The GSE126209 dataset, contains RNA sequencing data from OS and normal tissues, was downloaded from Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/). The FPKM values were transformed into TPM values and normalized by log₂(value+1) to further analysis. Afterwards, paired differential expression was analyzed using the R package limma and volcanic maps were draw by ggplot2 (|FC | > 1, and adj.p < 0.05).

Clinical and gene expression data of OS patients were obtained from the Tumor Alterations Relevant for Genomics-driven Therapy (TARGET) database. Patient samples (n=85) were divided into high and low groups based on the median of expression values, and survival curves were plotted by the “survival” package in R.

Putative RNF2 targets were predicted using the Gene Transcription Regulation Database (GTRD, http://gtrd.biouml.org) and the KnockTF Database (http://www.licpathway.net/KnockTF/index.php).

OS and adjacent non-tumoral bone tissues were collected from patients who were surgically treated at the Suzhou Municipal Hospital of Nanjing Medical University. Clinical samples were surgically collected, immediately fixed, dehydrated and embedded in paraffin. Written informed consent was obtained prior to sample collection, and this study was approved by the Ethics Committee of Suzhou Municipal Hospital.

A specific FISH probe targeting PRR7-AS1 was designed and synthesized by Ribobio Biotechnology (Guangzhou, China). The hybridization was performed in OS tissue and paired adjacent non-tumoral tissues as previously reported (15). All images were analyzed on a confocal laser scanning microscope (LSM 810, Carl Zeiss, Oberkochen, Germany) (16).

The human osteosarcoma cell lines, 143B and U2OS, were obtained from the Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China). U2OS cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, NY, USA) supplemented with 10% fetal bovine serum (FBS) (ExCell Bio, New Zealand) and 1% penicillin/streptomycin (NCM Biotech, Suzhou, China), and 143B cells were cultured in Eagle’s Minimum Essential Medium (EMEM) (Gibco) supplemented with 82% minimal essential medium (MEM), 15% FBS, 1% GlutaMAX, 1% sodium pyruvate and 1% penicillin/streptomycin. Cells were cultured in a humidified incubator with 5% CO2 and 95% air at 37°C.

The small interfering RNAs (siRNAs) targeting PRR7-AS1, RNF2, MTUS1 and negative control and pcDNA3.1-PRR7-AS1 were designed by GenePharma (Shanghai, China) and transfected into cells using Lipofectamine 2000 (Invitrogen, USA) as described previously (17). After 48 h, cells were collected for subsequent experiments. The siRNA sequences used in the present study were shown on Table S1 .

Total RNA was extracted from cells using the RNA isolater Total RNA Extraction Reagent (Vazyme, Nanjing, China), which was reverse transcribed into cDNA using the HiScript III RT SuperMix for qPCR (+gDNA wiper) kit (R323-01, Vazyme) as described previously (18, 19). RT-qPCR was performed using SYBR qPCR SuperMix Plus (Novoprotein Scientific Inc., Shanghai, China) and an Applied Biosystems 7500 Real Time PCR System. The primers were shown on Table S2 .

Osteosarcoma cells transfected with si-NC/si-PRR7-AS1 were lysed using a radioimmunoprecipitation assay (RIPA, Beyotime, Nantong, China) containing 1% protease inhibitor phenylmethylsulfonyl fluoride (PMSF) (17). After quantified using a bicinchoninic acid (Beyotime) kit, the protein samples were separated by SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) and transferred onto polyvinylidene difluoride (PVDF) membranes. The membranes were blocked with 5% skim milk, and then incubated with the anti-RNF2 antibody (Proteintech, Wuhan, Hubei, China) and anti-Tubulin (Beyotime) at 4°C overnight. Secondary antibodies combined with horseradish peroxidase at room temperature were incubated and band signals were visualized by an enhanced chemiluminescent substrate and quantified by Image-Pro Plus (Media Cybernetics, San Diego, CA, USA).

Cell Counting Kit-8 (CCK8) assay: cells were seeded into 96-well plate with 2.5×10³ cells/well. CCK8 (Beyotime) was added into each well at 0, 24, 36, 72, 96 h. Cell viability was detected by measuring the optical density at 450 nm (OD450) using a microplate reader (Bio-Rad Model 680, Richmond, CA, USA).

Colony formation assay: cells were seeded into six-well plate with 1.0×10³ cells/well and cultured for 14 days. Fresh medium was replaced every 5 days. After 14 days, culture medium was removed, and cells were fixed with methanol and stained with 0.1% crystal violet (Beyotime). Visible colonies were imaged and counted.

