A total of 15 pairs of chordoma tissues and adjacent non-cancerous tissues were collected from 15 chordoma patients in the Department of Orthopedics, the First Affiliated Hospital of Soochow University. All of the enrolled patients had not received chemo- or radiotherapy before surgical excision. The tissues were frozen in liquid nitrogen after resection and stored at -80°C until analyses. All participants signed their informed consents. This study was approved by the ethics committee of the First Affiliated Hospital of Soochow University.

The human chordoma cell lines U-CH1 and U-CH2 were obtained from the American Type Culture Collection (Manassas, VA, USA) and cultured in IMDM-RPMI 1640 (4:1) supplemented with 10% FBS (Gibco/Thermo Fisher Scientific, Shanghai, China), 100 U/ml penicillin and 100 mg/ml streptomycin (Biological Industries, HertzliyaPituach, Israel). Cells were cultured in a humidified incubator with 5% CO₂ at 37°C.

The full-length coding sequence of lncRNA MDFIC-7 (lnc-MDFIC-7-wt) and its mutant (lnc-MDFIC-7-mut) were inserted into the pmirGLO empty vector to generate pmirGLO-MDFIC-7-WT and pmirGLO-MDIFC-7-mut, respectively. The ARF6 3′-UTR (3′-UTR-wt) and its mutant (3′-UTR-mut) were inserted into the pmirGLO-vector to generate pmirGLO-ARF6-3′UTR-wt and pmirGLO-ARF6-3′UTR-mut, respectively. The pGLO vector, also named pmirGLO, was obtained from Promega (Madison, WI, USA).

For knockdown of MDFIC-7, we used the following shRNA sequence: 5’- GATCCCCGAGAGAGAATTAAAGTCTATTCAAGAGATAGACTTTAATTCTCTCTCTTTTTGGAAA-3’; the control shRNA sequence was 5’- TGCTGAAATGTACTGCGCGTGGAGACGTTTTGGCCACTGACTGACGTCTCCACGCAGTACATTT-3’. The shRNAs were cloned into lentivirus vectors from Addgene (Watertown, MA, USA) to generate Lv-sh-MDFIC-7 and Lv-sh-NC.

The mock, mimic, and inhibitor for miR-525-5p were inserted into lentiviruses. We inserted the full length ARF6 (NCBI Reference Sequence: NM_001663.4) into a lentivirus vector to generate Lv-ARF6, and Lv-vector was used as a negative control.

All lentivirus constructs were synthesized and packaged by GenePharma (Shanghai, China).

U-CH1 or U-CH2 cells were seeded in dishes or plates and grown to 70% confluence. Plasmids were transfected using Lipofectamine 2000 agent (Invitrogen). The titer of the packaged lentivirus used in experiments was approximately 1.0 × 10⁹ infectious units per milliliter. Cells were incubated with lentivirus for 48 h, and then cells were collected for examination or used in further experiments.

Total RNA was extracted from tissues and cells using TRIzol reagent (Thermo Fisher Scientific), and 0.5 μg of total RNA was used for reverse transcription to generate complementary DNA (cDNA) using SMART MMLV Reverse Transcriptase (Takara Bio, Inc., Dalian, China). qRT-PCR was performed with the AceQ qPCR SYBR GreenMaster Mix kit (Vazyme, Nanjing, China) on a CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, Inc., Hercules, CA, USA). GAPDH mRNA was used as an internal control for quantification of lncRNAs and mRNAs, and the small nucleolar RNA (snRNA) of U6 was used to normalize the relative abundance of miRNA-525-5p. The 2-ΔΔCt method was used to determine gene expression. All experiments were performed in triplicate. The primers used in experiments are provided in the Supplementary Table S1 .

