Fig 2: METTL14 enhances the stability of USP38 mRNA via YTHDF2.A. The relative expression of YTHDF2 in human normal uroepithelium cell line (SV-HUC-1) and BCa cell lines (T24, 5637, J82 and SW780) was evaluated by RT-qPCR. (n = 3) Each bar represents mean; error bars represent SD. P values were determined by one-way ANOVA followed by Dunnett’s test. B. The overexpression efficiency of pcDNA3.1-YTHDF2 was verified through RT-qPCR (left panel). (n = 3) Each bar represents mean; error bars represent SD. P values were determined by Student’s t test. Western blot of YTHDF2 protein level was shown (right panels). C. RT-qPCR was conducted to assess the relative expression of USP38 before and after the overexpression of YTHDF2 in BCa cells. (n = 3) Each bar represents mean; error bars represent SD. P values were determined by Student’s t test. D. RIP assays were applied to explore the relation between YTHDF2 and USP38 mRNA in BCa cells. (n = 3) Each bar represents mean; error bars represent SD. P values were determined by Student’s t test. E. RT-qPCR was used to assess the relative level of USP38 mRNA after the treatment of ActD with or without the overexpression of YTHDF2 in BCa cells. (n = 3) Each point represents mean; error bars represent SD. P values were determined by Student’s t test. F. Western blot analysis was applied to measure the protein level of YTHDF2 before and after the overexpression of METTL14 in BCa cells. G. The enrichment of USP38 mRNA in YTHDF2 antibody precipitates was measured in RIP assays before and after METTL14 overexpression. (n = 3) Each bar represents mean; error bars represent SD. P values were determined by two-way ANOVA followed by Tukey’s test. H. The knockdown efficiency of sh-YTHDF2 was tested by RT-qPCR (left panel). (n = 3) Each bar represents mean; error bars represent SD. P values were determined by Student’s t test. Western blot analysis of YTHDF2 protein level in YTHDF2-silenced cells was shown (right panels). I. The relative level of USP38 mRNA was measured via RT-qPCR in the presence of pcDNA3.1-METTL14 (and sh-YTHDF2) in BCa cells. (n = 3) Each point represents mean; error bars represent SD. P values were determined by one-way ANOVA followed by Dunnett’s test. **P < 0.01. See also S4 Data.

Fig 3: Effect of YTHDF2 on TMZ resistance in GBM cells. (a) The T98G and LN229 cells were treated with TMZ at different concentrations of 0, 150, 300, 600, 1200 and 2400 µM for 48 or 72 h. The IC50 value in each group was obtained. (b) The T98G and LN229 cells were infected with LV-YTHDF2 or LV-NC followed by 200 µm TMZ treatment for 0, 24, 48, 72 and 96 h. A CCK8 assay was performed to assess cell viability. (c, d) After infected with LV-YTHDF2 or LV-NC, the T98G and LN229 cells were treated with TMZ at 200 µm for 48 h. The cell proliferation and apoptosis were evaluated by an EdU assay and flow cytometry assay. (e) The T98G/TR and LN229/TR cells were infected with LV-sh-YTHDF2 or LV-sh-NC followed by 200 µm TMZ treatment for 0, 24, 48, 72 and 96 h. The cell viability was measured by the CCK8 assay. (f, g) The T98G/TR and LN229/TR cells infected with LV-sh-YTHDF2 or LV-sh-NC were treated with 200 µm TMZ for 48 h. The cell proliferation and apoptosis were detected. (h, i) T98G cells were infected with LV-YTHDF2 or LV-NC, and T98G/TR cells were infected with LV-sh-YTHDF2 or LV-sh-NC. Cells were then injected subcutaneously into the flanks of nude mice, and the mice were subsequently treated with 25 mg kg-1 TMZ by intraperitoneal injection every 2 days for 21 days. The intraperitoneal injection of saline was performed on the mice that had no TMZ injection. The experimental mice were grouped as follows (n = 7, per group): LV-NC + saline, LV-YTHDF + saline, LV-NC + TMZ and LV-YTHDF + TMZ; or LV-sh-NC + saline, LV-sh-YTHDF2 + saline, LV-sh-NC + TMZ and LV-sh-YTHDF2 + TMZ. After TMZ injection, the tumor size was measured every 2 days for 21 days. At day 21, the tumors were collected and photographed. The cell data are from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 vs. control or LV-NC or LV-sh-NC. NC, negative control; TMZ, temozolomide.

