Fig 1.
Plasma samples preserved in the presence of DMSO were pooled as described in Materials and Methods. Exosome pellets were obtained by differential centrifugation and removal of contaminant plasma proteins was improved by the use of sucrose gradient. Density fractions in the range (1,07–1,20 ± 0.02) g/ml were further processed for the label-free quantitative (LFQ) proteomics analyses by Liquid-Chromatography Tandem Mass Spectrometry (LC-MS/MS). Identified proteins quantified in at least two replicas (AP, IP) were submitted to hierarchical clustering, and those with Protein Abundance Profiles (PAPs) showing Pearson correlation R ≥ 0.7 (p ≤ 0.05) were selected (APs, IPs) for complete clustering, which included non-filtered control sample proteins (Ctr). Potential biomarkers belonging to E. granulosus were specifically identified in clusters of APs or IPs. Human proteins in APs and IPs (APhs, IPhs) were analyzed to obtain the interaction networks of enriched GO Biological Process and Reactome Pathways for each of APhs and IPhs (String Networks); a graph overlying both APhs and IPhs (Dynet graph) was built, which helped distinguishing between shared and specific proteins among the most central nodes. By analysing these outcomes, we could identify central proteins specifically enriched in exosomes from active or inactive CE, such as SFKs and TGF-β in active CE and Cdc42/Rac in inactive CE.
Table 1.
Description of pool samples.
Fig 2.
Biochemical characterization of EVs isolated from plasma pools from healthy subjects. (A) About 20% of sample volume from gradient fractions EXO/MV 5 was fixed for TEM analyses. (B) The diameter size distribution of particles is shown in percentage of total particles number and (C) in box-plot diagram. The median value of exosome diameter was 102 nm, while that of microvesicles was 158nm. (D) WB analysis of exosome markers CD81 (1:400), Synthenin (1:400), TFR 1 (1:1000), exposition time 1 min.
Fig 3.
Venn diagram summarizing protein datasets results.
APhs: selected human proteins from exosomes of patients with active CE; IPhs: selected human proteins from exosomes of patients with inactive CE; Ctr: control; Top100_ExoCarta: the 100 proteins most frequently identified in exosome proteomics studies.
Fig 4.
(A) We clustered together the proteins in the APs and IPs datasets plus all the proteins identified in the two control replicas. From left to right: we divided the hierarchical dendrogram obtained (Cluster 3.0) in 15 clusters (circled in red in the dendrogram) (named cl1-cl15) by the values of the Pearson correlation coefficients which are reported next to each cluster number. The median normalized abundance profile of the proteins belonging to each cluster is shown for each sample (A5-7, I5-7, and C5-7 indicate active, inactive and control respectively, the subscripts 5–7 referring to fraction numbers). The String Enrichment analysis was performed for each cluster considering only: high-confidence interactions (> 0.7); known interactions: curated database, experiments; protein-protein interaction (PPI) probability and false Discovery rate (FDR) < 10E-3. Shared proteins identified in all samples (clusters cl1-cl2) resulted enriched in vesicle-mediated transport and platelet degranulation; proteins of cl3, identified in APs and Ctr only, were enriched in transport and immune system processes; the leukocyte-mediated immunity and vesicle-mediated transport pathways were enriched in clusters specific of APhs (cl6-cl7) and IPhs (cl13-14), with FDR values of 1E-18 and 1E-8 respectively. Control sample proteins fell in cl5 and cl11. E. granulosus proteins fell into the clusters corresponding to APs-selective (cl6-cl7) and IPs-selective proteins (cl13-cl14). (B) For each cluster, the number of proteins previously identified in the investigated literature are reported. The protein content-based EV characterization (numbered from 1 to 5 as reported in MISEV2018, [28]); the ones recognized to be involved in exosome release [52, 53], and we extended this definition also to the E. granulosus (ECHGR) Rab-11a; the proteins most frequently identified in exosome studies (top 10/20/100 genes from Exocarta [49]); the proteins already identified in CE-related EVs [54]; and finally, the remaining “other proteins”.
Table 2.
List of E. granulosus proteins identified.
For each protein the presence in active (APs) or inactive (IPs) samples, the cluster annotation, the previous identification in EVs from HCF (HCF-EVs) and bibliographic references are reported.
Fig 5.
Comparative analysis of proteins detected in fraction 6 of Active, Inactive, and Control samples.
