Nucleic acid loading and fluorescent labeling of isolated extracellular vesicles requires adequate purification

Extracellular vesicles (EVs) are nanosized vesicular structures released by cells to communicate with one another. The growing interest in the (patho)physiological function and potential pharmaceutical application of these vesicles is accompanied by a vast number of new research groups entering this research field and a plethora of different protocols to separate EVs from non-vesicular components. This lack of uniformity often generates conflicting or difficult-to-compare results. Here we provide a comparative analysis of different EV isolation strategies, discussing the purity of the final isolate and highlighting the importance of purity on downstream experimental readouts. First, we show that ultracentrifugation (UC) of B16F10 melanoma cell-derived conditioned medium co-purifies proteins or protein complexes with nuclease activity. Such contaminants should be taken into account when aiming to apply EVs as delivery carriers for exogenous nucleic acids. Second, three commonly used purification strategies (i.e. precipitation, UC and density-gradient centrifugation) were evaluated for their ability to remove non-incorporated fluorescent dye (i.e. the lipophilic PKH67 dye), important when probing EV interactions with cells. For both types of impurities, endogenous and exogenous, density gradient purification outperforms the other evaluated methods. Overall, these results demonstrate that the implementation of stringent purification protocols and adequate controls is of pivotal importance to draw reliable conclusions from downstream experiments performed with EV isolates.