Tissues and cells were lysed using RIPA lysis buffer (Thermo Fisher Scientific). Equal amounts of protein (30 μg) were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and then transferred onto a PVDF membrane (Bio-Rad, Hercules, CA, USA). The membranes were blocked with 5% non-fat milk overnight at 4°C. Membranes were incubated with primary antibodies at 4°C overnight and then probed with secondary antibody at room temperature for 1 h. The antibodies used in this study are as follows: ARF6 (#ab131261) and PDK1 (#207450) primary antibodies were purchased from Abcam; Glut1 (#12939), HK2 (#2867) and LDHA (#3582) primary antibodies were purchased from Cell Signal Technology; p44/42 MAPK (ERK1/2) (#9102), phospho-p44/42 MAPK (ERK1/2) (Thr202/Tyr204) (#8544) and GAPDH (#MB9231) primary antibodies and horseradish peroxidase (HRP)-labeled goat anti-rabbit or -mouse secondary antibody (#BS13278 or #BS12478) were bought from Bioworld Technology. All other agents were purchased from Sigma (St. Louis, MO, USA).

The proliferative capacity of chordoma cells was determined by CCK-8 assay (Beyotime Institute of Biotechnology; Shanghai, China). Approximately 2 × 10³ U-CH1 cells were seeded into each well of a 96-well plate. At the indicated time points, 10 μl CCK-8 solution (Sigma) was added into each well. After 2 h incubation at 37°C, the optical density value of each well at OD450 nm was measured by a microplate reader (Thermo Fisher Scientific) and growth curves were plotted. The experiments were conducted in quadruplicate and repeated three times.

Colony formation was determined by soft agar assay. Briefly, 5000 U-CH1 or U-CH2 cells infected with the indicated lentivirus for 48 h were seeded in medium containing 0.4% soft agar on top of a layer consisting of 0.6% soft agar medium. Cells were cultured at 37°C with 5% CO₂ for 14 days. Cells were then fixed with 4% formaldehyde and stained with 1 µg/ml ethidium bromide. Colonies (>10 cells) were counted under a microscope in 10 fields per well.

To examine the regulation between lncRNA MDFIC-7 and miR-525-5p, U-CH1 or U-CH2 cells were co-transfected with pmirGLO-MDFIC-7-WT, which contains the predicted miR-525-5p binding site, or pmirGLO-MDFIC-7-mut, together with miR-525-5p mimic, inhibitor or mock. To investigate the effect of lncRNA MDFIC-7 and miR-525-5p on ARF6, cells were co-transfected with pmirGLO-ARF6-3′UTR-WT and pmirGLO-ARF6-3′UTR-mut, together with miR-525-5p mimic or pcDNA-MDFIC-7. Transfection was performed using Lipofectamine 2000 agent (Invitrogen). After transfection for 48 h, the luciferase activity was calculated using the Dual-Luciferase Assay Kit (Promega) on a GloMax 20/20 Luminometer. Firefly luminescence was normalized to the Renilla luminescence according to the manufacturer’s instructions.

To investigate the interaction between lncRNA MDFIC-7 and miR-525-5p, RIP analysis was conducted using the Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (Millipore, Billerica, MA, USA). Briefly, transfected U-CH1 cells were lysed in RIP lysis buffer. Cell lysates were then incubated with anti-Ago-2 (#SAB4301150, Sigma) or anti-IgG (#I4131, Sigma) in RIP buffer overnight at 4°C. The levels of lncRNA MDFIC-7 and miR-525-5p in the immunoprecipitated RNAs were examined by qRT-PCR.

To measure the glycolytic capacity and mitochondrial function of chordoma cells, the Seahorse Bioscience XF96 Extracellular Flux Analyzer and Seahorse XF Glycolysis Stress Test (#103020-100, Agilent) and Cell Mito Stress Test kits (#103015-100, Agilent) were used according to the manufacturer’s instruction.

Nude mice of BALB/c were obtained from the Model Animal Research Center of Nanjing University (Nanjing, China). A total of 10 male mice (4 to 6 weeks old) were randomized in 2 groups (n=5 per group). Approximately 4×10⁶ U-CH1 or U-CH2 cells infected with lentivirus (Lv-sh-MDFIC-7 or Lv-sh-NC) were inoculated into the right flank by subcutaneous injection. Tumors were measured once every 5 days for 35 days using a vernier caliper, and tumor volume was calculated using the formula: length × width² × 0.5. After 5 weeks, the mice were sacrificed and the tumor tissues were collected for weighing and gene expression analysis. The small animal euthanasia equipment was used for laboratory mice euthanasia, and the Euthanasia protocols were followed according to the guideline established by the American Medical Veterinary Association (AMVA) to minimize animal pain and suffering. The animal experiments were approved by the Committee on the Ethics of Animal Experiments of the First Affiliated Hospital of Soochow University.

