. 2B). Flow cytometry assay revealed that the decrease of MALAT1 increased the apoptosis rate of 143B and MG-63 cells (Fig. 2C). Furthermore, Transwell assay demonstrated that MALAT1 knockdown markedly suppressed the migratory and invasive rate of 143B and MG-63 cells compared with cells transfected with Scramble (Fig. 2D and E). These results indicated that MALAT1 knockdown inhibited cell proliferation, migration and invasion, and induced apoptosis in OS cells.

The effects of MALAT1 on EMT-related markers were determined using western blot assay. The results revealed that knockdown of MALAT1 increased the protein level of E-cadherin, and decreased the protein levels of N-cadherin, Vimentin and Snail in 143B and MG-63 cells transfected with si-MALAT1 compared with the Scramble group (Fig. 3A and B), which suggested that low levels of MALAT1 may inhibit EMT of OS cells.

Firstly, it was demonstrated that the expression of miR-590-3p was decreased in OS tissues compared with adjacent healthy tissues (Fig. 4A), and miR-590-3p was downregulated in HOS, U2OS, 143B and MG-63 cells compared with hFOB1.19 cells (Fig. 4B). After transfection of NC or miR-590-3p into 143B and MG-63 cells, the overexpression efficiency of miR-590-3p was determined by RT-qPCR (Fig. 4C). CCK-8 assay indicated that the proliferation of 143B and MG-63 cells was significantly inhibited in the miR-590-3p group compared with the NC group (Fig. 4D). Flow cytometry assay demonstrated that the apoptosis rate of both 143B and MG-63 cells transfected with miR-590-3p was increased (Fig. 4E and S1A). Transwell assay revealed that miR-590-3p overexpression markedly suppressed the migration and invasion of 143B and MG-63 cells (Fig. 4F and G; Fig. S1C and E). In addition, the protein level of E-cadherin was increased, and the protein levels of N-cadherin, Vimentin and Snail were decreased in 143B and MG-63 cells transfected with miR-590-3p compared with the NC group (Fig. 4H). These data suggested that miR-590-3p overexpression inhibited cell proliferation, migration, invasion and EMT, and induced apoptosis in OS cells.

Bioinformatics software starBase v2.0 predicted the binding sites between MALAT1 and miR-590-3p (Fig. 5A). Luciferase reporter assay showed that miR-590-3p overexpression decreased the luciferase activity of MALAT1-wt in 143B and MG-63 cells, but failed to inhibit the luciferase activity of MALAT1-mut1+2 (Fig. 5B). Luciferase activity was also reduced in the MALAT1-mut1 and MALAT1-mut 2 groups, because both binding sites are active. Therefore, a single mutation in one of the binding sites allows miRNAs to bind to the unmutated site. Then, the specific binding between MALAT1 and miR-590-3p was validated using RIP and pull-down assays. RIP analysis showed that MALAT1 was notably enriched in the miR-590-3p group after precipitation with AGO2 antibody compared with the NC group (Fig. 5C). In addition, pull-down analysis revealed that MALAT1 was significantly enriched by binding to Bio-miR-590-3p, but not Bio-NC (Fig. 5D). Taken together, these results demonstrated that MALAT1 directly bound to miR-590-3p.

Next, the overexpression efficiency of MALAT1 was determined in 143B and MG-63 cells transfected with Vector or MALAT1 (Fig. S2A). The results showed that knockdown of MALAT1 markedly increased the expression of miR-590-3p in 143B and MG-63 cells compared with those in the Scramble group, whereas overexpression of MALAT1 notably reduced the expression of miR-590-3p in OS cells compared with in the control group (Vector) (Fig. 5E). Furthermore, the expression levels of MALAT1 and miR-590-3p were negatively correlated in OS tissues (Fig. 5F). These results indicated that MALAT1 negatively regulated miR-590-3p.

To further explore the relationship between MALAT1 and miR-590-3p in OS, 143B and MG-63 cells were transfected with Scramble, si-MALAT1, si-MALAT1 + anti-NC and si-MALAT1 + anti-miR-590-3p, respectively. Firstly, the knockdown efficiency of anti-miR-590-3p was detected by RT-qPCR in 143B and MG-63 cells transfected with anti-NC or anti-miR-590-3p (Fig. S2B). Then, the results indicated that miR-590-3p was upregulated after knockdown of MALAT1, whereas miR-590-3p was markedly reduced after transfection with si-MALAT1 and anti-miR-590-3p (Fig. 6A). Knockdown of MALAT1 significantly suppressed cell proliferation (Fig. 6B), promoted apoptosis (Fig. 6C and S1B), and inhibited migration and invasion (Fig. 6D and E; Fig. S1D and F) in 143B and MG-63 cells. In addition, 143B and MG-63 cells transfected with si-MALAT1 had higher E-cadherin protein level and lower N-cadherin, Vimentin and Snail protein levels compared with the Scramble group (Fig. 6F). However, these results caused by MALAT1 knockdown could be reversed by inhibiting miR-590-3p.