3.1. Exosomal Proteomes Construction

Exosomal proteomes for the three biofluids (CSF, serum and plasma) were obtained from three online databases: EXOCARTA, Vesiclepedia and EVpedia (all accessed on 12 January 2020). All databases comprise protein, lipids and nucleic acids data and are all manually curated. EXOCARTA was the first of these databases to be launch and includes data only from exosomes (http://www.exocarta.org/) [148,149,150,151]. Vesiclepedia was launched later and, now, it gathers information from all classes of extracellular vesicles among them exosomes, microparticles and microvesicles (http://microvesicles.org/) [152,153]. EVpedia is the only one of the three databases which includes eukaryotic and prokaryotic extracellular vesicles data (http://student4.postech.ac.kr/evpedia2_xe/xe/) [154,155]. In the work described here, only human proteomes were considered.

The exosomal protein cargo for each biofluid was refined through several steps (Figure 1), considering the following: number of entries for the specific biofluid, number of entries for H. sapiens, number of entries for H. sapiens and exosomes, number of ID entries for gene names of H. sapiens and exosomes. Finally, repeated gene names were removed to obtain the final biofluid gene list. The EVpedia output was in the form of Uniprot IDs and thus it was necessary to convert this data to gene names. To achieve this the Uniprot Retrieve/ID mapping tool was used to convert the Uniprot IDs to gene names (https://www.uniprot.org/uploadlists/; accessed on 14 January 2020).

The proteome list for each biofluid includes every protein/corresponding gene name present in at least one of the databases. In the case of CSF, only data from EVpedia was collected since no information was available in the other databases.

Additionally, the proteomes obtained from the databases were complemented by including information from a literature search on PubMed (https://PubMed.ncbi.nlm.nih.gov/; accessed on 17 January 2020) using the keywords: “exosomes and Alzheimer´s disease and CSF or Serum or Plasma” (Supplementary Table S4). The protein name was converted to gene name, as described above. Thus, the final proteome lists for each biofluid included information collected from the databases and from the literature search (Supplementary Tables S5–S7). These final lists will be herein referred to as the distinct biofluid proteomes.