The DIANA tool was used to predict the potential interaction between miRNAs with lncRNA MDFIC-7 and the potential target mRNAs of miRNA-525-5p.

The results are shown as mean ± standard deviation (SD). The data were analyzed using GraphPad Prism 9 (CA, USA) and SPSS software (version 19.0, SPSS Inc., NY, USA). Statistical significance was tested by two-tailed Student’s t-test for two group comparisons and one-way analysis of variance (ANOVA) test with post-hoc analysis contrasts for multi-group comparisons. p<0.05 was considered significant.

We collected tumor tissues and paired adjacent non-tumor tissues from 15 chordoma patients. H&E staining revealed pathological characteristics of the tumor which that distinguished from the adjacent non-tumor tissues ( Figure 1A ). To examine the expression profiles of lncRNAs in tumor tissues compared to adjacent non-tumor tissues, we performed lncRNA sequencing (lncRNA-seq) analysis on three pairs of randomly selected tissue samples. On the basis of the lncRNA fold change, p-values and bioinformatics analysis ( Figure 1B ), a total of 4900 differentially expressed lncRNAs were identified. These lncRNAs were shown in the heatmap in Figure 1C . The top 30 upregulated lncRNAs were screened out and shown in Figure 1D . Meanwhile, the top 30 downregulated lncRNAs were also screened out and shown in Figure S1 .

The most significantly upregulated lncRNA was lncRNA MDFIC-7, and we further validated the expression of lncRNAs by qRT-PCR in the chordoma and adjacent normal tissues. The results showed that lncRNA MDFIC-7 was significantly upregulated in tumor tissues compared to adjacent non-tumor tissues ( Figure 1E ). We evaluated the expression of the tumor proliferation biomarker Ki-67 in tumor and non-tumor tissues and identified a significant correlation between lncRNA MDFIC-7 and Ki-67 expression in chordoma tissue ( Figures 1F, G ).

To investigate the role of lncRNA MDFIC-7 in regulating the tumor progression of chordoma, we first inhibited lncRNA MDFIC-7 expression in U-CH1 and U-CH2 cells using lentivirus expressing shRNA-MDFIC7. The efficiency of lncRNA MDFIC-7 inhibition was examined by qRT-PCR and the results confirmed reduced expression of lncRNA MDFIC-7 in cells infected with lentivirus targeting MDFIC-7 (Lv-sh-MDFIC-7) compared with the control (Lv-sh-NC) ( Figure 2A ). Knockdown of lncRNA MDFIC-7 by lentivirus significantly suppressed the proliferation of chordoma cells, as determined by CCK-8 and colony formation assays ( Figures 2B–D ). In addition, knockdown of lncRNA MDFIC-7 resulted in decreased expression of the cell growth biomarkers PCNA and CDK2 ( Figure 2E ). These results demonstrated that lncRNA MDFIC-7 inhibition suppressed the proliferation of chordoma cells and suggested that lncRNA MDFIC-7 functions as a tumor promoter in chordoma by enhancing cell proliferation.

