Fig 2: Senescence induces increased secretion of sEV-associated EphA2.(a) Comparative proteomic analysis of sEVs secreted from control and DXR-induced senescent RPE-1 cells by using mass spectrometry. The number of peptides detected in sEVs secreted from control and DXR-induced senescent RPE-1 cells are plotted for the identified proteins. Blue and red plots represent the data for the proteins significantly enriched in sEVs secreted from control and DXR-induced senescent RPE-1 cells, respectively. Green plots represent the data for the other proteins. (b) Immunoblotting of EphA2, Alix and β-actin in the sEV fraction and WCL of control and DXR-induced senescent RPE-1 cells. The same number of sEVs was loaded in each lane. Dot plot represents the relative density of sEV-associated EphA2 analysed by ImageJ. (c) Immunoblotting of indicated proteins in the sEV fraction, CM and WCL of DXR-induced senescent RPE-1 cells expressing non-targeting shRNA (shNT) or Rab35 shRNA (shRab35). Dot plot represents the relative density of sEV-associated EphA2 analysed by ImageJ. (d) Immunoblotting of indicated proteins in the sEV fraction and WCL of pre-senescent control and senescent TIG-3 cells. Senescence was induced by serial passage, oncogenic Ras expression or DXR-treatment. The same number of sEVs was loaded in each lane. Dot plots represent the relative density of sEV-associated EphA2 analysed by ImageJ. Replicates are biological replicates (n=3). Error bars indicate s.d. *P<0.05 (two-tailed t-test).

Fig 3: EphA2 phosphorylation facilitates its sorting into sEVs in senescent cells.(a) EphA2 immunoprecipitates prepared from control and DXR-induced senescent RPE-1 cells were immunoblotted with anti-phosphotyrosine and anti-EphA2 antibody. Dot plot represents the relative density of phospho-EphA2/EphA2 analysed by ImageJ. (b) Immunoblotting of indicated proteins in the sEV fraction and WCL of DXR-induced senescent RPE-1 cells expressing wild-type or phospho-resistant EphA2 tagged with FLAG. Dot plot represents the relative density of sEV-associated EphA2 analysed by ImageJ. (c) Relative PTP1B activity in control and DXR-induced senescent RPE-1 cells. (d) Relative PTP1B activity in DXR-induced senescent RPE-1 cells expressing empty vector or ectopic PTP1B. (e) Immunoprecipitates prepared from control and DXR-induced senescent RPE-1 cells expressing empty vector or ectopic PTP1B were immunoblotted with anti-phosphotyrosine and anti-EphA2 antibody. Dot plot represents the relative density of phospho-EphA2/EphA2 analysed by ImageJ. (f) Immunoblotting of indicated proteins in the sEV fraction and WCL of DXR-induced senescent RPE-1 cells expressing empty vector or ectopic PTP1B. Dot plot represents the relative density of sEV-associated EphA2 analysed by ImageJ. Replicates are biological replicates (n=3). Error bars indicate s.d. *P<0.05 (two-tailed t-test for a–d,f and one-way ANOVA with the post-hoc Dunnett's two-tailed test for e).

Fig 4: Identification of non-competing anti-EphA2 mAbs.(A) Histograms for EphA2-positive cell lines HEC-1-A and PC-3 first stained with unlabeled EphA2 mAb 1C1 or 3035, an IgG control, or buffer only, followed by staining with either labeled 1C1 (1C1-A488) or 3035 (3035-A488). A488 fluorescence is detected by conducting flow cytometry on 10,000 live, single cells. (B) Percent staining by 1C1-A488 or 3035-A488 relative to the HEC-1-A or PC-3 samples first incubated with buffer only.

Fig 5: Internalization interplay between two anti-EphA2 mAbs monitored by immunofluorescence.Surface-bound 3035 (red color) and 1C1 (green color) were concurrently internalized in PC-3 cells for 0 or 1 hr at 37°C and processed for confocal immunofluorescence microscopy with the lysosomal marker LAMP1 (blue color). Scale bar, 45 micron. White arrows illustrate colocalization. Representative staining from one of multiple independent experiments is shown.