AbstractBackgroundExtracellular vesicles (EVs) participate in the spread of pathological proteins between cells, contributing to the progression of neurodegenerative disorders, including Alzheimer’s disease (AD). As cerebrospinal fluid (CSF) reflects the biochemical environment of the brain, EVs from CSF constitutes a promising source to investigate novel biomarkers and therapeutic targets. We aimed to explore the unbiased proteome profile of CSF‐derived EVs from individuals with AD by label‐free quantitative liquid chromatography mass spectrometry (LC‐MS/MS).MethodEVs were obtained from 500 µL of CSF from sex and age‐matched cognitively healthy donors (HC, n = 10, mean age = 73±1.4) and patients with AD dementia (AD, n = 10, mean age = 76.2±0.9) by a high‐throughput size exclusion chromatography method (SmartSEC). Resultant EV fractions were characterized by nanoparticle tracking analysis (NTA). EVs were lysed and proteins were trypsinized overnight via S‐Trap filters (Protifi). The peptides were separated on an Evosep nano LC system and analyzed by the TimsTOF Pro (Bruker) mass spectrometer. Peptides were identified using PEAKS Online software (1% FDR). DAVID Bioinformatics Resources 6.8 was used to determine the gene ontology (GO) enrichment. We normalized log2 peptide intensities (quantiles) and determined the differential protein expression across groups by a linear mixed effects (fixed effects = diagnosis, random effects = peptide sequence) model using the MsqRob package in R. P‐values were adjusted using Benjamini‐Hochberg.ResultNTA showed homogeneous EV size profiles (mean size = 113.4±1.41nm). We identified 2096 proteins with at least 1 proteotypic peptide. GO enrichment analyses showed “extracellular exosome” as the most statistically significant enriched term (p = 1.8×10−212), confirming the enrichment of EVs. Statistical inference analysis revealed 18 differentially expressed proteins between both diagnostic groups (adj.p<0.05), all of them increased in the AD group. (Fig 1). Some of the greatest changes include CLUS (log2FC = 0.61, adj.p = 8.38×10−4, Fig.2), SPB1 (log2FC = 1.15, adj.p = 1.23×10−2, Fig.3), LRP1 (log2FC = 3.43, adj.p = 1.23×10−2), KLK7 (log2FC = 2.59, adj.p = 1.23×10−2), SAP3 (log2FC = 2.09, adj.p = 3.1×10−2) and SHPS1 (log2FC = 1.03, adj.p = 4.25×10−3), which are proteins that have been previously associated with inflammation.ConclusionOur results reveal novel protein changes in CSF‐derived EVs from individuals with AD, which could be explored further to their applicability in distinguishing the different clinical stages across the AD continuum.