# Molecular Evaluation of Different Enrichment Methods for Extracellular Vesicles from Healthy Subjects’ Biobanked Serum

## TL;DR

This study compares different methods for isolating extracellular vesicles from small amounts of stored human serum to determine which methods best preserve their molecular content for research.

## Contribution

The study evaluates and compares EV enrichment methods from biobanked serum, focusing on their impact on downstream proteomic and miRNA analyses.

## Key findings

- qEV1-70 nm was the most efficient method for proteomics analyses.
- ExoRNeasy provided the broadest miRNA coverage among the tested methods.
- All methods enriched small EVs with tetraspanin surface markers, but with varying concentrations.

## Abstract

Extracellular vesicles (EVs) from human body fluids are valuable tools for biomarker discovery and for exploring the mechanisms underlying various pathologies, including infectious diseases. The translation of EV research into clinical practice is however hindered by the variability in EV pre-clinical investigations. Therefore, standardisation of analytical procedures and reporting policies is essential. Human serum is a key biological matrix for biomarker discovery and is commonly stored within biobanks. Here, we investigated different strategies for EV enrichment from small volumes of biobanked serum and evaluated their impact on EVs’ downstream analyses. EVs were obtained from 250 μL of biobanked serum using ultracentrifugation (UC), size-exclusion chromatography-based methods (ExoSpin-ES, qEV1-35 nm, and qEV1-70 nm), or ExoRNeasy (ER). The resulting EVs were subsequently characterised for morphology, concentration, surface phenotype, and multi-omics profiles. All methods successfully enriched small EVs expressing tetraspanins on their surface, although at different concentrations. The most efficient method for proteomics analyses was qEV1-70 nm, followed by ES, which was more susceptible to contamination by serum proteins. EV-miRNA cargo was effectively profiled in UC-, ES-, and ER-EVs, with the latter providing the broadest miRNA coverage. Our results support the feasibility of using biobanked serum for EV-based research and further highlight the importance of selecting appropriate EV enrichment methods, since they influence both miRNA and protein cargo characterisation.