Fig 4: YTHDF2 activated NF-?B signalling through inhibiting TNFAIP3 expression. (a, b) T98G and LN229 cells infected with LV-YTHDF2 or LV-NC, or co-infected with LV-YTHDF2 and LV-TNFAIP3. The nuclear and cytoplasm was separated, and p65 protein level in the nuclear was measured. (c, d) T98G/TR and LN229/TR cells were infected with LV-sh-YTHDF2 or LV-sh-NC, or infected with LV-sh-YTHDF2 and transfected with si-TNFAIP3, following by the nuclear p65 protein level detection. (e) The LV-YTHDF2 was co-infected with LV-TNFAIP3 or LV-NC in T98G and LN229 cells, or cells were infected with LV-YTHDF2 and treated with JSH-23 (NF-?B inhibitor, 20 µm) for 24 h. Next, 200 µm TMZ was added to cells and cultured for 48 h. Cell proliferation was determined by an EdU assay. (f) LV-sh-YTHDF2 was infected into T98G/TR and LN229/TR cells followed by si-TNFAIP3 or si-NC transfection with or without treatment of JSH-23 (NF-?B inhibitor, 20 µm). After incubation with TMZ (200 µm) for 48 h, the percentage of EdU-positive cells was calculated. (g) After co-infection of LV-YTHDF2 and LV-NC or LV-TNFAIP3, or treated with 20 µm JSH-23 after infection of LV-YTHDF2, the apoptosis of T98G and LN229 cells was evaluated by flow cytometry. (h) The apoptosis of T98G/TR and LN229/TR cells was assessed after infection of LV-sh-YTHDF2 and transfection of si-NC or si-TNFAIP3 with or without 20 µm JSH-23 treatment. Data are from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 vs. NC; #P < 0.05 vs. si-TNFAIP3; NC: LV-NC or si-NC; Scale bar: 100 µm.

Fig 5: YTHDF2 decreased the mRNA stability of EPHB3 in an m6A manner. (a, b) The LV‐YTHDF2 or LV‐NC was infected into T98G and LN229 cells, and the LV‐sh‐YTHDF2 or LV‐sh‐NC was infected into T98G/TR and LN229/TR cells. The protein and mRNA level of EPHB3 was measured by Western blotting and qRT‐PCR. (c) RNA immunoprecipitation (RIP) was performed in T98G/TR and LN229/TR cells using anti‐YTHDF2 antibody. (d) FISH in conjugation with fluorescent immunostaining analysis was performed in T98G/TR and LN229/TR cells. (e) T98G/TR and LN229/TR cells were infected with LV‐sh‐YTHDF2 or LV‐sh‐NC followed by treatment of 5 μm Actinomycin‐D for 0, 3 and 6 h. The relative mRNA level of EPHB3 was measured. (f) T98G and LN229 cells infected with LV‐YTHDF2 or LV‐NC were treated with 5 μm Actinomycin‐D for 0, 3 and 6 h followed by the detection of EPHB3 expression. (g) Correlation analysis of YTHDF2 and PTEN expression in GBM tissues (r = −0.3157, P < 0.05). (h) The sequence of luciferase reporter gene vector EPHB3‐3′UTR (EPHB3‐WT) and EPHB3‐3′UTR with mutant m6A sites (EPHB3‐Mut) were established and exhibited. (i, j) A dual‐luciferase reporter gene assay was performed in T98G, LN229, LN229/TR, and T98G/TR cells to detect the interaction between YTHDF2 and EPHB3. Data are from three independent experiments.*P < 0.05, **P < 0.01 vs. LV‐sh‐NC or LV‐NC or anti‐IgG. Scale bar: 20 μm.