Fig 2: Particle number estimation and size of EVs derived from supernatants from 3T3-L1 cells and rat plasma.(A) The total number of particles from 3T3-L1 cells supernatants was estimated by nanoparticle tracking analysis. (B) Area under the curve from the total number of particles from 3T3-L1 cells supernatants. (C) The mean size of EVs from 3T3-L1 cells supernatants. (D) Western blot for CD63, ANXA2, and CD81 in EVs of 3T3-L1 cell supernatants. (E) Relative density of protein of CD63. (F) Relative density of protein of ANXA2. (G) Relative density of protein of CD81. (H) The total number of particles from rat plasma was estimated by nanoparticle tracking analysis. (I) Area under the curve from the total number of particles from rat plasma. (J) The mean size of EVs from rat plasma. Differences were tested by the Mann–Whitney U test. Data are presented as means ± SE.

Fig 3: Molecular features of gene subtypes in liver cancer. (A) Venn diagram showing differences in CNVs, methylation and mRNA expression between iC1 and iC2 subtypes. (B) Proportions of ANXA2 gain and loss in iC1 and iC2 subtypes. (C) Proportions of ANXA2 hypermethylation and hypomethylation. (D) Box plots showing the differences in ANXA2 expression between iC1 and iC2 subtypes. (E) Kaplan-Meier survival curves for ANXA2 expression. (F) Proportions of CHAF1B gain and loss in iC1 and iC2 subtypes. (G) The proportion of CHAF1B hypermethylation and hypomethylation. (H) Box plots showing the differences in CHAF1B expression between iC1 and iC2 subtypes. (I) Kaplan-Meier survival analysis results for CHAF1B expression.

Fig 4: Pre-SEC yield efficiency after concentration of SG1 and SG2 by different methods. (a) Total protein (µg) obtained from SG1 and SG2 of MDA-MB-468-derived EVs isolated by Pre-SEC and subsequently concentrated with UC, Pre, or UF. (b) Protein pattern visualized by Coomassie Blue staining from SG1 and SG2 and concentrated by UC, Pre, or UF. (c) Western blot of EV protein markers in samples of SG1 and SG2. (d-h) Relative quantity (as measured by relative intensity, using Exo-NM albumin intensity as reference (n = 3)) of EV protein markers CD9 (d), TSG101 (e), ANXA2 (f), ACTN 4 (g), and MFG-E8 (h) from SG1 and SG2, concentrated by either UC, Pre, or UF. (i) EV abundance calculated as EV marker intensity/total protein ratio

Fig 5: Proteomic analysis of MDA-MB-468 EV subgroups. (a) Venn diagram representing the number of proteins identified for SG1 and SG2. (b) Comparison of the number of proteins identified in SG1 and SG2 with a dataset of shared proteins contained in the EVs of 60 cancer cell lines described by Hurwitz et al. (NCI-60[stringent]). (c) Volcano plot showing the differentially expressed proteins in SG1 and SG2. The significance cut-off was set to a FDR < 0.05 and absolute log2-fold change > 1 (n = 3). The red dots represent enriched proteins in SG1 and blue dots represent enriched proteins in SG2. (d) After flotation of SG1 and SG2 vesicles in an iodixanol gradient, ten fractions were collected, and an equal mass of protein was loaded in a Tricine-SDS-PAGE. The immunoblots showed that EV markers were in the regions of less density of the gradient. Note that SG1 had enriched labeling of FLOT1, MFG-E8, ANXA2, and CD9 in fractions 1 and 2