Fig 2: Upregulation of EPHB4 enhances the proliferation, migration and invasion abilities of OSCC cells. A. PCR revealed that EPHB4 expression was successfully up-regulated in SCC9 and UM1 cells after lentiviral transfection. B. Western blot analysis confirmed that EPHB4 was up-regulated in SCC9 and UM1 cells after lentiviral transfection. C. After upregulating the expression of EPHB4, the proliferative ability of SCC9 and UM1 cells was enhanced. D. Overexpression of EPHB4 enhanced the colony-forming ability of SCC9 and UM1 cells. E-F. Transwell assays demonstrated that the overexpression of EPHB4 promoted the migration and invasion ability of SCC9 and UM1 cells (scale bar, 200 μm).

Fig 3: EPHB4 is highly expressed in OSCC compared with ANC tissues. A. Typical IHC results representing low and high EPHB4 expression in OSCC and ANC samples are shown (scale bar, 100 μm). B. Kaplan-Meier analysis revealed that patients with higher EPHB4 expression had a lower survival rate and worse prognosis. C. Western blot analysis confirmed that EPHB4 expression in OSCC cell lines (UM1 and SCC9) was significantly higher than that in NOK cells.

Fig 4: Influence of EphrinB2/EphB4 signaling on angiogenic markers in WJ-MSCs under serum-deprivation. (a) The mRNA expression levels of EphB4 and EphrinB2 were determined by qRT-PCR analysis in control and serum-deprived WJ-MSCs. GAPDH was used as the endogenous control (n = 4). (b) WJ-MSCs were cultured for 48 h under serum-deprivation in the presence of 8 μM of EphB4 inhibitor, NVP-BHG712. Total EphB4 and phosphorylated EphB4 (p-EphB4) expression levels were quantified by Western blotting to validate the action of NVP-BHG712 (n = 3). Representative blot images are shown. Band intensities were first quantified relative to GAPDH, which was used as a loading control. Then, the relative expression of p-EphB4 with respect to total EphB4 was calculated and plotted. (c) mRNA expression level of ADAMTS13, VEGF and PDGF, under serum-deprivation condition in presence of NVP-BHG712 was analysed by qRT-PCR (n = 3). (d) Protein expression pattern was detected by Western blotting (n = 3) for the above mentioned markers, representative blot images are shown. (e) WJ-MSCs were transfected with ADAMTS13 or NC siRNA and exposed to serum-deprivation stress for 48 h. Total EphB4 and phosphorylated EphB4 (p-EphB4) expression levels were quantified by Western blotting, and plotted; representative blot images are shown (n = 3). (f) WJ-MSCs were treated with NVP-BHG712 for 48 h under serum-deprivation and total ERK and phosphorylated ERK (p-ERK) expression levels were evaluated by Western blotting (n = 3). Representative blot images are shown. Band intensities were first quantified relative to GAPDH, which was used as a loading control. Then, the relative expression of p-ERK with respect to total ERK was calculated and plotted. (g) WJ-MSCs were treated with 30 μM of ERK pathway inhibitor, FR180204, for 48 h under serum-deprivation and mRNA expression levels of ADAMTS13 and angiogenic markers, VEGF and PDGF were analysed by qRT-PCR (n = 3). (h) Protein expression pattern for ADAMTS13, VEGF and PDGF under the same treatment was detected by Western blot analysis and representative blot images are shown (n = 3). Serum-deprived condition has been denoted as no-serum (NS), negative control as NC and control as CT in the figure. Each bar represents mean ± SEM (*p < 0.05, **p < 0.01). Statistical comparisons were assessed using one-way ANOVA followed by Bonferroni’s multiple comparison test. Original blots are presented in Supplementary Fig. S7i–m.

Fig 5: EPHB4 interacted with HMGB1 and activated NF-κB signalling. A. Mass spectrometry experiments detected 22 proteins that may specifically bind to EPHB4 in cells compared with control cells. B. HMGB1, which was identified to have a high binding efficiency to EHPB4, is marked with a red box. C-D. An endogenous EPHB4-IP assay verified that HMGB1 was immunoprecipitated by an anti-EPHB4 antibody, while an endogenous HMGB1-IP assay showed that EPHB4 could be immunoprecipitated by HMGB1 in turn, which jointly substantiated the mutual immunoprecipitation between EPHB4 and HMGB1. E. Western blot assays verified that when EPHB4 expression was down-regulated, HMGB1 expression decreased, and the expression of P-P65 and P-IKBα in the NF-κB signalling pathway was also inhibited. F. Conversely, P-P65 and P-IKBα were significantly increased in EPHB4-overexpressing cells. G. Protein stability experiments showed that HMGB1 stability decreased following the downregulation of EPHB4.