Abstract
Robust, well-characterized methods for purifying small extracellular vesicles (sEV) from blood are needed before their potential as disease biomarkers can be realized. Here, we compared isolation of sEV from serum by differential ultracentrifugation (DUC) and by exclusion chromatography using commercially available Exo-spin™ columns. We show that sEV can be purified by both methods but Exo-spin™ columns contain copious additional particles recorded by nanoparticle tracking analysis, invalidating its use for quantifying yields. DUC samples contained higher concentrations of exosome specific proteins CD9, CD63 and CD81 and electron microscopy confirmed that most particles in DUC preparations were sEV, whereas Exo-spin™ samples also contained copious co-purified plasma lipids. MACSPlex bead analysis identified multiple exosome surface proteins, with stronger signals in DUC samples, enabling detection of 21 of 37, compared to only 10 in Exo-spin™ samples. Nevertheless, the pattern of expression was consistent in both preparations, indicating that lipids do not interfere with bead-based technologies. Thus, both DUC and Exo-spin™ can be used to isolate sEV from human serum and what is most appropriate depends on the subsequent use of sEV. In summary, Exo-spin™ enables isolation of sEV from blood with vesicle populations similar to the ones recovered by DUC, but with lower concentrations.