Evidence for the need to evaluate more than one source of extracellular vesicles, rather than single or pooled samples only, when comparing extracellular vesicles separation methods

Abstract

To study and exploit extracellular vesicles (EVs) for clinical benefit as biomarkers, thera-peutics, or drug delivery vehicles in diseases such as cancer, typically we need to separate them from the biofluid into which they have been released by their cells of origin. For cultured cells, this fluid is conditioned medium (CM). Previous studies comparing EV separation approaches have typically focused on CM from one cell line or pooled samples of other biofluids. We hypothesize that this is inadequate and that extrapolating from a single source of EVs may not be informative. Thus, in our study of methods not previous compared (i.e., the original differential ultracentrifugation (dUC) method and a PEG followed by ultracentrifugation (PEG + UC) method), we analyzed CM from three different HER2-positive breast cancer cell lines (SKBR3, EFM192A, HCC1954) that grow in the same culture medium type. CM from each was collected and equally divided between both protocols. The resulting isolates were compared on seven characteristics/parameters including particle size, concen-tration, structure/morphology, protein content, purity, detection of five EV markers, and presence of HER2. Both dUC and PEG + UC generated reproducible data for any given breast cancer cell lines’ CM. However, the seven characteristics of the EV isolates were cell line-and method-dependent. This suggests the need to include more than one EV source, rather than a single or pooled sample, when selecting an EV separation method to be advanced for either research or clinical purposes.