Comparative Proteomics Reveals Cell-Type-Specific Heterogeneity of Human Neural Cell-Derived Extracellular Vesicles

Extracellular vesicles (EVs) are increasingly explored as biomarkers and as drug delivery vehicles for the brain, yet how EV cargo differs among neural cell types at baseline remains poorly defined. This gap limits both the interpretation of disease-associated EV changes and the rational choice of EV source for therapeutic use. Here, we compared EVs from human neuronal-like SH-SY5Y cells (N-EV), astrocytes (AS-EV), and HMC3 microglia (MG-EV) side by side using tandem mass tag (TMT)-based quantitative proteomics, with all samples analyzed in a single multiplexed batch. EV identity was supported by TEM, nanoparticle tracking analysis, and canonical MISEV2023 markers. Size, zeta potential, and protein content per vesicle were comparable across EV types, whereas N-EV showed the highest particle concentration and protein yield. Enrichment analyses and consensus clustering showed cargo consistent with parental-cell functions: neuronal connectivity and structural remodeling in N-EV, extracellular support and cellular interaction in AS-EV, and immune surveillance in MG-EV. Cross-comparison also revealed patterns beyond these canonical roles, including complement-associated cargo shared by N-EV and MG-EV and a Glial-EV signature of chromatin-associated and epigenetic regulatory proteins shared by AS-EV and MG-EV. We further defined cell-type-specific signature proteins; interaction networks of the larger sets were consistent with the enrichment results. Overall, these data provide a baseline comparative reference of EV proteomes from human neural cell models for interpreting disease-associated alterations, selecting EV sources for therapeutic use, and generating hypotheses on selective cargo sorting.