Influence of the isolation method on characteristics and functional activity of mesenchymal stromal cell-derived extracellular vesicles

Background aims Extracellular vesicle (EV) isolation methods are based on different physicochemical properties and may result in the purification of distinct EV populations. We compared two different isolation methods suitable for producing clinical-grade mesenchymal stromal cell-derived EVs (MSC-EVs)—ion exchange chromatography (IEX) and ultrafiltration (UF)—and evaluated their impact on the composition and functional properties of EVs. Methods EVs were purified from conditioned culture medium using an anion exchange resin (IEX) or Amicon filters with a 100-kDa cutoff (UF) (MilliporeSigma, Burlington, MA, USA). We assessed nanoparticle size and distribution by nanoparticle tracking analysis (NTA) and tunable resistive pulse sensing (TRPS) and morphology by transmission electron microscopy. We also measured protein, lipid and total RNA concentration and immunophenotyped both EV populations by flow cytometry (MACSPlex assay; Miltenyi Biotec, Bergisch Gladbach, Germany). Moreover, immunomodulatory activity was tested using a standardized macrophage polarization assay and T-cell stimulation assay. Finally, proteomic analysis and cytokine quantification were carried out to better characterize both EV populations. Results We found by both TRPS and NTA that IEX and UF yielded a comparable amount of total particles with similar size and distribution. In addition, a similar quantity of lipids was obtained with the two procedures. However, IEX yielded 10-fold higher RNA quantity and a larger amount of proteins than UF. MSC-EVs isolated from IEX and UF were positive for the exosome markers CD9, CD63 and CD81 and showed a comparable surface marker expression pattern. Both populations demonstrated immunomodulatory activity in vitro, as they prevented acquisition of the M1 phenotype in lipopolysaccharide-stimulated macrophages and inhibited acquisition of the activation markers CD69 and CD25 on T cells, but the IEX-EVs exerted a significantly greater immunomodulatory effect on both macrophages and T cells compared with UF-EVs. Proteomic analysis and gene ontology enrichment analysis revealed no major differences between the preparations. Finally, cytokine quantification revealed that IEX-EVs were more enriched in some crucial anti-inflammatory and immunomodulatory cytokines (e.g., IL-2, IL-10, transforming growth factor beta and vascular endothelial growth factor) compared with UF-EVs. Conclusions MSC-EVs isolated by IEX and UF displayed similar physicochemical, phenotypic and functional characteristics. In our conditions, both EV populations demonstrated important anti-inflammatory activity in macrophages and T cells. However, IEX-EVs were more potent than UF-EVs, which may indicate the superiority of this method for the production of clinical-grade EVs.