Transwell assay: transwell chambers (8 μm pore size; Millipore, Billerica, MA, USA) were added into 24-well plate. Cells at a density of 2.5 × 10⁴ cells/well in 300 µl of serum-free medium were seeded in the upper chamber, and 700 µl of medium containing 10% FBS was added in the bottom chamber. After incubation for 48 h at 37°C, the cells migrated to the bottom were fixed with methanol and stained with 0.1% crystal violet. Finally, five randomly selected fields per sample were imaged using a light microscope for counting migratory cells.

Four-week-old athymic BALB/c nude mice were habituated in a specific pathogen-free (SPF) environment. Animal experiments were approved by the Ethics Committee of Animal Experiments of Nanjing Medical University. Briefly, 100 µl of suspended 143B cells transfected with sh-PRR7-AS1 or sh-NC at 3×10⁷ cells were subcutaneously injected into the mouse axillary region (17). Tumor growth was regularly recorded. After 2 weeks, mice were sacrificed, and OS tissues were collected. The tumor volume was calculated using the following formula: tumor volume (mm³) = 0.5 × length (mm) × width² (mm²).

For the in vivo cell metastasis assay, 100 µl of suspended 143B cells transfected with sh-PRR7-AS1 or sh-NC at 6×10⁷ cells was injected into the mouse tail vein. After 2 months, mice were sacrificed, and lung tissues were collected. The number of metastatic lesions in the lung was counted.

Tumor tissues and lung metastasis lesions collected from mice were fixed with 4% paraformaldehyde and dehydrated in gradient concentrations of ethanol. After permeabilization in xylene, sections were embedded in paraffin, sliced into 5-µm-thick sections, deparaffinized with xylene and rehydrated in gradient concentrations of ethanol. Lung metastasis sections were stained with hematoxylin and eosin (H&E) for pathological examination.

After deparaffinization and rehydration of paraffin-embedded tissues, antigen retrieval was performed by boiling samples in 10 mM sodium citrate buffer (pH 6.0). The sections were blocked with 1% bovine serum albumin (BSA) in PBS and incubated with primary antibodies, and the sections were then incubated with secondary antibodies. Cell nuclei were stained with DAPI (Beyotime), and images were acquired using a Zeiss laser confocal microscope (LSM 810) as previously described (20, 21). The following antibodies were used: mouse monoclonal anti-Ki67 antibody (Abcam, ab238020, 1:100), mouse monoclonal anti-E-cadherin antibody (Abcam, ab231303, 1:100), mouse monoclonal anti-N-cadherin antibody (Abcam, ab76057, 1:100) and mouse monoclonal anti-vimentin antibody (Abcam, ab20346, 1:100).

Isolated and cultured cells were fixed with 4% paraformaldehyde for 20 min. Next, the cells were treated with 0.2% triton for 20 min and blocked with 1% BSA in PBS. Then the slides containing target cells were incubated with primary and Alexa-Fluor secondary antibodies (Thermo Scientific, Waltham, USA) orderly. In addition, the cells were stained with DAPI and observed under a fluorescence microscope. Sections were analyzed under a confocal laser-scanning microscope.

PRR7-AS1 was transcribed using the Ribo™ RNAmax-T7 Biotin Labeling Transcription Kit (Ambio Life). The kit used the Biotin RNA Labeling Mix as a substrate and utilized a DNA template containing the T7 promoter to synthesize RNA complementary to the antisense strand in the DNA template starting downstream of the T7 promoter. Then PRR7-AS1 was purified with the RNeasy Plus Mini Kit (Qiagen) and treated with RNase-free DNase I (Qiagen). Subsequently, transcribed PRR7-AS1 was biotin-labeled with the Biotin RNA Labeling Mix (Ambio Life). RNA pull-down was then performed by PierceTM Magnetic RNA-Protein Pull-Down Kit according to the manufacturer’s instructions (Thermo Scientific Pierce) and as previously described (22). The RNA-protein complexes were then subjected to mass spectrometry analysis, as previously described (23–25).

