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, therapeutics, 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, concentration, 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.

Characterization and function of extracellular vesicles in a canine mammary tumour cell line: Ultracentrifugation versus size exclusion chromatography

“…A previous study performed on N2a neuroblastoma cells, showed trough NTA that cells cultured in serum free conditions shed more EVs than cells cultured with EV‐depleted serum 28 . In our study, the absence of difference in particle concentration between EV from FBSf and 5% EV‐dFBS medium may also be due to different cellular behaviour (N2a vs. CIPp), different culture conditions or by the impossibility to distinguish EV from non‐EV particles 5,29 . Anyhow, based on this data and as already mentioned in other studies, 5,27 removal of FBS from CCM is strongly suggested to avoid any biases when working with CCM‐derived EVs.…”

“…This latter lack of effect may be related to FBS co-isolated proteins, which could have interfered with EVs, or to starving which might have changed EV composition or quantity. 5,28,29 Another technical issue in EV functional studies is the EV/recipient cells treatment ratio, for which standardization and overlapping with in vivo situation is arduous. 38 In our assays, EV concentration might have been too low or different from physiological concentration.…”

“…Co‐isolated particles might have also influenced the proliferation assays, where we evidenced a slight increase in cell proliferation only when treating cells with EVs isolated from FBSf medium compared to controls, but not when treating cells with EVs from 5% EV‐dFBS. This latter lack of effect may be related to FBS co‐isolated proteins, which could have interfered with EVs, or to starving which might have changed EV composition or quantity 5,28,29 …”

Proteomics profiling identifies extracellular vesicles’ cargo associated with tumour cell induced platelet aggregation

“…Immunoblotting was performed as previously described [19], using 10 μg of EVs or cell lysates. Primary antihuman antibodies used were to: PDGFRβ (1 μg/ mL) (R&Dsystems, Cat.…”

“…EV size and concentration was measured using NanoSight NS500 as we previously described [19]. EVs were captured at 30 frame/s speed and six 60 second videos were recorded.…”