Fig 2: METTL3 silencing reduced m6A modification in OS cells and thereby suppressed HDAC5 expression.A, B m6A level in 50 pairs of OS tissues and adjacent tissues was detected by m6A quantitative analysis. B m6A level in human OS cell lines (U2OS, HOS, SAOS2) and normal osteoblasts (hFOB1.1) was detected by m6A quantitative analysis. E, F HDAC5 expression in human OS cell lines (U2OS, HOS, SAOS2) and normal osteoblasts (hFOB1.1) was determined using RT-qPCR and western blot. G The correlation between HDAC5 and METTL3 expression in OS tissues was analyzed using Pearson correlation analysis. H The m6A modification level of HDAC5 mRNA in OS cells after intervention with METTL3 expression was detected using MeRIP-qPCR. I–K HDAC5 expression and stability in OS cells after intervention with METTL3 expression was measured using RT-qPCR and western blot. N = 50. The cell experiment was repeated three times independently. Data in panels A, B/D–F/H–K are presented as mean ± standard deviation. Data in panel C were analyzed using paired t-test and data in panels A/D were analyzed using t-test. Data in panels B/E–F/H–J were analyzed using one-way ANOVA, and data in panel K were analyzed using two-way ANOVA, followed by Tukey’s multiple comparisons test, *p < 0.05, **p < 0.01.

Fig 3: METTL3 silencing suppressed OS proliferation in vivo via the HDAC5/miR-142-5p/ARMC8 axis.The xenograft tumor model in nude mice was established with U2OS cells stably low-expressing METTL3. A Tumor volume recorded every week. B After the nude mice were euthanized on the 35th day, the tumor weight and representative images were presented. C The positive rate of Ki67 was detected using immunohistochemistry. D, E METTL3 level was determined using RT-qPCR and immunohistochemistry. F m6A level was quantitatively analyzed. G, H HDAC5 level was determined using RT-qPCR and immunohistochemistry. I, J miR-142-5p and ARMC8 level was determined using RT-qPCR. K ARMC8 level was detected using immunohistochemistry. N = 6. The cell experiment was repeated three times independently. Data in panels A–C/E, F/H are presented as mean ± standard deviation. Data in panels B–K were analyzed using t-test. Data in panel A were analyzed using two-way ANOVA, followed by Tukey’s multiple comparisons test, **p < 0.01.

Fig 4: HDAC5 overexpression partially reversed the inhibitory effect of METTL3 silencing on OS cell proliferation.pcDNA 3.1 HDAC5 (pc-HDAC5) was transfected into U2OS cells, with pc-NC as negative control. After 48 h, A, B HDAC5 expression in U2OS cells was measured using RT-qPCR and western blot. Then, pc-HDAC5-transfected U2OS cells were treated with si-METTL3#2 for 48 h. C–E The proliferation of OS cells was measured using CCK-8 assay (C), colony formation assay (14 days) (D), and EdU staining (E). The cell experiment was repeated three times independently. Data are presented as mean ± standard deviation. Data in panels A, B were analyzed using t-test. Data in panels D, E were analyzed using one-way ANOVA, and data in panel C were analyzed using two-way ANOVA, followed by Tukey’s multiple comparisons test, *p < 0.05, **p < 0.01.

Fig 5: miR-142-5p targeted ARMC8.A The target genes of miR-142-5p were predicted through Starbase, miRWalk, Targetscan, and miRDB databases. B The binding site of miR-142-5p and ARMC8 in the Starbase database. C, D The binding of miR-142-5p and ARMC8 in OS cells was verified using dual-luciferase assay and RNA pull-down. E ARMC8 mRNA level in 50 pairs of OS tissues and adjacent tissues was determined using RT-qPCR. F ARMC8 mRNA level in human OS cell lines (U2OS, HOS, SAOS2) and normal osteoblasts (hFOB1.1) was determined using RT-qPCR. G The correlation between ARMC8, miR-142-5p, HDAC5, and METTL3 in OS tissues was analyzed using Pearson correlation analysis. H ARMC8 mRNA level in OS cells after intervention with METTL3 and HDAC5 expression was measured using RT-qPCR. N = 50. The cell experiment was repeated three times independently. Data in panels C-D/F-H are presented as mean ± standard deviation. Data in panel E were analyzed using paired t-test. Data in panels F/H were analyzed using one-way ANOVA, and data in panels C, D were analyzed using two-way ANOVA, followed by Tukey’s multiple comparisons test, *p < 0.05, **p < 0.01.