To analyze the mechanism by which lncRNA MDFIC-7 regulates proliferation of chordoma cells, we examined the subcellular location of lncRNA MDFIC-7 in U-CH1 and U-CH2 cells. The nucleus and cytoplasm of cells were fractionated and the total RNA fractions were isolated and examined by qRT-PCR. The results showed that lncRNA MDFIC-7 is mainly distributed in the cytoplasm of chordoma cells ( Figure 3A ). It was suggested that lncRNA MDFIC-7 could act as a competing endogenous RNA (ceRNA) in human chordoma. The potential target miRNAs of lncRNA MDFIC-7 were predicted by online tools, including DIANA and StarBase 3.0. The results revealed that miR-525-5p contains a putative binding site for lncRNA MDFIC-7, suggesting miR-525-5p as a potential target of MDFIC-7 ( Figure 3B ). We found that the expression of miR-525-5p was decreased in tumor tissues compared with adjacent non-tumor tissues of chordoma patients ( Figure 3C ). We examined the binding between lncRNA MDFIC-7 and miR-525-5p using RIP, and the results demonstrated an interaction between lncRNA MDFIC-7 with miR-525-5p ( Figure 3D ). To further examine the interaction between lncRNA MDFIC-7 with miR-525-5p, we performed the dual luciferase assay. First, the overexpression effect of miR-525-5p mimic by lentivirus infection was detected by RT-qPCR in U-CH1 and U-CH2 cells ( Figure 3E ). The dual luciferase assay results revealed that the luciferase activity of the wild-type MDFIC-7 reporter (pGLO-MDFIC-7-WT) was significantly repressed after the introduction of miR-525-5p mimic, but miR-525-5p mimic had no impact on the mut-lncRNA MDFIC-7 reporter ( Figure 3F ). Furthermore, the activity of the wild-type MDFIC-7 reporter (pGLO-MDFIC-7-WT) was significantly enhanced after the introduction of miR-525-5p inhibitor, but miR-525-5p inhibitor had no impact on the mut-lncRNA MDFIC-7 reporter ( Figure 3F ). These data indicated that lncRNA MDFIC-7 may play a potential role in enhancing tumor progression of human chordoma via sponging miR-525-5p.

To further investigate the specific function of miR-525-5p in chordoma cell proliferation, we performed colony formation assay to determine the effect of miR-525-5p on cancer cell growth. MiR-525-5p overexpression inhibited cell growth of U-CH1 and U-CH2 cells, and this effect was blocked by overexpression of lncRNA MDFIC-7 ( Figures 3G, H ). In addition, the mRNA levels of cell proliferation biomarkers PCNA and CDK2 were decreased by miR-525-5p overexpression and this effect was blocked by overexpression of lncRNA MDFIC-7 ( Figure 3I ). These results demonstrated that lncRNA MDFIC-7 acts as a molecular sponge of miR-525-5p and indicated that lncRNA MDFIC-7 and miR-525-5p may play oncogenic and tumor suppressor roles, respectively, in regulating chordoma progression.

We next examined the potential target genes of miR-525-5p by bioinformatics analysis using the online tools DIANA and StarBase3.0. The 3′UTR in the mRNA of ARF6 was predicted to contain a complementary binding site for miR-525-5p ( Figure 4A ), suggesting that ARF6 might be a potential target of miR-525-5p. The expression of ARF6 was evaluated and found increased in chordoma tissues compared to adjacent non-tumor tissues ( Figure 4B ). Dual luciferase reporter assay revealed that miR-525-5p mimic transfection inhibited the activity of the luciferase reporter harboring the wild-type ARF6 3′UTR in chordoma cells, while co-transfection with lncRNA MDFIC-7 blocked the inhibitory effect of miR-525-5p mimic on luciferase activity ( Figure 4C ). However, these effects were abolished with the reporter construct in which the ARF6 3′UTR was mutated ( Figure 4C ). Additionally, RT-qPCR and Western blot assays revealed that miR-525-5p mimic transfection inhibited the expression of ARF6 mRNA and protein in U-CH1 and U-CH2 cells, while co-transfection with lncRNA MDFIC-7 reversed the effect of miR-525-5p overexpression on ARF6 mRNA and protein expression ( Figures 4D, E ). Notably, transfection with lncRNA MDFIC-7 alone enhanced both ARF6 mRNA and protein expression ( Figures 4D, E ). These results showed that ARF6 is a direct target of miR-525-5p, and its expression is positively regulated by lncRNA MDFIC-7.