(A) Venn diagram showing the distribution of the proteins detected in fractions 6 datasets. (B) Principal component analysis (PCA) plot of samples based on fraction 6 datasets, i.e. the median intensity values of the proteins identified in fractions 6 of the replicas of each sample. The position of conditions in the two generated main components is plotted. Variability explained by the first and second components of the PCA is indicated between brackets.
Fig 6.
String Networks of APhs and IPhs.
The String App on Cytoscape 3.1 platform was used to obtain the interaction networks of human proteins identified in the active (APhs) or inactive (IPhs) datasets respectively. STRING database Version 11.0 (string-db.com) was set as follows: high confidence ≥ 0.7; only known interaction: curated database; experiments. The node degree is visualized by node size and informs about the number of interactions of that protein; the betweenness centrality is visualized by node colour, with colours from orange to blue indicating increasing protein relevance in the network signalling; the edge score (confidence of protein-protein interaction) is visualised by direct proportional to the edge size. For STRING Enrichment analyses, the protein-protein interaction (PPI) probability threshold was set to 1.0E-16 and the FDR threshold to 1.0E10-3. Some of the enriched terms reported in Tab.2 are displayed by means of a colour surrounding the node, as follows. Specific APhs terms: Antigen processing and presentation of exogenous peptide antigen via MHC class II, lilac; Interleukin-4 and Interleukin-13 signalling, fuchsia; positive regulation of innate immune response, purple; regulation of ERK1 and ERK2 cascade, orange. Specific IPhs terms: gene and protein expression by JAK-STAT signalling after Interleukin-12 stimulation, bright green; Interferon-gamma-mediated signalling pathway, positive regulation of NF-kappaB transcription factor activity and regulation of TNF production, light green. Common terms: immune effector process, yellow; integrin-mediated signalling pathway, brown; interspecies interaction between organisms, blue; leukocyte mediated immunity, pink; wound healing: sky blue.
Fig 7.
The APhs and IPhs networks were superimposed in one graph by the use of the Dynet application, to highlight patient group-specific and shared central proteins. Proteins identified only in samples from individuals with active CE are coloured in red, proteins identified only in inactive CE samples in green, and proteins identified in both samples in white. The node degree (number of interactions) is visualized by node size and the edge score (confidence of node interaction) is directly proportional to the edge size. On the right side are clearly visible the central hubs of APhs network, i.e. the proteins Src, TGF-β and the integrins αM, αV, αL (ITGAM, ITAV, ITAL).
Fig 8.
The STRING networks corresponding to the GO Term “Immune effector” for APhs (81 proteins; FDR = 9.45E-35) and IPhs (55 proteins; FDR = 7.62E-24) are shown. Besides the different number of proteins involved in the two networks, the term results highly enriched in both datasets but the signalling is based on different proteins: Src, Lyn and certain Integrins (ITGAV, ITGAM, ITGAL) in APhs vs Cdc-42, several Rab proteins and Annexin 2 in IPhs.
Fig 9.
Networks of interspecies interaction between organisms.
The networks of GO Biological Process Term “Interspecies interaction between organisms” for APhs (46 proteins; FDR = 4.37E-14) and IPhs (39 proteins; FDR = 2.97E-15) are shown. The two networks are very similar in the protein number and interaction confidence, but the signalling is based on different proteins: Src, Lyn, several Integrins (ITGAV, ITGB1, ITGB3, ITGA2), TGF-β1 in APhs vs Ccd-42, Annexin 2 in IPhs.
Table 3.
A selection of GO Biological Process and Reactome Pathways enriched in APhs and/or IPhs.
For each Process or Pathway listed, the number of genes belonging to the Term and the false discovery rate (FDR) probability are reported. The colours used in Fig 6 are used to highlight the specific GO terms and are reported in the last column.
Fig 10.
Validation by WB of some exosome markers.
We validated the presence of SFK Lyn and Src specifically in active CE samples on a completely independent EXO preparation. On the left are reported the PAPs of EXO markers (Alix, Tfr1, HspA8) and SFK Lyn and Src identified by proteomics as reported in S4 Table. On the right, the WB performed on fraction 6 of new EXO preparation performed from pools of plasma samples (Tab.1) collected at the outpatient clinic for CE of San Matteo Hospital Foundation (Pavia, Italy). These plasma samples were not processed before freezing (not centrifuged at 10,000 x g), and were stored in absence of DMSO. We were able to confirm the presence of EXO markers such as Tfr-1 and HspA8 in all samples, while the presence of Src and Lyn was detected in samples from patients with active CE, at much lower levels in controls, but not in samples from patients with inactive CE.