Figures & data
Table 1. Clinical Characteristics of study subjects in the discovery and validation set.
Figure 1. Validation of EV sample quality.
(A) The workflow of the isolation of lEVs, and sEVs. (B) Electron microscopy (EM) shows typical round morphology of sEVs (left) and lEVs (right) from pleural effusion. (C)-(E) The size distribution of sEVs and lEVs from TPE and MPE group (n = 3) by NTA analysis, which indicated that most of sEVs were smaller than the lEVs in diameter, data presented with mean+SD., and the size distribution of sEVs and lEVs between TPE and MPE groups were similar (p > 0.05, data presented in table s1) (F) Presence of CD81, CD63, CD9 (EV markers), and TSG101, and non-VE marker of Calnexin and GM130, analysed by western blotting. An equal protein amount of 20 μg was loaded for all samples.
Figure 2. Characterization of metabolic profiles from samples of lEVs, sEVs, and pleural effusions (PFs).
(A) Score plots of PCA based on the detected 579 metabolites in the three groups. (B) Venn diagram showing the number of differential metabolites between these three groups. These comparisons were based on the data analysed from100 µL PFs, and the lEVs and sEVs subsequently isolated from 30 mL PFs. (C) (D) (E) Composition of the differential metabolites, based on the number of differential metabolites found in the comparisons of lEVs versus PFs, sEVs versus PFs, and lEVs versus sEVs, respectively.
Figure 3. Enrichment of metabolites in lEVs and sEVs compared to pleural effusions (PFs).
(A) Venn diagram showing the number of enriched metabolites in TPE-lEVs, MPE-lEVs, TPE-sEVs and MPE-sEVs. (B) Heat map of enriched lipids in lEVs or sEVs compared to in PFs, the shades of the colour represents lipid levels (black, is the mean level; yellow, and blue indicates higher level, and lower level, respectively). (C) Histogram of enriched metabolites of amino acids, acyl-carnitines; data were presented by mean±SEM, y-axis shows the metabolites, while x-axis shows the value of log2 (fold-change), and fold-change is the metabolite average level in lEVs or sEVs divided by its average level in pleural effusion. Because we only displayed enriched metabolites in the lEVs or sEVs samples, all values on the x-axis were positive, and the left (red) columns and right (blue) columns show enrichment in lEVs or sEVs, respectively. #: only enriched in the TPE-sEVs; *: only enriched in the MPE-lEVs. Cer: ceramide, PE: phosphatidylethanolamines, DG: diacylglycerol, FA: fatty acid, PC: phosphatidylcholine, SM: sphingomyelin, PS: phosphatidylserine, CE: cholesterol ester.
Figure 4. Differential metabolites in comparison of TPE and MPE from samples of lEVs and sEVs.
(A) Venn diagram of differential metabolites between TPE and MPE from samples of lEVs and sEVs, respectively. (B), (C) Heat maps of these differential metabolites in the samples of lEVs and sEVs, respectively. Depletion is depicted in blue and enrichment in yellow. #: the commonly altered metabolites in lEVs and sEVs. (D) Metabolite enriched pathway analysis based on differential metabolites (except lipids) found in lEVs samples. *: p-value of this pathway is less than 0.05. AA: amino acids, AC: acylcarnitine, PC: phosphatidylcholine, PI: phosphatidylinositol, SM: sphingomyelin, and LPE: lysophosphatidylethanolamine.
Figure 5. Network of interactions between clinical parameters and differential metabolites identified in sEVs (A) and in lEVs (B) subgroups.
Correlation analysis was based on the levels of differential metabolites and clinical parameters. Only correlations with absolute values of correlation coefficients greater than 0.5 and p < 0.05 were left. Each hexagon and dot is a clinical parameter and differential metabolite, respectively. Dot colours indicate the species of the differential metabolites between TPE and MPE (yellow dots: TG and DG; red dots: PC, LPC, PI, PG, PE and LPE; green dots: SM, Cer, and HexCer; light blue dots: acylcarnitines; blue dots: amino acids and related metabolites). The line colours of red and green display positive and negative relationships, respectively.
Figure 6. Identification of metabolic biomarkers candidates for distinguishing TPE and MPE in lEVs.
(A) PLS-DA score plot based on TPE and MPE groups in the discovery set. (B) Venn diagram displays metabolites with VIP values larger than 1 on two principal components (VIP1 and VIP2), and differential metabolites (with p < 0.05, FDR<0.05 and FC>2) when TPE compared with MPE. (C) Relative concentrations of defined biomarkers candidates of leucine, phenylalanine, PC 35:0 and SM 44:3 in the discovery set and validation set. The bottom and top of the box indicate the 25th and 75th percentile, respectively; the line within the box indicates the median. The whiskers below and above the box indicate the 10th and 90thpercentiles, respectively. (D) (E) The area under the receiver-operating characteristic curve of the biomarker panel, based on the combination of biomarkers candidates of leucine, phenylalanine, PC 35:0 and SM 44:3, in the discovery set and validation set, respectively.
Figure 7. Metabolic pathways of differential metabolites between EVs and pleural fluid.
☆, #, &, Δ present the metabolites were significantly changed in the samples of TPE-lEVs, MPE-lEVs, TPE-sEVs, and MPE-sEVs when compared to their corresponding samples of pleural fluid, respectively. The colour of red and blue present the metabolite was increased or decreased in the EVs samples, respectively. PS: phosphatidylserine, PC: phosphatidylcholine, PE: phosphatidylethanolamines, PI: phosphatidylinositol, PG: phosphatidylglycerols, DG: diacylglycerol, TG: triacylglycerol, FA: fatty acid, Cer: ceramide, SM: sphingomyelin, TCA cycle: tricarboxylic acid cycle.
Supplemental material