To further evaluate the correlation between ARF6 and lncRNA MDFIC-7 expression, we examined the expression of ARF6 mRNA and lncRNA MDFIC-7 in tumor and adjacent non-tumor tissues by RT-qPCR. A significant correlation between lncRNA MDFIC-7 and ARF6 expression in chordoma was found ( Figure 5A ). The expression of ARF6 mRNA was dramatically inhibited by lncRNA MDFIC-7 knockdown in U-CH1 and U-CH2 chordoma cells ( Figure 5B ). To validate whether ARF6 is a downstream targeted molecule of the lncRNA MDFIC-7/miR-525-5p axis, ARF6 was overexpressed using lentivirus in U-CH1 and U-CH2 chordoma cells and its expression was confirmed by RT-qPCR and Western blot ( Figures 5C, D ). Furthermore, we evaluated the effect of the lncRNA MDFIC-7/miR-525-5p axis and ARF6 on cell proliferation using colony formation and RT-qPCR assays. Our results showed that inhibition of lncRNA MDFIC-7 expression or transfection of miR-525-5p mimic repressed the colony forming activity of chordoma cells, but this effect was reversed by co-expression of ARF6 ( Figures 5E–G ). Moreover, the inhibition of PCNA and CDK2 expression induced by lncRNA MDFIC-7 downregulation or miR-525-5p mimic was also reversed by co-expression of ARF6 ( Figures 5H, I ). These results indicated that lncRNA MDFIC-7/miR-525-5p and ARF6 play important roles in regulating the proliferation of chordoma cells.

Previous studies showed that ARF6 is a downstream factor in the Kras/ERK signaling pathway and promotes proliferation and the Warburg effect in pancreatic cancer cells (34). To investigate whether ARF6 plays a similar role in chordoma cells and whether this function was modulated by the lncRNA MDFIC-7/miR-525-5p axis, we examined the status of ERK1/2 and the expressions of a series of glycolytic genes that encode key proteins with important roles in metabolizing glucose into lactate that are directly related to aerobic glycolysis, such as glucose transporter type 1 (GLUT1), hexokinase 2 (HK2), pyruvate dehydrogenase kinase 1 (PDK1) and lactate dehydrogenase A (LDHA) in U-CH1 and U-CH2 cells. The results showed that inhibition of lncRNA MDFIC-7 reduced activation of ERK1/2 ( Figure 6A ) as well as the expressions of GLUT1, HK2, PDK1 and LDHA mRNAs and proteins ( Figures 6B–F ). However, these inhibitory effects were reversed by ARF6 overexpression ( Figures 6A–F ). These findings suggest the role of lncRNA MDFIC-7/miR-525-5p/ARF6 axis in ERK signaling activation and aerobic glycolysis.

We further evaluated the changes in aerobic glycolysis in U-CH1 and U-CH2 cells with lncRNA MDFIC-7 knockdown cells using the Seahorse XF analyzers. The ECAR was significantly decreased in lncRNA MDFIC-7 knockdown cells, indicating that inhibiting lncRNA MDFIC-7 suppressed the glycolytic process in chordoma cells ( Figures 6G, H ). Notably, the effect of glycolysis was reversed by overexpression of ARF6 ( Figures 6G, H ). The OCR value reflects glucose mitochondrial oxidation. Consistent with the ECAR results, lncRNA MDFIC-7 knockdown enhanced OCR in chordoma cells, and the promotion of OCR was reversed by ARF6 overexpression ( Figures 6I, J ). These results demonstrated that lncRNA MDFIC-7/miR-525-5p plays an important role in regulating aerobic glycolysis of chordoma cells through modulating ARF6 expression.

To further investigate the effects of the lncRNA MDFIC-7/miR-525-5p/ARF6 axis on tumorigenicity in vivo, we generated a xenograft tumor mouse model using U-CH1 cells infected with lentivirus to downregulate lncRNA MDFIC-7. The tumors in the Lv-sh-MDFIC-7 group were noticeably smaller than the control Lv-sh-NC group ( Figure 7A ). The data of tumor volume and weight in various groups were shown in Figures 7B, C . We evaluated the gene expression of miR-525-5p, ARF6, PCNA and CDK2 in the xenograft tumor tissues by RT-qPCR ( Figures 7D, E ). We also examined the expression of a series of glycolysis-related genes, including GLUT1, HK2, PDK1 and LDHA genes, in the tumors. Consistent with the in vitro data, inhibition of lncRNA MDFIC-7 expression suppressed the mRNA expression of GLUT1, HK2, PDK1 and LDHA in the xenograft tumor tissues ( Figure 7F ). These data validated that the lncRNA MDFIC-7/miR-525-5p/ARF6 axis plays an important role in regulating cell proliferation and the Warburg effect in chordoma in vivo and modulates chordoma progression.