ChIP was performed using the EZ-Magna ChIPTM A/G Chromatin Immunoprecipitation Kit (Millipore) as previously described (26). Briefly, transfected OS cells were lysed in 1% methanol and sonicated, which resulted in 500-bp DNA fragments. Cell lysate was incubated with the anti-RNF2 antibody, anti-H2AK119Ub antibody and anti-IgG. The chromatin supernatant was then incubated in 20 μl of protein A/G MagBeads at 4°C overnight. The protein-DNA complex was eluted and purified, and the obtained DNA samples were subjected to RT-qPCR. The primers used for ChIP-qPCR were follows:

RIP was performed using the Magna RIPTM RNA-Binding Protein Immunoprecipitation Kit (Millipore) as previously described (16). Briefly, OS cells were lysed in RIP lysis buffer, and 100 µl of cell lysate was incubated with the anti-RNF2 antibody and anti-IgG. A protein–RNA complex was captured and digested with 0.5 mg/ml proteinase K containing 0.1% SDS. The magnetic beads were repeatedly washed with RIP washing buffer to remove non-specific adsorption. Finally, the extracted RNA was subjected to RT-qPCR.

All data are expressed as mean ± standard deviation ( x ¯ ± s ). GraphPad Prism 9.0 (GraphPad Software, CA, USA) was used for statistical analyses. Differences between two groups were compared by Student’s t−test, and those among three or more groups were compared by one−way analysis of variance (ANOVA). Overall survival was estimated by the Kaplan–Meier method, and the log-rank test was employed to evaluate differences. P<0.05 was considered as statistically significant.

A total 19 lncRNAs were differentially expressed in osteosarcoma tissues (n=5) compared with paired normal tissues based on the GEO database (GSE126209) (|FC| > 1.0 and adj.p < 0.01) ( Figure 1A ), and 2 of them (PVT1 and PRR7-AS1) had a significant effect on the overall survival of OS patients (n=85) based on the TARGET database ( Figure S1 ). It has been found that PVT1 (plasmacytoma variant translocation 1) could promote human OS malignant biological behaviors (27), but the role of PRR7-AS1 in osteosarcoma remains unknown.

PRR7-AS1 was significantly upregulated in OS tissues (n=5) ( Figure 1B ), and high PRR7-AS1 expression was associated with poor prognosis in OS patients ( Figure 1C ). Afterwards, the FISH assay verified the overexpression of PRR7-AS1 in OS tissues compared to non-tumor samples ( Figures 1C, D ), and it also demonstrated that PRR7-AS1 was mostly expressed in the nucleus ( Figures 1C–F ). These results suggested that PRR7-AS1 is upregulated in OS and mainly involved in functional regulation in the nucleus.

The transfection efficiency of si-PRR7-AS1 were verified by RT-qPCR ( Figure 2A ). Both CCK8 and colony formation assays found that knockdown of PRR7-AS1 significantly inhibited the proliferative capacity of 143B and U2OS cells ( Figures 2B–E ). In addition, transwell assays revealed a lower migration capacity of OS cells transfected with si-PRR7-AS1 than si-NC ( Figures 2F, G ). These findings indicated that PRR7-AS1 promotes the proliferation and migration of OS cells in vitro.

A xenograft model in nude mice was established to explore the in vivo function of PRR7-AS1. During the experimental period, tumor growth was significantly lower in the sh-PRR7-AS1 group compared to the controls. After 2 weeks, mice were sacrificed, and OS tissues were collected. Compared to the control group, both the volume and weight of OS tissues were significantly lower in the sh-PRR7-AS1 group than empty vector group ( Figures 3A–C ). In addition, the lower Ki-67-positive cells was shown in the sh-PRR7-AS1 group too ( Figures 3D, E ).

We established an OS lung metastasis model in nude mice through injecting 143B cells into the mouse tail vein as previously reported (28). Compared to the controls, the number of lung metastases was significantly lower in nude mice injected with 143B cells transfected with sh-PRR7-AS1 ( Figures 3F, G ). The same results were obtained with H&E staining ( Figure 3H ). Furthermore, increased E-cadherin as well as decreased N-cadherin and vimentin expression were obviously in the sh-PRR7-AS1 group ( Figure 3I–L ).

Collectively, these findings demonstrated that PRR7-AS1 promotes the proliferation and metastasis of OS cells in vivo.

To further explore the molecular mechanism underlying the carcinogenic activity of PRR7-AS1 in OS, an RNA pull-down LC-MS/MS assay was performed to identify potential proteins that interact with lncRNA PRR7-AS1 in OS cells ( Figure 4A ).

According to the label-free quantitation (LFQ) method, proteomic analysis identified 18 overlapping proteins from three independent RNA pull-down experiments ( Figure 4B ; Table S3 ). Figure 4C shows the top 10 proteins listed from high to low. Among them, RNF2 (ring finger protein 2) caught our attentions. As a member of the polycomb group (PcG), RNF2 is widely involved in tumor development through epigenetic regulation (29). RNA pull-down and RIP experiments in 143B and U2OS cells further confirmed the interaction between PRR7-AS1 and RNF2 ( Figures 4D–F ). Strikingly, it was noticed that RNF2 expression levels remained unaltered after knockdown of PRR7-AS1 compared to control ( Figures 4G, H ). These findings suggest that PRR7-AS1 only binds RNF2, but does not affect its expression.

We selected U2OS cells for western blotting and immunofluorescence experiments of RNF2. The protein expression of RNF2 was significantly reduced after knockdown of RNF2 ( Figures 5A, B ). Immunofluorescence analysis showed that RNF2 protein was primarily localized in the nucleus rather than the cytoplasm and could be significantly knocked down ( Figures 5C, D ), which further verified that RNF2 protein was involved in gene regulation. To explore the oncogenic features of RNF2 in OS, we synthesized two RNF2 siRNAs and verified their transfection efficiency ( Figure 5E ). Knockdown of RNF2 significantly reduced cell viability and the number of colonies in 143B and U2OS cells, indicating inhibition of proliferation ( Figures 5F–I ). Moreover, the number of migratory cells was reduced in OS cells transfected with si-RNF2 compared to si-NC ( Figures 5J, K ). Therefore, these findings suggested that RNF2 facilitates the proliferation and migration of OS cells in vitro.

It has been reported that RNF2 contributes to inhibit gene transcription as a component of the polycomb repressive complex 1 (PRC1) (30). The KnockTF database predicted a total of 14,096 differentially expressed genes after knockdown of RNF2, and 2,778 genes were upregulated (|FC| >1.5). Through screening the ChIP-Seq datasets extracted from the publicly available Sequence Read Archive (SRA), GEO and ENCODE databases by GTRD, a total of 11,407 genes were identified to bind the RNF2 transcription factor, including 23,655 binding peaks. Venn diagram analysis yielded 272 candidate genes that not only bound to RNF2 but also were regulated by RNF2 ( Figure 6A ). Five of the target genes have been previously reported to have an oncogenic effect on OS cells ( Figure 6B ). RT-qPCR were performed to further confirm the regulatory effect of RNF2 on these target genes. MTUS1 (microtubule associated scaffold protein 1) was significantly upregulated in both 143B and U2OS cells after knockdown of RNF2 ( Figures 6C, D ). Therefore, RNF2 might target MTUS1 in OS cells.

RNF2 exerts its biological function in silencing target genes through histone H2A lysine 119 monoubiquitin (H2AK119ub) (31). The RNF2-binding region predicted by the GTRD falls from -2000 to +100 bp of the MTUS1 transcription start point, which is its promoter region ( Figure 6E ). ChIP-qPCR results showed that RNF2 was significantly recruited in this region ( Figures 6F, G ). Moreover, MTUS1 mRNA expression was upregulated after PRR7-AS1 silencing in 143B and U2OS ( Figure 6H ). Knockdown of PRR7-AS1 reduced the recruitment of H2AK119ub to the MTUS1 promoter ( Figures 6I, J ). Taken together, these findings demonstrated that PRR7-AS1 is required for RNF2 to inhibit the transcription of MTUS1 through H2AK119ub. Furthermore, it is obvious that overexpressing PRR7-AS1 decreased the MTUS1 mRNA level and knockdown of RNF2 reverses this decrease ( Figure 6K ). This finding demonstrated that PRR7-AS1 inhibit the expression of MTUS1 dependent on RNF2.

MTUS1 is a tumor suppressor gene that has been validated to inhibit the proliferation and migration of OS cells. The transfection efficiency of si-MTUS1 were verified by RT-qPRCR ( Figure 7A ), si-MTUS1 2# was used in the following experiments. 143B and U2OS cells were co-transfected with si-MTUS1 and si-PRR7-AS1/si-RNF2. Interestingly, the proliferative and migration capacities of co-transfected cells were significantly higher than those only transfected with si-PRR7-AS1 or si-RNF2 ( Figures 7B–G ). These findings indicated that knockdown of MTUS1 reverses the inhibited proliferation and migration of OS cells caused by knockdown of PRR7-AS1 and RNF2.